Professional Documents
Culture Documents
AREAS
PROCEEDINGS OF THE INTERNATIONAL
CONFERENCE ON WATER RESOURCES OF
ARID AND SEMI ARID REGIONS OF
AFRICA (WRASRA), AUGUST 3–6TH 2004,
GABORONE, BOTSWANA
Water Resources of Arid Areas
Edited by
D.Stephenson
Civil Engineering Department, University of Botswana,
Gaborone, Botswana
E.M.Shemang & T.R.Chaoka
Department of Geology, University of Botswana,
Gaborone, Botswana
Preface xii
Keynote address
Africa’s water resources are threatened by population growth with the resultant increase
in water demand, the stresses of water use for various activities, desertification, global
warming and climate change, and other interventions in the water cycle by man. These
effects are more pronounced in the Arid and Semi-Arid regions of Africa in particular
and the world in general. It is therefore vitally important that the water resources in Arid
and Semi-Arid regions are developed and managed in a sustainable and integrated
manner.
Integrated management of water resources in the arid and semi-arid regions of Africa
requires a spectrum of efforts from local and community stakeholders to national and
transboundary river basin management. This conference aims at sharing the best practices
of water use and conservation around the globe.
The main objective of this conference was therefore to bring together educators,
researchers, practitioners, managers, policy makers and NGO’s from Africa in particular
and the world in general involved in various aspects of water resources in arid and semi
arid regions. The more specific objectives of the conference were to
(i) Assess the current state of the art of water resources management in arid and semi arid
regions with particular emphasis on African regions.
(ii) Address the future water stress due to limited water resources, population growth,
increasing demand and pollution and other related risks, resulting in insufficient water
supply.
(iii) Promote dialog and interaction between different disciplines and professions.
(iv) Forster insights into issues of global sustainable development and set concrete targets
to meet the need for drinking water and water borne sanitation in arid and semi arid
countries of Africa and the world in general.
We received an overwhelming response to our call for papers. We received over 120
abstracts and each abstract was reviewed and more than two thirds of the abstracts were
accepted. Authors were then requested to submit full text of papers. The full texts of the
papers were reviewed by the conference organizing committee and 68 papers were finally
accepted for conference.
The papers in this book “Water Resources of Arid and Semi-Arid Regions of Africa”
constitute the conference proceedings. This book is subdivided into seven sections.
Section 1 deals with problems in obtaining data. Section 2 deals with groundwater
recharge: natural and artificial; Section 3 deals with Socio economic aspects of water
demand management; Section 4 deals with geophysical, GIS and remote sensing
techniques for groundwater exploration; Section 5 deals with climate change and its
impact on water resources; Section 6 deals with vulnerability and risk assessment and
Section 7 water management.
This book will be of interest to researchers and practitioners in the field of surface
water hydrology, groundwater hydrology, environmental engineering, agricultural
engineering and earth sciences, as well as those engaged in water resources planning,
development and management in arid and semi arid areas. Graduate students and those
wishing to conduct research in hydrology, environmental science and engineering and
water resources will find the book to be of value.
Dr A.R.Tombale
Permanent Secretary
Ministry of Minerals, Energy and Water Resources, Botswana
Keynote address
Sustainable water management in arid and
semi-arid environments
W.Kinzelbach, P.Bauer, P.Brunner & T.Siegfried
Institute for Hydromechanics and Water Resources Management, Swiss
Federal Institute of Technology, Zurich, Switzerland
Water Resources of Arid Areas—Stephenson, Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN 04 1535 913 9
1 INTRODUCTION
Fresh water is a scarce resource on a worldwide basis. This becomes apparent when
looking at the basic items of the global freshwater balance (Postel et al., 1996). Of the
110,000km3/a of precipitation on the landmass of the earth, 50,000 are returned to the
atmosphere via evapotranspiration by the planet’s natural plant cover. Another 21,000 are
used by man-made ecosystems (18,000km3/a by rain fed agriculture and 3,000km3/a by
irrigated agriculture). This shows that agriculture and natural vegetation are already fierce
competitors for the available freshwater. Of the accessible runoff of 13,000km3/a about
4,000 are appropriated by mankind. 70% of those go into irrigated agriculture. This
means that a global water crisis would above all be a global crisis in food production.
Compared to the 13,000km3/a available the abstracted 4,000 appear small. One should,
however, not forget that these figures are averaged in time and space and therefore hide
the real problem, e.g. droughts and floods. One can still use the ratio a of withdrawals
and available renewable resources as an indicator. Due to the variability of the quantities
involved, it is the experience that a value of a>0.4 already reflects severe scarcity. On a
global scale a=0.31 is found (Alcamo et al., 2003). This indicates that scarcity on a global
level is a reality today, with the arid world already experiencing very severe scarcity
problems.
Water resources of arid areas 4
2 SUSTAINABILITY
– the model is used for the testing of management strategies and/or for optimization, and
finally
– the connection is made to the field of socio-economics.
The North Western Sahara is underlain by one of the world’s largest aquifer systems,
which covers approximately 1,000,000km2 and consists of two major aquifers, the deeper
Continental Intercalaire (CI) and the shallower Complexe Terminal (CT) (Fig. 1). Their
water resources are being utilized by the three countries Algeria, Tunisia and Libya
mainly for irrigation purposes. The system nowadays hardly receives any recharge. At
most 30m3/s are estimated as recharge
to seawater intrusion. Salt water can also be mobilized from lower saline aquifers such as
the Turonian. With a reduced pressure overburden this confined aquifer can infiltrate at a
larger rate into the CT from below.
A numerical model of the system has been built which demonstrates that with the
required total pumping rate by 2050 large areas of the presently strongly pumped regions
will face a piezometric decline with economically infeasible drawdowns of more than
250m below ground level (area with sawtooth pattern in Fig. 4). At the same time the
constraints for water quality locally can no longer be fulfilled.
The groundwater model was then coupled with optimization algorithms to find
allocation patterns that conform to demand, drawdown and quality constraints in time
while minimizing overall provision costs. The wells in an optimal scheme spread out over
the area to equilibrate distribution cost with pumping cost, which depend on drawdown.
They further spread to the CI from the CT. Two variants were analyzed. In the first, the
existing pumping locations were used and the pumping rates at those constituted the set
of decision variables. At the Chott cells, gradient constraints were introduced to prevent
gradient reversal and thus preserve the productivity of the oases. On the whole the costs
are exploding over time, with the running cost of water increasing by a factor of about 30
in 50 years (Siegfried, 2003). In a second variant, pumping at any location was allowed
with the costs being the only criterion for choosing a specific cell. The results show that
compared to the first variant much better abstraction schemes are possible with
considerably lower running cost (and total costs) over the next 50 years. However
interesting such a scenario is, it would require a complete renewal of infrastructure.
Realistically, only a gradual transition from today’s pumping well distribution to a more
favorable one in the future will be feasible.
The model demonstrates that it is possible to minimize pumping cost to reasonable
levels and provide water for the next 50 years. This time however must be used to
develop alternatives. All optimization runs were carried out ignoring national borders in
order to assess benefits from cooperative management. As the results demonstrate,
cooperation between the three countries involved brings considerable advantage in the
exploitation of the resource. Nevertheless, in the long run the conservation of the oasis
culture requires heavy subsidies as the substitution between the production factors water
and capital progresses.
The Okavango River flows from the Benguela plateau of Southern Angola in south-
eastern direction through the northern tip of Namibia and then into Botswana, where it
forms an inland
Sustainable water management in arid and semi-arid environments 9
The third example studies a region in China’s arid west. The Yanqi basin is formed by
the lowlands of the Kaidu River and Lake Bostan (Fig. 8). The area has been used
intensively for agriculture over the past 50 years. The main products are grapes, cotton
and red peppers.
Water resources of arid areas 12
6 CONCLUSIONS
In arid countries the problems of sustainability in the water sector are prominent. On a
worldwide basis the three subjects discussed are the most widespread. They show several
common features. Water management in the arid and semi-arid environment must include
salt management. Modern tools such as remote sensing, geophysics and modeling
hydrological science help even in regions with weak infrastructure to quantify the
implications of human interaction and to give advice to decision makers on the
sustainability of water management practices. Models summarize the state of affairs and
are the only means to make predictions. They are bound to be crude and simulations will
always be idealized. Still, they can serve as points of reference. A further common
feature is that sustainable solutions require the system boundary to be taken sufficiently
large, often transgressing political boundaries. While science can give some decision
support, the decisions for or against sustainability are made in the political arena.
REFERENCES
Alcamo, J., Doll, P., Henrichs, T., Kaspar, F., Lehner, B., Rosch, T. & Siebert, S. 2003.
Development and testing of the WaterGAP 2 global model of water use and availability.
Hydrological Sciences Journal-Journal Des Sciences Hydrologiques, 48(3):317–337.
Bastiaanssen, W.G.M., Menenti, M., Feddes, R.A. & Holtslag, A.A.M. 1998a. A remote sensing
surface energy balance algorithm for land (SEBAL). 1. Formulation. Jnl. of Hydrology, 212–
213:198–212.
Bastiaanssen, W.G.M., Pelgrum, H., Wang, J., Ma, Y., Moreno, J.F., Roerink, G.J. & van der Wal,
T. 1998b. A remote sensing surface energy balance algorithm for land (SEBAL). 2. Validation.
Jnl. of Hydrology, 212–213:213–229.
Bauer, P., Gumbricht, T. & Kinzelbach, W. 2004. A large-scale coupled surface water/ground
water model of the Okavango Delta, Botswana. Water Resources Research, submitted.
Dong, X., Jiang, T. & Jiang, H. 2001. Study on the pattern of water resources utlilsation and
environmental conservation of Yanqi Basin. In: G.Li (Ed.), Development, Planning and
Management of Surface and Groundwater Resources. IAHR congress proceedings. Tsinghua
University Press, Beijing, China: 333–340.
Ellery, W.N., Ellery, K., Rogers, K.H., McCarthy, T.S. & Walker, B.H. 1993. Vegetation,
hydrology and sedimentation processes as determinants of channel form and dynamics in north-
eastern Okavango Delta, Botswana. African Jnl of Ecology, 31:10–25.
ERESS 1972. Etude des Ressources en Eau du Sahara Septentrional. Rapport sur les Résultats du
Projet, Conclusions et Recomm endations, UNESCO, Paris.
Gumbricht, T., McCarthy, T.S. & Bauer, P. 2003. Microtopography of the Okavango Delta using
correlation between land cover and elevation. Earth Surface Processes and Landforms, in press.
Gumbricht, T., McCarthy, T.S. & Merry, C.L. 2001. The topography of the Okavango Delta,
Botswana, and its tectonic and sedimentological implications. South African Jnl. of Geology,
104:243–264.
Herman, A., Kumar, V.B., Arkin, P.A. & Kousky, J.V. 1997. Objectively Determined 10 Day
African Rainfall Estimates Created for Famine Early Earning Systems. International Journal of
Remote Sensing, 18(10):2147–2159.
Hutchins, D.G., Hutton, L.G., Hutton, S.M., Jones, C.R. & Loenhert, E.P. 1976. A summary of the
geology, seismicity, geomorphology and hydrogeology of the Okavango Delta, Geological
Survey Botswana, Gaborone.
Sustainable water management in arid and semi-arid environments 17
ABSTRACT: Currently slug tests in South Africa are used with two
objectives in mind: (i) to get a first estimate of the yield of a borehole
(relationship obtained by Viviers et al., (1995) and (ii) to estimate the K-
value (or T-value) of the aquifer in the vicinity of the borehole. The paper
shows that the use of currently available slug test interpretation methods
to analyse slug tests in fractured rock aquifers to estimate a T or K-value
is problematic. The estimated value is dependent on the flow thickness
(thickness of the part of the aquifer in which flow occurs due to the slug
input). If this thickness of flow is known, the estimated K-value is more
representative of that of the fracture zone. By using the total thickness of
the formation for the estimation of the K-value in slug test analysis, the
estimated K-value (and thus KD-value) does not represent the T-value of
the formation.
1 INTRODUCTION
In performing a slug test, the static water level in a borehole is suddenly lowered or
raised. This is usually done by lowering a closed cylinder into a borehole. The cylinder
replaces its own volume of water within the borehole, thus increasing the pressure in the
borehole. As the equilibrium in the water level is changed, it will recover or stabilise to
its initial level. If the rate of recovery or recession of the water level is measured, the
transmissivity or hydraulic conductivity of the borehole can be determined (Kruseman
and De Ridder, 1994).
In South Africa slug tests are conducted for the following two reasons:
● To estimate the hydraulic conductivity (K) of the aquifer in the vicinity of the borehole
and
● To get a first estimate of the yield of a borehole (Vivier et al., 1995).
Water resources of arid areas 22
Vivier et al. (1995) performed slug tests on 32 boreholes, of which the maximum yield
was known and they then derived empirically the following formula (there is a 93%
correlation between the actual yield and the yield estimated with the formula):
Q=117155.08t−0.83
(1)
The Campus Test Site is underlain by a series of mudstones and sandstones from the
Adelaide Subgroup of the Beaufort Group of formations in the Karoo Supergroup (Figure
2). There are three aquifers present on the Site. The first, a phreatic aquifer, occurs within
the upper mudstone layers on the Site. This aquifer is separated from the second and main
aquifer, which occurs in a sandstone layer of between 8 and 10m thick, by a layer of
carbonaceous shale with a thickness of 0.5 to 4m. The third aquifer occurs in the
mudstone layers (more than 100m thick) that underlie the sandstone unit.
Water resources of arid areas 24
Parameter Value
T of formation* (m2/d) 19
K of fracture zone (m/d) 3600
T of fracture zone (m2/d) 576
K of matrix (m/d) 0.17
T of matrix** (m2/d) 3
*Average for fracture+matrix, obtained from
Cooper-Jacob fit to late drawdown values.
**For 20m thickness.
A major characteristic of the main aquifer is the presence of a horizontal fracture that
coincides approximately with the centre plane of the sandstone layer, and which
intersects all 11 boreholes with significant yields on the Site, of which UO5 is one. The
remaining 14 boreholes all have very insignificant yields. The fracture zone thickness is
approximately 10mm, but the adjacent 200mm of sandstone is also highly permeable.
Figure 5 shows a graph of the data from a constant rate test conducted on UO5 at a
rate of 1.25L/s. Measurements were also taken in the observation borehole UO6. These
pumping test data were analysed with a numerical 3D model (Van Tonder et al., 2001),
and the following parameters were estimated in Table 1.
The thickness of fracture zone (referred to in Table 1) was obtained from tracer tests
and the borehole video, and is 0.16m. The hydraulic parameters given in Table 1 are
regarded to be accurate (Van Tonder et al., 2001). It would now be interesting to analyse
the data of a slug test (Figure 6) conducted on borehole UO5 and compare the estimated
values with the values given in Table 1.
The 90% recovery occurred after about 9 seconds, and using Equation (1) the yield of
borehole UO5 is estimated as 5.3L/s. The tested blow yield of borehole UO5 was 6L/s
during drilling.
The Bouwer and Rice method (1976) was applied to the data in Figure 6. The Bouwer
and Rice equation reads:
(2)
where: rc=radius of the unscreened part of the borehole where the head is rising
rw=horizontal distance from the borehole centre to the undisturbed aquifer
Re=Radial distance over which the difference in head h0 is dissipated in the flow
system of the aquifer
Slug tests in fractured rock formations 27
3 DISCUSSION
(or T=360m2/d) is estimated from the slug test, which is neither the T-value of the
fracture zone nor the T-value of the matrix.
● For a flow thickness of 0.16m (i.e. the thickness of the fracture zone), a K-value of
541m/d is estimated with the Bouwer and Rice (1976) slug test method. This
estimated K-value is more representative of the K-value of the fracture zone.
● The average T-value of the formation, which is important for management purposes,
was estimated as 19m2/d from the constant rate pump test. It is impossible to estimate
the T- or K-value of the aquifer formation via a slug test.
4 CONCLUSIONS
The use of the current available slug test interpretation methods to analyse a slug test in a
fractured rock aquifer to estimate a T- or K-value is problematic. The estimated value is
dependent on the flow thickness (thickness of the part of the aquifer in which flow occurs
due to the slug input). If this thickness of flow is known, the estimated K-value is more
representative of that of the fracture zone. By using the total thickness of the formation
for the estimation of the K-value in slug test analysis, the estimated K-value (and thus
KD-value) does not represent the T-value of the formation.
REFERENCES
Bouwer, H. & Rice, R.C. 1976. A slug test for determining hydraulic conductivity of unconfined
aquifers with completely pr partially penetrating wells. Water Resources Research, 12:423–428.
Cooper, H.H, Bredehoeft, J.D., & D Papadopulos, I.S. 1967. Response of a finite-diameter well to
an instantaneous charge of water. Water Resources Research, 3:263–269.
Kruseman, G.P. & de Ridder, N.A. 1994. Analysis and Evaluation of Pumping Test Data. 2nd ed.
International Institute for Land Reclamation and Improvement. Publication 47. Wageningen, the
Netherlands: 237–247.
Vivier, J.J.P., Van Tonder, G.J. & Botha, J.F. 1995. The use of slug tests to predict borehole yields:
correlation between the recession time of slug tests and borehole yields. In Conference
Proceedings: Groundwater’95: Groundwater Recharge and Rural Water Supply, Midrand, South
Africa.
Van Tonder, G.J., Botha, J.F., Chiang, W.H., Kunstmann, H. & Xu, Y. 2001. Estimation of the
sustainable yields of boreholes in fractured rock formations, Special issue of Journal of
Hydrology: No 241.
Flow simulation model performance
assessment using entropy approach
A.M.Ilunga
Civil Engineering, University of the Witwatersand, South Africa
D.Stephenson
Civil Engineering, University of Botswana, Gaberone, Botswana
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
ABSTRACT: Hydrological data (e.g. rainfall, river flow, etc) are used in
water resources planning and management for planning reservoir and
operation. However, it happens sometime that the appropriate site where a
hydraulic structure (e.g. reservoir) should be built has no available data
due for example to inaccessibility to erect a flow gauging station, etc. This
is particularly a problem in arid areas. Very often hydrologists make use
of simulation models to estimate the flows data series at the very site from
the nearby stations and using some physical characteristics of the
catchment area. In this paper, a merely methodology is proposed to
evaluate the performance of simulation models in terms of entropy (e.g.
reduction of the uncertainty of flows) before and after applying a model to
the site. This reduction should be above a certain threshold value for the
model to be retained as performing well. An example is illustrated
through RAFLER model, which is used to simulate yearly flows at
Braamhoek.
1 INTRODUCTION
data series), while in this paper the same approach is applied to cases where no data
series is available at all at the target site. In very recent paper (see, Ilunga and
Stephenson, 2003b) the methodology for evaluating the model performance was roughly
used but it was not explained systematically as done in this paper. It should be noted that
the reduction evoked above should be above a certain threshold value for the model at
hand to be retained as performing well. An example is illustrated through RAFLER
model, which is used to simulate yearly flows at Braamhoek.
(1)
where K: is a function of the base used or the scale factor (bits for base 2, napiers for base
e, decibels for base 10), i=1, 2,…, n and pi is the probability of occurrence of the event i.
It can be shown that the value of H(X) reaches its maximum when all variate values xi
are equally likely, that is, when the outcome has maximum uncertainty (Amorocho &
Espildora, 1973). In this case the entropy becomes
Hmax (X)=log n
(2)
The theoretic entropy definition was extended to hydrology. Hence entropy is considered
as a measure of the degree of uncertainty of random variable hydrologic processes
(Amorocho & Espildora, 1973). Since the reduction of the uncertainty by means of
making observations is equal to the amount of information gained, the entropy criterion
indirectly measures the information content of a given series of data (Harmancioglu et al.,
1994).
It arises that the distribution of the variable can be unknown a prior although some of
its properties may be known, e.g. mean, variance, normality condition. These proprieties
(information) enable to determine the distribution of the variable, which maximizes the
entropy function. In this way the distribution is consistent with the available information,
Flow simulation model performance assessment using entropy approach 31
but retains maximum uncertainty within the feasible domain and thus ensures the least
bias; that is the principle of maximum entropy (POME) introduced by Jaynes in 1968.
This principle has been applied intensively in hydrology in the last two decades.
Amorocho and Espildora (1973) suggested that the mutual information (between the
observed values and the simulated ones) could be used as a criterion in the selection of
hydrological models; e.g. rainfall-runoff prediction. Note that the mutual information
concept is derived from entropy notion and for more details; the reader is referred for
example to the above-mentioned paper. Later the directional information transfer index
(DIT) appeared as a generalization of the mutual information Yang and Burn (1994) and
was used for dependency evaluation between streamflow gauging station pairs. Recently,
it is argued that since mutual information is used for model performance assessment, its
generalization i.e. DIT can be extended to model performance evaluation (Ilunga &
Stephenson, 2003a).
The above considerations are valid when the estimated values have to be compared to
the observed ones. In that respect statistical criteria such as root mean square error, etc
can be also used to crosscheck the results (Ilunga and Stephenson, 2003a). However it
becomes difficult to use these considerations when missing values are encountered in the
data series. Thus Panu (1992) introduced the notion of reduction of uncertainty of the
hydrological variable before and after infilling the data series. The reduction of
uncertainty Re d(%) at a given site as defined by Panu (1992) can be given as follows:
(3)
where Hcc and Hcomp are entropy values before and after infilling the data series
respectively. It should be noted that this concept was applied to cases of consecutive
missing data values, e.g. hydrological data exist before and after the missing values.
4 PROPOSED METHODOLOGY
Panu (1992) used expression (3) for infilling data problems, in other words some data
exist before infilling process. In this paper the same expression (3) is proposed for cases
where no available flow data exist at all at the site. It is more natural to say that a case
where no data is available, the uncertainty is higher than a case where data exist. Thus it
is assumed that the uncertainty should be maximum (e.g. if all hydrological events would
occur equally likely) at a site where no data is known. Thus, in this case expression (3)
can be re-written as (Ilunga & Stephenson, 2003b):
(4)
Water resources of arid areas 32
Braamhoek is situated in the Free State, in South Africa. The catchment area is about
62km2. Neither rainfall data nor stream flow data is available at this particular site. Thus
it was possible to simulate flows at Braamhoek using rainfall data from the nearby sites;
viz at Van Reenen (MAP=1002mm/month); at Moorside (MAP=839mm/month) and at
Baldergow
Flow simulation model performance assessment using entropy approach 33
(MAP=887mm/month). The monthly rainfall data (1916–2002) were obtained from the
Weather Bureau, South Africa.
Braamhoek was 66.06% by this model. This value is the equivalent of information
inferred about the site using RAFLER model. This model could be thought to perform
well. Thus RAFLER model could be used for flow prediction at Braamhoek with regard
to the total yearly flows. Nonetheless the model needs to be tested on other flow regimes
for that specific site.
7 CONCLUSION
The focus of this paper was to give a methodology for evaluating the performance of
simulation models using entropy approach. The methodology has been tested with
RAFLER model on Braamhoek site where records were simulated. Recall that this
methodology was roughly used in Ilunga and Stephenson (2003a), but without presenting
systematically the steps involved as been done in this paper. The computations from the
entropy criterion showed that RAFLER model could be used for simulating the total
yearly flows at Braamhoek when a threshold value of 50% is considered for the reduction
of uncertainty before and after simulation. Investigation should also be done on other
flow regimes.
REFERENCES
Amorocho, J. & Espildora, B. 1973. Entropy in the assessment of uncertainty in hydrologic systems
and models. Water Resources Research, 9(6):1511–1522.
Harmancioglu, N.B., Alpaslan & Singh, V.P. 1994. Assessment of the entropy principle as applied
to, water quality monitoring network design. Stochastic and Statistical Methods in Hydrology
and Environmental Engineering., 3:135–148.
Ilunga, M. & Stephenson, D. 2003a. Performance of hydrological data infilling techniques using
entropy approach: Expectation maximization algorithms. 11th South African National
Hydrology Symposium, Port Elizabeth, South Africa: 6.
Ilunga, M. & Stephenson, D. 2003b. Entropic measures for comparing flow simulation models at
Bedford site. Paper submitted to the J. Hydrology, Elsevier.
Panu, U.S. 1992. Application of some entropic measures in hydrologic data infilling procedures. In:
Singh, V.P. & Fiorentino, M. (Eds) Entropy and energy dissipation in water resources, Kluwer
Academic Publishers, The Netherlands: 175–192.
Shannon, C.E. & Weaver, W. 1949. The Mathematical Theory of Communication. University of
Illinois Press Urbana, Chicago, London.
Singh, V.P. 1998. Entropy as a decision tool in environmental and water resources. J. Hydrology ,
Indian Association of Hydrologists, 21(1–4):1–12.
Singh, V.P. & Florentino, M. 1992. A historical perspective of entropy applications in water
resources. In: Singh, V.P. & Fiorentino, M. (Eds) Entropy and energy dissipation in water
resources, Kluwer Academic Publishers, The Netherlands: 21–61.
Singh, V.P. & Krstanovic, P.F. 1987. A stochastic model for sediment yield using the principle of
maximum entropy. Water Resources Research, 23(5):781–793.
Stephenson, D. 2002. “Modular kinematic model for runoff simulation”. In: V.P.Singh, &
D.K.Frevert (Eds). Mathematical models of small watershed hydrology and applications. Water
Resources Publications, LLC, pp. 183–218, Chapter 7.
Yang, Y. & Burn, H. 1994. “Entropy approach to data collection network design”. J. Hydrology,
94:307–324.
Flow simulation model performance assessment using entropy approach 35
1 INTRODUCTION
The DGS Hydrogeology division has been collecting data in the areas of monitoring,
borehole archive and research projects. Data collection in monitoring started in 1983 and
since then boreholes from various groundwater projects have been added to the
Data collection experiences in water level monitoring, borehole archive and research projects 37
2 MONITORING
Water level measurement by manual dipping and change of rainfall charts are done on
monthly basis in selected boreholes. The problems recorded in the data collection sheets
of ten monitoring networks are depicted in Figure 1.
However, this schedule is not followed due to several problems such as poor accessibility
due to the country’s hostile environment and its vastness. Since the present monitoring
network is determined by the groundwater potential projects carried out, most of the
monitored areas are in remote areas far from the main office and are laden with heavy
Water resources of arid areas 38
sands and muddy soils from heavy thunderstorms. This means that a four wheel drive
vehicle is a necessity. However, in some cases these vehicles are not available and this
results in data gaps for some months. The other problem is late availability of transport
which results in the late changing of water level/rainfall charts and water sampling. This
renders the data useless as it results in the superposition of more than one line on the
charts making it difficult to read. In some cases, other stakeholder dealing with water
connect these monitoring boreholes to water supply out of emergency to supply people
with water and this creates data gaps in the records as the borehole will not be accessible
and cannot be used. In other cases, other water authorities drill production boreholes
adjacent to these monitoring boreholes and the pumping creates interference and as such
the borehole will be taken out of the network. Some of the boreholes are in private
property and in some cases there is no access as the gates are locked. In other situations
the dipper is reported to be stuck for the whole year such as in borehole 6736 in
Lethakeng/Bothapatlou monitoring network.
4.1 Recording
One of the major problems is the recording system. There are several problems associated
with data recording and these include the following:
– Unclear or no comments, this include things such as reporting same borehole dry and
blocked/ collapsed for different months or reporting “no hole for dipper”
– Water level taken from a borehole with unknown number
– Incorrect entry of measured values
– Measured values being different from the chart recorded value.
This unclear recording resulted in data being discarded hence creating data gaps.
4.4 Equipment
Lack of proper preparation for the field also creates data collection problems. In some
cases, it is reported that dipper or its light was not working and this results in partial or no
data collection. Due to lack of timely quality check it becomes difficult in ascertaining
the source of the mal/non-functioning of the equipment, whether it is the sensor or the
batteries. The other problem was from the mechanical water level recorders being used,
these were perceived as a better replacement of manual dipping, and however, they came
with their own shortcomings. In most cases, the monitoring equipment used was designed
and manufactured in Europe where the environment is completely different from the semi
arid conditions of Botswana. This has lead to tremendous reduction of the lifespan and
the poor performance of these instruments. Some of the problems experienced include
– Stuck pens rendering the data
– No marking on the chart resulting in blank chart.
5.1 Knowledge
In most cases, the personnel operating some of the equipment such as water level
recorders, dippers, sampling pumps and rain gauges lack the technical know-how
necessary to implement first line maintenance. This result in late or no acquisition of data
and hence data gaps develop while the equipment is sent for maintenance or replacement.
The lack of knowledge sometimes results in the equipment not being calibrated or set up
properly and this indicated off scale water level and rainfall curves.
5.2 Availability
The personnel used for data collection are at technician and artisan level. In most cases
these personnel are shared among the various on going research projects and the normal
monthly monitoring. This results in one of these activities suffering due to non-
availability of the personnel for a certain period and as such data gap will be inevitable.
NBA records borehole data on daily basis. This includes registering privately drilled
boreholes; entering borehole data in the database including plotting boreholes on
hardcopy maps and storing rock chip samples.
problem is the use of outdated topographical maps for plotting boreholes which results in
difficulty in borehole location verification. The lack of physical verification of registered
boreholes is also one of the problems. This is due to the fact that boreholes are drilled
almost daily all over the country and due to its vastness; it is difficult to cover all the
drilled boreholes.
6.3.1 Availability
The personnel used for data collection are at technician and artisan level. In some cases
these personnel is shared between NBA and the normal monthly monitoring. This results
in one of these activities being suspended for a certain period and creating backlog.
The Hydrogeology division has been running various projects ranging from groundwater
potential survey to hardcore research projects such as Groundwater Recharge Evaluation
Studies (GRES) and the Kalahari Research Project. These projects are multi-disciplinary
and use different equipment and collect different data sets. In most cases these projects
are carried out jointly with external partners and therefore timely bound. These projects,
especially those run in-house experience a lot of problems.
8.1 Knowledge
In most cases these projects use specialised equipment that requires special operating
level. However, most of the personnel have never been exposed to this equipment and
this can result in loss of data or collection of unreliable data. For, example various
software used to operate Skye data loggers are still unknown to a good number of
hydrogeological technical personnel within Hydrogeology Division.
8.2 Availability
The personnel used for data collection are at technician and artisan level. In most cases
these personnel is shared among the various on going research projects and the normal
monthly monitoring. This lack of technical level staff impacts negatively on the running
of these projects. This results in the project using unqualified staff to fill the gap;
however, that has serious implications on the quality of the collected data. In other
situations, there is a need to have specialised personnel such as a welder to develop a
specialised piece of equipment on site. This can result in delays especially if that person
is unavailable or is occupied with other departmental work.
8.3 Equipment
Some of this necessary specialised research equipment needs special care and due to
harsh conditions prevailing in these remote areas, a lot of time is lost when the equipment
breaks down since it must be sent overseas to be fixed. This also results in loss of data
especially temporally dependent data.
9 CONCLUSION
On the basis of this experience, we conclude that the following aspects are vital for a
successful and reliable hydrogeological data collection effort in the semi-arid Botswana
environment.
– Routine analysis of the archived data should be a must rather than an option, in order to
eliminate useless data before it accumulates in large amounts within records.
– Routine training programmes for technicians on field equipment should be designed to
enable technicians to keep abreast with the changing technology that is specific for
hydrogeological applications.
– Increasing manpower capacity by recruiting personnel with basic hydrogeological
monitoring and database knowledge in order to facilitate data collection and reduce
data losses arising from lack of knowledge. This will increase the reliability of the
collected data.
– The general public and other stakeholders need to be informed about the importance
and relevance of hydrogeological research, borehole archiving and monitoring
Water resources of arid areas 42
activities, so that they can allow such activities in their private properties such as
farms.
– Manufacturers must be encouraged to design field equipment suited to the hostile semi
arid and saline conditions of Botswana, so that the durability of the field equipment
can be guaranteed.
– Periodical inspection of water level monitoring boreholes should be a must in order to
curtail issues of “dry” or “blocked” boreholes, hence maintain a continuous and an
accurate water level record.
– A comprehensive process map of water level monitoring program which include
recording of the environmental status or changes in the vicinity of the monitoring
borehole such as new pumping borehole.
– Conduct a routine water sampling of observation boreholes.
Currently the Hydrogeology Division is engaged in improvement of data collection and
archiving through implementation of the following programs.
– Acquisition of digital water level and rainfall recorders and accessories to replace
mechanical ones and manual dipping. This will reduce human errors and improve data
quality.
– Development of proper databases and checking data immediately from the field to
ensure that issue of unclear comments and data anomalies are reduced hence
maintaining good quality data.
– Development of process maps to improve the quality of the data being collected and
being entered into the databases. This will ensure that all factors are considered before
a inexplicable conclusions such as “dry” boreholes are reached.
– Field programs are being carried out to review borehole location maps. This is to ensure
that borehole locations are correct and indeed the plotted boreholes do exist.
Regular data collection even if it is not part of a specific study, helps to build a picture of
the general behavior of the system. The data collected provides valuable comparisons and
context when the system is studied in more detail. However, all this will not be possible if
the data collected is wreaked with a lot of problems.
REFERENCES
1 INTRODUCTION
Efficient and sustainable use of available water resources is paramount for a peaceful,
sustainable and equitable development of any region.
Some of the challenges to integrated and sustainable water resource management
include:
Water resources of arid areas 44
Changing land use, land degradation by erosion, deteriorating water quality and
competing water demands by stakeholders.
There exists, therefore, a strong demand for an integrated water allocation and
decision support system. The backbone of such a system must be scientifically sound to
be accepted and trusted by stakeholders. Efforts to improve catchment management and
to impose a sustainable water resources management are of economic and political
importance for any country.
Variable semi-arid climate characterized by precipitation patterns unfavorably
distributed in space and time and high evaporation rates reaching up to 100% of the
incoming monthly precipitation is a challenge facing water resources management in arid
and semi-arid lands (ASALs).
Kitui district in Kenya is an example of an ASAL environment where water resources
management issues are particularly important and sensitive. Sources of water are nearer
to the people in the wet season, but as the seasonal rivers dry up, distance to water points
get as far away as 25–30km.
The main problems in water development and management in this district include:
● unreliable rainfall and inadequate supply to meet the demand,
● the available water resources are unevenly distributed and inaccessible to all,
● traditional farming methods lack water conservation principles,
● most of the water projects have since been abandoned, and
● High rates of potential evaporation on the available water resources.
There is great potential for rainwater harvesting. This is dependent on proper
understanding of the patterns of precipitation both in space and time. This would be
useful in understanding drought characteristics in order to develop strategies to capture,
store and redistribute the available water.
Droughts have been the phenomena of great concern throughout the continent of
Africa, because of the devastating effects they have inflicted on the economies of some of
the countries in the continent. Kitui District located in Eastern Kenya is no exception and
is an example of one of the most vulnerable areas to the effects of drought.
Droughts are usually classified as meteorological, hydrologic or agricultural
depending on the variable under investigation. Definition of droughts has also been given
on the basis of theory of runs and stationary structure of time series, Yevjevich (1967).
The most important variable in meteorological drought is rainfall, in hydrological
drought is availability of water in rivers, lakes, reservoirs, and ground water resources;
and in agricultural drought is the soil moisture content to sustain the crop growth.
Drought analysis involves investigation of duration, magnitude or severity, frequency
and regional spread of the event. There have been limited investigations on
meteorological droughts in Kenya. However, substantial work exists on the drought
characteristics for the various agro-climatic regions of South Africa; Dyer and Tyson
(1977), Zucchini and Adamson (1984), Dent et al. (1987).
The study investigates the duration, magnitude or severity as well as frequency aspects
of drought within Kitui District.
Rainfall characteristics in semi-arid Kitui district of Kenya 45
(1)
ai is the weighting factor of the ith gage with record Pi and N is the number of index
gages while Px is the rainfall to be estimated at x.
The different methods differ in their methods of estimating ai’s where i=1,
2,…………………N
2.2 Methods
(2)
where,
Zi is variable i in the standardized form
Fi represents the common orthogonal vector (factor)
ui is the unique factor for variable i
ai1=standardized multiple regression coefficient of the variable i on the common factor
1 (factor loading). The unique term diui=0 since principal component analysis does not
consider the unique component of the variance.
Details of this method are available in many referencesincluding Drosdowsky (1993),
Ogallo (1989) and Basalirwa (1991) among others.
The results of the PCA, onset/cessation of rainfall, dry spell lengths and frequency are
presented in the following sections.
Water resources of arid areas 48
SR 36.18 11 69 17.31
duration
9137010
LR onset 82.35 0 122 25.21
LR 143.1 124 163 10.04
cessation
LR 62 18 149 30.68
duration
SR onset 306.9 287 319 9.462
SR 349.4 328 366 12.45
cessation
SR 46.17 14 76 19.04
duration
9137003
LR onset 93.7 64 129 16.35
LR 138.4 122 152 6.905
cessation
LR 46.05 11 83 20.02
duration
SR onset 307.8 289 324 10.26
SR 342.3 336 361 8.42
cessation
SR 37 16 71 15.02
duration
9138000
LR onset 97.92 78 144 14.44
LR 124.5 99 172 16.38
cessation
LR 27.63 8 47 12.15
duration
SR onset 316 296 341 12.51
SR 353.8 336 366 11
cessation
SR 41.13 6 60 15.66
duration
Rainfall characteristics in semi-arid Kitui district of Kenya 51
The two methods were found to be comparable. For example, Kitui Water Office:
Average onset of long rains is on day 82.36 and cessation on day 126.3 using the two
methods.
On the other hand onset is on day 84.37 and cessation on day 137.5 using the pentad
method. These results are also presented in Table 2.
Water resources of arid areas 54
4 CONCLUSION
Three rainfall regions were delineated. The longest spell is in 1997 (107 days) during the
long rains. The same year also showed the shortest spell (7 days) during the short rains
Average onset for long rains is on day 82, cessation was on day 126.
The years 1980, 1985, 1990, 1995 were severe during long rains as well as during
short rains.
Average onset for long rains is on day 82, cessation was on day 126 using the two
methods.
The largest seasonal totals had return periods ranging between 15–25 years while the
smallest seasonal totals had return period on average of 1 year, and indication of frequent
drought.
This information is crucial if water management and planning is to meet a particular
demand in a specified duration of water stress.
REFERENCES
Basalirwa, C.P.K. 1991. Raingauge network designs for Uganda, Ph.D Thesis, Univ. of Nairobi,
Kenya.
Basalirwa, C.P.K. et al. 1995. The climatological zones of Tanzania based on rainfall
characteristics. Water Resources Engineering Research Report, University of Dar-es-Salaam.
Castell, R.B. 1966. The Scree test for the number of factors. Multivar. Behav. Res., 1:245–276.
Rainfall characteristics in semi-arid Kitui district of Kenya 59
Dyer, T.G.J. and Tyson, P.D. 1977. Estimating above and below normal rainfall periods over South
Africa. Journal of Applied Meteorology 16(2):145–147.
Dent, M.C., Schulze, R.C., Wills, H.M. & Lynch, S.D. 1987. Spatial and temporal analysis of the
recent drought in the summer rainfall region of Southern Africa, Water SA, 13(1):37–42.
Drosdowsky,W.1993. An analysis of Australian seasonal rainfall anomalies 1950–1987I: J climat.
13:1–30.
Mhita, M.S 1990. The onset and cessation of rains andr importance for cropping strategies in
Tanzania.
Ogallo, L.A. 1988a. The spatial and temporal clusters of the East African Seasonal Rainfall
anomalies derived from Principal component analysis. J.Climatol. 6:1–23.
Ogallo, L.A. 1989. The spatial and temporal patterns of East African Seasonal Rainfall derived
from Principal component analysis. J.Climatol. 9:145–167.
Yevjevich, V. 1967. An Objective Approach to Definitions and Investigations of Continental
Hydrologic Droughts. Hydrol. Paper 23, Colorado State University, Fort Collins, Colorado.
Zucchini, W. and Adamson, P.T. 1984. The occurrence and severity of droughts in South Africa.
WRC Report No.91/1/84, Department of Civil Engineering, University of Stellenbosch, South
Africa.
Quantification of the impact of irrigation on
the aquifer under the Vaalharts Irrigation
Scheme
R.G.Ellington, B.H.Usher & G.J.van Tonder
Institute for Groundwater Studies, University of the Free State,
Bloemfontein, South Africa
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
The Vaalharts Irrigation Scheme was initiated approximately 55 years ago. It is the
largest irrigation scheme in South Africa at approximately 32000ha. Vaal River water is
transferred via an extensive canal system from Warrenton into two subsequent canals,
namely the North Canal and West Canal.
This research entailed a detailed groundwater investigation including drilling of
additional boreholes, aquifer parameter determination from slug, pump and tracer testing,
Quantification of the impact of irrigation on the aquifer 61
groundwater monitoring of water levels and water quality for more than a year, and the
construction of water and salt balances using empirical and numerical modelling
techniques.
2 METHODOLOGY
2.1 Hydrocensus
An initial field recognisance study was undertaken. Literature indicated that 41 diamond-
prospecting boreholes had been drilled during the 1970’s. It was hoped that these
boreholes could provide access to the aquifer. During a hydrocensus inspection to locate
these boreholes, all were found to either be destroyed by farming practices, or have been
blocked by stones.
A second hydrocensus across the Vaalharts Irrigation Scheme to obtain an indication
of the groundwater response was conducted. The boreholes located during this
hydrocensus were equipped with mono-pumps, and had been encased in concrete. In
total, 22 boreholes were discovered. Only pumped samples were possible from these
boreholes.
where:
l=distance along the flow path, v=the average linear water velocity, t=time,
C=concentration at time t, C0=initial concentration and D=the coefficient of molecular
diffusion, for the solute in the porous medium.
The expected salt leaching from the irrigation scheme was obtained from du Preez et
al., (2000), who used the Aragues model and the Soil Water Balance to determine salt
leaching from the scheme. These values were used as input in the empirical and
numerical models for salt balance calculations.
separate hydraulic conductivities applied to them, based upon knowledge of the geology
and tested aquifer parameters. For instance, areas with a higher degree of gravels were
assigned a higher hydraulic conductivity for that area. In other areas, where fractures with
significant yields were encountered, an increased hydraulic conductivity was assigned.
The drain package in MODFLOW was applied to the North Canal and West Canal
areas during the simulations. The drains applied in the model accurately represent the
processes occurring naturally in the Vaalharts system. The subsurface drains were
positioned 2mbgl to simulate natural conditions.
For the mass transport simulations, an initial concentration of 500mg/l TDS was
assigned, while the input concentration from irrigation was obtained from the salt
leaching models described by du Preez et al., (2000).
3 RESULTS
3.1 Geology
From the drilling a consolidated geological model was constructed using the Rockworks
(Version 3.4.1.6, Rockware Incorporated) program. The geology model shows thicker
shales to the northern side of the Vaalharts Irrigation Scheme, with thinner shales in the
south of the scheme. In addition, while the model indicates calcretes throughout the
scheme, the calcretes are more pronounced in the southern half of the scheme. The
southern half of the scheme’s geology is represented more by gravels and clays. This may
be due to possible erosion and deposition as the Harts River meandered during the
Vaalharts’ geological history.
cultivated area the 50th percentile value for the groundwater of 2.2mg/l N is considerably
lower than expected. As nitrate is often used as a tracer to highlight the effect of
cultivation on water quality (e.g Pulido-Bosch et al., (1999)) such a low value would
seem to indicate that vertical migration of salts from the cultivated lands to the
groundwater is less pronounced than previously expected. This is confirmed by the low
Water resources of arid areas 68
potassium values, another key constituent in fertilizers, where 95% of the measured
values fall under 15mg/l.
Consideration of the major parameters using interpretive diagrams such as Piper plots,
showed no dominant anion or cation, although some of the cations do tend towards the
Na+K field. Comparison with surface waters and geology indicated a Mg- enrichment in
the Dwyka shales, and sulphate largely from incoming Vaal River water used for
irrigation. The in situ water quality, as determined through hydrochemical logging,
exhibited only minor variations with depth. No significant evidence of stratification of
poorer water quality was observed in any of the boreholes.
For the purposes of the salt balance water qualities of various water types were obtained
from DWAF, previous reports and measured in this project. Combination of the water
balance, these concentration values and the output of salt leaching models reported by du
Preez et al., (2000), allowed salt loads to be calculated. The following was used to for the
salt balance.
There are several permutations of these options but these give similar results regarding
the overall salt balance. Of importance is the recharge salt addition to the groundwater
system. Using the median value of approximately 98000t/year of salts added to the
Quantification of the impact of irrigation on the aquifer 69
groundwater and the assumption that the net storage in aquifer remains relatively constant
over time, the expected net increase in TDS should be in the order of 14mg/l.
The main aim of the study was to ascertain the impact of many years of irrigation on the
groundwater resource. More specifically, emphasis was placed on the assessment of
previous hypotheses related to the aquifer system and salt migration within this system.
Short of addressing the upstream Vaal River practices and means by which to ensure
cleaner water entering the Vaalharts system, the Vaalharts itself needs to be addressed.
The logical approach to ensure lower salinity water enters the groundwater in the
Vaalharts system is to make use of less water. Since agriculture accounts for nearly 70%
of all water withdrawn from rivers, lakes, and underground aquifers for human use, the
greatest potential for conservation lies with increasing irrigation efficiencies (Clarke,
1991). What is needed in the Vaalharts is a more efficient manner of irrigation, where
less water is applied per unit area, and therefore fewer salts enter the groundwater via
leaching. A more efficient means of irrigating would be drip irrigation, with a field
application efficiency of 95%, which is 40% to 60% more efficient than gravity systems
(Postel, 1997). The installation of drip irrigation in the Vaalharts would increase the
efficiency of irrigation, thereby reducing the volumes of water needed. This would
simultaneously decrease the mass of salts applied to the Vaalharts Irrigation Scheme, and
reduce the tonnage leaching to groundwater and, eventually, entering the Harts River.
REFERENCES
Annandale, J.G., Benadi, N., Jovanovic, N.Z. & Du Sautoy, N. 1998. SWB: A user friendly
irrigation scheduling model. Soils and Crops towards 2000 Congress, South African Society of
Crop production, Alpine Health, Kwazulu-Natal.
Aragues, R.M. 1996. Conceptual irrigation return flow hydrosalinity model. In K.K.Tanji (ed.).
Agricultural Salinity Assessment and Management. Am. Soc. Of Civ. Eng., New York.
Clarke, R. 1991. Water: The international crisis. London, Earthscan:. 193.
Birsoy, Y.K. & Summers, W.K. 1980. Determination of aquifer parameters from step tests and
pumping data. Groundwater, 18:137–146.
Chiang, W.-H. & Kinzelbach, W. 2000. Processing Modflow (PMWIN), Version 5.1. The Institute
for Groundwater Studies, University of the Free State, Bloemfontein, South Africa, 9300.
Du Preez, C.C., Strydom, M.G., Le Roux, P.A.L., Pretorius, J.P., Van Rensburg, L.D. & Bennie,
A.T.P. 2000. Effect of Water Quality on Irrigation Farming along the Lower Vaal River: the
Influence on Soils and Crops. WRC Report No. 740/1/00. Water Research Commission.
Harbaugh, A.W. & McDonald, M.G. 1996. User’s documentation for MODFLOW-96, an update to
the U.S. Geological Survey modular finite-difference ground-water flow model: U.S. Geological
Survey Open-File Report 96–485, 56p.
Herold, C.E. & Bailey, A.K. 1996. Long Term Salt Balance of the Vaalharts Irrigation Scheme.
Water Research Commission.
Gombar, O. & Erasmus, C.J.H. 1976. Vaalharts Ontwateringsprojek, Technical Report GH2897.
Department of Water Affairs.
Leap, D.I. & Kaplan, P.G. 1980. A single-well tracing method for estimating regional advective
velocity in a confined aquifer theory and preliminary laboratory verification. Water Resources
Research, 23(7): 993–998.
Postel, S. 1993. Water and Agriculture. Water in Crisis. New York, Oxford University Press. pp.
55–66.
Pulido-Bosch, A., Bensi, S., Molina, L., Vallejos, A., Calaforra, J.M. & Pullido-Leboeuf, P. 1999.
Department of Hydrogeology, University of Almeria. Canada, Spain.
Riemann, K. 2002. Aquifer parameter Estimation in Fractured Rock Aquifers using a combination
of Hydraulic and Tracer Tests. PhD thesis. Institute for groundwater studies.
South African Weather Service. 2002. Climate data. http://www.weathersa.co.za/
Quantification of the impact of irrigation on the aquifer 71
van Tonder, G., Bardenhagen, I., Riemann, K., van Bosch, J., Dzanga, P., Xu, Y. 2002. Manual on
Pumping Test Analysis in Fractured-Rock Aquifers. WRC Report No. 1116/1/02. Water Res.
Comm.
Theme B:
Groundwater recharge:
natural and artificial
Groundwater development—identification of
artificial recharge areas in Alla, Eritrea
K.S.Viswanatham, Filmon Tesfaslasie & Michael Asmellash
Water Resources Department, Government of Eritrea, Asmara, Eritrea,
N.E. Africa
Arun Kumar
Department of Earth Sciences, University of Asmara, Asmara, Eritrea,
N.E. Africa
S.A.Drury
Department of Earth Sciences, The Open University, Milton Keynes, UK
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
The advent of polar-orbiting satellite remote sensing has provided hydrogeologists with a
sophisticated and reliable tool for rapidly assessing natural resources of an area with
reasonable accuracy. The focus of this paper is on the analysis of remotely sensed data
combined with ground truth to delineate geological and geomorphologic patterns and
their effect on groundwater occurrence and movement.
The area selected for the study lies south east of Asmara, the capital city of Eritrea, in
Sub-Sahelian Africa (Fig. 1). Alla-Ghadien and its vicinity are well known for
horticultural activities. Irrigation is from open dug wells and in a limited way from bore
Water resources of arid areas 76
wells. The continual exploitation of water resources has resulted in gradual decline of
water levels and also a decrease in well yields (Habteab 2000). The area lies within a
depression at the edge of the Red Sea escarpment, and is an intermontane basin enclosed
by several outstanding ridges. Ghadien is the administrative center of five villages, Alla,
Belesto, Adi-Asambo, Adi-Moya and Ghadien itself. It is located 16km northeast of
Dekemhare town. The elevation of the area ranges from 1680m to 1936m above mean
sea level. The horticultural activities in the area commenced during the Italian
colonization. In the early times, sufficient groundwater was found at shallow depths (five
meter); however at present it is difficult to get water in some areas, even at depths of 30
meters (Habteab, 2000). The total population of the Ghadien administrative center is
about 2773, with a total cultivated area of 2888.24 hectares. The average annual rainfall
is 463mm. Based on the information of the Ghadien administrator the number of wells is
about 600, of which 200 are dry (Habteab, 2000). Some of the existing problems in the
area are—acute shortage of groundwater, lowering of water table, high density of wells
with close spacing, over exploitation of groundwater, mismanagement of the existing
water, lack of assessment of groundwater potential, extensions of irrigated areas,
well yields. The main objective of the present work is to identify the existing
hydrogeological problems, to test the water quality at random sites and to evaluate water
use and management. With the help of image interpretation and field checks, areas for
future groundwater development and suitable site for check dam and subsurface dam
have been suggested.
The remotely sensed data employed are from the NASA EO-1 satellite’s Advanced
Land Imager (ALI) instrument (NASA, 2001). The deployment aims at testing the
suitability of 9 spectral bands in the visible to short wave infrared, with 30m resolution,
and a single panchromatic band with 10m resolution, for the eventual replacement of the
Landsat-7 Thematic Mapper. The chosen bands include 6 similar to that of the TM
instrument, with an additional 3 that augment possibilities for vegetation and mineral
discrimination. Sharpened by the 10m panchromatic band, various combinations of the
spectral bands offer unprecedented (for Eritrea) opportunities for resource assessment.
2 PHYSICAL SETTING
Eritrea is located on the western flank of the Red Sea at about 12.5°–18° North, 36.5°–
43.5° East. Its border length in the north and west with Sudan is about 624km, that with
Ethiopia to the south is about 917Km, and that in the southeast with Republic of Djibouti
is about 104km. With a total area of about 124000km2, it has more than 350 islands and a
coastline of more than 1200km. The population of Eritrea is approximately 3.5 million.
Eritrea has five major river basins, namely Mereb-Gash, Setit, Barka-Anseba, Red Sea
and Danakil depression. All the rivers (except the Setit) and their tributaries are mostly
seasonal and intermittent. It has four physiographic regions namely the Central
Highlands, Western Lowlands, Eastern Lowlands and Coastal Lowlands. The average
temperature ranges from 3°C–28°C in the highlands and 20°C–48°C in the lowlands.
Eritrea being an arid and semi-arid country is not endowed with rich water resources.
It has a vulnerable environment due to recurrent and devastating droughts, being part of
Sahelian Africa. Rain-fed agriculture is the main occupation of most rural people. The
majority of the population depends on groundwater as the main water-supply source.
Rainfall is torrential in nature, i.e. high intensity and short duration, and is monsoonal.
Annual precipitation ranges from 300mm to greater than 800mm in the Central Highlands
and Southwestern Lowlands, 200mm to 300mm in the Northwestern Lowlands, and 100
to 200mm in the Eastern and Southeastern Lowlands.
Situated close to the Red Sea escarpment, the Alla-Ghadien area receives intermittent
orographic rainfall, as well as that during the two monsoonal periods in March–April and
June–September.
3 PREVIOUS WORK
Drury and Berhe (1993) reported significant regional geological controls for groundwater
in Eritrea and their expression on satellite images. According to them, the main potential
for groundwater developments occur as fracture systems, carbonates with enhanced
permeability, granites (which develop deep, coarse and porous soils and extensive joint
Water resources of arid areas 78
systems), fissile rocks in shear zones, deep basins of unconsolidated sediments in the
mountains and coastal plains, igneous intrusion which act as natural barriers, and
outcrops of lava and laterite. Many such features have been identified on images of
Eritrea. Some of them were considered to be targets for locating bore wells in difficult
terrains of northwestern Eritrea, and drilling them met with a success rate of 85%. Drury
et al. (2001), focusing on the hydrogeological potential of fracture systems, indicated that
NNW-SSE Precambrian shear zones, normal faults roughly parallel to those earlier
structures and prominent East-South-East-West-North-West dilatational fractures offer
considerable scope for groundwater development. Prospect geophysical profiling across
several of these structures in both lowland and highland terrains revealed conductive
features believed to relate to saturated zones in large, regionally extensive fractures.
Zerai and Solomon (1993) identified five main hydrogeological units:
(i) Unconsolidated sediments with variable inter granular permeability.
(ii) Volcanic rocks (basalts) with fracture and fissure permeability.
(iii) Fissured and karstic carbonate aquifers.
(iv) Metamorphic and intrusive rocks with localized low to moderate permeability along
fractured and weathered zones.
(v) Aquitards and aquicludes and groundwater barriers (acid to intermediate volcanic).
4 GEOLOGY
Knowledge and understanding of the geological events of an area are important for
groundwater investigations. The geology of Eritrea is made up of a Precambrian
basement complex comprising high-to low-grade metamorphic rocks and associated
intrusives, which are overlain by predominantly Mesozoic sedimentary rocks and
Tertiary to Quaternary volcanic and sedimentary rocks (JICA Report 1998 Drury and
Berhe 1993). Precambrian granites are exposed in the eastern, western and northern parts
of Alla-Ghadien. The southern part of the study area is a wide and flat plain between
actively rising ridges of metamorphic rocks. The geological events in Eritrea are
summarized as follows:
(1) Precambrian: Formation of the crystalline basement complex and its associated
intrusive rocks;
(2) Paleozoic: Peneplanation of the basement complex and deposition of sparse
sedimentary rocks;
(3) Jurassic: Transgression-regression of the Mesozoic sea, which deposited lower
sandstone, Adigrat sandstone and Antalo limestone during subsidence towards the
Indian Ocean;
(4) Upper Eocene-Miocene: Uplift forming domes, extensional fault systems, basaltic
flood volcanism and opening of the Red Sea rift system;
(5) Miocene Period: Formation of upper sandstone; and
(6) Quaternary Period: Formation of alluvial, eluvial and colluvial unconsolidated
sediments.
Groundwater development—identification of artificial recharge areas in Alla, Eritrea 79
5 HYDROGEOLOGY
The reconnaissance survey done in Ghadien and Alla areas shows that the plain is
covered by Quaternary alluvial and colluvial sediments and with subordinate outcrops of
granitic intrusive rocks. There are two types of aquifers in the study area: alluvial
sediments and weathered and fractured acidic granitoids. Moreover major and minor
lineaments exist in the area, many being fracture zones or igneous dykes, which naturally
serve as conduits for or barriers to groundwater flow respectively. That the NNW-SSE
basaltic dikes serve as groundwater flow barriers is confirmed by high yields of wells
around the dikes. Minor structures trending roughly NE-SW are found to serve as a
conduit for the water flow, confirmed by the high yield of the wells in the Ghadien area,
where the structures are dominantly observed.
In 1994 and 2000 the Water Resources Department (WRD) did a well inventory in
Alla-Ghadien area, encompassing location, static water levels and well design. Based on
the well inventory data in 1994 there were 144 wells, the average well depth was 15m+
with static water level (SWL) 14.4m. In the year 2000 the average well depth was 17
meters & SWL 16.51. Therefore the lowering of water level in 4 years is 2 meters. In the
Alla-Ghadien area, hand dug wells are very closely located, from approximately 20m to
100m apart. An ad-hoc assessment of the catchments of 1465km2 area based on rainfall
for Alla has shown that the total groundwater recharge is 643Ha.m, while the total annual
groundwater draft is 1062Ha.m leaving a negative balance of −419Ha.m
(K.S.Viswanatham 2002). The above figure indicates “mining” of water.
The Alla-Ghadien area could therefore be classified as an overexploited area. Areas
where groundwater resource assessment shows stage of groundwater development more
than 100% and both pre and post monsoon groundwater levels show a significant long-
term decline are classified as over exploited areas.
6 METHODOLOGY
An integrated approach that involves interpretation of remotely sensed data and ground-
based ancillary investigations has been implemented for the Alla-Ghadien study.
Groundwater zonation was prepared using various thematic maps at 1:50,000 scales,
which include geology, geomorphology, lineament trends and land use/landcover. Field
survey was done to correlate the image characteristic to ground feature to confirm the
interpretation. Specific field data, such as well inventories of SWL, yield, depth,
diameter, quality of water and drilling logs were collected. Images (bounded by UTM
7500000–7516000 E 1666000–1682000 N) of the Alla-Dekamhare area were from an
EO-1 overpass in April 2001 and combine ALI bands 5, 4 and 3 as red, green and blue
components, sharpened by the use of the ALI panchromatic band to modulate intensity.
This combination is optimum for expressing vegetation cover, but does discriminate
some lithologies and soil types, as well as revealing small-scale topographic features.
Water resources of arid areas 80
The various water-bearing and movement properties, and the controlling parameters in
the study area based on image interpretation and ground check are as follows:
Fluvial/alluvial sediments and alluvial fans,
Tertiary laterite,
Granite,
Tertiary basaltic rocks, as lavas and dykes.
Geologically, the area is dominated by granitic terrain. The alluvial areas of Alla and
Ghadien are eastward sloping plains surrounded by hills with mostly steep slopes. The
alluvial sediments have a yellowish colour and define a roughly triangular area. Tertiary
laterites are indicated by grayish white color in irregular shapes along the streams and on
the plains. Areas of bedrock comprise granites, which are traversed by NNE to SSW
basaltic dikes. In the imagery, the granites are bluish to dark blue in color and occupy
topographical ridges. The granitic exposures form circular and semi-circular shapes, for
example Bazit Hill near Bazit village. There is a possibility of marble being present as
roof pendants in the granitic masses at the northern flank of the area (represented by light
to dark gray tones with signs of bedding), which have to be checked and confirmed in the
field. Topographically the area is surrounded by hills with undulating slopes, and almost
plain on the center. These features could be observed north west of Ghadien. Dikes show
as linear features mostly concentrated north of Ghadien and roughly north of the Alla
plains.
The major land use of the area is classified based on color, shape and texture as fallow
land, barren rocky terrain, dense vegetation (mainly horticulture gardens), river sediments
and sparse vegetation. The Ghadien, Sesah and Bazit Rivers drain the area. The drainage
pattern is distributed and dendritic. All the rivers flow eastwards. Some flow features
follow a structurally controlled direction. Low drainage density in the alluvial plains,
which probably indicates high rates of infiltration, suggests good groundwater prospects
in parts of them.
8 FIELD CHECKS
From the fieldobservations, the area can be divided into high- and low-potential zones. In
the upper Ghadien river successful hand dug wells are being pumped for 4–7 hours per
day and irrigating 6–10ha of horticulture gardens, particularly citrus bushes. In addition,
this high-potential area has been investigated by the geophysicist from the Ministry of
Agriculture, and recommended to be a potential area for development. In the satellite
image alluvial fans and fills clearly represent the upper Ghadien zone.
Groundwater development—identification of artificial recharge areas in Alla, Eritrea 81
The low-potential zones, as defined by the well inventory data, are represented in the
satellite image by granitoid and granitic ridge with adjoining alluvial cover. The farmers
are pumping water for 2–3 hours per day from wells there, some of which have been
abandoned due to poor yields or non-availability of water. The geophysical surveys have
also confirmed the absence of suitable aquifers at deeper levels in such low potential
zones.
A preliminary interpretation on the regional geology was attempted, based on the
imagery. In order to corroborate this interpretation, a quick field trip was made. The field
observations revealed additional information on the geology, which necessitated some
modification to the preliminary assumptions about the geological set-up. The alluvial fans
as interpreted from the imagery are observed in the field to be a thin cover of alluvium on
granitic rocks. The possibility of carbonate rocks/Marble in the North West part of the
area is ruled out as granites varying in color, composition and texture represent these
outcrops. Only few small patches of marble, which are not mapable, have been reported,
which could be the extension of carbonates that supply abundant spring water in the
Maihabar area 15km to the north. Beside this, cherts and conglomerates are observed on
the Northwest part (assigned as basement metasediments), diorite on the Northeast, and
metabasalt on the Southeast part of the study area (Fig. 2).
In the Alla area, four bore wells drilled in the range of 47 to 50 meters are reported to
yield 1.4 to 2.2 liters per second. The depths of the alluvial cover at those sites are from
13 to 21 meters, followed by granites. In the Bazit area, there are 3 boreholes, each
having depths of 50 meters, and yields of 1.4lps, 2.5lps and 5Lps. In the Ghadien area,
there is only borehole with a depth of 50 meters and yield of 2.5Lps. The alluvial cover is
21 meters. In Chuhot area 3 boreholes, with depths of 36 meters, 48 meters and 50 meters
respectively proved to be highly successful, with yields of 4Lps, 5Lps and 5.5Lps.
The alluvial formations in Alla plain provide suitable sites for sub-surface dams,
infiltration galleries and check dams. This is due to fine sediments predominantly quartz
gravels and sand derived from granitic terrain. The natural barriers namely the NNE-
SSW dikes have to be taken into consideration while constructing the sub-surface dams
and check dams for the suitability of the structures.
Areas can be delineated into recharge and discharge areas depending on whether water is
added to or abstracted from the zone of saturation. In the case of the water table aquifer,
usually the areas occupying higher elevations with deeper water tables constitute the
recharge areas while the
Groundwater development—identification of artificial recharge areas in Alla, Eritrea 83
ACKNOWLEDGEMENTS
The authors acknowledge with thanks Mr. Ghebremichael Hagos Director General for
giving an opportunity for the investigations and permitting to publish this paper. The
authors also thank Mr. Ghebremichael Temenewo and Mr. Michael Negash for
discussions from time to time, which helped to improve the paper. Ms. Meron Teshome
is acknowledged for her support in preparing the maps.
REFERENCES
Drury, S.A and Berhe, S.M. 1993. Remote Sensing and Water Exploration in Eritrea WRD
Eritrea/EIAC/ GREADCO and Open University UK. (Unpubl) 6 pp.
Drury, S.A and Berhe S.M. 1993. Accretion Tectonics in Northern Eritrea revealed by remotely
sensed imagery. Geol Mag, 130(2):177–190.
Drury, S.A, R.J.Peart M.E. & Andrews Deller. 2001. Hydro geological potentials of major fractures
in Eritrea. Journal of African Earth Sciences, 2(2):163–177.
Habteab.T. 2000. Groundwater Depletion in Alla Commercial Farm. Department of Environment,
Ministry of Land Water and Environment (Unpubl).
Karanth, K.R. 1994. Groundwater Assessment, Development and Management, New Delhi, Tata
McgrawHill Publishing Company Limited: 720 pp.
JICA Report. 1998. Study on Groundwater Development and Water Supply for Seven Towns in
Southern region of Eritrea.Water Resources Department and Sanyu Consultants Inc., Japan.
NASA, 2001. EO-1Science Validation Team Home Page,
http://eol.gsfc.nasa.gov/science/SVTAuth.cfm
Viswanatham, K.S. 2002. Water Resources Development Management of Critical Areas in Eritrea.
Journal of Applied Hydrology, XV(4), Oct:21–25.
Zerai Habteab. 1996. Groundwater and Geothermal Resources of Eritrea with the emphasis on their
chemical quality: Journal of African Earth Sciences, 22:415–421.
Subterraneous injection of nutrient rich
groundwater to the coastal waters
K.K.Balachandran
National Institute of Oceanography, Regional Center, Cochin
Joseph Sebasgtian Paimpillil
Center for Earth Research & Environment Management, Cochin, India
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
The west coast of India is environmentally more sensitive than the east coast primarily
because it is bordering one of the most sensitive ecosystems in the world, the Arabian
Sea. The environmental property of the northern Arabian Sea is unique which manifests
in rich biological production throughout the year through different processes and thus,
Water resources of arid areas 86
explain for the Arabian Sea ‘Paradox’ Mathupratap et al (1996). The symptoms are there
to show considerable impact of deterioration of estuarine waters on the coastal ecosystem
Nair et al (1991), Naqvi et al (2000), Jayakumar et al (2001). The emerging industrial
establishments and human settlements along the west coast of India, thus necessitates a
critical evaluation of the nature and quantum of inputs to the Arabian sea as well as their
regional assimilative capacities. If there is a possible threat to the well being of the living
resources of EEZ of India, then the coastal waters of southwest coast of India, and in
particular, Cochin region is the prime location prone to trigger it. The booming city of
Cochin has population of nearly 1.5 million Anonymous (1998) and 60% of the chemical
industries of Kerala are situated in this area Cochin backwaters are the largest of its kind
on the west coast of India with an area of 256Km2. The 16 major and several minor
industries situated in the upstream region of the backwaters discharge nearly
0.105Mm3d−1 of effluents Anonymous (1996). The fertilizer consumption in Kuttanad
region (the main agricultural field draining to Cochin backwater) alone is reported to be
20,239ty−1 Anonymous (1996). The backwater
monsoon periods, the southwest coastal waters remained oligotrophic and surface
chlorophyll a typically ranges from 0.1 to 5.3mgm−3, while primary productivity ranges
from 100 to 360mgCm−2d−1. Recent studies as the one discussed here contradict these
findings and show that even after the monsoon period, fresh injection of nutrients by
hitherto unknown processes fertilize the coastal waters that are either permanent or quasi-
permanent in nature. One of the major mudbank regions (Fig. 1 A, B) of southwest coast
of India was selected for observation that indicates episodic introduction of nutrients into
the coastal waters during periods when mud banks are passive.
During the typical pre-monsoon (February) months, the nitrogenous nutrients in water
remained low except for the southern transects centered on Chethi and Alleppey. The
phosphate concentrations did not show any spatial or vertical variation in the water
column, but higher concentrations of ammonia, nitrate and silicate were observed at
selected regions starting in the near shore regions and extending offshore (Fig. 2 A–D).
The Nitrate-N concentrations point towards a clear source between Chethi and
Pazhayangadi, where it peaked up to >8µM and decreased towards offshore. A similar
trend was observed for ammonia-N with the source centered on Chethi (at about 15m
depth). It may be assumed that the ammonia released were either rapidly utilized by
phytoplankton or oxidized within the system itself where the waters were saturated with
dissolved oxygen. Distribution of silicate-Si was similar to that of nitrate (4–10µM),
higher than the corresponding values reported for the waters of Southeastern Arabian
Sea. The input of these nutrients supported high primary production up to 14mg/m3 of
chlorophyll a (peak column production of
Water resources of arid areas 88
out a definite source to these high nutrients during this period, as the fresh water
discharge was at the minimum.
During post monsoon (November), homogenous mixed layer prevailed in the entire
region. While the physical characteristics were more or less stable, there was
considerable variability in the nutrients and in chlorophyll a concentration (Fig.2 A–D).
A marked decrease in sub-surface dissolved oxygen (2.8–4.8ml/l) was the characteristic
feature of this period, which was concomitant with enriched nitrite (0.5–2.0µM),
phosphate (0.4–2.8µM) and silicate (0.5–14µM). The ammonia (1–7µM) and nitrate (1–
6µM) were also elevated at some regions along southern transects. The enriched
particulate organic carbon (>3.5mg/l) and Chlorophyll a (14.8mg/m3) were also the
notable features of this period. It is likely that chlorophyll a values were proportionate to
carbon production indicating a strong positive relationship binding it with nutrient related
factors rather than seasonal or diurnal fluctuation. The elevated nitrite and phosphate
levels around Cochin may be due to the input from the backwaters. Higher values of
nitrite, POC and chlorophyll a towards the southern offshore waters off Pallana were
conspicuous and the regions with high nitrite had nitrate levels up to 6µM and the low
levels of ammonia had ruled out the nitrification as a significant process responsible for
nitrite accumulation. The remarkable co-existence of nitrite with nitrate strongly
suggested that the nitrite production should mostly be due to assimilatory reduction. This
was further substantiated by the high concentration of chlorophyll a (4–9.8mg/m3) on
these transects.
The N/P ratio in the coastal waters was below 15 during November (Fig. 3), possibly
due to the disproportionate release of P from mudbank sediment. However, a band of
N/P>15 funneling out from Alleppey region was indicative of an ‘external source’ of
nitrogenous compounds into the coastal waters. A comparison of long-term (decadal)
trend in the chlorophyll data of this region showed “greening” of near shore waters
Devassy (1983). This suggests that phytoplankton standing crops had increased
historically, possibly in response to watershed nutrient inputs. These sources of nutrients
deserve identification as it was traced to a region, far away from any river mouths.
The current observations in general indicated the presence of a nutrient source
between Chethi and Pallana. This region has mud banks but the release of nitrogenous
compounds cannot be accounted from sediments. The injection of nutrients was in non-
monsoon months when mud banks were passive and a new influence of Vembanad Lake
on the coastal waters is very clear. One of the recent estimate shows that in spite of
receiving 42.4×103mold−1 of inorganic phosphate and 37.6×103mold−1of inorganic nitrate
from Periyar side of the estuary, the export to the coastal waters is only 28.2×103mold−1of
inorganic phosphate and 24×103mold−1 of inorganic nitrate Naik (2000) and the lake acts
as a sink for the nutrients, flushing out only a portion of the pollution load that it receives.
Increased human population along the coastal belt has also resulted in concomitant
increases in widespread use of septic tanks and nutrient inputs to coastal waters,
particularly from regions occupying limestone beds. It has been found that domestic
wastewater from septic tanks provide more nitrogen than that due to precipitation or use
of fertilizers. The situation is exacerbated in the present study region, as more than 70%
of households in these coastal belt and adjacent areas of Vembanad Lake do not have
proper sanitation facilities. Significant amounts of nutrients from fertilizer applied in
agricultural fields (approx. 94kg/ha) leach out into waterways, groundwater and to the
coastal bays inducing coastal fertilization due to direct discharge into coastal ocean and
through ground water seepage.
Subterraneous injection of nutrient rich groundwater to the coastal waters 91
3 CONCLUSIONS
The nutrient fluxes into coastal region were influenced by fluxes from Cochin backwater
and by the mud bank formation. The present study isolates a possible link between
Vembanad Lake that supplies primary nutrients to the adjacent coastal waters and
precondition it for rich primary production during non-monsoon months. The causative
factors discussed are indicative of existence of a subterranean flow connecting Vembanad
Lake to the adjacent coastal waters through the submerged porous lime shell beds.
Continuous nutrient entry through such process is bound to upset coastal water
productivity pattern. If the existence of the subterraneous channels linking Vembanad
Lake to the adjacent coast is proved, it might even re-construct the historical evidence
that the subterraneous flow plays a decisive role in the formation of mud banks along this
region. A sub aqueous injection of nutrients into the coastal waters through this region is
possible even after the rainy season. This assumption need further study to establish
cause and affect mechanisms and quantify actual trends created by increased nutrient
loading.
REFERENCES
Anonymous, 1996. Pollution potential of industries in coastal areas of India. Coastal Pollution
Control Series: central Pollution Control Board Report. COPOCS/9/1995–96.
Anonymous. 1998. NEERI- carrying capacity based developmental planning of Greater Kochi
Region. Phase I Report.
Devassy, V.P. 1983. Mahasagar, Bull Bull Nat. Inst. Oceanogr.7:101–105.
Hema Naik, 2000. Budgets for Periyar estuary, Kerala. Presented at Regional Training Workshop
on Biogeochemical Budgeting and Socio-Economic modeling for Coastal Scientist.
APN/SASCOM/LOICZ, 18–22 September, Colombo.
Jayakumar D.A., Naqvi S.W.A., Narvekar P.V. & George M.D. 2001. Methane in coastal and
offshore waters of the Arabian Sea. Mar. Chem. 74:1–13.
Mathupratap N.M., Prasanakumar S., Bhattathri P.M.A, Dileepkumar M., Reghukumar S., Nair
K.K.C. & Ramaiah N. 1996. Mechanism of the biological response to winter cooling in the
north eastern Arabian Sea. Nature, 384:549–551.
Nair C.K., Balchand A.N. & Nambisan N.P.K. 1991. Heavy metal speciation in sediments of
Cochin estuary determined using chemical extraction techniques. Sci.Total Environ. 102:113–
128.
Naqvi S.W.A., Jayakumar D.A., Narvekar P.V., Naik H., Sarma V.V.S., D’Souza W., Joseph S. &
George M.D. 2000. Increased marine production of N2O due to intensifying anoxia on the
Indian continental shelf. Nature, 408:346–349.
Qasim, S.Z., Wafar, M.V.M., Sumithra Vijayaraghavan, Joseph P., Royan. & Krishna Kumari, L.
1978. Ind. J. Mar. Sci.,7:84–93.
A new method for the estimation of episodic
recharge
J.Bean, G.van Tonder & I.Dennis
Institute for Groundwater Studies, University of the Free State,
Bloemfontein, South Africa
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
There can be no doubt that the South African water industry has been profoundly
transformed over the last 10 years, with millions of rands invested in water infrastructure
aimed at ensuring that every South African has access to fresh drinking water. In drier,
more isolated, inland areas of the country, this has often meant that available
groundwater resources must be exploited. As such, government and non-government
organisations have invested in research associated with developing new assessment
techniques so that these resources can be managed sustainably. In common with all these
strategies is the need for recharge processes to be understood, and if possible, quantified.
An understanding of site recharge behaviour is far more important than many
geohydrologists realise, and goes beyond estimating the average proportion of rainfall
entering a given aquifer. For example, from a planning viewpoint, groundwater ingress
into a mine is seldom problematic to mine management, providing it is constant;
problems occur when unpredicted increases occur, such as those associated with the
sudden entry of recharge water into surrounding aquifers. Thus, through understanding
the episodic nature of recharge in semi-arid and arid areas, and therefore the thresholds
that must be exceeded before recharge occurs, geohydrologists are better able to provide
predictive advice for their clients. This paper discusses a new stable isotope-based
A new method for the estimation of episodic recharge 93
technique, the Modified Amount Effect (MAE) Method. This was developed during the
study, which provides insight into episodic recharge processes.
2.1 General
This technique provides insight into episodic recharge processes by estimating the
proportion of preferential pathway-to-matrix-derived flow entering an aquifer, and the
amount of rainfall required to initiate recharge via the respective flow paths.
Significantly, the proportion of bypass flow can be determined without undertaking
expensive and time-consuming unsaturated zone studies, both factors often of primary
concern when undertaking recharge investigations in developing countries.
Where PPflow=the proportion of recharge derived from preferential flow, and dGWL, dEWL-
U, and dMWL-U represent the d excess in δ2H (‰) for GWL, EWL-U, and MWL-U,
respectively.
It should be appreciated that the calculated value PPflow is sensitive to:
● Variations in the orientation of the LMWL and EWL-U.
Water resources of arid areas 94
● The recharge threshold. At low recharge thresholds (i.e. recharge occurs rapidly in
most years), particularly in more temperate areas, evaporation effects may not be
represented in the stable isotopic composition of groundwater data. In these areas,
transpiration, and not evaporation, probably has a greater potential to reduce the
recharge flux to site aquifers.
● The source of recharge water. The suggested method assumes that recharge water is
derived solely from precipitation, with no contribution from an adjacent surface water
body where pre-recharge evaporation has occurred.
however, by considering CRD and long-term average values together. For example, the
CRD Method indicates that, for an aquifer under equilibrium conditions, the recharge
threshold is approximately 35mm/month. Since, on average, the recharge threshold
cannot be lower than this amount, it must represent the average lower recharge threshold.
Thus, the respective average lower recharge thresholds can be calculated once the
isotopic composition of rainfall for an equivalent depth of 35mm has been estimated from
amount effect data.
Once the lower and average long-term thresholds for both preferential pathway (RTlow-
pp and RTave-pp) and matrix-medium recharge (RTlow-uzm and RTave-uzm), the upper recharge
thresholds RThigh-pp and RThigh-uzm can also be calculated, i.e.
RThigh=2.RTave−RTlow
3 CONCLUSIONS
Four recharge thresholds can be identified using the MAE Method; the low and high
recharge thresholds that must be exceeded before recharge occurs via preferential
pathways or the matrix, respectively. These represent threshold limits, the low value only
of importance following successive months of wet weather, the high value representing
the rainfall that must be received to restore an aquifer system to equilibrium after
prolonged dry spells. Once these thresholds are known, the recharge history of a site can
be modelled using available rainfall data by adapting the CRD Method. An important
finding of modelling undertaken during this investigation is that in those semi-arid to arid
areas where most recharge water enters, the aquifer via the matrix, the period of time that
elapses between successive rainfall events that exceed the matrix recharge threshold often
extends to scores of years. This has significant resource management implications for
much of the region, as it indicates that the current approach of basing allocations on
average recharge estimates is only justified if sufficient groundwater is available for use
over the entire period between recharge events.
The MAE Method was found to be sensitive to the recharge history of the site, the
returned recharge estimate significantly higher when calculated immediately after
recharge via the matrix had occurred. This is not to say that these estimates were
incorrect (indeed they were representative of site recharge processes at the time of
sampling), but that rainfall in the preceding months should be considered prior to
sampling. In general however, sampling should be undertaken near the end of the dry
season, which in the summer-dominant rainfall areas of Southern Africa is between
Water resources of arid areas 98
September and November (allowing for a 30 to 60 days lag time between rainfall and
subsequent recharge).
REFERENCES
Alison, G.B., Barnes, C.J., Hughes, M.W. & Leaney, F.W.J. 1984. Effect of climate and vegetation
on oxygen-18 and deuterium profiles in soils. Isotope Hydrology 1983. IAEA Symposium 270,
September 1983, Vienna.
Prioritisation of the impacts of pollutants on
groundwater flow systems in South Africa
I.Dennis, B.Usher & J.Pretorius
Institute for Groundwater Studies, University of the Free State,
Bloemfontein, South Africa
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
Groundwater pollution can occur, as a result of various activities of man. With increased
human settlement and economic development, a range of undesirable waste products are
produced which can end up in the environment. According to the National Water Act
(Act No. 36, 1998), pollution is defined as the direct or indirect alteration of the
physical, chemical or biological properties of a water resource so as to make it—
1. Less fit for any beneficial purpose for which it may reasonably be expected to be used;
or
2. Harmful or potentially harmful—
● to the welfare, health or safety of human beings;
Water resources of arid areas 100
2 METHODOLOGY
2.1.1 Sources
Sources, in this context, refer to the origin of the substances (inorganic species, organic
compounds or microbial agents) that are causing, or may potentially cause, the pollution.
The term
Table 1. Systems for classification of groundwater
contamination sources.
Classification Examples
system based on
Way of release Discharge sources, transport
sources
Loading history Spill or continuous
Location Above ground surface, below
surface
Degree of Point (or line) and non-point
localization sources
Origin Industrial sources, mining
sources
Likelihood of For example petrol service
occurrence stations found more often
than chemical manufacturing
plants
is used very broadly over a range of scales and may describe physical entities (e.g. a
pond, a tank, a pipeline); human activities (e.g. mining, irrigation, wastewater treatment);
the site at which potential pollutants are stored, used or disposed (e.g. wastewater
treatment works, cemeteries, fuel filling stations) or even large scale phenomena (e.g.
atmospheric deposition).
Prioritisation of the impacts of pollutants on groundwater flow systems in South Africa 101
Source of pollutant plays a large role in whether the pollutant will reach the
groundwater table and if it does the rate at which the pollution will enter the groundwater
system. There are also several existing methods for classifying the sources of
groundwater pollution. A simplified classification based on that of Nonner (2002) was
used to classify South African sources (see Table 1).
2.1.2 Pollutants
Pollution refers to levels of hazardous substances in the environment over and above
what would ordinarily be found in the absence of local activities. Groundwater pollution
therefore refers to the occurrence of substances (inorganic species, organic compounds or
microbial agents) in concentrations above those that would naturally be found in an
aquifer. The substances themselves, both chemical and microbial, are called pollutants.
There are various ways in which to group or classify groundwater pollutants. Each of
these has major classes which can then be broken down into smaller categories. The
choice of system and level of detail of the classification is dictated by the purpose of the
classification for the sake of this investigation pollutants were classified according to:
● Fate in the environment
– Degradable pollutants, which can be rendered harmless by natural processes and
need therefore cause no permanent harm if adequately dispersed or treated; and
– Persistent pollutants, which eventually accumulate in the environment and may be
concentrated in food chains.
– Pollutants may also be divided by their behaviour in water into:
(a) Soluble pollutants, which includes most inorganic species and some organics.
(b) Insoluble substances, which are small enough to be carried through the aquifer
matrix, including microbial pollutants and colloidal inorganic pollutants.
(c) Non-aqueous phase liquids (NAPLs), which are organic compounds that do
not dissolve readily in water and remain as a separate liquid phase. These are
further subdivided into Light Non Aqueous Phase Liquids (LNAPLs) and
Dense Non Aqueous Phase Liquids (DNAPLs).
● Human health impacts
– Non-harmful substances, which have no observed effects on human health.
– Toxic substances, which cause various effects on the body from short-term exposure
or long term accumulation, ranging in severity depending on the dose e.g. nausea,
rashes, kidney failure or neurotoxic effects.
– Carcinogenic substances, which are known to cause cancer.
– Pathogenic substances, which are known to cause diseases in humans.
● Other aspects that are taken into account:
– Duration of pollution—if the pollution results from a single (once-off) spill, the
impact will probably be smaller than that resulting from continuous pollution.
– The vulnerability of the aquifer represents the intrinsic characteristics that determine
the sensitivity of an aquifer to the adverse effects resulting from the imposed
Water resources of arid areas 102
pollutant (Lynch et al., 1994). Factors taken into account include depth to
groundwater, recharge, aquifer media, soil media, topography and impact of the
vadose zone.
give the assessor a guideline of the risks. The assessment should take a couple of hours to
complete. The next tier (LEVEL 2) is an intermediate assessment. The first step in the
intermediate assessment is to collect all relevant data. Data requirements include aquifer
and contaminant parameters, as well as health information. General information will be
obtained from databases, but it is sometimes necessary to have site-specific data. The
confidence attached to this assessment should be medium to high. Both the second and
third tiers include risk assessments based on a fuzzy logic methodology. Figure 1 is a
schematic representation of the tiers and the function performed on each level of
assessment.
borehole or wellfield, through which contaminants are reasonably likely to move and
reach such a borehole or well field.
In many cases it is difficult to protect the whole area, therefore various zones are
established within the area.
3.1 The wastewater treatment works within the City of Cape Town
prioritization of sources and contaminants on a regional scale
The City of Cape Town (CCT) is located in the Western Cape Province on the south-
eastern corner of South Africa. A major portion of the CCT consists of the area known as
the Cape Flats, which has an elevation of between 20 and 45m above sea level. CCT has
a mean annual rainfall of 515mm/annum and an average temperature of 16.7°C. It is a
winter rainfall area. The current population of the CCT is estimated at 3.2 million with
the highest population density occurring on
Table 2. Source prioritization for CCT (incomplete
list).
Source prioritisation (from highest to lowest
risk)
On-site sanitation
Petrol service stations (underground storage
tanks)
Cemeteries
Stormwater/sewer systems
Agriculture (general and crop cultivation)
Feedlot/poultry farms
Wastewater treatment
the Cape Flats and there are approximately 90000 consumers on informal sites. There are
21 wastewater treatment plants within the CCT. According to TIER 0 the wastewater
treatment works are rated as the 7th highest pollutant source within the CCT. Due to the
length of the complete list only the 7 highest potential polluters have been documented in
Table 2.
Typical contaminants found at wastewater treatment works include ammonium,
nitrate, potassium, phosphate, chloride, sulphate and faecal pathogens. Micro-organisms
were not included in the investigations and will therefore not be included in the
prioritization list. The prioritization of the above-mentioned chemicals is listed in Table
3.
The results of the risk assessment for nitrates indicate there is a 68% chance that there are
going to be negative impacts on the environment (including human health) as a result of
groundwater becoming polluted with nitrates as a result of the wastewater treatment
works.
Wellhead protection zones were calculated for boreholes in the Cape Flats. Table 6
contains the information needed for the calculations and Table 7 defines the protection
zones and gives the radius of protection zones.
Pollution of South Africa’s urban aquifers presents a threat to the sustainability of this
water resource. Man’s activities, use of chemicals and generation of wastes tend to
concentrate potential sources of pollution in the urban areas. The threat caused by
undesirable substances is recognized in this country, but the understanding of the extent
of the problem in South Africa’s urban catchments is poor.
This paper therefore briefly outlined a risk-based methodology to prioritise and
determine the impacts of pollutant sources and pollutants. The methodology takes the
following into account:
● Characteristics of pollutant sources
● Characteristics of pollutants
● Human health impacts of pollutants
● Vulnerability of South African aquifers
● Duration of pollution
The methodology was then applied to determine:
● National list of priority chemicals and sources
● Regional list of priority chemicals and sources for the large South African urban areas
● Local risk assessments to determine the risks of certain pollutants
● Delineation of protection zones
The results are intended to help groundwater practitioners and water authorities in
assessing the likely transport, fate an impact of pollutants in the subsurface in an urban
environment.
It is recommended that the following aspects receive more attention in future research
investigations:
● Based on the paucity of groundwater-related microbial data encountered in this project,
the inclusion of these aspects in urban groundwater management must be regarded as a
priority.
● Petroleum products, industrial thinners and mineral oils and other non-aqueous phase
liquids represent a category of potential pollutants that have been largely overlooked
by regulatory agencies and legislature, despite their harmful effects at small
concentrations.
● A general lack of data on groundwater pollution from pesticides is evident. This is due
to: (i) surface waters are the main source of water supply in the country; (ii) cost and
difficulty to measure organic contaminants; (iii) private companies are often sensitive
to make public data related to pollution problems. Therefore there is a need to
investigate pesticides in groundwater.
Water resources of arid areas 108
REFERENCES
Lynch, S.D., Reynders, A.G. & Schulze, R.E. 1994. Preparing input data for a national-scale
groundwater vulnerability map of Southern Africa. Water SA, 20(3):239–246.
National Environmental Management Act. Act 107 of 1998, Pretoria, South Africa.
National Water Act. Act 36 of 1998, Pretoria, South Africa.
Nonner, J.C. 2002. Chapter 3: Sources of groundwater contamination. In: A. Zaporozec (ed.)
Groundwater contamination inventory: A Methodological Guide.UNESCO, IHP-VI, Series on
Groundwater No. 2. 23–38.
Understanding problems of low recharge and
low yield in boreholes: an example from
Ghana
A.J.E.Cobbing & J.Davies
British Geological Survey, Wallingford, Oxon, UK
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
The Afram Plains area is located in the Eastern Region of Ghana, in the Volta River basin
between latitude 6°30′ and 7°30′N and longitude 1°00′W and 0°15′E (Figure 1). The area
is about 4285km2 in extent, and lies between lake water level at 76m and 300m above
mean sea level. The topography is subdued, with the main feature being a low northeast
to southwest trending ridge 200–300m high. The Afram Plains supports savannah
vegetation that is being progressively cleared for agricultural use. Coarse tussock-grass
with a few stunted trees covers the low-lying lakeside plain and dense bush with large
trees covers the better-drained ridge area. Since initial settlement in 1930, the rural
population of the Afram Plains has increased rapidly following the construction of the
Akosombo Dam in the 1960s. Between 1970 and 1984 census data show a 250% increase
in the farming population, attracted by fertile soils and improving infrastructure. There
are now more than 140 villages on the Afram Plains. Access to the area is poor, with the
principal route by ferry across Lake Volta. The main town is Donkorkrom, which has a
hospital, secondary school and post office.
Water resources of arid areas 110
Before construction of the Akosombo Dam, village water supplies were obtained from
the perennial Afram and Volta Rivers, seasonal flows and pools along ephemeral
tributary streams and shallow water-filled dugouts. Rainfall on the Afram Plains is
seasonal, with an average of about 1200mm/year falling almost entirely between April
and October. Surface drainage is mainly ephemeral, storm water draining by sheet flow
as short-lived floods. The seasonal rainfall and limited surface water storage result in
acute water shortages during the November to March dry season. Reliance on
unprotected pools and dugouts for water supply results in water-washed and diarrhoeal
disease, and much time and effort in water collection. Guinea worm infections
occasionally occur in the Afram Plains.
Obosum Beds that underlie much of the southern Afram Plains have yet to be studied in
detail. Present geological knowledge has been derived from rapid reconnaissance
surveys, several deep exploration boreholes and a number of shallow groundwater
boreholes.
Many of the boreholes drilled were dry whilst other nominally successful boreholes
showed a progressive decline in yield to fail after two to three years of use, especially in
the west of the area. Due to their short period of land tenure, communities have yet to
develop coping strategies to manage the limited water available during dry periods.
Water resources of arid areas 112
Populations attracted to the area by the groundwater supply have no effective alternative
water source if borehole yields fail after several years of use.
The British Geological Survey (BGS) first worked on the hydrogeology of the Afram
Plains in 1985–86, in a project examining shallow wells and boreholes in the
Donkorkrom area (Buckley, 1986). Beginning with a visit to the area in February 2000,
the BGS collaborated with the Afram Plains Development Organisation (APDO),
WaterAid, DANIDA, Legon University (Accra) and other partners in a two-year project
specifically aimed at investigating the hydrogeological problems of the area. The BGS
work was funded by the British Department for International Development (DfID). The
project was timed to coincide with the drilling of 36 village water supply boreholes on
the Afram Plains, funded by WaterAid working with the APDO. In addition, DANIDA
funded the drilling of a further 5 deep (>100m) exploration boreholes, the first four of
which were sited and geologically logged by BGS hydrogeologists. Studies carried out by
the BGS in collaboration with local partners included:
● A reconnaissance geological and hydrogeological survey, and the creation of a GIS
base map of the area.
● The geophysical survey of four of the five deep exploration borehole sites using
frequency domain electromagnetic induction (EM34).
● The geological logging of rock chip samples produced during drilling, and the
recording of penetration rates and drill stem yields.
● The test pumping of boreholes, including the demonstration of bailer tests and low
yield “whale” pumps.
● The sampling of groundwaters for hydrochemical analysis of major and minor ions,
and isotopes.
● The geophysical logging of the deep exploration boreholes was carried out by
DANIDA.
BGS undertook 11km of EM34 surveys at the first four exploration borehole sites, using
10m, 20m and 40m inter-coil separations. Readings were made in both vertical and
horizontal orientations. The survey results were correlated with the geological logs from
the exploration boreholes. Geophysicists from the University of Ghana, Legon, undertook
max-min EM and electrical resistivity geophysical traverses along the main road in the
eastern Afram Plains (Banoeng-Yakubo and Armah, 2001). The result of these studies
demonstrate that geophysical surveys can be used in the eastern Afram Plains to
differentiate between near surface shale, siltstone, sandstone and conglomerate bands, as
well as delineate possible fault zones. In the west of the area, re-cemented sandstones up
to 60m thick form a low permeability homogeneous layer below an ancient weathered
surface. Thin water bearing fracture or weathered zones beneath this layer cannot be
detected using EM34 or VES equipment.
Useful geological and hydrogeological data that can be gathered during the drilling of a
borehole includes:
● Geological data
● Penetration rate data
● Flow data
● Hydrogeological data
Rock chip samples produced during drilling were collected at 1m intervals. Weathered
zones (colour changes) and fracture zones (calcite and quartz mineralisation) enabled
identification of water bearing zones. The chip samples were placed in a marked half pipe
and photographed to produce pseudo-core logs. This procedure allowed zones of water
inflow to be correlated with changes in lithology, and deductions regarding the nature of
groundwater occurrence to be made. The rate of drill penetration and flow rate,
determined at water strike zones and at the end of each drilling rod can be correlated with
changes in lithology and weathered zones. Photo logs can show the nature of the
weathered zones. The results obtained from exploration borehole showed that the rock
types present are generally tight and fine-grained, with water being produced from
horizontal weathered zones and along lithological boundaries rather than near-vertical
fractures. In the western half of the area, the presence of a thick duricrust weathered zone,
stopping recharge to underlying aquifer systems, was recognised.
6 BOREHOLE GEOPHYSICS
A suite of geophysical logs was obtained from six boreholes in the study area, i.e. the five
deep exploration boreholes together with a water supply borehole located at the APDO
office in Tease. The calliper logs show the fracture zones, which can be correlated with
the drillers report and the chip sample logs. The fracture zones are also indicated by
lower resistivity measurements.
Water resources of arid areas 114
Pumped fluid logging of the boreholes clearly shows that fluid inflows occur at
discrete fractured or weathered horizons, and that most of the water obtained from the
boreholes is derived from these features. The discontinuous nature of the fracture systems
that supply water is illustrated by drilling at the APDO office in Tease: in 2000, a 70m
deep borehole (“hole no. 28”) was drilled in an attempt to provide a water supply for the
office. This borehole proved to be dry and was backfilled. In 2001 a further two
boreholes were drilled within 20m of this hole, one to 54m and a deep exploration
borehole to 152.8m. Both of these boreholes yielded water.
Pumping test data in fractured aquifers is more difficult to interpret compared with
intergranular systems. There is often a distinct change between early and late time
drawdown rates, due to the effect of fracture dewatering. This can allow erroneous
interpretations to be made, particularly if pumping tests are carried out over only short
periods of time. Pumping test interpretation requires specific training, and pumping tests
have sometimes been done on the Afram Plains merely as required by the contract,
without the pumping test information being used to inform the borehole completion. BGS
developed a simple bail test, which allows field personnel on the Afram Plains to decide
in a general way whether or not to equip a borehole, without going through the lengthier
and more complex process of a pumping test (Davies and Cobbing, 2002). There are
cases however where the bail test is inconclusive and the borehole requires a pumping
test. Bail tests are
Understanding problems of low recharge and low yield in boreholes 115
8 HYDROCHEMISTRY RESULTS
Water samples for hydrochemical analysis were obtained from 29 boreholes and wells
during the 2001 visit. Samples were taken from sources after several minutes of pumping
where possible. Measurements of pH, specific electrical conductance (SEC), temperature
and bicarbonate were taken at each site. Filtered acidified and non-acidified samples were
obtained from each source for laboratory analysis. A GPS was used to locate the areal co-
Water resources of arid areas 116
ordinates of each sample site. Stable isotope analysis (δ2H and δ18O) was carried out on
twelve samples by mass spectrometry. The results of these determinations plot close to
the world meteoric water line. There is some evidence for the possible mixing of lake-
derived waters with aquifer waters in some areas. The major and minor ion analyses
show that most determinants are within World Health Organisation (WHO) Guide
Values, with the exceptions of boron and sodium that are a problem in the unfractured
shale and sandstone area. Nitrate and ammonium levels in a few boreholes were evidence
for anthropogenic pollution, which can occur because water is able to move relatively
rapidly through fractures. The fluoride concentration in one sample exceeded WHO
Guide Values.
A five-fold hydrogeological division of the rocks of the Afram Plains can be produced,
based on the conclusions of Bannerman (1990) and Acheampong (1996), and taking the
current study into
10 DISCUSSION
In regions of seasonal or low rainfall with ephemeral drainage patterns rural settlements
may be totally dependent upon groundwater supply during the dry part of the year. Such
is the present and future shortage of land in many areas that communities once settled in
water poor areas are difficult to move. Therefore understanding of groundwater resources
is a vital factor for long-term development plans of such marginal areas. Groundwater
development in the Afram Plains has followed a pattern that is typical of areas underlain
by low permeability rocks in sub-Saharan Africa. Reconnaissance level geological and
hydrogeological surveys were first undertaken with limited drilling more than thirty years
ago. Some borehole drilling by the VRA was undertaken at the time of population
resettlement following the building of the Akosombo Dam and consequent flooding in
the 1960s, but these boreholes have fallen into disuse following lack of maintenance.
NGO-led water supply programmes, undertaken by World Vision International, a
Catholic Church Group and WaterAid, funded the drilling of some 370 boreholes on the
Afram Plains during 1984–2001. During these programmes the economic design and
construction of boreholes, and borehole drilling “success rates” were emphasised. A
borehole was judged a success if “wet” at the completion of drilling. The hand pump
equipped boreholes were expected to supply 250 people with at least 20 litres per capita
of water per day. In the Afram Plains the acceptable yield minimum is about 121min−1,
due to the low borehole yields obtained. The high borehole “failure rate” (40%) has led to
further study of the distribution of fracture and near surface weathered zones, these being
perceived as the best groundwater bearing targets. Although many boreholes have been
drilled, the geology of the area, groundwater occurrence, and the nature of the water
resource remain poorly understood. This problem is exacerbated by the failure of
apparently successful boreholes after 3–4 years of use.
Table 2. Summary of the hydrogeology of the five
hydrogeological units.
Description Ground Ground Ground Field Technology Comments
of rock/ water water water techniques
hydrogeology targets potential quality
unit
Obosum Massive Weathered ** Good. Weathered Boreholes Good
Beds— conglomerate zones and Presence conglomerate 60–100m recharge,
Upper and sandstone fracture of NO3- gravel often best sites
Voltaian zones. N and visible at located in
System Success rate NH4 surface: valleys.
~66% wet indicates EM34—used Boreholes
38%≥30l/min pollution to locate should be
in fractures and drilled to
heavily sandstones/ below
used conglomerate present day
Water resources of arid areas 118
depth of along
weathering fractures,
Fractures
poorly
defined.
Description Groundwater Groundwater Groundwater Field Technology Comments
of rock/ targets potential quality techniques
hydrogeology
unit
Unfractured Weathered * Poor to saline. Low lying Boreholes— Poor to
shale and grey zones and low altitude 50–100m moderate
sandstone fracture lake side recharge to
zones. areas. tight
Success rate EM34— formation
~50% wet moderate to except where
14%≥30l/min high conglomeratic
conductivities, bands area
used to locate present.
fracture zones Boreholes
VES— should be
indicates drilled to
depth of below present
weathering day lake
level.
Fractured shale Weathered *? Poor to Low lying Boreholes— Unknown
and sandstone zones and saline? low altitude 50–100m
fracture lakeside areas.
zones. EM34—
Success rate moderate to
Unknown due high
to lack of data conductivities,
used to locate
fracture zones
VES—
indicates
depth of
weathering
KEY: Groundwater potential: *Low; **Moderate; ***High.
Note: Groundwater Potential is an overall function of groundwater storage, groundwater yield and
groundwater residence time (length of time groundwater remains in the unit, i.e. rate of groundwater
throughflow). It indicates both the available yields and the length of time these are available for: i.e.
high, moderate or low yields, available only during the wet season and immediately afterwards, or year-
round. See below for more detail. EM34 conductivity response: High>50mmhos/m; Moderate 20–
50mmhos/m; Low <20mmhos/m.
Yield: High >1l/s; Moderate ~0.5l/s; Low <0.2l/s.
Note: Where groundwater residence times are long, groundwater availability is likely to be less
vulnerable to variations in seasonal rainfall—e.g. one year of drought.
Where few data are available locally, the interpretations given here are preliminary, and should
be updated as new data are provided by continuing groundwater development work.
Water resources of arid areas 120
The main features of the aquifer model for the Afram Plains as a whole are as follows:
● Groundwater is thought to occur in discrete fracture systems or zones of weathering.
● The geological units have different hydrogeological characteristics but all are relatively
low yielding.
● In the west of the Afram Plains in particular, the aquifer units may not be adequately
recharged during successive wet seasons, leading to the progressive mining of
groundwaters that leads to the failure of boreholes with time. Old water is often
present in the fracture systems.
● In areas where recharge of surface water occurs, rapid movement through near surface
weathered zones and fracture systems can lead to rapid transport of contaminants into
boreholes below sanitary seal zones, as indicated by high ammonium and nitrate levels
discerned in the central village borehole water sources on the Afram Plains.
● Water bearing weathered zones may be too deep and discrete to be determined using
geophysical survey methods.
● Drilling deep boreholes to below the present day lake level may allow interception of
fracture and weathered systems that can potentially be recharged by lake water. This
process of recharge from the lake remains to be proven.
● The collection of accurate geological and hydrogeological data is vital for better
understanding of the aquifer systems present. The use of currently available data is
hindered by a lack of accurate site locations. The interaction of geological factors such
as lithology, diagenesis, recent weathering, ancient weathering, tectonism with ancient
and modern water level changes needs to be understood.
12 CONCLUSIONS
The water supply problems on the Afram Plains cannot be solved by borehole drilling
and groundwater development alone. The failure of boreholes after two or three years of
use is particularly serious since in that time communities come to rely on the groundwater
resource. Conjunctive use with rooftop rainwater catchment systems and small dams may
need to be considered as well as artificial recharge to aquifers. There is a need to
understand recharge mechanisms before borehole drilling commences. This project has
demonstrated the types of data that can be easily collected at little additional cost during
borehole drilling, and the uses to which such data can be put to the benefit of subsequent
water supply programmes. A regional summary of groundwater occurrence in this
“difficult” hydrogeological area has been built up, and presented in a format that can be
used in subsequent groundwater development. The general shift from centralised
groundwater development towards demand-driven, private organisation or NGO led work
in Africa has had some benefits in terms of sustainability, community involvement and
ownership issues, and the targeting of resources at the poorest communities. However,
the negative effect has been the non-collection, storage and sharing of basic groundwater
data, which leads to a lack of understanding in those areas where the groundwater
resources are limited or difficult to access. At present, data collection is frequently seen
Understanding problems of low recharge and low yield in boreholes 121
REFERENCES
Acheampong, S.Y.1996. Geochemical evolution of the shallow groundwater system in the Southern
Voltaian Sedimentary Basin of Ghan. A dissertation submitted in partial fulfilment of the
requirements for the degree of Doctor of Philosophy in Hydrology/Hydrogeology, Univ. of
Nevada, Reno.
Acheampong, S.Y. & Hess, J.W. 1998. “Hydrogeologic and hydrochemical framework of the
shallow groundwater system in the southern Voltaian Sedimentary Basin”. Hydrogeology
Journal, 6:527–537.
Ako, J.A. & Wellman, P. 1985. The margin of the West African craton: the Voltaian Basin. Journal
of the Geological Society of London, 142:625–632.
Anani, C. 1999. Sandstone petrology and provenance of the Neoproterozoic Voltaian Group in the
southeastern Voltaian Basin, Ghana. Sedimentary Geology, 128:83–98.
Bannerman, R.R. 1990. Afram Plains borehole drilling programme, hydrogeological survey for
WaterAid Ghana.
Banoneng-Yakubo, B. & Armah, T. 2001. Hydrogeological and geophysical test investigations in
the Afram Plains, Ghana. Department of Geology, Univ. of Ghana, Legon, for DANIDA-
CWSA Project, Eastern Region, Ghana.
Black, R. & Liegeois, J.-P. 1993. “Cratons, mobile belts, alkaline rocks and continental lithospheric
mantle: the Pan-African testimony”. Journal of the Geological Society, London, 150:89–98.
Buckley, D.K. 1986. Report on advisory visit to WaterAid projects in Ghana. British Geological
Survey Technical Report.
Davies, J. & Cobbing J. 2002. An assessment of the hydrogeology of the Afram Plains, Eastern
Region, Ghana. British Geological Survey. Technical Report CR/02/137N.
Grant, N.K. 1967. Complete Late Precambrian to Early Palaeozoic orogenic cycle in Ghana, Togo
and Dahomey. Nature, 215:609–610.
Kesse, A.O. 1988. The Mineral and Rock Resources of Ghana. A A Balkhema.
Shackleton, R.M. 1976. Pan-African structures. Philosophic Transactions of the Royal Society,
London. 280: 491–497.
World Vision 1995. The Conrad N Hilton Foundation Funded World Vision Ghana Rural Water
Project, Hydrogeological Report, Second Phase.
Spatial variation of groundwater recharge in a
semi-arid environment—Serowe, Botswana
L.M.Magombedze & B.Frengstad
Geological Survey of Norway, Trondheim, Norway
M.W.Lubczynski
International Institute of Geoinformation Science and Earth Observation,
Enschede, Netherlands
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
Serowe area, like the whole of Botswana, is characterised by a low rainfall pattern and
lack of surface water resources. People in this area depend mainly on groundwater.
Recently, the population of Serowe has increased tremendously, making it the largest
village in the country, thus widening the gap between demand and availability of water.
In this regards, it is important for sustainability and management purposes to determine
the renewable groundwater resources and how net recharge varies in space and time. This
Spatial variation of groundwater recharge 123
study focuses on estimation of net recharge rates using the chloride mass balance method
and the assesment of spatial variability of both net recharge and potential recharge.
Serowe study area is situated in the Central District of Botswana, at the eastern fringe
of the Kalahari Basin and is about 275km northeast of the capital, Gaborone (Fig. 1). It is
characterized by semi-arid climate with cool dry winters (May to September) and hot
moist summers (October to April). It receives an average annual rainfall of 447mm/year.
The most prominent feature in the area is the 90–150m high escarpment, which extends
NNE-SSW splitting the area into two hydrologically contrasting areas: eastern and
western. The western part slopes gently to the west and is covered by thick Kalahari
sands. The eastern part slopes steeply to the east and Kalahari sand cover is thin or absent
where the Ntane Sandstone or the basalt outcrops. The study area boundaries were
assigned after the boundaries of the numerical model set up by Wellfield Consulting
Services (WCS 1998, Lubczynski 2000). The northern boundary is a regionally stretching
impermeable graben. The eastern and south-eastern boundary is delineated along the
The net groundwater recharge is defined by Lubczynski (2000) as the amount of water
that reaches the groundwater minus groundwater evapotranspiration representing
discharge of groundwater by tree transpiration and evaporation from groundwater in the
form of upward convective water flux. Groundwater recharge is the most important factor
in evaluating the renewability of groundwater resources of regional aquifer systems in
arid and semi-arid environments and it is unfortunately the most difficult to quantify
(Allison 1988). Several methods of estimating recharge have been developed. These can
be divided into physically based, chemical and isotopic methods (Simmers 1997, Lerner
1990). Recently, also numerical models have been used to estimate groundwater
recharge. While making site-specific recharge measurements, one of the most important
problems to overcome is the spatial data presentation (Allison 1988, Lerner 1990). In this
study, two techniques i.e. kriging interpolation (Ahmed et al. 1995, Gieske 1999) and
spatial extrapolation with intergrated GIS recharge modelling technique were applied to
assess spatial variability of recharge.
(1)
Spatial variation of groundwater recharge 125
where RT=total recharge rate (mm/yr), TCl=average annual total chloride deposition at the
surface (mgm−2/yr); P=rainfall (mm/y); ClP=chloride content in rainfall (mg/l);
Clgw=chloride content in groundwater (mg/l); D=dry deposition of chloride measured
during dry season (mgm−2/yr).
Due to lack of total dry deposition data during this investigation, estimation of
recharge was based on the 1986/93 dry deposition of chloride (D) for Serowe of
442±124mgm−2/yr determined from rain gauge measurements (Selaolo 1998). Also
Gieske (1992) recommended a similar D of 400–500mgm−2/yr for Serowe area. Based on
127 measurements of chloride in groundwater collected in September 2001, site-specific
net recharge rates ranging from 2mm/yr to 74mm/yr were calculated according to
equation 1. The results are shown in Figure 2. Two high values were regarded as outliers
and have been omitted from further calculations. The method gave a harmonic mean
recharge rate of ~12mm/yr, which is within range of estimates from previous studies. For
example, SGS (1988) calculated net recharge rate of 11.7mm/yr from the chloride mass
balance method.
indicate places of locally high and low net recharge. Low net recharge in the north could
be explained by confinement by very thick basalt layer hampering net recharge. High
evapotranspiration resulting in high chloride concentration in groundwater is the likely
cause of locally low net recharge in the east. High net recharge in the northeast could be
due to the absence of a confining basalt layer. Generally, recharge is lower in the western
part than in the eastern part. This is due to the thick sand and basalt cover, deeper
groundwater table and also lower rainfall in the western part of the study area.
landform and depth to water table (Lerner et al. 1990). However, these factors influence
recharge with different weights. Therefore, GIS modelling, which involves combining
maps of different recharge attributes was used to come up with a recharge potential map,
qualitatively displaying the spatial distribution of groundwater recharge potential. The
modified index overlay method described by Bonham-Carter (1997) was applied in this
study. Each attribute map was subdivided into six classes, with most recharge suitability
assigned score 5 and 0 representing the least recharge suitability. Each attribute map was
then assigned a weight according to its significance in controlling recharge. The average
score of each pixel is defined as:
(2)
Water resources of arid areas 128
where S=the weighed score for each pixel, Wi=the weight of the ith map and Sij=the score
of the jth class of the ith map.
The recharge potential map of the study area was derived considering the following
factors influencing recharge potential:
– Basalt cover: Basalt lacks primary porosity and impedes vertical flow of water except
where it is fractured. Therefore high recharge suitability is expected where basalt is
absent (score 5) and least recharge where basalt is present (score 0).
– Kalahari thickness and depth to groundwater table: The thinner the Kalahari layer the
faster water passes the unsaturated zone and reaches the aquifer and the less chances
of water encountering duricrusts. The shallower the water table the faster water
reaches the saturated zone. Assigned scores are shown in Tables 1 and 2. A correlation
matrix showed that Kalahari thickness and depth to groundwater table are positively
correlated. Therefore, the two scored maps were crossed and combined in a new map
which was used for assigning recharge suitability scores.
– Soils: Porous and coarse textured soils have high infiltration capacity and low field
capacity and therefore enhance recharge processes. Soils with high silt and clay
content do not release water to the lower zone fast enough to allow recharge processes
to take place. Scores were assigned according to soil type and average infiltration rates
as described by De Wit & Nachtergaele (1990), see Table 3.
– Vegetation density: Areas with high vegetation density have also large density of root
network and are therefore less suitable for recharge. High vegetation density may also
facilitate infiltration by reducing runoff, but this is less important in case of Serowe
due to the generally high infiltration capacity of the Kalahari sands. A NDVI map
constructed from the Landsat TM 7 image of 24 April 1998 was classified into three
recharge suitability classes (Table 4).
– Slope: Slope of the landscape influences recharge rate. Generally, the steeper the slope
the more runoff and the less the amount of water infiltrating the soil. A slope
percentage map was assigned scores according to the FAO slope classification (Allen
et al. 1998) see Table 5.
– Lineament density: The influence of lineaments on recharge is greatest when the
fractures and faults are deep, continuous over some distance and are not filled with
secondary material. Not
Table 3. Scores for soils.
Soil type Average infiltration rate Score
(cm/hr)
Arenosols 25–33 5
Regosols 22–30 4
Luvisols 0.05–0.8 3
Lixisols 0.05–0.8 3
Vertisols <0.05 1
130–140 Medium 3
140–250 High 1
Kalahari 7 7 8 7
thickness/ depth
to water table
Vegetation cover 7 6 7 6
every lineament, but the majority of them can provide paths for infiltrating water.
The density of lineaments was assessed from a combination of a lineament map
obtained from WCS (1998) and lineaments derived from satellite TM5 data
(Table 6).
– Vegetation cover: Evapotranspiration and interception vary with vegetation cover.
More evapotranspiration is expected from woodlands than from shrubs and grass.
Therefore high recharge is associated with grasslands and low recharge with
woodlands. In this assessment the vegetation map prepared by Ecosurve (1998) was
used (Table 7).
changing scores and weights to fit the point distribution of recharge (Lubczynski &
Gurwin 2004). However, due to time constraints, this could not be done resulting in
deterministic recharge potential rather than recharge potential modelling. Though
verification data is missing for the far eastern and northern part of the study area, Model
3 is in best agreement with most of the point chloride mass balance results and was
therefore considered the most realistic. Figure 3 shows the spatial variation of recharge
potential based on Model 3 overlain with point net recharge values. In this map there are
places in the central part of the study area with inconsistencies, where the moderate
recharge values of the chloride mass balance fall in the very low recharge category of the
recharge potential map. This could be explained as inaccuracy of the lineament
assessment in terms of the size i.e. depth, width and openness which could not be
incorporated in the GIS map modelling. Also, it is difficult to assess which lineaments are
important for recharge and which ones are not. Other inconsistencies such as those in the
northeast, north and center where low chloride mass balance recharge rates fall in the
moderate to high recharge potential zones are most likely attributed to the subjective way
of scoring and weighting in map modelling. It could also be attributed to the
underestimation of recharge due to the presence of groundwater evapotranspiration in
that area.
3 CONCLUSIONS
– The chloride mass balance technique gives an insight into the spatial variation of net
recharge. Net groundwater recharge in the Serowe study area is spatially variable and
it generally ranges from 2mm/yr to ~30mm/yr with a harmonic mean of 12mm/yr.
– The recharge potential map obtained by GIS recharge map modelling provides a semi-
quantitative distribution of recharge. This technique is a very useful tool in
reconnaissance studies and in numerical model calibration, particularly when
quantitative data is limited, because it is able to integrate spatial data from various
sources. The zones do not necessarily show that recharge occurs or how much
recharge occurs but give an indication of where recharge is most likely to occur. Such
an assessment can therefore also be used as a planning tool taking into account local
hydrogeological knowledge and constraints in the development, management and use
of groundwater resources.
– The process of GIS modelling as presented is largely based on expert knowledge of the
modeller as well as knowledge of the area. More objective solutions can be obtained
by optimization of the scores and weights to best fit the recharge attribute zones with
the point measurements (Lubczynski & Gurwin 2004).
– The pattern of net recharge interpolated by kriging is more or less similar to the
extrapolated recharge pattern obtained by intergated GIS modelling of recharge
potential. The differences can be attributed to inaccuracies in both methods and also to
the fact that chloride mass balance gives net recharge while GIS map modelling shows
potential recharge.
Water resources of arid areas 132
REFERENCES
Ahmed, S., Sankaran, S. & Gupta, C.P. 1995. Variographic analysis of some hydrogeological
parameters: Use of Geological soft data. Journal of Environmental Hydrology 3(2).
http://www.hydroweb.com/
Allen, R.G., Pereira, L.S., Raes, D. & Smith, M. 1998. Crop evapotranspiration, guidelines for
computing crop water requirements, Food and Agriculture Organisation of the United Nations,
FAO Irrigation and Drainage paper 56, Rome, Italy.
Allison, G.B. 1988. A review of the physical, chemical and isotopic techniques available for
estimating groundwater recharge. In I.Simmers (ed.) Estimation of Natural groundwater
recharge. NATO ASI series, C222, Reidel, Dordrecht: 49–72.
Beekman, H.E., Selaolo, E.T. & Nijsten, G.J. 1996. Groundwater Recharge at the Fringe of the
Botswana Kalahari-The Letlhkeng-Botlhapatlou Area. Botswana Journal of Earth Sciences 3.
Cohen, W.B., Spies, T.A. & Bradshaw, G.A. 1990. Semi variograms of Digital Imagery for
analysis of conifer canopy structure. New York: Elsevier Inc.
Ecosurve. 1998. Vegetation mapping and ground truthing for Radar Imagery (Vegetation report).
Serowe. Ecosurve project.
Eriksson, E. & Khunakasem, V. 1969. Chloride concentration in groundwater, recharge rate and
rate of deposition of chloride in the Israel coastal plain, Journal of Hydrology 7:178–197.
Hendrickx, J.M.H. & Walker, G.R. 1997. Recharge from precipitation. In I.Simmers (ed.)
Recharge of Phreatic Aquifers in (Semi-) Arid Areas. Rotterdam: Balkema.
Gieske, A. 1992. Dynamics of groundwater recharge: A case study in the semi-arid eastern
Botswana. PhD thesis, Vrije Unversiteit Amsterdam, The Netherlands.
Gieske, A. 1999. Geostatistics for hydrologists, Principles and applications, Lecture notes, Adapted
from de Marsily, 1986, ITC, Enschede, The Netherlands.
Lerner, D.N., Issar, A.S. & Simmers, I. 1990. Groundwater Recharge: A guide to understanding
and estimating natural recharge. International contributions to Hydrogeology 8.
Lubczynski, M.W. 2000. Groundwater evapotranspiration—Underestimated component of
groundwater balance in a semi-arid environment—Serowe case, Botswana. In Oliver Sililo et al.
(eds), Groundwater: Past achievements and future challenges: 199–204. Rotterdam: Balkema.
Lubczynski, M.W., Gurwin, J. 2004. Integration of various data sources for transient groundwater
modelling—Sardon study case, Spain. Journal of Hydrology—in revision.
Selaolo, E.T. 1998. Tracer studies and groundwater recharge assessment in the eastern fringe of the
Botswana Kalahari, The Letlhakeng-Botlhapatlou Area. PhD thesis, Vrije Universiteit
Amsterdam, The Netherlands.
Swedish Geological Survey (SGS) 1988. Serowe Groundwater Resources Evaluation Project, Final
Report, Ministry of Mineral Resources and Water Affairs, Department of Geological Survey,
Lobatse, Botswana.
Wellfield Consulting Services (WCS) 1998. Serowe wellfield 2 extension project (TB10/3/10/95–
96), Main report, DWA, Gaborone, Botswana.
Wellfield Consulting Services (WCS) 2000. Serowe wellfield extension project, Groundwater
Modelling report, DWA, Gaborone, Botswana.
Quantification of artificial ground water
recharge
G.C.Mishra
Water Resources Development Training Centre, Indian Institute of
Technology, Roorkee, India
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
Sedimentary groundwater basins are mostly comprised of alternate layers of sand and
clay. Occurrence of a clay layer at the top surface prevents direct recharge from rainfall.
Natural and man-made surface drains in such a region are likely to carry away most part
of the rainfall as direct runoff. These drains, while conveying the runoff, can be
considered as a source of water for artificial recharge. Vertical shafts (recharge well) or a
pit may be constructed in the bed of the surface water body through the intervening clay
layer to facilitate recharge to the underlying confined aquifer (Sandford, 1938). If the
piezometric head in the aquifer stands below the water level in the drain, the recharge
would take place under the action of gravity. The recharge rate is governed by (i) the
difference in the hydraulic heads at the water body and in the confined aquifer under the
shaft, (ii) diameter and length of the shaft, (iii) transmissivity and storage coefficient of
Water resources of arid areas 134
the aquifer being recharged, and (iv) hydraulic conductivity of the coarse material the
shaft may be filled with. Appropriate methods of artificial recharge for different geo-
hydrological conditions have been described in detail (Todd, 1985; Oaksford, 1985). In
the present study, an analytical method is described to quantify time variant recharge
from a surface water body to a confined aquifer through a vertical shaft or a recharge well
under the action of gravity.
In the paper, analytical solutions have been obtained applying unit response function
coefficients known as discrete kernels and convolution technique to quantify time variant
recharge from a surface water body to a confined aquifer through a vertical shaft and a
recharge well. The following cases have been dealt: (i) a vertical shaft marginally
penetrating into an aquifer and filled with coarse sand; (ii) a vertical shaft marginally
penetrating into an aquifer; (iii) a recharge well fully penetrating an aquifer. A shaft with
radius ranging from 1 to 2m filled with coarse sand can recharge at a significant rate
between 250 to 700m3/day. If the shaft is not filled with sand, the rate of recharge at large
time is twice that of when the shaft is filled with coarse sand. Recharge through fully
penetrating well is more than 10 times that of the recharge through a vertical shaft filled
with coarse sand.
3 ASSUMPTIONS
The assumptions made to quantify the recharge rate are:(i) the time span is discretised by
time steps of uniform size ∆t (day); within each time step the recharge rate is uniform; the
varying recharge is a train of pulses, (ii) an unsteady state is a succession of steady states,
(iii) within a time step, Bernoulli’s equation is applicable.
Quantification of artificial ground water recharge 135
4 ANALYSIS
4.1 Case 1: A vertical shaft penetrating marginally into the aquifer and
filled with a coarse material
A vertical shaft penetrating marginally into an aquifer can be treated as a recharge well of
zero penetration. Hantush (1961) has derived an analytical expression for evolution of
piezometric surfaces in response to continuous uniform pumping from a well with zero
penetration. The corresponding Hantush’s well function can be used to compute the
evolution of rise in piezometric surface due to a unit pulse recharge. The response of a
linear system to a unit pulse perturbation has been designated as discrete kernel
coefficient (Morel-Seytoux, 1975). Morel-Seytoux and Daly (1975) have demonstrated
the use of kernel coefficients in solving complex ground water flow problems.
The rise in piezometric surface is expressed in terms of varying recharge and kernel
coefficients derived from Hantush’s well function using a convolution technique. Let the
time span be discretised by time steps of uniform size ∆t and let the time varying
recharge through the shaft be treated as a train of pulses. Let R(γ) be the recharge during
γth time step. Let δp(m, ∆t) be the rise in piezometric surface at the well face at time m∆t
due to unit recharge (unit pulse input) that occurs during the first time step only. The
expression for δp (m, ∆t), the kernel coefficient, is given in Appendix-1. The rise in
piezometric surface, s(rw, m∆t), at the recharge well face at time m∆t due to variable
recharge, R(γ), γ=1, 2, …, m, is given by:
(1)
The hydraulic head at time m∆t at the bottom of the shaft is summation of the initial
height h2 and the rise in piezometric surface, s(rw, m∆t). Applying Darcy’s law, the
recharge during mth time step is given by:
(2)
in which, kf=hydraulic conductivity of the coarse material the shaft is filled with. The
term within the bracket is the hydraulic head difference dissipated in length L of the shaft.
Solving for the recharge during the mth time step from (2)
Water resources of arid areas 136
(3)
4.2 Case 2: A vertical shaft penetrating marginally into the aquifer, with
no filling material
An analytical expression for recharge is derived applying Bernoulli’s equation (vide
Streeter and Wylie, 1981). Accounting for the entry, exit, and friction losses and applying
Bernoulli’s equation
(4)
(5)
in which, ce=coefficient of entry loss, f=friction loss factor, L=length of the shaft, γw=unit
weight of water, g=acceleration due to gravity (m/sec2), v=velocity of water in the shaft
(m/sec) during mth time step, R(m) recharge volume (m3) during mth time step, ∆t=time
step size (day). The first term in the right hand side of equation (4) accounts for entry
loss, the second term accounts for friction loss in the shaft and the third term is the
expansion loss at the exit of the shaft. s(rw, m∆t) is the rise in piezometric surface at the
recharge well consequent to the recharge taken place. Incorporating (1) and (5) in (4) the
following quadratic equation in R(m) is obtained:
(6)
(7)
in which,
For m=1, c=h2−h1. R(m), m=1, 2,…n, can be found in succession starting from m=1. The
recharge rate during mth time step is equal to R(m)/∆t (m3/day).
Water resources of arid areas 138
5 RESULTS
The kernel coefficients are generated assigning values to the aquifer parameters. The
thickness of the intervening clay layer is taken as 10m; the hydraulic conductivity of the
packed porous medium is 10 times the hydraulic conductivity of the aquifer medium and
is assumed to be 380 m/day. The friction factor f=0.02; entry loss coefficient ce=0.05;
bl/rw. The corresponding non-dimensional recharge rate, R(m)/∆t T(h1−h2), with
dimensionless time factor, , are presented in Figure 2. Since, the recharge
rate is governed by the difference in the hydraulic heads at the entry and exit points of the
shaft, and the head difference decreases with time, the recharge,
therefore, decreases with time. When the shaft is filled with the coarse material, the rate
of recharge at large time is half of the recharge rate that would occur without filling
material in the shaft. Recharge through fully penetrating well is more than 10 times that
of the recharge through a vertical shaft with a filler material.
Numerical results are presented for the following aquifer parameters for various radii
of the recharging structure: transmissivity, T=655.5m2/day; storativity, ;
thickness of clay layer, L=10m; initial hydraulic head difference, h1−h2=5m. The average
recharge rates during 120 days for different well radii are presented in Table 1. A vertical
shaft with 2m radius, 10m length filled with a filter material having hydraulic
conductivity of 380m/day, can recharge at an average rate of 700m3/day under an initial
hydraulic head difference of 5m.
6 CONCLUSIONS
Analytical methods are presented to estimate unsteady recharge, that can occur under the
action of gravity, through (i) a vertical shaft filled with coarse sand, (ii) a well penetrating
marginally into an aquifer, and (iii) a fully penetrating well. Application of unit response
function coefficient is illustrated while quantifying the recharge rate. A vertical shaft with
radius ranging from 1 to 2m filled with coarse sand can recharge at a significant rate
between 250 to 700m3/day.
7 APPENDICES
(1)
and
Let the time domain be discretised by time steps of uniform size ∆t. The unit pulse
response function of the system, δp(m, ∆t), is given by:
(2)
W(u) and Wn(u, nπrw/b) are improper integrals as the upper limit of integration is infinite.
W(u) is Theis’ Well function and can be computed using the polynomial and rational
approximation (Abromwitz and Stegun, 1970)
Wn(u,n πrw/b) is evaluated using Gaussian quadrature after converting the improper
integral into proper integral and changing the limit. The procedure is as follows.
Quantification of artificial ground water recharge 141
As x→ −1, the value of the integrand in the second integration is found as follows:
(1)
in which,
Water resources of arid areas 142
functions of first kind of zero and first order respectively; Y0(x) Y1(x)=Bessel functions
of second kind of zero and first order respectively; T=transmissivity (m2/day), and
ø=storativity of the upper aquifer; rw=radius of the well or shaft(m).
The integral in (1) is an improper integral as the upper limit of integration is infinite.
The improper integral is reduced to a proper integral as described below.
Expanding the exponential term, and applying L’ Hospital’s rule, it can be shown that as
v tends to −1, the integrand tends to 0. The integral I1 is a proper integral and can be
evaluated numerically using Gauss quadrature.
Similarly,
REFERENCES
Abramowitz, M. & Stegun, I.A. 1970. Handbook of Mathematical Functions. Dover Publications,
Inc, New York, 231pp.
Carslaw, H.S. & Jaeger, J.C. 1959. Conduction of Heat in Solids. New York, Oxford Univ. Press:
338pp.
Hantush, M.S. 1961. Drawdown around a partially penetrating well. J. Hydr. Div., ASCE,
87(HY4):83–98.
Hantush, M.S. 1964. Hydraulics of wells. Advances in Hydroscience, Ed. Ven Te Chow, Vol. 1,
340pp.
Morel-Seytoux, H.J. 1975. Optimal legal conjunctive operation of surface and ground water. Proc.
Second World Congress. Intl. Water Resour. Assoc., New Delhi, Vol. IV:119–129.
Model-Seytoux, H.J. & Daly, C.J. 1975. A discrete kernel generator for stream-aquifer studies.
WaterResour. Res., 11 (2):253–260.
Oaksford, E.T. 1985. Artificial Recharge: Methods, Hydraulics, and Monitoring. Artificial
Recharge of Groundwater. Ed. Takashi A. Butterworth Publisher: 69–127.
Sandford, H.J. 1938. Diffusing pits for recharging water into underground formation: chemical well
cleaning methods. American Water Works Association Journal, 30(11):1755–1766.
Todd, D.K. 1985. Groundwater Hydrology. New York, John Wiley & Sons: 458–493.
The architecture and application of the South
African Groundwater Decision Tool
I.Dennis & G.J.van Tonder
Institute for Groundwater Studies, University of the Free State, South
Africa
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
The purpose of the SAGDT is to assist the Department of Water Affairs and Forestry
(DWAF) and Catchment Management Agencies (CMA) in decision making with regard
to aquifer protection and management. The following applies to the SAGDT:
● Consists of a standard system of consistent methods/rules to guide planning and
decision making about water resources.
● Allows transparency, accountability and long-term goal-setting to be incorporated into
water resource management.
● Calculates the level of confidence of results obtained.
This paper discusses the legal requirements and policies that the SAGDT complies with
together with the application architecture to highlight the methodology used in the
design. Finally a simple case study is given to demonstrate the application of the tool.
The SAGDT is aligned with existing legal requirements, policies and DWAF activities
which are discussed in the sections below.
objectives of the groundwater network is to develop training material and workshop the
groundwater RDM process. The SAGDT has been included in this training and the
relevant components will be presented at the workshops.
Selection of the membership function is done by an expert on the field of study. Linear
membership functions are seldom used in practice in contradiction to sinusoidal
functions, which are very popular. In most cases risk analysis will involve more than one
input to be considered in the analysis.
Fuzzy logic makes it possible to generate a set of decision rules according to the
number of inputs, and these rules must then be evaluated by an expert in the field of
study. The number of rules generated is given by Equation 1.
n=2inputs
(1)
where n represents the number of rules generated. The rules consist of all possible binary
combinations of the respective inputs with a weight assigned to each rule representing the
risk. Table 1 shows the decision rules generated for 3 inputs. Instead of true and false the
terms favourable and unfavourable are used to make the rules easier to read.
The architecture and application of the South African groundwater decision tool 149
(2)
where
n=number of rules
DOM=degree of membership
Wn=Weight of rule n
Note that the minimum function must return the minimum value of all inputs for each
rule.
4 SYSTEM ARCHITECTURE
The sub-systems comprising the SAGDT are discussed in the sections that follow. Refer
to Figure 2 as reference for the sections to follow.
In some instances more than one object performing the same function exists, and they
differ only in the detail their attributes require as well as their respective confidence
levels. In building a scenario, an object tree will result due to the parent child
relationships of the objects used. The base object will always be a GIS object
representing the area of the internal world. This object tree is then passed to the fuzzy
logic engine to determine the risk assessment result, and further analysed to determine the
confidence levels associated with it.
A warning system exists so that the user will be notified when to do a more detailed
scenario. The warning system uses the confidence level, risk assessment result and policy
as inputs to
5 CASE STUDY
5.1 Geology
The region consists of mudstones, shales and sandstones from the Adelaide Subgroup of
the Beaufort Group within the greater Karoo Super Group. Post-Karoo dolerite sill
intrusions are present, which have to a large extend been eroded, exposing the underlying
sedimentary rocks. The surface dolerite on the Campus are highly fractured with little
ground cover and it is assumed that recharge over these areas is probably high.
weathered sections of the aquifer, diagonal fractures intersect the bedding plane fractures.
The sandstone containing the most horizontal fractures also forms the main water-
carrying formation.
5.3 Scenario
The scenario that will be evaluated for this case study is the determination of the
sustainable yield of UO5 when pumped continuously for 2 years. The assumptions used
in the case study are as follows:
● Only UO5 will be pumped at variable rates over the assessment period.
● The assessment period is 2 years.
● Three levels of evaluation will be done, that is each successive analysis uses objects
with higher confidence levels than the previous set of objects. This implies that more
detailed data are needed for objects used with higher confidence levels.
5.4 Results
Figure 5 shows the results obtained from the SAGDT for the specified scenario. From the
graphs it is clear that the higher the confidence level of the scenario, the more accurately
the risk of failing can be determined. There exists a good correlation between the 99%
Water resources of arid areas 154
risk of failure for each of the confidence levels evaluated, but the higher confidence
scenarios give opportunity for better management.
As a recommendation one could propose a 0.4L/s abstraction rate from the 89%
confidence scenario, which indicates a 20% risk of failure. From extensive field
investigations it has been proved that UO5 can be pumped for 6 months at 0.33L/s
without failing, which correlates well with the proposed recommendation.
It is important to note that the tool will produce an overestimate for the risk when the
confidence is low. This is why a warning system was implemented to make sure the user
is aware of the fact that the risk is too high according to policy and that a more detailed
analysis is required to confirm the high risk situation. This will prevent users from
making management decisions based on high-risk results with low confidence.
6 CONCLUSIONS
The SAGDT has proven to be a powerful groundwater management tool. The tool
provides a common framework for all groundwater practitioners in South Africa in which
they can perform groundwater risk assessments that relate to policy.
By employing fuzzy logic to do the risk analysis the user has the knowledge of an
expert captured in the application to assists in the decision-making process.
The SAGDT also acts as a groundwater educational environment, due to the extensive
groundwater dictionary and object help files available.
The architecture and application of the South African groundwater decision tool 155
REFERENCES
Constitution of the Republic of South Africa. Act No 108 of 1996, Pretoria, South Africa.
DWAF 1999. Water Resources Protection Policy Implementation—Resource Directed Measures
for the Protection of Water Resources Version 1.0 Volumes 2–6; Department of Water Affairs
and Forestry, Pretoria.
DWAF 2003. Resource Directed Measures—Module 1—Introductory module; draft edition August
2003, Department of Water Affairs and Forestry, Pretoria.
DWAF 2003a. National Water Resource Strategy; current draft edition, Department of Water
Affairs and Forestry, Pretoria.
DWAF 2004. Groundwater resource directed measures training manual. Sponsored by Fetwater,
Department of Water Affairs, Pretoria.
National Water Act. Act 36 of 1998, Pretoria, South Africa.
Water Services Act. Act No 108 of 1997, Pretoria, South Africa.
The development of a groundwater
management tool for the Schoonspruit
dolomitic compartment
B.H.Usher
Institute for Groundwater Studies, University of the Free State,
Bloemfontein, South Africa
S.Veltman
Department of Water Affairs and Forestry, Geohydrology Division, Free
State Region, Bloemfontein, South Africa
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
2.1 Overview
The setting can be described in more detail as the compartment is categorised as
Transvaal Highlands with elevation changes of more than 100m over a 40-km distance.
The topography slopes downward from the northeast to the southwest. The Pretoria
Formation in the north forms the water divides, in the north of the compartment, between
the Vaal and Limpopo rivers. The Schoonspruit compartment falls within the surface
water drainage area C24, drained by the Schoonspruit, and circular depressions can be
found in the area that show elements of karstic evolution. Most of the rainfall occurs from
November to February. The average rainfall for the area is 606mm and the average
evaporation is in the vicinity of 1900mm.
2.2 Geology
The geology of the area can best be described by differentiating between the main
geological systems. In general the geology is known as dolomites of the Malmani
Subgroup that plunge regionally northward and are overlain by the Pretoria Group.
Outcrops of the Witwatersrand Supergroup appear along the southern boundary of the
dolomites (Fleisher, 1981). The Malmani Subgroup is described as dolomite, banded iron
formation, chert and shale. This series consists mostly of layered strata of calcium
Water resources of arid areas 158
magnesium carbonates (CaMgCO3), some layers massive and some with chert bands.
Secondary limestone also occurs in the dolomites and is widely mined for the
manufacturing of cement. Dolomites in this area are generally easily weathered and form
undulating landscapes. (Kok, 1972)
The Subgroup is further described as representing the dolomitic sequence and is
largely concealed by overburden throughout the study area and therefore difficult to trace
(Polivka, 1987). The majority of outcrops and aquifers in the area are associated with the
Malmani Subgroup (Kotze, 1994).
2.3 Geohydrology
The dolomitic aquifer of the Schoonspruit dolomitic compartment consists of four
different formations (Polivka, 1987). Of these the chert-rich formations, Monte Christo
and Eccles, are better aquifers compared to the chert-poor formations, Oaktree and
Lyttleton, and boreholes drilled on fault intersections also gave high yields (>10l/s)
(Kotze, 1994). The strata dip northward and are overlain by the Pretoria Group. Average
borehole yields differ for the different formations and range from 11l/s in the Eccles to
3l/s in the Lyttletone formation (Polivka, 1987). Agriculture has the most important
influence on the compartment’s water quality. As such, factors such as nitrate pollution
are of particular importance.
3 GEOHYDROLOGICAL DESCRIPTION
The most important aspect for the geohydrological description was the determination of
an appropriate water balance methodology for the compartment. Details regarding the
verification of the compartment boundaries, current water quality and aquifer parameter
determination are contained in Veltman (2003).
hi=(b/s)/n.Rfj+F
(1)
incorporates the groundwater balance to a degree, although levels and not volumes are of
concern. The storativity and recharge values, attributed to the specific borehole’s reaction
to averaged rainfall over an area, are obtained. In the dolomitic aquifer these values are
not as related to fracture flow, since the dolomites characteristics cause variations in
groundwater levels to be smoothed over an area. The values are therefore indicative of
the aquifer characteristics. Further information gained from these simulations is the
effective depth of the aquifer and the threshold values of rainfall before recharge will take
place.
Table 1 summarise the information as acquired with these simulations.
The simulations provide valuable information for use in regional modelling of the
aquifer, and for determining the groundwater balance. However, for proper management
of the relationship between the system’s response and the flow of the Schoonspruit Eye
needs refinement.
The calibrated parameters for the system have been determined as ReN (7%), ReF%(44),
ThN(26mm) which the recharge threshold and ThF(43mm) which is the flood recharge
threshold. The Rf values all refer to the month lag included. The equation therefore
amounts to Equation 5:
Schoonspruit Flow
(Mm3/m)=(0.07*Rf24MMA/Rf120MMA*(Rf96MMA−26)*0.842) (5)
+(0.44*(IF((Rf120MMA−43)>0,
Rf120MMA−43))*0.842)−AbsGW
This equation is used in the groundwater management tool of the Schoonspruit dolomitic
compartment. The biggest advantage of this method is that abstractions can now be
incorporated into the simulation and predictions can be made with long-term predicted
rainfall. The effective recharge for the Schoonspruit Eye was determined as 13% for
2002.
Water resources of arid areas 164
4 MANAGEMENT TOOL
4.2 Users
The users of such a tool range from the groundwater user’s association to the regulators
and also groundwater consultants operating in the area. The tool needs to be versatile and
contain all the necessary geohydrological equations, yet at the same time to be user-
friendly. Equations for inclusion in this tool included the Schoonspruit Eye simulation
equation and incorporation of domestic and ecological requirements.
4.3 Input
The tool was constructed in such a way that only the latest rainfall and water quality need
to be included as time-variant data. Aspects such as the reserve requirements and
rainfall/recharge equations are built in.
quantities and qualities. Data input includes compulsory data inputs and optional data
inputs. The optional data are helpful if available, but the simulations are not dependent on
these cells to run. Navigational buttons to other sheets and data entry points are also
included. The Drinking Water Quality Classes for the different parameters are included
as fixed parameters.
On the Prediction sheet of the Tool all the calculations for the simulation of the
Schoonspruit Eye flow, therefore allocable volumes are done. The simulated flow
(Mm3/m) is then determined using the spring flow equation as only rainfall values, and
not equations, are now incorporated. Input data to this sheet is obtained from the Data
sheet.
Allocable volumes are determined with the simple equation of subtracting surface
water demand from the simulated flow, as this has already taken into account current
groundwater use. Figure 4 shows the spring flow and allocable volume graph.
5 DISCUSSION
The aim of the groundwater management tool was to provide a first order technical tool,
which is a practical and workable tool, for use by the WUA in determining allocable
volumes.
The following conclusions are made with regard to the groundwater management tool:
● Input and output parameters as outlined in this paper were used and proved to be
sufficient for defining quantity and quality concerns in the Schoonspruit dolomitic
compartment.
● Allocable volumes can be determined for the two zones using predicative rainfall data.
● The Schoonspruit Eye can be simulated using predicative rainfall data with the
following equation:
Schoonspruit Flow
3
(Mm /m)=(0.07*Rf24MMA/Rf120MMA*(Rf96MMA−26)*0.842)
+(0.44*(IF((Rf120MMA−43)>0,
Rf120MMA−43))*0.842)−AbsGW
Water resources of arid areas 166
REFERENCES
Bredenkamp, D.B., Botha, L.J., Van Tonder, G.J. & Van Rensburg, H.J. 1995. Manual on
quantitative estimation of groundwater recharge and aquifer storativity. Report no. TT 73/95.
Water Research Commission, Pretoria.
Bredenkamp, D.B. & Swartz, A. 1987. Reconstruction of the flow of springs by means of annual
recharge estimates. Technical report no. GH 3525. Department of Water Affairs, Directorate
Hydrology, Pretoria.
DARCY Groundwater Scientists and Consultants. 2002. A catchment management plan for the
Schoonspruit and Koekemoer Spruit catchments: A groundwater situation analysis. Department
of Water Affairs & Forestry, Bloemfontein.
The development of a groundwater management tool 167
Fleisher, J.N.E. 1981. The geohydrology of the dolomite aquifers of the Malmani Subgroup in the
SouthWestern Transvaal, Republic of South Africa. Technical report no. GH 3169. Department
Water Affairs & Forestry, Directorate Hydrology, Pretoria.
Kok, T.S. 1972. Wes-Transvaal en Noord-Kaap waterbeplanningstreek—geologie, fonteine en
myne in opvanggebied. Technical report no. GH 1758. Department of Mines, Geological
Survey, Pretoria.
Kotze, J.C. 1994. Summary of the Geology, Geohydrology, and Boundaries of the proposed
SGWCA, District Ventersdorp, Drainage Area C24. Technical report no. 3833. Department of
Water Affairs & Forestry, Directorate Hydrology, Pretoria.
National Water Act, Act No. 36 of 1998.
Polivka, J. 1987. Geohydrological investigation of the Schoonspruit compartment in the dolomitic
area of Ventersdorp. Technical report no. GH 3524. Department of Water Affairs, Directorate
Hydrology, Pretoria.
Schoeman & Vennote. 1996. Ventersdorp Oog Ondergrondse Staatswaterbeheergebied. Report no.
B0307/2. Department of Water Affairs & Forestry, Sub directorate Water Allocation, Pretoria.
Selaolo, E.T. 1998. Tracer Studies and Groundwater Recharge Assessment in the Eastern Fringe of
the Botswana Kalahari. Ph.D. thesis, Free University of Amsterdam. GRES Project Publication.
Van Tonder, G. & Xu, Y. 2001. A guide for the estimation of groundwater recharge in South
Africa. The Institute of Groundwater Studies, Bloemfontein.
Vegter, J.R. 2001. Groundwater development in South Africa and an introduction to the
Hydrogeology of groundwater regions. Report no. TT 134/00. The Water Research
Commission, Pretoria.
Veltman, S. 2003. A Methodology for Groundwater Management in Dolomitic Terrains with the
Schoonspruit Compartment as Pilot Area. Unpublished M.Sc thesis. University of the Free
State, Bloemfontein, South Africa.
Effects of mining and urban expansion on
groundwater quality in Francistown,
Botswana
Benjamin Mafa
Department of Water Affairs, Gaborone, Botswana
Horst Vogel
Department of Geological Survey, Lobatse, Botswana
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
Francistown is the oldest established town in Botswana. Born during the late 19th century
as a gold mining town at the confluence of the ephemeral Tati and Ntshe sand rivers,
Effects of mining and urban expansion on groundwater quality in Francistown, Botswana 169
Francistown is the commercial hub in the NE of Botswana. The city’s rapid economic
development, in particular since the 1970s, has caused its population to triple over the last
three decades to approximately 100,000 inhabitants. Today Francistown is the second
largest city in Botswana.
In the not too distant past, water demands were entirely met by groundwater locally
available from shallow alluvial and fractured volcanic rock aquifers. However, in the
1970s it was found that groundwater produced from the city’s public wells contained
elevated concentrations of nitrate. In addition, the available limited groundwater
resources could no longer meet the steadily rising demand for water. For these reasons
public water supply was shifted in 1982 to surface water from the Shashe dam, which is
located at a distance of approximately 30km to the SW of Francistown. The Shashe dam
was built during the 1970s to supply the copper-nickel mine in Selebi-Phikwe.
2 GEOLOGY
3 HYDROGEOLOGY
Very little detailed groundwater monitoring of the Francistown aquifers was undertaken
since the first abstractions in the early 1950s and since the recommendations made by
Water resources of arid areas 170
consultants in 1974 (Colquhoun et al., 1974) and in 1979 respectively (Gibb & Partners,
1987).
Groundwater consultants identified the major aquifer in Francistown as the
Penhalonga mixed formation about 1.5km wide extending for at least 7km downstream of
the Tati and Ntshe river confluence (Colquhoun et al., 1974). The most productive
aquifers were recognized as relatively shallow discontinuous zones of fracturing. These
fracture zones have a high transmissivity and draw from storage in the overlying
weathered rock and alluvium. They may be up to 4m thick and are usually semi-confined
by alluvial sediments and clayey weathered rock.
Confining layers composed of sandy horizons contain water and contribute leakage
into the underlying aquifer thereby acting as perched aquifers. Weathering appears to be
confined to certain horizons within the Penhalonga mixed formation where it appears to
be restricted to the easily weathered acid meta-volcanics. Indeed the river Tati is an
excellent outward expression of this feature since it also follows the geological strike of
this formation within these acid metavolcanics. The river tends to change its course
where it traverses more competent members of the Penhalonga mixed formation.
Groundwater also occurs in the sandy channels of the rivers Tati and Ntshe and this
perennial baseflow component may also be regarded as an aquifer. Upstream of the
confluence, the river Tati is 35 to 40m wide with the average thickness of the sand bed
being 1.7m. However, sand pockets of up to 3m deep exist and increase the saturated
storage of this aquifer. Downstream of this confluence, larger volumes of water can be
stored since the river becomes wider with widths ranging from 20 to 100m and deeper
sand beds of more than 2m in parts.
The study commenced with a census of all existing wells so as to establish their
distribution, usage, and availability for sampling. A Garmin 40 hand-held GPS
(http://www.garmin.com/) was used for coordinate acquisition. Similarly, all industries
and other sites that may have a negative impact on groundwater quality were mapped.
Boreholes that were found to be accessible in terms of water level measurement were
used together with the topographical elevation to infer groundwater flow directions. An
electrical dipper was used for water level measurements and a Trimble high-precision
GPS for ground elevation measurements as well as more accurate Cartesian coordinates.
The sampling of accessible boreholes involved the use of a Grundfos MP1
submersible pump (http://www.grundfos.com/) equipped with riser pipes of up to 90m.
All discharge water generated while pumping was released at least 30m away from the
borehole down gradient of the prevailing land slope.
The method of sampling was such that electrical conductivity and groundwater
reaction were measured continuously until both parameters had stabilized. Once they had
stabilized a groundwater sample was taken from the particular borehole. The sample
bottles were all made of plastic. Upon sampling, water reaction (pH), electrical
conductivity (EC), and dissolved oxygen (DO) were measured using hand-held meters
(http://www.wtw.com/). Bicarbonate and carbon dioxide (CO2) were
determined through titration.
Effects of mining and urban expansion on groundwater quality in Francistown, Botswana 171
The data obtained from the various fieldwork exercises and the hydro-chemical
laboratory analyses were used to produce several environmental geology maps. For this
to materialize, all data were transferred to the ArcView GIS software (Version 3.2)
environment (http://www.esri.com/) where the various data layers were put together to
produce the thematic maps. Data obtained from the chemical analyses were also used to
deduce redox conditions, to delineate redox zones, and to determine the predominant
redox processes.
5 RESULTS
Out of the total of 202 boreholes that were identified during the well census, only 48
could be sampled for groundwater. All the others were inaccessible because of collapse,
vandalism, or else, they had fallen dry.
The vast majority of the accessible boreholes were concentrated along the two rivers
Ntshe and Tati. However, groundwater yields were generally low. Several of the few
known borehole yields were below 2m3/h, hence their proximity to the rivers. Only very
few such as the monitoring boreholes at the abandoned and the new waste disposal site
were beyond the rivers.
The chemical analyses revealed that there was not much variation in groundwater
reaction (pH). Most of the samples had neutral pH levels around 7, which is normal for
groundwater. No groundwater sample showed acid conditions.
Magnesium (Mg2+), calcium (Ca2+), and bicarbonate were the most
important ions. Hence, Mg-Ca-HCO3 type of water was dominant. In some places, Na-
Mg-Ca-HCO3 type of waters were prevalent that also featured elevated concentrations of
nitrate , chlorine (Cl¯), and sulfate The concentration of total dissolved
solids (TDS) was less than 1000mg/L in these particular boreholes.
Over most of the built-up city area the groundwater was strongly influenced by
anthropogenic activities. This was evident from TDS levels greater than 1000mg/L, and
Cl¯, and constituted the dominant anions. In order to identify and delineate
distinct groundwater pollution zones, all chemical groundwater parameters were used as
environmental indicators and mapped individually.
The concentration of oxygen allowed to identify zones with different aeration status,
namely zones with aerobic (oxic) and those with anaerobic (probably reduced)
groundwater conditions. This was the starting point towards defining likely pollution
zones and also towards predicting redox states.
In order to allow for a sound investigation it was necessary to examine the main
indicator species for redox state, namely sulphate ferrous iron (Fe2+),
manganous manganese (Mn2+), nitrate nitrite and ammonium .
As was to be expected, a comparison between these species revealed that areas with high
levels of Fe2+ and Mn2+ had at the same time low levels of and . Equally,
areas rich in sulphates and nitrates coincided with zones high in dissolved oxygen (O2),
indicating oxidizing (aerobic) conditions, while zones high in ferrous iron and
Water resources of arid areas 172
manganous manganese overlapped with zones very low in dissolved oxygen, thus
indicating reduced (anaerobic) environments. The change from one zone to another was
gradual.
Important information in order to identify buffering systems was the presence of
carbon dioxide (CO2) and bicarbonate . The distribution of these two species
did not show a significant relationship to the redox state of the water. It rather was related
to the calcium (Ca) and magnesium (Mg) distribution.
The next parameter under consideration was the distribution of chloride (Cl¯). Due to
its significant mobility, Cl¯ was meant to point to possible pollution sources. Yet, only
few zones of high concentration could be identified.
The next step was to seek out possible pollutants, that is heavy metals and other trace
elements. All heavy metals that were detected in the study area showed distributions quite
different from each other and were possibly related to mine waste sites. Zinc (Zn2+)
however was not connected to mine dumps only; very strong concentrations of reduced
Zn were much wider spread.
The spatial distribution of the different pollutants revealed that they formed zones.
Thus, once the areas with oxic (aerobic) and anoxic (anaerobic) groundwater
environments and the spatial distribution of pollutants were identified, the study area
could be divided into different pollution zones (Fig. 1):
Effects of mining and urban expansion on groundwater quality in Francistown, Botswana 173
Zone III—Typical mining waste point. Elevated heavy metal and trace element
concentrations, very high arsenic concentration, but also high nitrate and sulphate levels.
Ferrous iron and manganous manganese strongly decreased.
Zone IV—Reduced zone with high Zn2+ and Cr concentrations; Fe2+ and Mn2+
enhanced. Some trace elements present.
Zone IVa—Disturbance within Zone IV with high oxygen concentration and some
elevated heavy metals and trace elements; Cl¯ also highly concentrated.
Zone V—Oxidised, very high sulphate and chloride concentrations; no nitrate; trace
elements in significant concentrations; some heavy metals present; controlled landfill.
Zone VI—Reduced, but no Fe2+ and Mn2+; Cl¯ highly concentrated; Mg2+ and Ca2+
enhanced as well as CO2 and ; some trace elements present; old and abandoned
landfill.
Zone VII—Oxidising conditions; no organic pollution but very diverse trace elements
present (some of them concentrated); no potential source of pollution could be identified
from the groundwater hazards map.
6 DISCUSSION
The results of this study revealed that groundwater in Francistown had become polluted
through three major sources, namely pit latrines (Zones I and II), mine tailings dumps
(Zones Ia, III, IV, IVa, and VII), and waste disposal sites (Zones V and VI).
bends ephemeral river flows are slowed down and the observed depth to the groundwater
was shallower here than elsewhere in Francistown. In the crest of the second river bend
there were big alluvial
6.2 Mine waste dumps—Zones Ia, III, IV, IVa, and VII
Several groundwater zones were indicative of pollution due to historic gold mining
activities. The strongest evidence came from the wider surroundings of the Lady Mary
mine, which is located in the SE corner of the study area (zone III). Two boreholes
located close to this abandoned mine site (strongly) violated international and Botswana
drinking water arsenic standards, which allow for a maximum of 10 ppb (µg L−1). Yet,
the groundwater in the two boreholes featured levels of 26 and 244µg (ppb) As L−1
Water resources of arid areas 176
source could be detected in these two zones it is suggested that remnants of old mine
deposits may still exist in these two areas.
A small mine dump within zone I (cf. Lehmann, 2001) caused raised concentrations of
zircon, tantalum, hafnium, cerium, niobium, bismuth scandium and titanium and it was
therefore separated out as mine waste zone Ia. Increased concentrations of Cl¯ and
clearly pointed to anthropogenic pollution. The oxygen and nitrate concentrations
at this site were strongly reduced but nitrite was increased. This indicated a change in
bacterial populations from nitrifying to denitrifying bacteria.
Given the obvious similarities in groundwater pollution between the above sites, they
were put in the same pollution risk group. They may be even more hazardous than zone
III since they are situated upstream from the built-up areas. Clearly, more investigations
need to be carried out and immediate attention must be given.
similar pictures emanated from the spatial concentrations of rubidium, thallium, silver,
uranium, molybdenum, lanthanum, zircon, titanium, sodium, bromide, and boron.
The observed slightly enhanced concentrations of CO2 and may have caused
the dissolution of Ca2+ and Mg2+ out of the carbonates. Probably as a result of this, the
concentrations of these two cations were slightly raised (cf. Christensen et al., 2001).
This could have influenced the buffering system of the sediments.
Both waste disposal sites also featured high Cl¯ concentrations, though the
concentration was much wider at the new (zone V) as compared to the old landfill site
(zone VI). This supported the assumption that there was no new input of pollution at the
old landfill site. So far, the new landfill is only used to deposit inorganic waste. Once it
will be used for other kinds of waste, different processes may set in.
Considering the semi-arid environment in Francistown it may be assumed that
pollution at both landfill sites is localized, will not move readily from place to place, and
is probably confined to the soil only. From this it would follow that the two landfills had
no significant adverse effect on groundwater quality in the study area. On the other hand,
natural remediation in the form of transporting pollutants to other places or through
bacterial degradation is also not likely to take place. Thus pollution would probably stay
as a hazard for a long time.
7 CONCLUSIONS
hazards emanating from the tailings may have gone unnoticed. Clearly, further
investigations are required.
Amongst the three major pollutants, landfills had the least impact on groundwater
quality. They are sited away from the main aquifer and within rock formations that yield
little groundwater. Because of the limited rainfall in the study area, pollutants within
these zones are likely to stay contained within the area. Only occasionally will they be
flushed out during the rainy season and become diluted.
8 RECOMMENDATIONS
Groundwater from a substantial number of boreholes was found to be not suitable for
human consumption. It is therefore necessary to determine which boreholes are used for
humans so as to discontinue their use. As a rule, the Francistown city council ought to
adopt a development strategy that places more emphasis on an environmental approach to
planning taking into account the existing water resources. For example, all new
infrastructures should be placed as far away as possible from the rivers because the
aquifers in the area are dependent on rainfall and river recharge. Activities such as the
recent aligning of the sewage pipelines along the riverbanks must in future be avoided by
all means. Such activities not only destroy a natural flood barrier but they may in fact
lead to serious water pollution.
Similarly, any new development must not include pit latrines. Since a sewage
reticulation system has been put in place throughout the city, it is necessary to educate the
residents on the need to connect to the sewerage and put an end to the use of pit latrines.
So far, connection to the sewerage system is on a voluntary basis and pit latrines (and
septic tanks) are currently still the main means of wastewater discharge in the newly
connected areas.
The study also confirmed that environmental and health hazards emanating from
abandoned mine tailings must be dealt with in a way that guarantees human safety and
environmental protection. The reported chemical “cocktail” conditions of tailings and the
observed trace element concentrations in some boreholes make this obvious.
The waste disposal (landfill) sites appeared to have been well sited in areas of low
groundwater yields. But continuous monitoring is necessary in order to determine the
dynamics of possible plume development so as to act upon possible groundwater
pollution. Further investigations are also necessary to determine the source of heavy
metals and other pollutants at the new landfill site.
REFERENCES
Colquhoun, B., O’Donnel, H. & Partners 1974. Redevelopment of the Francistown groundwater
studies report. Phases I, II and III. Australian Groundwater Consultants.
BOS 2000. Water quality—Drinking water—Specification. BOS 32, Botswana Bureau of Standards,
Gaborone, Botswana.
Christensen, T.H., Kjelsden, P., Bjerk, P.L., Jensen, D.L., Christensen, J.B., Baun, A. Albrechtsen,
H.J. & Heron, G. 2001. Biogeochemistry of landfill leachate plumes. Applied Geochemistry:
659–718.
Water resources of arid areas 180
Gibb, A. Sir & Partners 1987. Francistown Water Development. Pre-Investment Study. Appendices
B1 and B2. Water Resources. Water Utilities Corporation, Botswana.
Key, R. 1976. The geology of the area around Francistown and Phikwe, Northeast and Central
Districts, Botswana. District Memoir 3, 121p. plus maps, Dept. Geological Survey (DGS),
Lobatse, Botswana.
Lehmann, A. 2001. Conceptual map of the urban soils of Francistown. Draft map and explanations
with special reference to town planning and environmental quality. Report by the
Environmental Geology Division, Dept. of Geological Survey (DGS), 48p, Lobatse, Botswana.
Smedley, P.K. & Kinniburgh, D.G. (2001). Source and behaviour of arsenic in natural waters. In:
United Nations Synthesis Report on Arsenic in Drinking Water.
Vogel, H. & Kasper, B. 2002. Mine soils on abandoned gold mine tailings in Francistown. Report
by the Environmental Geology Division, Dept. of Geological Survey (DGS), 43p., Lobatse,
Botswana.
WHO (1998). Guidelines for drinking water quality. World Health Organization, 2nd ed., Volumes
1 and 2, Geneva, Switzerland.
In situ remediation potential for Southern
African groundwater resources
Sumaya Clarke, Gideon Tredoux & Pannie Engelbrecht
Water Programme, Environmentek, CSIR, Stellenbosh
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
Research into groundwater remediation methods has been intensified and various
techniques have been tested and applied. “Pump and treat” technologies have been
widely used in the USA. The success of this method has been questioned, considering its
excessive costs, (Simon et al., 2001). As an alternative, in situ technologies are being
developed and applied for removing contaminants in the aquifer. Literature references to
more than 100 successfully operating sites confirm that permeable reactive barriers can
remove a variety of contaminants including halogenated organic compounds, metals,
nitrates, acid mine drainage, phosphorous, chromium and gasoline/petrol derivatives
using various reactive materials in the barrier.
The currently operating sites range from household scale permeable reactive barriers
to industrial sites, to mining and wastewater treatment plants and municipal well fields
(Robertson and Cherry, 2003). Slowly degradable carbon sources are placed in barriers
perpendicular to the flow and such treatment occurs with a high success rate. The
Nitredox® plant in Vienna, Austria to treat nitrate, iron and manganese; has been
operated successfully for more than a decade. Various low cost, robust treatment
techniques like permeable reactive barriers and biological denitrification have proven to
be successful in Canada, New Zealand, Austria, France, and the USA. Cost implication of
implementing any in situ technology is important and need to be taken
into consideration. A cost estimation, performed for a town in the Northern Cape
Province, South Africa demonstrates the cost difference between in situ and ex situ
treatment methods.
Internationally a nitrate concentration of 10mg/L as N (45mg/L as NO3) is accepted as
guidelines for health risk. The maximum allowable level is set as 20mg/L as N (90mg/L
as NO3) in South Africa. Nitrate concentrations in groundwater are alarmingly high in
some parts of Southern Africa as shown in Figure 1. The northern provinces of South
Africa all have many groundwater sources with nitrate concentrations ranging from 251–
500mg/L. In some areas concentrations up to 1000mg/L occur, while the Southern
Kalahari has concentrations of up to 2000mg/L, particularly in the more saline areas,
(Marais, 1999, Tredoux et al., 2000).
In view of the prevalence of nitrate in groundwater, this paper focuses on in situ
denitrification as a viable treatment option for town and rural applications. It is crucial
that groundwater pollution be taken seriously, and that remediation and protection of the
groundwater resources available be considered as a priority in countries affected by
pollution. Surface water resources are limited, more particularly in arid and semi arid
regions, and with groundwater being unfit for use by inhabitants of these regions, a
serious threat is posed to the survival and growth of communities affected.
Permeable reactive barriers (PRB) are constructed across the flow path of the migrating
plume of contaminated groundwater. These systems are typically designed as a
continuous trench, filled
Water resources of arid areas 184
4 BIOLOGICAL DENITRIFICATION
used to suit site-specific requirements. The most successful configurations are those used
in Vienna and Nebraska. A list of operational systems is presented in Table 1.
The daisy configuration is shown in Figure 3. Most of the biological denitrification
systems use variations of this basic configuration. The area in the sketch labelled (IV)
represents nitrate polluted water. The “daisy” represents the area (in plan view)
progressively affected by the denitrification due to substrate (carbon source) injection.
The reaction takes place in zone I, followed by filtration of any by-products in zone II
while the nitrate free water (or water with a lowered nitrate concentration) is found in
zone III.
The Nitredox® system consists of one pumping borehole located at the centre of two
concentric circles of injection boreholes. It involves injection of an organic substrate
(outer ring) to enhance denitrification, but includes an additional phase of aerated water
injection for the oxidation and removal of iron (inner ring) once the nitrogen is removed.
The groundwater recovered from the
Table 1. Some pilot and field operational
denitrification sites and their experiences.
Method and Period NO3- Aquifer Carbon Injection/ Nitrate
location Nmg/L substrate barrier removed
PRB, Canada 5 yrs+ 5–57 Primary Sawdust/ Emplaced 58–91%
(1) woodchips barrier
PRB, New 5 yrs+ 5–15 Unconfined, Sawdust Emplaced 95+%
Zealand (2) sandy barrier
Electrokinetics/ Test Controlled Primary/ None: Abiotic Emplacement 84–87%
Fe-wall, period amounts secondary of wall and
USA(3) electrodes
NitrEI system, Many Up to Primary/ None: Electrodes Reduced
Canada (4) Currently 1000 unsaturated Electrochemical down to
operating zone electrodes 0.1mgN/L
sites
Daisy wheel, 40 Sand and Ethanol C and P 35%-c-
ISBD, gravel injection injection;
Nebraska (5) 90–100%
p-
injection
Nitredox, 15 years 14 Primary Ethanol P injection 75%
ISBD, Vienna aquifer
(6)
ISBD, line of 226–565 Chalk Ethanol 80%
injection
boreholes,
France (7)
C-Continuous injection, P-pulse injection. (1) Robertson and Cherry (1995, 2000), Blowes
et al., (1999), (2) Schipper & Vojvodic-Vukovic, 2001, (3) Chew and Zhang, 1998 and
Loo, 2000, (4), (5) Khan & Spalding (1998), (6) Braester and Martinell (1988), Jechlinger
et al., (1991), (7) Chevron et al., (1998).
Water resources of arid areas 186
central production borehole is partly free of nitrate but completely free of iron,
manganese or other by-products (Braester and Martinell, 1988). This method has been
applied to coastal aquifers and primary aquifers. Where biological denitrification is
implemented, it is important to know and monitor the permeability and porosity.
Clogging may result when the carbon substrate injection exceeds the amount required for
denitrification. The method has been applied mainly to primary aquifers where flow
dynamics are well understood.
In situ remediation potential for Southern African groundwater resources 187
5 OPERATIONAL SITES
6 OPERATIONAL EXPERIENCES
The PRB systems in Canada have been operational for more than 5 years and are used on
various scales including household, municipal, and huge water treatment plants.
In New Zealand, the reactive barrier had to be replaced after 5 years of operation as
reactive material had clogged parts of the aquifer.
In the USA, the electrokinetic methods worked better when combined with iron walls.
The biological denitrification used in Nebraska used both continuous and pulse
injection regimes. The continuous carbon source injection was more efficient in
denitrification but led to complete biofouling after 10 days. The system used inner
oxidation ring to remove possible nitrite, iron and manganese.
In Vienna, where the Nitredox® method is currently operational, clogging was
experienced. Pulse injection and reduction of the amount of ethanol (carbon source)
prevented clogging of the system.
In France, natural in situ denitrification was carbon limited. Remediation by carbon
source addition was selected to accelerate denitrification. Denitrification was achieved in
long time operation (450 days). Rates of denitrification were improved when trace metals
were supplied in conjunction with the carbon substrate.
The costs of implementing denitrification were estimated for Marydale in the Northern
Cape Province of South Africa. The results are shown in Table 2. In situ application was
compared with a conventional ex situ method. In Marydale, a well field containing 10
boreholes is used as the town water supply. Half of these boreholes, produce groundwater
of nitrate concentration above the maximum allowable 20mg/L, (Hofmann, 1997).
Microbiological sampling showed that coliform counts of 15/100ml in some boreholes
were three times as much as the SABS specification (Hofmann, 1997). No faecal
Water resources of arid areas 188
coliforms were detected. This gives an indication that no human or animal waste reaches
the boreholes.
Exploration boreholes drilled in the area revealed that a primary and a secondary
aquifer are present in the area. The secondary fracture system is not well understood, but
it is believed that the bed rock is not very permeable. The main water bearing unit is the
alluvial cover of more or less 12m thick. The aquifer material consists of sedimentary
layers containing primarily sand and silt. The Projected water demand of the town for
2005 is 142287m3/a (Shand and VSA, 1997).
Capital expenses for permeable reactive barriers (PRB) include excavation costs, wall
emplacement costs and dewatering prior to wall emplacement. Woodchips or sawdust
was considered as suitable permeable reactive barrier material as these are cheap and
slowly degradable carbon sources. The barrier size is based on the size of the well field
and the depth to bedrock. The largest contribution to capital costs for in situ biological
denitrification (ISBD) methods include borehole construction costs, purchasing of
injections pumps among other costs. Erecting infrastructure is a major expense for
conventional treatment plants.
Running costs were based on the projected annual water demand and estimated
chemical costs. The PRB method requires limited maintenance. In the calculation,
operation and running costs are included; however, they may not occur frequently for
methods like PRB. Operational and maintenance costs are relevant especially when
clogging or partial clogging of wells occurs.
Pump and treat methods and other ex situ methods (in this case ion-exchange)
generally cost an order of magnitude more than in situ methods. It is clear from this
information that rural communities for which funding is not always in surplus may
capitalize on this advantage as well as the ease of use of some of these methods. Proper
management and monitoring of sites are essential to detect potential clogging cases early
and to put remedial measures in place.
8 DISCUSSION
Field scale plants have proven in situ technologies to be successful. The nitrate removal
rate at most currently operating sites are high. It is evident from Table 2 that the
permeable reactive barrier method is the most cost effective method. Capital costs are
relatively low and it requires little or no additional treatment of groundwater after passing
through the system. Installation and running costs of ex situ treatment exceed that of in
situ methods.
Operational sites in the US, Canada and New Zealand showed that barrier material
replacement was only required after 5 years, while ex situ methods have set running
expenses per cubic meter of water. The largest full-scale in situ denitrification plant uses
the Nitredox® principle. This plant is located at Bisamberg, Vienna (Austria) and has
been operating successfully for more than a decade (Jechlinger et al., 1991). It uses
ethanol as the carbon substrate and the process is regulated to ensure that the raw water
nitrate, which exceeds 65mg/L, is reduced to approximately 35mg/L in the product water.
There are advantages as well as some disadvantages of in situ treatment technologies.
Some advantages of implementing such a treatment system include minimal exposure to
In situ remediation potential for Southern African groundwater resources 189
dangerous chemicals, job creation in rural villages, little or no treatment required at the
surface, possible treatment of other contaminants due to redox changes, costs savings in
comparison to conventional ex situ treatment plants, low maintenance costs, simple to use
technologies, and no need for electricity (PRB- method).
Disadvantages include possibilities of clogging of boreholes. This occurs when the
carbon dosage is in excess of the required amount. Sulphate reduction may occur when
carbon substrate dosage is too high and result in acetate production as a by-product of
microbial activity (Israel, 2004, unpublished data). Loss of hydraulic permeability of the
aquifer is possible if carbon addition is not effectively managed. Preferential flow of
groundwater can occur, where a great contrast develops between the treatment zone and
the rest of the aquifer and the path of least resistance is taken by the groundwater. Hence
monitoring of the above mentioned parameters is very important. Cautionary measures
include proper estimation of the required amount of carbon substrate, and monitoring the
effective porosity and permeability of the aquifer before, after and during treatment.
Management of implementation and monitoring are essential for success. It is
important to note that no microbes are added to initiate the process, as this would affect
the ecosystems that are already established at any specific site. There are many strains of
bacteria that occur naturally under the various environmental conditions, which are
capable of denitrification. Although some scientists may prefer to add appropriate
bacteria to initiate the process but the addition of a carbon source is sufficient to activate
resident bacteria.
9 CONCLUSION
In situ groundwater treatment methods are widely used and accepted in the US, Canada,
Europe and New Zealand. Literature shows that various in situ methods for a range of
heavy metals, organic compounds and other constituents have been successfully
implemented at field scale in these countries. In situ denitrification methods are also
viable treatment methods which are successfully implemented and currently operating.
The cost analysis performed in South Africa, showed that there is an order of magnitude
difference between the costs of ex situ and in situ treatment plants. Optimal conditions for
most in situ treatment methods include the following:
● Primary aquifer systems, or well understood secondary aquifers (with respect to flow
characteristics and porosity/permeability).
● Aquifer material can include sand, gravel, and chalk material.
● A known concentration of nitrate-nitrogen is important for estimation of the
appropriate quantity of carbon substrate.
● A maximum aquifer thickness of 20m for injection type methods (e.g. ISBD) and 10m
for emplacement methods (e.g. PRB).
● Monitoring of aquifer parameters (permeability, hydraulic conductivity etc.), chemical
changes (pH, Eh, etc.) and microbiological changes with time.
It is important to note that no foreign microbes are added, as this would affect the
ecosystems that are already established at any specific site. Although some scientists may
Water resources of arid areas 190
REFERENCES
Blowes, D.W., Ptacek, C.J., Benner, S.G., McRae, C.W.T., Bennett, T.A. & Puls, R.W. 2000.
Treatment of inorganic contaminants using permeable reactive barriers. Jnl. of Contaminant
Hydrology, 45:123–137.
Braester, C. & Martinell, R. 1988. The Vyredox and Nitredox method in situ treatment of
groundwater, Wat. Sci. Tech., 20(3):149–163.
Cartmell, E., Clark, L., Oakes, D., Smith, S. & Tomkins, J. 1999. Feasibility of In situ
Bioremedition of Nitrate in Aquifer systems, R & D Technical Report P277, WRC report no. EA
4683.
Chevron, F, Lecomte, P., Darmendrail, D. & Charbonnier, P. 1998. Rehabilitation de qualitè
physicochimique d’un aquifere contaminepar des nitrates d’origine industrielle- un example en
region Nord-Pas de Calais. L’Eau, L’Industrie, Les Nuisances, 208(31–35) (In French).
Chew, C.F. & Zhang, T.C. 1998. In situ remediation of nitrate contaminated ground water by
electrokinetics/ iron wall process. Water Science and Technology, 38(7):135–142.
EPA, 1995. In situ remediation technology status report: Treatment walls. Report No. EPA/540/K-
94/004. Office of Solid Waste and Emergency Response, US Environmental Protection Agency.
Israel, S., 2004, Subsurface Manipulation of the Nitrogen Cycle: In-Situ denitrification and its
potential for remediation of contaminated soil and ground water resources: Case Study:
Marydale, Northern Cape, MSc research, unpublished data, University of Stellenbosch.
Jechlinger, G., Schöller, F., Seidelberger, F., & Zibuschka, F. 1991. Denitrification In Situ. In:
Proc. of I.W.S.A workshop: Inorganic nitrogen compounds and water supply. Hamburg, 27–29
Nov:113–122.
Khan, I.A., & Spalding, R.F. 1998. Denitrification using a daisy well system. Presentation to
National Sanitation Foundation International Symposium, Safe Drinking Water in Small
Systems: Technology, Operations, and Econimics. Washington D.C., May 10–13.
Kruithof, J.C., Van Paasen, J.A.M., Hijnen, W.A.M., Dierx, H.A.L. & Van Bennekom, C.A. 1985.
Experiences with nitrate removal in the eastern Netherlands. Proc. Nitrates Dans les Eaux, Paris
22–24 October.
Loo, W.W. 2000. Electrokinetic treatment of hazardous wastes. Standard Encyclopedia of
Environmental Science and Technology, New York, McGraw Hill,: 14.69–14.84.
Mercado, A., Libhaber, M. & Soares, M.I.M. 1988. In situ biological groundwater. denitrification:
Concepts and preliminary field test. Wat. Sci. Tech., 20(3):197–209.
Ninham Shand and VSA Consulting, 1997, Geohidrologiese Ondersoek van die
Groundwaterbronne by Marydale, Noord-Kaap Provinsie, VSA Conculting pty. Ltd.
Robertson, W.D. & Cherry, J.A. 1995. In situ denitrification of septic system nitrate using reactive
porous media barriers: Field trials. Ground Water, 33(1):99–111.
Robertson, W.D., Ford, G. & Lombardo, P. 2003. Wood-Based Filter for Nitrogen Removal in
Septic Systems, (Submitted to: Journal of Environmental Quality), (unpublished).
Schipper, L.A. & Vojvodic-Vukovic, M. 2000. Nitrate removal from groundwater and
denitrification rates in a porous treatment wall amended with sawdust. Ecol Engineering,
14:269–278.
Schipper, L.A. & Vojvodic-Vukovic, M. 2001. Five years of nitrate removal, denitrification and
carbon dynamics in a denitrification wall. Wat. Res. Research, 35(14):3473–3477.
Tredoux, G, Talma, A.S. & Engelbrecht, J.F.P. 2000. The increasing nitrate hazard in groundwater
in the rural areas. Paper presented at WISA 2000, Sun City, RSA, May 2000.
Coastal aquifers intrusion at semi-arid region
of Turkey
L.Yilmaz
Technical University of Istanbul, Civil Engineering, Hydraulic Division,
Maslak, Istanbul, Turkey
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
1.1 Relationship between the level of the water table and the depth to the
saline wedge
A relation between the level of the water table and the depth to the saline wedge in an
unconfined aquifer under steady conditions of flow (Badon Ghijben, 1889; Herzberg,
1901, Davis, 1978) points out that Joseph DuCommun (1828) made similar observations.
Prior to the work of these pioneers it was thought that saline water occurred at a depth
close to sea level. The saline water close to the sea shore is defined by Badon Ghijben-
Herzberg equation, which was derived by a simple application of hydrostatics. The
weight of a column of fresh water of height hf+z is equal to the weight of a column of
saline water of height z. If rof and ros are the densities of fresh and saline water
respectively, it is given in equilibrium conditions
rosgz=rofg(hf+z)
(1)
or
(2)
If the relative densities of fresh and saline water are taken as 1.0 and 1.025 respectively,
then
Water resources of arid areas 192
z=40 hf
(3)
This expression is a good approximation in steady state conditions when the zone of
dispersion is only a small fraction of the saturated thickness of the aquifer. Since fresh
water is flowing along the interface some mixing will occur due principally to
microscopic and macroscopic dispersion. When the saline and fresh water mix in the
zone of dispersion then the diluted saline water becomes less dense and will rise along a
seaward path. The resulting mechanism is similar to thermal convection, the only
difference being that the gradients are caused by changes in density due to changes in
salinity instead of temperature (Cooper, 1964). This flow will advect some saline water
towards the sea. Therefore, in order to preserve the saline mass balance, a small flow of
saline water must occur in the landward direction. This flow creates a head loss, thus a
reduction in pressure at the interface and a reduction in the level of the interface. This
application gives the position and movement of a saline front in a coastal aquifer.
(4)
2 MAIN OBJECTIVE
If the relative densities of fresh and saline water are taken as 1.0 and 1.025 respectively,
then
z=40 hf
(6)
This simple expression gives a remarkably good first approximation to the depth below
sea level of the interface under steady state conditions when the zone of dispersion is
only a small fraction of the saturated thickness of the aquifer.
Water resources of arid areas 194
This concept can be further developed to determine the extent of the penetration of the
saline wedge inland. Many analyses can readily be developed, for example the
determination of the shape of the interface when the seepage surface is submerged
beneath the sea (Glover, 1964) and the shape of the saline upcone beneath a pumping
well in a coastal aquifer (Schmorak and Mercado, 1969; Sahni, 1972).
grid. The finite difference method is, perhaps, the most frequently used technique for
solving the flow equation. However, it is not often used to solve the advection-dispersion
equation because of a phenomenon known as numerical dispersion. The numerical
solution usually appears to advance the solute at a rate which is greater than is physically
possible. Finite difference schemes can be developed to minimize dispersion. However
they are liable to cause either overshooting or undershooting which appear in the solution
as oscillations. Van Genuchten (1976) analyzed and gave as a result that the finite
element schemes will usually yield more accurate solutions than finite difference ones.
There are some rules which can be helpful in minimizing the effects of dispersion. These
use a form of the Peclet number, Pe, and the Courant number, C. The grid should be
designed such that Pe(=dx/De)<4, where dx is some characteristic grid size and De is
some characteristic dispersivity.
The finite element method was first developed in the solid mechanics (1950). Then it
was used to solve the groundwater flow equation. The finite element method is an
integral (as opposed to differential) approach in which the regular grid of the standard
finite difference method is replaced by an irregular polygonal mesh which allows the
modeller to describe natural shapes more precisely. In groundwater the polygonal shape
is, almost always triangular. The finite element mesh can be adapted to describe the
irregular shape of the boundary and obtained an accurate description of rapidly varying
phenomena. In this approach the piezometric surface is approximated by a series of small
triangular surfaces which can be flat or curved. If the chosen basic functions are linear
then the surfaces will be flat and the variation of head within each element will be linear.
The point of intersection of the triangles is called a node and each triangle is called an
element. The equation is solved by a weighted residual technique of which the most
popular one is the Galerkin method. In this method the weighting functions are made
equal to the basic functions and the integration is then performed over each element and
summed to yield the contribution from all the elements that make up the solution domain.
The finite element method is a powerful and mathematically elegant technique but it is
difficult to program.
The method of characteristics was developed to solve hyperbolic partial differential
equations (Courant and Friedrichs, 1948) and was first used for flow through porous
media by Gardner et al. (1964). They proposed the method because they argued that
when flow velocities become large the dispersion equation is, essentially, hyperbolic. The
method has been extensively applied for solving the advection-dispersion equation and is
now the basis of one of the standard solute transport models (Konikow and Bredehoeft,
1978). The solutions are x=x(t), y=y(t) and C=C(t), where x and y are the coordinates in a
Cartesian system, C is the pollutant concentration and t is time. These are called the
characteristic curves of, in this case, the advection-dispersion equation. Once these
solutions are available then a solution of the advection-dispersion equation can be
obtained by following the characteristic curves. Gardner et al. (1964) state that “Each
point corresponds to one characteristic curve and values of x, y and C are obtained as
functions of t for each characteristic”. Essentially this is the Lagrangian approach of
classical hydrodynamics. It is particularly useful for making cross-sectional models of
saline intrusion.
There are many other methods for solving the groundwater flow and advection-
dispersion equations, as integrated finite differences (Tyson and Weber, 1963; Goodwill,
Water resources of arid areas 196
1980), boundary element methods (Liggett and Liu, 1983) and analytic elements (Strack,
1989).
3 RESULTS
The model requires substantial amounts of field data, the collection of which is both time
consuming and expensive. The hydrological and geological data are used for the area to
be modeled. These data are:
● Surface and subsurface geology.
● Piezometric levels for all the aquifers contained in the system.
● Aquifer characteristics and likely boundaries, soils, land use and vegetations.
Since aquifers are subject to recharge and pumping, data on the quantities and timing of
these will be required. Such data will include precipitation, evapotranspiration and
pumping. If irrigation is undertaken, then also rates of application and return flows and
river flows are required, including flows to and from the rivers to the aquifers if they are
not in direct hydraulic contact. Since the concern here is with saline intrusion then data
on salinity, both areally and vertically, will be required. If any of these data do not exist
or are too scanty, then a field programme will be required to collect them. In order to
collect and plot all these data an accurate topological map is essential, the scale of which
will depend on the size of the aquifer and the scale of the problem being studied. This
map should show the surface contours, surface water bodies, streams and man-made
watercourses such as irrigation canals and drainage ditches.
REFERENCES
Bear, J. 1972. Hydraulics of Groundwater. New York, McGraw-Hill Book Co., 567 p.
Bras, R.L. & Rodriguez-Iturbe, I. 1976. Evaluation of mean square error involved in approximating
the areal average of a rainfall event by a discrete summation, Water Resources Research, 12(2),
181–184, a.
Hubbert, M.K. 1940. The Theory of Ground-Water Motion, The Journal of Geology, 48(8), Part-I,
Nov.–Dec.
Pinder, G.F. 1982. Finite Element Simulation in Surface and Subsurface Hydrology, Gallagher,
Vol. 4.
Pinder, G.F. & Abriola, L.M. 1982. Calculation of Velocity in three space dimensions from
hydraulic head measurements, Groundwater, 20, 205–213.
Pinder, G.F. & Gray, W. 1982. Finite Elements in Water Resources, edited by P.Holz, V.Meissner
& C.A.Brebbia (eds), Berlin, Springer Verlag: 4.
Rushton, K.R. & Redshaw, S.C. 1979. Seepage and Groundwater Flow, Wiley, Winchester, UK,
339 pp.
Sarma, S.V.K. & Silva, T.C. 1987. Hydraulic response to pumping in free aquifers, ABAS, 11, 26–
32.
Sarma, K.V.S. & Antonio, A.P. 1997. Decontamination of pollutants from aquifers using the
concept of induced flow from adjacent rivers. Intl Conf. on Large Scale Water Resources
Projects, Oct. 20–23. Kathmandu, Nepal, EI 17–24.
Coastal aquifers intrusion at semi-arid region of Turkey 197
Sumer, B. 1980. The Determination of Water Quality at the Sapanca Lake, TUBITAK Project No.
QA6–4, Sakarya.
U.S. Environmental Protection Agency, 1988. Model Assessment for Delineating Wellhead
Protection Areas, Office of Groundwater Protection, Washington DC, 210 pp.
Yilmaz, L., Agiralioglu, N. & Saltabas, L. 1999. The determination of the water-use capacity of the
Sapanca Lake in Turkey. Proc. Intl. Conf. on Water, Environment, Ecology, Socio-economics
and Health Engineering (WEESHE), Oct. 18–21, Seoul National University, Seoul, Korea,
Water Resources Pubs., LLC, 162–166.
Evaluation of groundwater recharge rates in
the Kizinga catchment in Dar es Salaam
region
Y.B.Mkwizu
Lawyers’ Environmental Action Team (LEAT), Dar es Salaam, Tanzania
H.H.Nkotagu
Department of Geology, University of Dar es Salaam, Dar es Salaam,
Tanzania
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
The high demand for freshwater in the Dar es Salaam City, suggests clearly that, surface
water can no longer meet the projected total demand. The second best alternative remains
on groundwater. Even with the presence of surface water, groundwater can still be
preferred on the basis of easier protection from pollution, better dependability during
drought periods, and on the supplying costs. It is necessary however that groundwater
resource is used with proper management focusing on both its quality and quantity.
Evaluation of groundwater recharge rates in the Kizinga catchment in Dar es Salaam region 199
Evaluation of groundwater recharge rates and areas is an important step towards thorough
understanding of its quality and quantity.
1.2 Objectives
The purpose of the study was to evaluate groundwater recharge rates in the area.
Specifically, the research aimed at:
Determining the mean annual To determine the source of
groundwater recharge rates in the study groundwater recharge in the study
area. area.
Water resources of arid areas 200
2 METHODOLOGY
Data that incorporate various hydrogeological units and taking into account all flow
components such as discharge, infiltration, subsurface inflow to and outflow from the
basin’s aquifer and abstractions through pumped wells were collected. The main data/and
data sources were:
– Boreholes and wells and their hydrologic information drilled in the study area up to
1999 from the Ministry of Water, borehole drilling unit-Ubungo
– Monthly rainfall data, maximum and minimum temperature and evaporation for Dar es
Salaam from Tanzania Meteorological Agency, Dar es Salaam office
– Runoff measurements from rivers Kizinga from 1967 to 1980, From Ministry of Water,
RBM-Ubungo
Table 1. Chloride concentration of rainwater in the
study area in (mg/l).
Location Mar Apr May Mean
Ukonga 3.1 2.8 2.6 2.83
Kiwalani 3.0 1.9 3.9 2.93
Yombo 3.2 2.4 2.8 2.80
Temeke 3.7 2.6 3.4 3.23
Mbagala 3.6 2.8 3.0 3.13
Pugu 3.2 1.8 2.4 2.47
Gongolamboto 3.0 2.8 3.2 3.00
Total 22.8 17.1 21.0 20.39
Mean 3.26 2.44 3.04 2.91
Evaluation of groundwater recharge rates in the Kizinga catchment in Dar es Salaam region 201
The fieldwork involved rainwater sampling in different sites within the study area for the
rainfall period of March, April and May in the year 2000.
Another fieldwork activity was to collect water from boreholes in some selected
locations in the study area.
Chloride determination from water samples was undertaken following standard
method as reported by (APHA 1985). The results obtained from rainwater samples are
given in Table 1.
3 RESULTS
The comparison between hydraulic head and topographical height shows that the
direction of groundwater flow is roughly to the drainage pattern which follows the
gradient of the land surface. Groundwater movement in the study area is therefore
approximately to the north east which is generally the direction of Kizinga river (Figs 1
& 2). The above observations on the groundwater movement in the study area suggest
that groundwater starts flowing from Pugu and Kisarawe hills towards low lying plains of
Yombo; Changombe; Temeke; Mbagala; Mtoni and Kurasini. This shows that
precipitation within Pugu and Kisarawe hills is the major source of groundwater recharge
in the study area.
Evaluation of groundwater recharge has significant implications for not only the study of
groundwater quantity, but also water quality. Infiltrating water can carry contaminants
from
Evaluation of groundwater recharge rates in the Kizinga catchment in Dar es Salaam region 203
The annual actual evaporation in the study area always exceeds annual rainfall.
However a plot of mean monthly actual evaporation and rainfall for the one year period
(Fig. 5), shows that there are few months when rainfall exceeds actual evaporation. These
months are mainly, March; April; and November and it is expected that during these
months, groundwater recharge takes place.
Where: Q is the quantity of water (m3/s), K is the hydraulic conductivity (m/s), I is the
hydraulic gradient and, A is the area (m2).
The value obtained was 6036076.153×109mm/year and when extrapolated to the entire
basin’area of 191km2, this amount of annual groundwater in flow in mm per year was
calculated such that,
Recharge rate=(6036076.153×109mm/year)/191×1012mm=31.6mm/year.
Evaluation of groundwater recharge rates in the Kizinga catchment in Dar es Salaam region 207
The major contribution in recharging the study area has been found to be the faults on the
slopes of Pugu and Kisarawe hills, which are quite permeable and that the direct
infiltration of rainwater is the main source of groundwater recharge in the study area.
The average value of annual groundwater recharge rates has been found to be
52.8mm/year after combining all the three methods used. This value is approximately
10.1×106m3/year and it is 4.7% of the long term mean annual precipitation of 1124mm. A
projection of maximum groundwater abstraction showed an increase of more than 100%
after every two years. However the natural groundwater recharge is expected to remain
constant, and therefore the maximum production rate will be more than natural recharge
rate within a period of few years to come. This relation when projected over a long period
may result in negative consequences such as depletion of groundwater supply, reduced
Evaluation of groundwater recharge rates in the Kizinga catchment in Dar es Salaam region 209
stream flow, deterioration of water quality and more importantly land subsidence may be
expected in the near future.
On the other hand there is no comprehensive legal mechanism to ensure proper
management of groundwater resources. The Water Utilization Act has not provided a
holistic approach to the management of water resources especially with regard to the
management of groundwater resources. There is no specific provision for groundwater
abstraction under the Act and instead groundwater exploitation has been provided for its
management in the same way as surface water. There is no separate provision under the
Act to regulate groundwater-drilling operations. However, abstraction of groundwater of
more than 22,700 litres per day requires a water right issued under the Act. The Act has
not demanded for operators of wells or boreholes to submit data and records to the water
authorities. It is therefore possible that, commercial drilling of groundwater of substantial
scales has been conducted without adequate monitoring and controls.
Clear understanding on the aquifer parameters and recharge rates for other parts of Dar
es Salaam and the country at large is fundamental before embarking on further
exploitation of the resources. Studies on the subject however are very limited to academic
purposes and have not been able to comprehensively provide a clear understanding on the
aquifers that provide the resource. In general, the studies that have been conducted so far
indicate a negative trend in the status of groundwater in different parts of the country.
The absence of adequate data and legislation has been impinging on effective
management of groundwater resources.
The Kazimzumbwi forest which is in the Pugu and Kisarawe hills is suffering a
massive deforestation from illegal harvesting of forest products. The capacity of the
faults to serve as groundwater recharge will consequently be affected. It is possible
therefore that, in the near future the rate of natural recharge will decrease while that of
abstraction will keep on increasing. It is recommended that, the faults on the slopes of
Pugu and Kisarawe hills, which serve in recharging the area be conserved and protected
from disturbance to ensure continuation of safe and enough groundwater supply. Possible
alternatives for freshwater supply should be used. Measures that reduce water wastage
must be introduced and encouraged. More studies on groundwater recharge and aquifer
performance need to be conducted in the study area and others around Dar es Salaam
using other methods so as to have a complete understanding on the aquifer system in Dar
es Salaam city.
ACKNOWLEDGEMENTS
We would like to thank all those people who contributed to the preparation and
completion of this work. Our special thanks go to the Department of Geology of the
University of Dar es Salaam for the facilities they have provided to us during our
research. Many thanks go to MHO program through the Faculty of Science University of
Dar es Salaam for the financial support in undertaking the research. We wish to
acknowledge the assistance of the Staff of the institutions in that time; Ministry of Water,
Ubungo Maji; DAWASA headquarters and Tanzania Meteorological Agency Dar es
Salaam office for supporting us with necessary data. We are grateful to the Management
Water resources of arid areas 210
of Lawyers’ Environmental Action Team (LEAT), for their support on access to the
organizational facilities and information during the preparation of this paper.
REFERENCES
APHA, 1985. Standard Method for the Examination of Water and Wastewater. 16th Edition.
American public Health Association, Washington, D.C.
Matondo, J.I. 1978. A kinematic conceptual model for Kizinga basin for Estimation of Hydrological
variables. M.Sc. Thesis, University of Dar es Salaam.
Ministry of Energy and Minerals (MEM) 1979. Coast/Dar es Salaam regions water master plan.
Dar es Salaam.
Nkotagu H.H. 1996. Hydrological and Isotopic characterization of a fractured basement
groundwater flow system in Semiarid Area of Dodoma, Tanzania. Znge: Berlin, Techn. Univ.
Diss.
Serviceplan, 1997. Report on the evaluation of groundwater sources of Dar es Salaam. Supporting
Report B. Dar es Salaam.
Theme C:
Socio-economic aspects
KNUST experiences in capacity building in
the water and sanitation sector
S.N.Odai, F.O.K.Anyemedu, S.Oduro-Kwarteng & K.B.Nyarko
Department of Civil Engineering, KNUST, Kumasi, Ghana
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
The human resources requirements of the water and sanitation sector are largely similar
to those of other professions. Factors militating against capacity building in the sector
have been mainly due to lack of funds and lack of understanding of the urgency of the
need to improve the sector. In many developing countries of Africa, improvement in the
water supply and environmental sanitation sector has lately been recognised to have a
direct positive impact on public health (Ghana Government, 2003). The development of
the sector is seen as crucial to the successful control and eradication of communicable
diseases in general. As a result, governments, policy makers, non-governmental agencies,
Water resources of arid areas 214
and external support agencies have begun directing attention to issues related to these
sector in an effort to improve the overall health of citizens by facilitating easy
accessibility to water and sanitation.
Until the last decade, the progress and growth in the water supply and environmental
sanitation sector has been little and very slow. One of the major factors hampering the
desired rate of progress and growth in the sector has been the lack of adequate personnel
and professionals with the requisite skills, expertise and experience to lead and manage
the sector. In fact, during the 1991 UNDP symposium at IHE-Delft, on A Strategy for
Water Sector Capacity, it was acknowledged that capacity building in the water supply
and environmental sanitation sector was essential for the development, growth and
sustenance of the sector at the local, national and even sub-regional levels (KNUST,
2002a).
In the wake of this symposium, it was felt that the training and re-training of
professionals for and within the water supply and environmental sanitation industry must
become part of the central focus of academic institutions that have long traditions of
providing quality leadership training as well as professional expertise. The UNESCO-
IHE Institute for Water Education in Delft and the Department of Civil Engineering of
the Kwame Nkrumah University of Science and Technology (KNUST) in Kumasi
quickly took the initiative to develop a programme for supporting the restructuring and
strengthening of the water supply and sanitation sector in Ghana and the sub-region. The
programme was preceded with a needs assessment of the situation in several
neighbouring countries, to assess the percentage coverage of sanitation and water supply
in urban and rural communities, and the statistics were stunning.
Following this needs assessment, the DCE of KNUST in collaboration with
UNESCO-IHE in 1996 initiated a programme for developing human resource capacity in
the water supply and environmental sanitation for a wide range of beneficiaries, spanning
from sector professionals to low-level operators and sometime even uneducated water
board members. The high-quality training programme in the water supply and
environmental sanitation sector was conceived under the project name “Water and
Environmental Sector Capacity Building and Sustainable Development in Ghana and the
Region” designed by the KNUST-Kumasi and UNESCO-IHE. The project aims at
providing the sector within the West African sub-region with the necessary skills,
knowledge and expertise to meet the demands, challenges and opportunities anticipated
with the projected growth in the region.
This paper presents the experiences of KNUST in capacity building in Ghana and the
sub-region.
The Kwame Nkrumah University of Science and Technology (KNUST) was established
in 1951 to train Scientists and Technologists for both Ghana and other African countries.
Academic programmes are run at the undergraduate and post-graduate levels by different
faculties, schools and institutes. Training in water-related disciplines is offered in several
departments in the university.
KNUST experiences in capacity building in the water and sanitation sector 215
Presently, the DCE, the Chemistry Department, and the Geological Engineering
Department collaborate in delivery of education and training in water quantity and
quality. The programmes of the Department of Civil Engineering cover both surface and
groundwater resources and water quality analyses, while the Chemistry Department is
involved mainly in water quality analyses, and the Geological Engineering Department
involved in groundwater resource development. In addition, the Agricultural Engineering
Department looks at irrigation and water for food, the Biological sciences Department
works on environmental science, while the Physics Department masters in groundwater
and limnology.
Since 1996, the DCE and UNESCO-IHE have been developing human resource
capacity in water supply and environmental sanitation, targeted at all levels of sector
professionals (sometimes including persons with low-level education). Since the
establishment of WSESP which has the primary focus of postgraduate training for the
sector, several, short courses such as Public Private Participation, Water Treatment,
Wastewater Treatment, Solid Waste Management, Urban Water Transportation and
Distribution, etc., have been offered annually.
The institutional capacity of KNUST to act as a capacity building centre in the water and
sanitation sector is discussed under the following sub-headings.
3.2 Facilities
The facilities at KNUST for training purposes include refurbished classrooms, a
computer laboratory that gives access to each participant, a refurbished laboratory that
allows for water and wastewater quality analyses for training and research purposes.
Students and lecturers in the DCE, together with other departments in the School of
Engineering, have access to a well-equipped library, apart from the main library of the
University. In addition, a collection of specialised books is at the disposal of the staff and
participants in WSES project. A 30-room hostel has been built purposely for the MSc
participants. These rooms are also available for use by short course participants when
these are organised during the vacation periods. The university now has Internet
connectivity for research, and a website has been created for the project.
Water resources of arid areas 216
The Water Supply and Environmental Sanitation Project is aimed at capacity building for
sustainable development and growth in the water supply and sanitation sector in Ghana
and the West African sub-region. It seeks to strengthen the sector through training of
high-level personnel for institutions and organisations that have a stake in the water
supply and environmental sanitation industry and professionals with active careers in the
sector, in addition to training of low-level personnel.
In Ghana, the programme targets the following sector organisations and their
professionals:
● Ghana Water Company Limited (GWCL),
● Community Water and Sanitation Agency (CWSA),
● Environmental Protection Agency (EPA),
● Ministry of Works and Housing MWH),
● Ministry of Local Governments and Rural Development (MLGRD),
● Ministry of Health (MoH),
● Ministry of Environment and Science (MES),
● Water Research Institute (WRI),
● Consulting firms and contractors,
● Water-related industries,
● Small Towns Water Operators and Managers.
KNUST experiences in capacity building in the water and sanitation sector 217
2003 10 9 1
Total 53 47 6
– Telecommunication
– Structural Building works
Table 3. Statistics of municipal engineering
trainings.
Year Total no. of participants
February 2003 10
August 2003 23
January 2004 15
Total so far 48
Small towns’ water supply operators’ course: This course is organised with the primary
aim of upgrading the technical and managerial capacity of water board members and the
system operators to enable them achieve sustainability. Needs assessment showed that
there are more than 2000 people in the 210 districts in Ghana needing the first round
training. The objectives of the training are (KNUST, 2003)
– To update the knowledge, technical and management skill, and attitude of operating
staff of small town water systems for effective and efficient operation and
maintenance, and management of the water systems.
– To train water and sanitation board members to effectively oversee and manage the
water system and deal with consumer complaints and requests satisfactorily.
The global objective is to attain sustainability, which will subsequently lead to improved
health conditions and productivity.
The attendance statistics is shown in the table below, and it is impressing to note that
the number of participants has increased since we started the tailor-made courses.
The course is organised for duration of one week and the courses offered are
– Water supply system operation and maintenance
– Borehole pumping system operation and maintenance
– Managing system information
– Water and health
– Roles of the water board
– Budgeting and tariff setting.
Water resources of arid areas 220
5 CONCLUSION
In Ghana, statistics show that there is low capacity in the water and sanitation sector. Our
experience in capacity building shows that organizations are aware of their need but
sometimes they need help to articulate such needs. The capacity building programme of
sector professionals and low-level personnel will be strengthened, and sustainability
enhanced. It is anticipated that the 210 districts in Ghana will eventually benefit from our
programmes. The department is gradually shifting to tailor-made programmes since the
short courses do not proof to be financially sustainable and looking for clients who will
patronise the course places extra demands on us. The recent thinking of tailor-made
programmes is catching on since there is usually money available for particular
organizations to build capacity. We therefore develop programmes, which we discuss
with donors and such organizations; upon approval of the courses we then prepare
teaching materials and the cost estimates. This approach seems to becoming popular with
us because of the high response of participants and the monetary value. Thus we can
ensure sustainability in this process of capacity building in the water and sanitation sector
in Ghana.
REFERENCES
Ghana Government, 2003. Ghana’s poverty reduction strategy. National Development Planning
Commission, Accra, Ghana. 112–113.
KNUST, 2002a. Brochure for MSc programme in Water Supply and Environmental Sanitation.
KNUST, Kumasi, Ghana.
KNUST, 2003. Evaluation report on operation and maintenance. Department of Civil Engineering,
KNUST, Kumasi, Ghana.
KNUST, 2002b. Proposal for municipal engineering and infrastructure management course.
Department of Civil Engineering, KNUST, Kumasi, Ghana.
Odai, S.N., Andam, K.A., & Trifunovic, N. 2004. Strategic partnerships for sustainable water
education and research in developing countries. Proc. Int. Conf. on Water Resources of Arid and
Semi Arid Regions of Africa, Gaborone Botswana, 3–6 August 2004.
Strategic partnerships for sustainable water
education and research in developing
countries
S.N.Odai
Department of Civil Engineering, KNUST, Kumasi, Ghana
K.A.Andam
Vice-Chancellor, KNUST, Kumasi, Ghana
N.Trifunovic
UNESCO-IHE, Delft, The Netherlands
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
The development of the water and sanitation sector is seen as a crucial process to the
successful control and eradication of communicable diseases in general. Increasingly,
more governments are realising that improvement in the water supply and environmental
sanitation sector has a direct positive impact on public health. There are stories about
communities where charging for the provision of potable water caused the inhabitants to
go back to other sources, e.g. nearby streams, which are not potable. This practice
resulted in some persons suffering from diarrhoea, while others suffered from guinea
Water resources of arid areas 222
2 STRATEGIC PARTNERSHIPS
To overcome some of these challenges, the DCE has entered into partnerships with
several institutions and external support agencies for water education and research. This
is in recognition of the modern trend of education, which takes advantage of virtual
classrooms and uses expertise around the world as guest lecturers.
Strategic partnerships for sustainable water education and research in developing countries 223
2.1.3 Technicians
This group is made up of those charged with the operation and maintenance of facilities
used in the direct provision of IWRM services such as water treatment plants,
maintenance of water supply system, etc.
Cap-net helped establish this partnership and they continue to support WANet
financially for its activities in getting the institutions involved to build the capacity and
the materials they need to kick-start their trainings. WANet is however faced with the
challenge of offering training in English and French.
theme “shelter and sanitation for the homeless” (1999–2005), and the other with the
University of Bristol under the theme “sustainable water delivery for the poor” (2003–
2006). These links are established based on the mutual consent and interest of the two
institutions involved since they will have to make their expertise and facilities available
for use by the other institution.
Under the link arrangements, staff from UK institutions may serve as lecturers in
Ghana, with the cost borne by BC. On the other hand, staff from Ghana benefit from
sitting in some courses, having discussions with experts, having access to their library
materials and electronic journals. The staff exchange grants staff from both institutions
the mutual benefit of learning of new areas of research.
Ghanaian students on exchange to a UK institution benefit from having access to
facilities for research, while students from UK benefit from best practices in Ghana. Over
50 Ghanaian students and 40 British students have benefited from this partnership
exchanges. This students exchange programme has been at the BSc level, and recently
extended to PhD students. MSc students are yet to be included in the programme.
● To combine the strengths of all partners and enhance the capacity of each partner in
order to produce joint products, such as, deliver capable professionals in the sector,
find innovative solutions for sector challenges, and build institutional capacity for
better efficiency.
In the process of combining strengths and levelling the capabilities of the individual
partners, joint products in the field of education, training, communities of practice, staff
exchange, and collaborative research will be developed in a multidisciplinary manner.
These shall be demand-responsive and duly accredited.
The list of the benefits of such partnerships is endless. Some of them are
● Combining expertise from various institutions to do research and publish papers
● Combining strengths to prepare lecture materials
● For very expensive experiments, if one institute has the equipment the other institutions
can have access to them for their work
● There is leadership in research, since one institute may have all strength in a particular
area; and the other institution may depend on such an institution for direction
● Mutual benefit of learning of new areas of research
● Encourages distance and electronic learning
● Financial assistance usually available for research and training
● Knowledge from the north is made available to the south
● Best practices developed in the south are made available to the north.
The challenges that come with such partnerships are numerous but not destructive. Some
of them are mentioned below.
● Each institution must strive to attain excellence and international
recognition/accreditation
● Strive to stay modern by improving ICT equipment and providing or having access to
video-and tele-conferencing facilities
● Make your strengths available for other institution to benefit from
● Source funding for the partnership
● Difference in languages of present and potential partners.
The present partnership is expected to grow into an international collaboration within
which partners will complement each other’s effort as done in the aviation industry, e.g.,
the alliance between KLM, Northwest and Kenya Airways.
Water resources of arid areas 226
5 CONCLUSIONS
REFERENCES
Donkor, E. & Nyarko, K.B. 2002. Establishing nodal resource center in West Africa for capacity
building in integrated water resources management. Department of Civil Engineering, KNUST,
Kumasi, Ghana.
Ghana Government 2003. Ghana poverty reduction strategy. National Development Planning
Commission, Accra, Ghana: 112–113.
KNUST 2002. Brochure for MSc programme in Water Supply and Environmental Sanitation.
KNUST, Kumasi, Ghana.
Odai, S.N., Anyemedu, F.O.K., Oduro-Kwarteng, S. & Nyarko, K.B. 2004. KNUST experiences in
capacity building in the water and sanitation sector. Proc. Int. Conf. on Water Resources of Arid
and Semi Arid Regions of Africa, Gaborone Botswana, 3–6 August 2004.
UNESCO-IHE 2002. PoWER and knowledge for sustainable development.
http://www.ihe.nl/power/%20knowledge.htm
Assessing demand for clean and safe domestic
water in eastern Zimbabwe
E.Manzungu, M.Machingambi & R.Machiridza
Department of Soil Science and Agricultural Engineering, University of
Zimbabwe, Harare, Zimbabwe
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
ABSTRACT: This paper assesses rural people’s demand for clean and
safe domestic water in two districts in eastern Zimbabwe. It explores the
role played by physical, socio-economic and cultural factors in
influencing rural people’s willingness to pay for domestic water. A semi-
structured questionnaire was administered in January 2002 to
representatives of some 239 randomly selected households in Chitakatira
and Nyanyadzi wards. Willingness to pay for clean and safe domestic
water was found to be influenced by the availability of other water
sources, rainfall received in the catchment area, perceived safety of a
water source, age and occupation of respondents. Season and gender did
not significantly affect respondents’ willingness to pay. Demographic
characteristics influenced willingness to pay for clean and safe water,
which should be taken into account when implementing cost recovery
policies in the domestic water sector. For completeness, effective demand,
best illustrated by ability to pay, should be determined.
1 INTRODUCTION
In 1996 it was estimated that 2.5 million Zimbabweans had no access to safe water
(Chenje and Johnson, 1996). The situation was worse in the rural areas where only 64%
of the population had access to safe water compared to 99% in the urban areas. A recent
survey confirmed that water in the rural areas was largely unsuitable for human
consumption due to bacterial contamination (Moyo and Mtetwa, 2000). The situation has
deteriorated in the last 5 years because of severe economic problems, worsened by the
withdrawal of donor support. Since independence in 1980 the donor community has
heavily financed Zimbabwe’s Rural Water and Sanitation Programmes. Reduction in
funding has resulted in poor maintenance of water supply facilities forcing rural
communities to revert back to unsafe water sources (NAC, 1997).
Water resources of arid areas 228
Local communities are increasingly being called upon to contribute in cash and kind to
the operation and maintenance of the domestic water sources, a development linked to the
Economic Structural Adjustment Programme that the Government of Zimbabwe started
in 1991. This World Bank/International Monetary Fund-supported programme advocated
for cost recovery in many areas including social services. Provision of domestic water
was not spared. For example, the enactment of both the Water Act and the Zimbabwe
National Water Authority Act in 1998 incorporated policies like cost recovery and
economic water pricing. This signified a policy shift towards the concept of treating
water as an economic good (Manzungu, 2001). However the concept, borrowed from the
international community, as captured in the World Water Vision (Cosgrove and
Rijsberman, 2000), is characterised by a number of inconsistencies (Savenije and van der
Zaag, 2002). In a country like Zimbabwe, where 75% of the rural population is regarded
as poor, (GOZ, 1995) there are legitimate grounds to ask whether such a policy best
serves the population.
A recent survey found that, at the local level, there were mixed signals regarding
people’s willingness to pay for water (Machingambi and Manzungu, 2003). Respondents
wanted the cost of water point establishment and repairs shared between the community
(69%), the government (11%) and the donors (5%). Sixty-three percent of the
respondents wanted the government to take the responsibility of establishing water
points. Close to half (43.9%) indicated that they had individually contributed towards the
establishment of the water points they were currently using. There was also a willingness
to participate in the maintenance of most domestic water sources except in the piped
water scheme apparently because of the high costs involved. The question is: Does this
willingness to participate in operation and maintenance of domestic water facilities
translate to a demand for clean and safe water by rural communities in Zimbabwe?
This study sought to determine whether there was a demand for clean and safe water
among the rural people by assessing their willingness to pay for domestic water in the
Lower Odzi subcatchment in Chimanimani and Mutare districts in eastern Zimbabwe. In
many respects this area typifies most rural areas in the country. In the study the demand
for safe and clean water was assessed using the contingent value method (Pearce,
Markandya and Barbier, 1989). The method is based on eliciting, from respondents,
valuations/bids, which to some extent reflect the strength/ depth of feeling i.e. degree of
concern about access to clean and safe water on the basis of a hypothetical market. The
hypothetical market is taken to include, not just the good itself, but also the institutional
context in which it would be provided, and the way in which it would be financed. The
respondent is asked to indicate whether or not they would be willing to pay (WTP) a
“starting-point bid/price (SPP)”. An iterative procedure then follows: the SPP is increased
to determine whether or not the respondent would still be willing to pay the increment in
the price. The last accepted bid, then, is the “maximum willingness to pay (MWTP)”.
Besides WTP, respondents were asked about their ability to pay for clean and safe water.
Assessing demand for clean and safe domestic water in eastern Zimbabwe 229
was used to determine which of the factors; season, age, region, gender, occupation and
access to piped water affected respondents’ WTP regarding the establishment and repairs
of different water points. The Moses test was used to determine whether the observed
variation between variables was due to the influence of some of the aforementioned
factors. The Kruskal-Wallis test was used to assess if the means of the variables in each
category were the same at 5% significance level.
3 RESULTS
In some cases, there were no significant differences in the observations between the two
wards that were studied. For this reason some data is presented in a collated form. Where
there were differences the data is presented separately for the two wards.
Respondents did not make a distinction between clean and safe water although the
general perception was that clean and safe water was free of bacteria. Table 3 gives
perceived characteristics of clean and safe water.
Assessing demand for clean and safe domestic water in eastern Zimbabwe 231
The perception of whether water was clean or safe was influenced by the source of the
water and the season in which that water was used as shown in Table 4. In the wet
season, across all water sources, water was perceived to be unsafe.
infrastructure. About 80% of the respondents said it was the responsibility of the
government to ensure that the communities had enough water.
Domestic water provision was said to involve some cost by 28.9% of the respondents.
Of these, 4.8% put the costs as ranging between Z$50 and Z$400 (Z$4000=US$1). This
was however misleading since the figures coincided with the monthly water bills for
respondents, especially those staying at rural service centres. The ability to pay different
sums of money for water for the month is as shown in Figure 1.
Ability to pay, it should be noted, is a function of affordability. It, however served, as
a good indicator of respondents’ demand since it showed the price respondents would
want their water supplied at. It was observed that generally the number of respondents
decreased with the increase in the amount to be paid. However Z$10 was the most
common WTP figure. Ability to pay was linked to the SPP and MWTP, which was
affected by a number of factors such as the source of the water, the season and the
treatment water was subjected to.
The following hypotheses were then formulated to test the assumption of normality on
equal variance using the Levene’s test.
H0:
δ12=δ22=δ32=δ42=δ52=δ62 (2)
Assessing demand for clean and safe domestic water in eastern Zimbabwe 233
(homogeneous
variance)
H1:
δ12≠δ22≠δ32≠δ42≠δ52≠δ62 (3)
(heterogeneous
variance)
Accept H0 if p>0.05
Further analysis was undertaken using non-parametric tests, namely the Mann-
Whitney, the Moses and the Kruskal-Wallis.
Deep 6.7 5.9 5.4 12.6 22.6 23.4 0.4 2.1 0.8 0 0 0.8 0.4 0 0.4
well
Borehole 4.6 5 4.2 34.8 31.8 33.1 1 7 4.3 1.3 2 1 1.7 2.5 0.8 1.3 1 7
River 9.6 8.8 7.9 17.5 18 19.3 0.4 0.4 0.4 0.8 1.3 1.7 0 0 0
Dam 7.9 7.9 6.3 18.7 17.9 18.9 0 0.4 0.4 0.8 1.3 2.1 0.4 0.4 0.4
Piped 2.5 2.5 2.5 48.9 47.6 45 2.5 2.9 4.2 8.2 8.7 9.5 2.1 2.5 2.8
scheme
Canals 0.4 0.4 0.4 2.1 2.1 2.1 0 0 0 0.4 0.4 0.4 0 0 0
& water
taps
Key: S—Summer; W—Winter; D—Drought; MWTP—
Maximum willingness to pay.
Rainfall: There was a significant effect of rainfall received in a particular area on the
demand for clean and safe water. Chitakatira generally had lower SPP and MWTP values
compared to Nyanyadzi. Water was therefore perceived to be more valuable in the
generally drier Nyanyadzi than in the wetter Chitakatira. The ANOVA test on the means
showed that except for the SPP and MWTP to repair broken down pipes, all other
variables were significantly affected by the amount of rainfall received. The Mann-
Whitney test confirmed that the amount of rainfall had an effect on all variables except
the SPP for repairing a broken down pipe. The Moses test showed that all variables,
except the SPP and MWTP for establishing a new borehole, had major differences within
them because of the rainfall factor. The amount of rainfall received in a particular
location can therefore be said to influence respondents’ WTP.
Source of water: Water from boreholes, piped water schemes, canals and water taps,
shallow and deep wells had a modal MWTP figure of Z$10 whilst that of dam and river
water was Z$0. Piped water was the most popular water source followed by borehole,
shallow well, deep well, river, dam, canals and water taps.
Season: The measures of association between season and the different variables were
also very small (ranging between 0.015–0.098) confirming minimal association between
season and the SPP and MWTP values. However, drought had the highest mean followed
by winter and summer. Table 5 shows the distribution of respondents’ MWTP across the
seasons.
Water treatment: Treatment of water marginally changed the proportion of
respondents willing to pay more than Z$40 across all the water sources, especially for
piped, river and dam water. When treatment was factored in, the modal MWTP figures
remained the same for the respective water points although the MWTP figures rose to
Z$999. The proportion of respondents willing to pay Z$40 or less actually increased for
water sources such as the piped scheme and river, whilst it decreased in the case of water
from boreholes, shallow and deep wells. The modal MWTP figures for different water
sources did not change. The distribution of respondents’ MWTP with and without
treatment is shown in Tables 6 and 7.
Gender: The ANOVA, Mann-Whitney and Moses tests revealed that gender did not
significantly affect the SPP and MWTP values of the respondents. Except for the SPP
and MWTP of establishing a new borehole and repairing it, all other variables showed
that males had higher mean WTP figures than their female counterparts.
Assessing demand for clean and safe domestic water in eastern Zimbabwe 235
Age: The means for the different variables were found to have small variances with
the 30–45 and 60–75 year age groups having the highest SPP and MWTP figures whilst
the dependent (15–30 and 75+ year) age groups had the lowest values. ANOVA tests
showed that the SPP and MWTP for a new deep well were affected by age. The Kruskal-
Wallis test showed that the SPP for a new water source, as well as the SPP and MWTP
for the repair of a borehole, were different for
Table 6. Comparison of respondents’ MWTP for
treated water.
Treated water—MWTP range
(Z$/bucket—25 litre container)
Source of 0 1–20 21–40 41–100 >100
water
Shallow 6.7 32.9 2.5 1.3 0.4
well
Deep well 5.4 22.1 2.1 1.7 0.4
Borehole 4.6 32.5 3.7 2.5 1.3
River 7.1 20.6 0.8 2.1 0.4
Dam 5.9 19.8 0.8 2.1 0.4
Piped 2.9 44.7 5.1 9.2 2.4
scheme
Canals & 0.4 2.5 0 0.4 0
water taps
respondents with the different ages. It can therefore be concluded that age affected these
variables particularly the 15–30 and 75+ year age groups.
Occupation: Traditional leaders’ MWTP was Z$1 for borehole water, Z$5 for piped
water and Z$0 for all the other water points as well as for all repairs. Communal farmers,
Water resources of arid areas 236
local government officials and students had higher SPP/MWTP figures for most variables
in decreasing intensity than traditional leaders, Agritex and ZINWA officials. ANOVA
tests revealed that the SPP and MWTP for establishing a new deep well, SPP and MWTP
for a new borehole, SPP and MWTP for repairing a borehole, and SPP and MWTP for
repairing a broken down piped scheme, had significant variances due to occupation of the
respondent. There was minimal negative association between occupation and the SPP and
MWTP for a new water source, SPP and MWTP for a new borehole, SPP and MWTP for
a new deep well, and the SPP and MWTP for borehole repairs. The SPP and MWTP for
establishing a new deep well, SPP and MWTP for a new borehole, SPP and MWTP for
repairing a borehole and SPP and MWTP for repairing a broken down piped scheme were
shown to be significantly affected by occupation according to the Kruskal-Wallis test.
Occupation had a weak association with these variables. This could be explained by the
fact that more respondents did not have a stable income hence their responses masked
those of respondents with stable sources of income.
Access to piped water: The Levene test for equality of variances on the impact of
access to piped water on respondents’ SPP/MWTP showed that there was homogenous
variance, which implied that access to piped water did not affect differences in the
SPP/MWTP. However, the Mann-Whitney test showed that access to piped water
affected WTP for all water sources although respondents with piped water had lower
SPP/MWTP figures compared to those without piped water except for borehole water.
In order to establish whether demand for a better water service delivery existed in the
communities, respondents were further asked whether they were willing to contribute
towards the establishment of a new water source that would save women time compared
to an old source, new borehole and deep well.
Investment in a new water source for women’s needs: Respondents’ SPP and MWTP
were not much different for a new water source that would save women time compared to
an old source across the seasons. However, during drought the SPP/MWTP figures were
higher than in winter and summer. The Levene test showed that variances in the SPP and
MWTP values observed were due to the effect of the amount of rainfall received.
Parameter estimates showed that Chitakatira had lower SPP and MWTP values than
Nyanyadzi. Agritex/ZINWA officials, traditional leaders and local government officials
had a decreasing effect that is lower SPP and MWTP values compared to communal
farmers and students. Gender did not affect the respondents’ SPP and MWTP values
although males had higher SPP and MWTP figures than females. The Mann-Whitney test
showed that access to piped water affected the SPP/MWTP. Respondents with piped
water had higher SPP/MWTP than their counterparts without.
For the establishment of a new borehole, Chitakatira had lower WTP figures compared
to Nyanyadzi, which decreased the SPP and MWTP values. Traditional leaders had
lowest WTP figures whilst the Agritex/ZINWA officials had the highest SPP and MWTP
figures. Drought had higher SPP and MWTP values followed by winter then summer. In
this case females were found to have higher SPP and MWTP values than males. Access
to piped water was shown not to affect respondents’ SPP/MWTP for establishing a new
borehole.
Occupation, age and region had significant effects on the SPP and MWTP for the
establishment of a new deep well. Traditional leaders still had a decreasing effect on the
SPP and MWTP whilst students had the highest SPP and MWTP values. The 15–30 year
Assessing demand for clean and safe domestic water in eastern Zimbabwe 237
age group had the lowest SPP and MWTP values. Chitakatira had lower WTP values than
Nyanyadzi, which lowered the SPP and MWTP values. Gender did not significantly
affect the SPP/MWTP values although males had higher SPP and MWTP values than
their female counterparts. Season did not have a significant effect on the SPP and MWTP
values for the establishment of a new deep well.
4 DISCUSSION
The evidence gathered in this study showed interesting perceptions held by rural people
in relation to WTP for clean and safe domestic water. At a general level it can be said that
there is no substance in the assertion that poor people do not want to pay for water. In
Zimbabwe poor people have already begun to meet operational and maintenance costs in
domestic water sources (Machingambi and Manzungu, 2003), and in publicly owned
irrigation schemes, contrary to claims that the government maintained these schemes
(Manzungu, 1999). Worldwide it has been documented that poor people tend to pay the
highest amounts for domestic water (Cosgrove and Rijsberman, 2000). Where payments
are not forthcoming the problem may be a lack of money rather than willingness to pay
(Machingambi and Manzungu, 2003). Poor community mobilisation methodologies may
also be another reason (Global Water Partnership, 2000).
The study has also provided insights into specific issues concerning the supply of
domestic water in rural areas, which may be of interest to policymakers and practitioners.
It was clear that there was a high awareness of the potential danger caused by
consumption of water containing bacteria. Respondents characterised clean and safe
water as being free of bacteria. There was also a realisation of the likely causes of the
contamination. This was shown by the fact that water was perceived to be generally
unsafe in the wet season (hence lower WTP figures than for the dry season). Piped water
had the highest WTP figures as it was rated the safest. River water was rated the most
unsafe; it had the largest proportion of respondents not willing to pay anything for it in
summer. Treating water had the effect of increasing MWTP figures.
In some cases physical scarcity of water also affected SPP and MWTP. This explains
why Nyanyadzi, the drier of the two regions, had respondents who were willing to pay
higher amounts of money for their water than their counterparts in Chitakatira. The
impact of physical scarcity of water on shaping the management of water resources is
increasingly being acknowledged internationally. It is not absolute scarcity of water that
is a problem but an economic scarcity regarding the availability of finances for the
development and management of water resources (IWMI, 2000). This explains the
paradox of a country like Zambia with more water resources than South Africa, but has a
greater percentage of the population suffering from water scarcity more than the latter.
Some commentators have also argued that water scarcity can lead to better adaptive
capacities, which may mean the adoption of more intensive water uses (Turton and
Ohlsson, 1999). This underlines the importance of analysing the role of social, cultural
and economic factors in influencing willingness to pay for water.
Socio-economic and cultural factors affected respondents’ WTP for water. The effect
of the respondents’ economic circumstances on WTP for water was illustrated by the fact
that economically dependent individuals’ (15–30 and the 75+ year age groups) were not
Water resources of arid areas 238
interested in contributing towards their water use. Their WTP was not affected by
whether the water was safe or unsafe since they showed no interest in paying for water.
Generally traditional leaders were not willing to pay for water because they considered
themselves as owners of the water, underlining the role of cultural factors in influencing
the demand for water. Perceptions about who owns water also affected WTP
(Machingambi and Manzungu, 2003). However safety was a fundamental factor in
influencing respondents WTP as even the traditional leaders who were unwilling to pay
for water from any other water source wanted to pay for the “safe” piped water.
Females were found to have lower WTP than males probably due to the fact that they
normally do not handle finances in the home. They therefore tended to be more
conservative regarding money issues than the males. However females had low WTP
figures in relation to investing in new water sources to reduce labour upon women. The
influence of physical, socio-economic and cultural factors on the WTP provided a basis
for respondents to portray their degree of concern about access to reliable, safe and clean
water, the ideal institutional context in which water could be provided and the way in
which it would be financed. Addressing such issues constitutes a more holistic
intervention in water issues affecting respondents, rather than merely focusing on cost
recovery.
5 CONCLUSION
ACKNOWLEDGEMENTS
The authors wish to thank the Water Research Fund for Southern Africa (WARFSA) for
providing the grant that made the study possible. Mr Chimedza of the University of
Zimbabwe’s Department of Statistics is greatly acknowledged for the assistance with data
entry and analysis.
REFERENCES
CSO, 1992. Central Statistics Office, Census 1992, Provincial Profile: Manicaland. Government
Printers.
Chenje, M. & Johnson, P. (eds.) 1996. Water in Southern Africa, SADC/IUCN/SARDC. Harare,
Print Holdings.
Cosgrove, W.J. & Rijsberman, F.R. 2000. World Water Vision: Making Water Everybody’s
Business. London. Earthscan Publications Ltd.
Global Water Partnership, 2000. Towards Water Security: A framework for Action. GWP,
Stockholm.
Assessing demand for clean and safe domestic water in eastern Zimbabwe 239
ABSTRACT: Palestine consists of the West Bank and the Gaza Strip. The
proclaimed state of Palestine has a land area of 6657km2. Water is
considered an essential factor of life and needs to be developed in arid
countries. Reuse of treated wastewater for irrigation as supplementary
irrigation will increase the irrigated area in Palestine and replace fresh
water.
1 INTRODUCTION
Palestine consists of the West Bank and the Gaza Strip. The proclaimed state of Palestine
has a land area of 6657km2. Water is always considered as an essential factor of life and
development in arid and semi-arid countries. In Palestine the total per capita water
consumption is 139m3.
The total available water for Irrigation is 239 M.C.M. which is responsible for
irrigating only 330000 dunums out of 2314.000 dunums cultivated that can be irrigated if
water is available i.e. 5% of the total cultivated land.
The average rainfall is 450mm and unfortunately there isn’t any water harvesting
structures i.e. dams, most of this rainwater flowing towards the Dead Sea or the
Mediterranean Sea as waste. So harvesting this water in individual farmer land and using
this water for supplementary irrigation to irrigate olive trees, almonds, grapes and cereals
will be of a great impact on the Palestinian land for feed production. It should be noted
that there are few farmers who practice supplementary irrigation for production of
vegetables that are planted in summer as individual initiative. The quantity and quality of
production that they have is extremely tangible.
Since most of the land in Palestine is planted by olive, grape, and cereals,
supplementary irrigation should be introduced and practiced where the production of
wheat via irrigation by treated wastewater was three times that under rain fed planting
project implemented in a pilot project.
The role of supplementary irrigation for food production in a semi-arid country-Palestine 241
Reuse of treated wastewater for irrigation as supplementary irrigation will increase the
irrigated area in Palestine and will replace the fresh water that can be used for domestic
purposes.
As it was mentioned before, Palestine is a semi-arid country, where the average rainfall is
450mm. The availability of water is questionable. Furthermore, the availability of water
for agriculture is reducing in a tangible way due to the following:
1. The normal increase in growth rate, the population of the country is increasing, so the
demand for domestic water is also increasing. This will affect the availability of water
for agriculture.
2. Since rainwater is the only source of water, the quantity of rainwater (rainfall) has been
decreasing in the recent years.
3. There is a huge conflict on water issues at this stage between the Palestinians and the
Israelis since Israel occupied Palestine. It should be mentioned here that during early
negotiations in the peace process, four main issues have been delayed since 1992; they
are Jerusalem, refugees, water and borders. Still after 8 years of negotiations, there
hasn’t been any significant movement on these issues. So the quantity of water that
can be available for the Palestinians will probably not be increased.
4. The quality of ground water wells especially in Gaza and Jericho becomes saline and
shortly it cannot be safely available for agriculture.
From the above, it seems that extra availability of water for additional irrigated area or
even to sustain the irrigated area is not an easy task.
Total cultivated area in the West Bank is 2100.00 dunums, but the irrigated area is
110000 dunums. From the small experience (pilot project) for this field as well as other
country experience i.e. Syria. It has been proven that the production of crops under
supplementary irrigation is 3 times higher than under rain fed crop, in addition to the
increase in the quality of the product. So if supplementary irrigation has been practiced
we can easily increase the production of rained crops to three times or twice. This will
play a major role in providing food for the people and even exports can take place and the
net income of the country will be increased.
3 BACKGROUND
It is foreseen that the world’s food production has to be doubled in the next 25 years, and
thus, the agriculture continues to be an important sector in the 21st century. Meanwhile,
the agriculture sector remains the largest user of the water resources, and it is evident that
there is a decline of agricultural water due to increasing demands from cities, industries,
and hydropower utilities in the developing countries such as Asia. Much of the water has
to come from irrigation water savings.
Water resources of arid areas 242
Population and economic growth in many developing countries of Asia have created
serious problems, such as the shortage of food, the scarcity of water, and the deterioration
of the environment.
Some of the irrigation and drainage projects have been seriously criticized due to their
high-cost and low-efficiency for the construction and maintenance. The concept of
maximum yield is now changing to optimum yield for creating an efficient irrigation
schedule. The water saving is the most sustainable conservation, because it reduces the
new construction needs to meet the increased water demand. The major issues of
agricultural water are how to increase withdrawals about 15–20% by water saving, how
to increase storages 10–15% by new irrigation facilities, and how to conserve the water
quality of irrigation.
4 SUPPLEMENTAL IRRIGATION
4.1 Definition
ICARDA defines supplemental irrigation (SI) as; the addition of essentially rain fed crops
of small amounts of water during times when rainfall fails to provide sufficient moisture
for normal plant growth, in order to improve and stabilize yields. Accordingly, the
concept of SI in areas having limited water resources is built on three bases:
First: water is applied to rain fed crops, that would normally produce some yield
without irrigation;
Second: since precipitation is the principal source of moisture for rain fed crops, SI is
only applied when precipitation fails to provide essential moisture for improved and
stabilized production and;
Third: the amount and timing of SI are not meant to provide moisture stress-free
conditions rather to provide minimum water during the critical stages of crop growth to
ensure optimal instead of maximum yield.
The management of supplemental irrigation is seen as a reverse case of full or
conventional irrigation (FI). In the latter the principal source of moisture is the fully
controlled irrigation water, while the highly variable limited precipitation is only
supplementary. Unlike FI the management of SI is dependent on the precipitation as a
basic source of water for crops grown.
Water resources for supplemental irrigation are mainly surface, but shallow ground
water aquifers are being increasingly used lately. Non-conventional water resources are
of a potential for the future, but an important one emerging is water harvesting (Dwas
2001).
SI of about 180, 125 and 75mm respectively. Generally, optimal SI amounts range from
75mm to 250mm in areas with annual rainfall between 500 to 250mm, respectively.
Determining the optimal amount under various conditions will be discussed later (Oweis
2001).
When rainfall is low, more water is needed but the response is greater, but increases in
yield are remarkable even when rainfall is as high as 500mm. The response was found to
be higher when rain distribution over the season is poor. However, in all rain fed areas of
the region it was found that some time in the spring there is usually a period of stress,
which threatens, yield levels. This soil moisture stress usually starts in March, April of
May, if total annual rainfall received is low, average or high respectively (Oweis 2001).
In Syria average wheat yields under rain fed conditions are only 1.25t/ha and this is
one of the highest in the region. With SI the average grain yield was up to 3t/ha. In 1996
over 40% of rain fed areas were under SI and over half of the 4 mil tons national
production was attributed to this practice. Supplemental irrigation does not only increase
yield but also stabilizes farmer’s production. The coefficient of variation in rain fed
production in Syria was reduced from 100% to 10$ when SI was practices. This is of
special socio-economic importance since it affects farmer’s income (Oweis 2001).
4.2.1 Introduction
Historical Palestine is located between the Mediterranean Sea and the Jordan River, as
well as to the Red Sea from the south. The present proposed Palestinian state consists of
West Bank and Gaza Strip. The other part of Palestine is occupied by Israel in 1948. This
study focuses on the West Bank and the Gaza Strip. The proclaimed state of Palestine has
a land area of 6657 square kilometres (Kateeb 1993). Population senses has been taken
place recently by the Bureau of Statistics early 1998. It is reported that the population of
the West Bank is 1571571 and Gaza Strip is 963026 where the total population of the
Palestinian people is 2534598 people.
Ground water is the main water source in the country. It is recharged by rainfall.
Rainfall varies from 100mm in the south east to 800mm in the north. The average rainfall
is 550mm (Sbeih 1995). Where the average rainfall in Jordan Valley is from 100mm to
270mm/year (Zaru 1992), and in Gaza is 200–400mm/year (Abu Safieh 1991).
Not all the rainwater is available to the Palestinian due to Israeli Military orders.
Water is abstracted from the ground water through 340 wells in the West Bank and 1781
wells in Gaza. In addition to that springs contribute a lot, where half of the irrigation
water in the West Bank is due to springs.
The quality of the available water varies from almost rain water to brackish water. In
the Jordan Valley where it is the lowest point in the elevation in the world where
temperature is very high in this area especially in summer. As example, the chloride
content is reaching 68mg/l and the SAR reaches 11.7 where the TDS reaches 5000PPM.
Still the utilization of this saline water is not as efficient and environmentally safe as it
should be where further utilization of this water could play a major role in developing the
area where still the irrigated area consists of not more than 6% of the cultivated area in
the West Bank.
Water resources of arid areas 244
It should be mentioned that not only saline water does already exist and utilized
unproperly, but it also seems to be that the additional water that can be allocated for
irrigation is also saline water which is going to be from:
1. The Eastern aquifer to be used in Jordan Valley.
2. From the treated waste water from different cities and villages in the West Bank.
On a global basis at least 60% of all water abstracted at present is used for agricultural
production. In Palestine 70% of all water consumed is due to agriculture.
Here in Palestine, agriculture is considered to be one of the main national income.
Agricultural production contributes 47.61% of the total national income in 1970.
The potential for irrigation to raise both agricultural productivity and the living
standards of the rural poor has long been recognized. Irrigated agriculture occupies
approximately 17% of the world’s total available land but the production from this land
comprises about 34% of the world total.
In Palestine, irrigation is considered to be the spinal chord of plant production for the
following reasons:
1. Palestine is considered as a semi arid region where some of the crops cannot be grown
without irrigation (example, citrus).
2. In the Jordan Valley, which constitutes the main agricultural production for the
country, irrigation is a must due to low rainfall and high temperature.
3. With irrigation the same plot of land can be planted up to three times per year while it
cannot be planted more than two times with dry farming.
4. Different varieties and crops can be planted in any region due to the availability of
water i.e. more flexibility of planting several crops at different regions in different
times of the year.
5. Job creation: Since the labour requirement per irrigated dunum is more than double
that of job required per dry farming per one season. This has now become more vital
due to continuous of closures of the West Bank and Gaza Strip where the number of
labourers that are working in the Palestinian part that occupied in 1948 is sharply
reduced.
6. Agricultural production is much higher for irrigated farming than for dry farming per
dunum per season. As example average tomato production per dunum is as follows:
– Dry farming: 2–3 ton per dunum per season.
– Irrigated (open land) 6–8 ton per dunum per season
– Irrigated (greenhouse) 12–16 ton per dunum per season
7. Net income per dunum of dry farming does not exceed $150 while from irrigated area
the net income can exceed $1500 per dunum.
8. Especially in Palestine, where the horizontal expansion in agriculture by increasing the
total cultivated area due to the Israeli occupation, and shortage of water. The vertical
expansion could be the main parameter to play with. Irrigation will be the main
element in this formula. So that providing extra water for irrigation to irrigate as much
as possible of the cultivated area is a must. This implies that Palestinian should use
any drop of water. Regardless the quality of that water practically and efficiently:
Table 2 shows the irrigated area in each district in Palestine where the total irrigated area
in 1993–94 was 217000 dunum (PSBS 1996).
The role of supplementary irrigation for food production in a semi-arid country-Palestine 247
Where there are another 7000 dunums, which are already irrigated.
Where about 44000 (PCBC 1991) dunums of this area is currently irrigated. So the total
additional area that could be irrigated in the West Bank is (210900−44
000+(535100−935000)= 608500 dunums.
It should be mentioned that the Jordan Valley produces more than 59% of the
vegetables produced in the West Bank. It also produces 100% of the bananas produced in
Palestine.
Still the term supplementary irrigation is not even used formally and officially in
Palestine. Until this time there is not any plan of implementing any project of
supplementary irrigation. This is mainly due to the lack of qualified staff at the Ministry
of Agriculture as well as to the lack of great interest to agriculture from M.O.A. due to
the following reasons:
1. The lack of responsibility of the Palestinian Authority on most of the agricultural land
due to the occupation.
2. The lack of finance and funding to development projects.
Nevertheless, there are individuals who attempt to use supplementary irrigation, an
example of that are few farmers in Sinjel town in the Ramallah area.
This village is located just between Ramallah and Nablus cities, situated 20km to the
north of Ramallah. The total agricultural area in the village excess 4000 dunums, out of
these areas. About 1000 dunums are plain and flat.
The role of supplementary irrigation for food production in a semi-arid country-Palestine 249
This 1000 dunums is planted with vegetables in summer and cereals in winter. All of
this area is rain fed, there are no source of water for irrigation since this area is located
close to the village (houses), it is easy for the farmers to bring water by mobile tanks.
Usually the farmers in summer, bring some water and store them in a container
(barrel) of 200 liter capacity each, since the ownership of land is between 3–5dunums, the
number of barrels used are 6–8.
In summer farmers used to plant vegetables, at the time of planting the seedlings,
farmers used to irrigate the seedling by a bucket. Farmers used to mix the fertilizer water
and irrigation at the time of planting the seedlings. Later on, after 20 days the second
irrigation with fertilizer is applied. The third one and the last one are provided with
fertilizer before flowering. The total amount of water applied per each plant is not more
than 1 liter, for a dunum of 1000 plants, 1000 liter is applied 1 cubic meter of water
applied for the whole season per one dunum. While for the irrigated area the minimum
irrigation water requirement is 70m3/dunum per the season.
In this village, Sinjel, and through my investigation, in the year 2000 I found 3 farmers
who are using this approach technology, when I asked one of them what is the result that
you will expect, he broadly replied:
1. The quality of agricultural product that I used to obtain for the last two years where I
used to use supplementary irrigation is much better than the product of my neighbor in
the same plot of land in the village, so the price per 1kg. That I got is much higher
also.
2. The total production is much higher than that of my neighbor, i.e. I got 4 tons each per
dunum, my neighbor got 2 tons of squash per dunum.
3. The period of production that I have is much bigger than that of my neighbor has, this
means that total income that I gained is much higher. I used the produce vegetables for
2 months, while my neighbor only one month, i.e. the harvesting period is much
higher when supplementary irrigation used.
I informed this farmer that I am working on an irrigation project coordinator for an NGO
that provides funds for farmers. Since this farmer believes that he was happy from his
production since he has only 3dunums and all of his family working in this plot of land,
he did not ask what service that since that we offered, this totally indicated that he is
happy, and he did not need any further assistance. At that time there was visiting
irrigation professor from Canada. This professor told me that we should use him as a
model to encourage people using appropriate technology.
Another example of using supplementary irrigation is found in Hebron where a farmer
from Al Tamimi family, who has a grape field and luckily a pipe water pass through his
field and used to get some water from this pipe and provide some water for his grape. In
winter since the rainfall in Hebron is not exceeding 300mm, as well as in July.
Water resources of arid areas 250
It is well known in Hebron, that the quality of grape of that man is the best in Hebron,
since Hebron is of the biggest producing city (country) in Palestine.
Since the municipality constructed a pilot treatment plant, it thought of planting crops
using the treated effluent. This was funded by American Near East Refugee Aid
(ANERA). Three crops were selected by the Agriculture Department to be planted for the
first time in Palestine using treated wastewater:
● Artichokes on 150m2—planted on October 31, 1993.
● Onion frozen production on 500m2—planted on November 6, 1993.
● Wheat on 1000m2—planted on November 22, 1993.
● Drip irrigation as well as sprayers were used.
Several treatments were made as follows:
● Irrigation with wastewater used, fertilization was used.
● Same as above, but without application of fertilization.
● Irrigation not used but fertilization was used.
● No irrigation and no fertilization (dry land farming).
All the agricultural practices were used (pesticides, ploughing, seed control, etc.
Table 3 shows the production of each kind of treatment. The impact of using treated
wastewater appears clear.
Notes:
1. Time of planting was October 1993; all the crops received rainfall during the growing
period.
2. Time of harvesting was June 2, 1994.
3. Production with irrigation with treated wastewater with fertilization was five times
without irrigation and fertilization.
4. Production increased the soil when irrigated with treated wastewater where fertilization
was applied on both cases (irrigated and non-irrigated).
The role of supplementary irrigation for food production in a semi-arid country-Palestine 251
Since the ownership of land is very small in size i.e. from 5–10 dunums, supplementary
irrigation can be easily implemented for vegetables, trees and to cereals to some extent
constructing of small ponds of 40–50m3 capacity, i.e. this pond can be located on a 14–18
meter square area. This pond can be located on the lowest point in elevation of the
individual land. This land serves two farmers if agreed upon where it can be sited on the
border of each farmer land.
Distributing of water to the plant can be done manually by lifting the water and
distributing it to the plants by a bucket. Another way of distributing this water that this
water can be lifted manually from the pond and poured into a barrel that can be located
on the dip of the pond with 1/2meter raised over the surface so water can be distributed to
the plant by gravity through pipe line. The farmer can distribute the water pipe from the
plant to another. These methods can be implemented
Table 4. Results of Al Beireh Pilot Wastewater
Treatment, 1994.
Production
(kg/dunum)
Crop Kind of Seed Hay
treatment
Wheat 870 Irrigation with 687.5 1375
type fertilizer
Irrigation 656.70 1373
without
fertilizer
Rainfed with Rainfed with 537.5 1187.5
fertilizer fertilizer
Rainfed without 500 1531.25
fertilizer
Wheat anber Irrigation with 864 1656
type fertilizer
Irrigation 824 1212
without
fertilizer
Fertilizer Rainfed with 600 1000
fertilizer
Rainfed without 236 336
fertilizer
easily with zero operation cost. Since only the farmer himself can conduct this job easily,
another method of distributing water is by using a small pumped electricity is available
since the head required is very small.
In the case of cereals water can be distributed easily by establishing ponds, so water
can be discharged into the farm then water can flow by gravity. In order to reduce the
Water resources of arid areas 252
cost of pumping farms can cooperate between themselves when each farmer can
construct his pond on the highest point in elevation on his land. His pond can receive
water from his neighbour’s field and so on.
9 REVENUES
Assume a plot of land of 5 dunums planted with vegetables. The production of vegetables
of rainfed per dunum is 3 tons/dunum, the production of dunum with supplementary
irrigation is 4.1 ton.
The price per ton is $200 for The income per supplementary irrigation
rainfed crops. is 4×250=$1000.
The price per ton for The net income due to supplementary
supplementary irrigation is $250. irrigation will be 1000?=400 per dunum.
S the income per rainfed 5 dunums×400=2000 per session per 5
dunum=3×200=$600. dunums.
6. The additional water that will be available for the Palestinians will be either from (a)
Eastern aquifer, (b) Jordan River, or (c) Treated wastewater. Where all of this water is
saline water, where there are another source such as the mountain aquifers, but this
seems to be difficult to be secured soon.
7. The early possible of expansion in irrigation will be in Jordan Valley where the
existing water wells and the future water that might be available is saline.
8. Since the treated water is in the full control of the Palestinians, more attention and care
should be paid in order to better and safe utilize of this water for developing the
agricultural sector in Palestine, and this water can be used for supplementary
irrigation.
9. The productivity of one cubic meter of water with supplementary irrigation is much
higher than that of irrigated land since the water prepared by irrigated dunum is 7
times more than the required for supplementary irrigation.
10. The existing irrigated area is already exhausted since this land used to be planted two
or three times a year where the other land used to be cultivated once a year even it
kept fallow on some years.
11. Palestinian Agricultural Ministry and Palestinian Water Authority should recognize
the situation and consider supplementary irrigation as a major element for food supply.
REFERENCES
1 INTRODUCTION
Any resource, such as water, when used by more than one user in a single catchment or
river basin, tends to attract conflicts about how it is shared and distributed.
This paper focuses to a large extent on a highly committed small river catchment;
namely, the Mswenzi River Catchment that forms part of a greater basin, which
eventually drains off into the Zambezi River and into the Indian Ocean. The Mswenzi is a
tributary of the Shuru Shuru River in the northwestern region of Zimbabwe. The Shuru
Shuru River in turn is a tributary of the Mupfure River that drains northwesterly and into
the Zambezi River. The study catchment has a total surface area of approximately
Conversion of priority water rights to proportional water permits and conflict management 255
158km2 with the Mswenzi River stretching a distance of 22km. The study catchment
generates on average an estimated 70mm of surface water per year, or 11.0*106m3/year.
Hence, this paper looks into the operationalization of water rights in the Mswenzi river
catchment and gives an analytical recommendation of converting the old water rights
used under the Zimbabwe Water Act (1976) to new water permits as recommended under
the new Zimbabwe Water Act of 1998. A water right, for the purpose of this study, has
been defined as a right to use beneficially a certain volume of water expressed in absolute
volumetric units per time unit, whilst a water permit is a permit or allowance for the use
of water, which specifies and restricts the use of water allocated. At the same time the
paper attempts to bring out the conflicts involved in the Mswenzi catchment between
upstream and downstream commercial water users. The paper discusses the legislative
policies by looking at the old and new Water Acts and highlights the important principles
that will govern conflict management and help spells out recommendations for water
authorities and catchment management institutions.
In Zimbabwe, the Water Act of 1976, which was repealed by the new Water Act of 1998,
vested all public water in the President and private water was water belonging to the
owner of land on which it was found. The right to use water was dependent on the type of
use. For primary use no right was required. Access of water for non-primary use was
based on the prior appropriation doctrine, where an appropriated right was based on the
application of the appropriated public water to some beneficial use. The granting of any
water right was the exclusive function of the Administrative Court sitting as the Water
Court. The right would only be granted if public water was available and if it could be
ascertained that the water would be put to beneficial use. The right granted was
dependent on the date on which the application for the right was made. This date
determined the applicant’s priority in the use of water applied for.
The new Zimbabwe Water Act of 1998, which replaces the old Water Act (1976)
sought to bring about equal and fair distribution of the available water resources in the
national interest for the development of the rural, urban, industrial, mining and
agricultural sectors. The major principles of the new Water Act were that all water would
now be owned by the State and any use of it other than for purposes of primary use
should be approved by the State. All stakeholders should be involved in decision-making
processes and contribute to sustainable management of water resources. Water resources
would be managed at catchment and sub-catchment levels, and the environment would
also be considered a legitimate user of water. One of the important changes of the Water
Act that is brought out in this study is that the priority date system of first come first
served of water allocation was abolished to enable the principle of equitable access to
water and sharing of water at all times. A fractional allocation system is now the
recommended allocation system for non-primary water use.
Water resources of arid areas 256
The Mswenzi River Catchment has an area of approximately 160km2, and on average
generates 11.0*106m3/a of blue water. Water generated in the catchment is based on the
annual unit runoff and the catchment area, where the water generated (m3/a), is a product
of the two.
Mean unit runoff for the catchment is 70mm/year and the calculated catchment area is
158.15km2. This gives an amount of 11.0*106m3/a as water generated in the catchment.
The Mswenzi River catchment has a total of 17 existing water rights owned by mainly
commercial farmers, who have built reservoirs with a total capacity of 7.3*106m3.
Among the 17 water rights, the total commitment level of the catchment is 65%. This is
water used, (m3/a) as a fraction of water generated (m3/a). The major dams in the
catchment are Balwearie and Tawstock dams, which have a combined capacity of
4.9*106m3, and serve a total of 8 water rights to various farmers in the catchment. The
other 9 water rights are served from smaller dams along the Mswenzi River. Table 1
shows existing water rights and current users in the catchment as well as the priority date
for each property.
The procedures for the conversion of water rights to water permits were that, (i) the
priority date attached to each water right be removed, (ii) volumes of flow and storage
rights as allocated in the old system remain the same and shall be used as permits until
such a time as the water authorities see fit to amend or revise the permit, and that (iii)
water rights be converted according to the applicants ability to beneficially use the water.
In the conversion exercise a simple formula has been recommended for the purposes
of the study catchment, which follows:
Permit (m3/a),
P=[S1+S2+S3…]+F1+F2…
years.
Handley Cross, P=[2244*103m3/a]
for a validity of 20 years.
Cornucopia, P=[523*103m3/a] for a
validity of 20 years.
Merchiston,
P=[(611+20+5+20)*103m3]
=656*103m3 for a validity of 20 years.
The new permits as recommended for the study catchment after the conversion exercise
is shown in the Table 2.
The case singles out the dispute between upstream and downstream commercial farmers
holding water rights in the Balwearie and Tawstock Dams where the latter is
downstream. Downstream
Table 2. Water permits for Mswenzi river.
Water Property River Abstraction Dam Validity
permit permit
number (103m3/a)
7573 Kasama Mswenzi 45 Weir 20 years
tributary
8881 Dodington Mswenzi 136 Farm dam 20 years
tributary
10626 Rem. of Mswenzi 1430 Balwearie 20 years
Luton
10156 Balwearie Mswenzi 1620 Balwearie 20 years
2276 Strathspey Mswenzi 1100 Tawstock 20 years
12398 Handley Mswenzi 2244 Suri 20 years
cross Suri/Tawstock
10364 Cornucopia Mswenzi 523 Tawstock 20 years
12007 Merchiston Mswenzi 656 Farm dam
Total 7754
farmers who held earlier water right priorities in Tawstock Dam were outraged that they
were not receiving sufficient water from upstream Balwearie Dam and as a result
jeopardized their operations. As earlier applicants, Tawstock farmers were entitled to
water first, which meant that Balwearie farmers had to open the outlet gates at Balwearie
Dam and release water for Tawstock farmers before the former could store and use any
water.
In 1982 An investigation came about as a result of a submission for a decision made to
the Administrative Court by Tawstock farmers who possessed water rights no, 2276,
Conversion of priority water rights to proportional water permits and conflict management 259
10364 and 9101 of Tawstock Dam and who are referred to as the applicants. The
applicants were concerned that the holders of water rights no, 10156, 10626 and 10659
(Balwearie Farmers) of Balwearie Dam who are referred to as the respondents, were
unable to pass sufficient water from their storage works to satisfy the downstream
priorities. The applicants maintained that the reason for this was that the outlet pipe of
Balwearie Dam was of insufficient internal diameter for this purpose and therefore the
operation of their prior rights was jeopardized.
In this case the two dams have a similar capacity and are separated by a mere 2km
where each dam has three participants in the utilization of the stored water. Two parts of
water right no. 2276 have the earliest priority, after which water right no. 10156 has its
turn. Then the remaining part of water right no. 2276 followed by the other two
participants is satisfied. The other two participants in water right no. 10156 then follow
each with separate priorities. This rather complex situation involves 6 separate water
rights and 9 priorities.
An agreement was reached at the Administrative Court between the applicants and the
respondents. It was agreed that:
● A siphon of 12 inches diameter be constructed and installed together with an outlet
pipe of not less than 12 inches diameter in Balwearie Dam. Both devices were to be
used to pass water that flows into Balwearie Dam and down to Tawstock Dam to meet
the entitlements of holders of water rights no, 2276, 9101 and 10364 together with the
primary requirement of 85 liters per second.
● Not less than 425l/s will be released from Balwearie Dam, and the construction and
installation of the siphon and the gauging weir immediately upstream of the
headwaters of Balwearie Dam shall be carried out by Balwearie farmers so as to be in
full operation.
5 SUMMARY
The case has described the way in which the operation of priorities of this complex
situation worked under the old system and how the issue of satisfying earlier priorities
was resolved. In arbitration (a conflict management tool), the Tawstock farmers
submitted their argument before the Administrative Court who acted as the judge and a
solution was reached whereby both the Tawstock and Balwearie farmers signed an
agreement.
6 RECOMMENDATIONS
It is recommended that measures be taken to establish actual water use rather than assume
values of water rights so that true commitment levels are achieved. It is also
recommended that conflict management play a more important role in water resource
management so as to empower local water authorities to handle such situations involving
disputes over water allocation. It is hoped these recommendations will lay the platform
for increased participation, negotiation and dialogue for better basin management.
Water resources of arid areas 260
7 CONCLUSION
The main objective of this study was to establish how existing water rights in a small
catchment were operated under the old system based on the 1976 Water Act and to
describe the conversion process with the requirements of the permit system. The results
from the study show that most storage rights were operated with staff gauges installed in
the basin that enabled the stored volume and abstraction to be determined for any
reservoir level.
The study also showed that senior water rights consumed water impetuously without
much consideration for downstream users and therefore new users found it difficult to
receive a full entitlement of water allocated. The new system now allows new users the
opportunity to be given an abstraction permit for their beneficial use therefore
disregarding priority.
The study revealed through a questionnaire that there was little cooperation and
communication over data between upstream and downstream users that often resulted in
disputes over water allocation. Of the seven farmers interviewed, six of them said they
were not aware of the water reform. It is important therefore that water authorities and
catchment agencies seriously consider the issue of enhancing dialogue and cooperation
between different users and assist in the issuing of water permits to improve the
management of water at catchment and basin level.
The results of the study showed that the Mswenzi is a highly committed catchment
with a total use of flow and storage water rights totaling 6900*103m3/a for all the water
right holders in the catchment. The Mswenzi generates on average per year,
11070*103m3/a of water, bringing the water commitment of the catchment to 62.3%. The
study showed that in the conversion process from the old water rights into new water
permits, the permits would have to use the same volumes as previously granted for their
water rights and discard the priority date. Therein catchment councils have the obligation
of amending or revising the water permit according to beneficial use of the permits and
accommodation of new entries among other criteria.
Perceptions of the commercial farmers in the catchment have shown that the majority,
almost 80%, of the big stakeholders in the catchment are unaware of the principles of the
water reform and how their new water permits will be operated. Conflicts over water
allocation have emerged under the old system due to misperceptions and lack of adequate
data and it is anticipated that the new permit system will mitigate the grounds for
conflicts in the future.
REFERENCES
Ashton, P. 1999. Southern African water conflicts: Are they inevitable or preventable? In: Water
for Peace in the Middle East and Southern Africa, Geneva, Green Cross International, 2000.
Biswas, A.K. 1993. Management of International Waters: Problems and Perspectives, Water
Resources Development, 9.
Heun, J. 1998. Water resources planning and analysis, Lecture Note, Dept. of Civil Engineering,
University of Zimbabwe, Harare.
Huffaker et al. 2000. The Role of Prior Appropriation in Allocating Water Resources into the 21st
Century, Water Resource Development, 16.
Conversion of priority water rights to proportional water permits and conflict management 261
Jaspers et al. 1999. An external review of the Mupfure Catchment Integrated Water Management
Project, Prepared for the Royal Netherlands Embassy, Harare.
Lang, H. 1997. Options for a New Water Rights System. Draft Unpublished Paper, October.
Manzungu, E. 1999a. Conflict management in the Umvumvumvu Catchment, In: Water for
Agriculture in Zimbabwe: Policy and Management Options for the Smallholder Sector,
University of Zimbabwe, Harare.
Manzungu, E. 1999b. Strategies of smallholder irrigation management in Zimbabwe, PhD Thesis,
Wageningen University, Netherlands.
Manzungu, E, Senzanje, A & Van der Zaag, P. 1999. Water for Agriculture in Zimbabwe: Policy
and Management Options for the Smallholder Sector. University of Zimbabwe Publications,
Harare.
Mazvimavi, D. 1998. Water Resources Management in the Water Catchment Board Pilot Areas,
Phase I: Data Collection, CASS Technical Paper, NRM Series; CPN 95/98, University of
Zimbabwe, Harare.
Natsa, T.F. 1999. From priority date to fractional allocation: Towards equitable distribution of
surface water resources in Zimbabwe. MSc Thesis, University of Zimbabwe, Harare.
Resolve Inc. et al. 2000. Participation, Negotiation and Conflict Management in Large Dams
Projects, Final version, Cape Town, Republic of South Africa.
Van der Zaag, P & Nyagwambo, L. 1998. Water Allocation Criteria for the Mupfure Catchment,
Final Document, December 1998, Harare.
Van der Zaag, P. 2001. Water Law Lecture Notes, Department of Civil Engineering, University of
Zimbabwe, Harare.
Wallensteen, P & Swain, A. 1997. Comprehensive assessment of the freshwater resources of the
world. International fresh water resources: Conflict or cooperation, Stockholm, Sweden.
Wolf, A.T. 2000. Indigenous Approaches to Water Conflict Negotiations and Implications for
International Waters, Published in: International Negotiation: A Journal of Theory and
Practice, December 2000.
Zimbabwe 1996. Zimbabwe Water Act 1976, (Chapter 20:22), Government Printers, Harare.
Zimbabwe 1999. Zimbabwe Water Act 1998, (Chapter 20:24), Government Printers, Harare.
Impacts of water development in arid lands of
Southern Africa: socio-economic issues
J.P.Msangi
University of Namibia, Windhoek, Namibia
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
supplies to meet both animal and human requirements in Southern Africa. In some areas,
irrigation water has been provided to overcome inadequate and/or unreliable rainfall
while in others domestic water supply schemes have been constructed including dam
building and borehole construction. The main justifying reason for such undertakings is
always that water in its natural state is seldom in a position to satisfy the requirements
which include public water supply for domestic use and/or livestock production;
regulated flow for hydro-electric power production; adequate supplies for industrial
processing and irrigated agriculture. Water development projects have been recorded to
cause adverse impacts to the environment the world over, Southern Africa included. Such
impacts are known to reduce the performance of the economy and undermine the
sustainability of projects implemented to off set the difficult situations found in the arid
lands. Such impacts need to be identified and/or predicted during the project planning
stage so that appropriate mitigative measures can be taken into consideration before and
after the project is implemented. Environmental impact issues in Southern Africa include
high population concentrations (both human and animal) attracted by the putting up of a
reliable water source; soil erosion; agricultural and chemical pollution from irrigated
fields as well as over exploitation of groundwater aquifers which may lead to collapse
and eventual destruction of the aquifer. Others include denying down stream populations
and habitats fresh water supply through damming or excessive abstractions to meet
upstream water demands. Through environmental impact studies, the multisectoral nature
of water resource development can be taken into consideration during the planning and
implementation phases of water resource development projects. Strong partnerships and
indigenous knowledge considerations are necessary to make sure all aspects of the
resource are included in such studies.
2 INTRODUCTION
Most always, the purpose of water development is to provide adequate supplies of water
to meet various water demands. This is necessitated by the fact that in most instances,
Impacts of water development in arid lands of Southern Africa 265
water in its natural state is not in a position to satisfy the numerous demands placed upon
it, which may include:
– Public water supply for domestic use;
– Hydro-electric power production;
– Irrigated agriculture development;
– Water for livestock production;
– Industrial water for processing or for cooling machines;
– Water for sewage treatment;
– Fishing;
– Navigation;
– Recreation.
Water development projects are known to cause both positive and adverse environmental
impacts to the environment that need to be identified and/or predicted prior to project
implementation. Examples range widely depending on the nature and scale of the project,
its location and the type and level of technology required for its sustainability. A good
example is where a project involving groundwater recharge using wastewater must take
into consideration the danger of pollution from the wastewater from industrial and
residential sources. The cost of treatment of the water before using it to recharge ground
water must be considered included in the assessment. Similarly, the impacts of other land
use management activities taking place in the project area are known to affect the water
projects. Activities of unprotected catchment areas including unplanned deforestation
and/or overgrazing would produce sediments which would reduce the capacity and
adversely affect the life span and operation of a down stream reservoir greatly
undermining the project performance and disrupting its sustainability such as happened in
Kisongo dam in Arusha and Imagi dam in Dodoma, Tanzania (Msangi, 1987; Kitheka,
1993; Christiansson, 1981). On the other hand a well-managed cultivation system
(including terracing and aforestation) of a catchment area will prolong the life span of the
reservoir making it possible to meet multiple demands from its waters. All the occurring
impacts need to be identified and/or predicted during the project planning stage so that
appropriate mitigative measures can be considered before a project is implemented.
Needed also is post planning project monitoring to assess the impacts resulting from the
undertaking.
In the process of considering whether to carry out a major water project or not, the first
main concern is the extent of environmental change that will result from the construction
and particularly after the project is implemented. This change will be both to the physical
environment as well as to the people and other living organisms inhabiting that
environment. The changes to the physical environment will influence the social-
economic environment just like the people’s response to the physical change will affect
the physical environment. Thus both types of change must be considered when impacts
are predicted and later on monitored. Most of the changes will be permanent and in some
Water resources of arid areas 266
cases cumulative. Alternatives form part of the pre-implementation phase when impacts
are weighed and compared to provide careful and sufficient consideration of all possible
impacts, both negative and positive ones. The assessment process for water development
projects mainly addresses environmental impacts resulting from the project itself and
land use management practices occurring in the river catchment area. These impacts
could include high surface runoff; soil erosion and increased sediment flow and raised
concentrations that are a result of induced land use changes in a river catchment where a
project is proposed. Moreover, lack of knowledge or lack of apprehension of the
consequences of overgrazing the catchment areas contributes to the shortening of the life
span of a project (Msangi, 1996). Changed ecological conditions, such as the creation of
an ideal habitat for disease carrying flies, snails and mosquitoes, have also made areas
around a reservoir created behind a dam undesirable and unhealthy (Kaduma, 1977;
1972). Other impacts may include changes to groundwater levels, changes to river flow
and flow peaks, flooding or drying up of a river (Christelis & Struckmeier, 2001). Others
could be introduction of agricultural chemicals and fertilizer residues by surface runoff
from farmlands where irrigated agriculture is part of the project or where the project will
induce such undertakings. An indirect impact on an undertaking emanates from the
attitudes of the population towards the proposed project as well as attitudes held on water
resource use and management. In some parts of the dry lands in Southern Africa, attitudes
centering on cattle numbers as wealth are most likely to override environmental
conservation so that once water becomes available, the number and intensity of grazing
increases without due regard to carrying capacity of the range (Msangi, 1992; 1996;
Darkoh, 1989; Ellis & Swift, 1998). Such attitudes and other related social practices
should form part of the assessment during and after implementation of a water
development project.
The actual impacts of a water resource project depend on the purpose, scale and
location of the project. For a small water supply project for example, the positive impacts
will include the expected socio-economic benefits such as drinking water and water for
other domestic uses. Raised health and sanitation standards and the general well being of
the people will also be included. An indirect impact will include elevated economic
standards, as people enjoying good health will be able to work to produce more and thus
generate some economic returns. On the other hand negative impacts will include over
grazing of the land around watering points and beyond, along animal tracks and adjacent
land etc (Darkoh, 1989; Msangi, 1991, 1996; Stone, 1991). For a large water supply
project with pipelines and house connections and staggered animal watering points, the
adverse impacts on the physical environment would be minimized or controlled
completely. Thus alternatives ought to be considered carefully before implementation.
A water development project involving the construction of large structures such as
dams and canals will produce a varying range of impacts on both the physical and socio-
economic environments. Possible adverse impacts could include the displacement of
people and animal populations as happened when Lake Kariba and other large dams in
Southern Africa (Kaduma, 1997) and others including loss of flora and fauna; changes in
groundwater conditions, triggering seismic activity due to the presence of a large body of
water; deterioration of the health status of the environment through creation of ideal
habitats for disease vectors; lowered water quality from rotting inundated vegetation and
altered river flow characteristics (Kaduma, 1997). Conversely, the advantages of a large
Impacts of water development in arid lands of Southern Africa 267
scale water development would include creation of new habitats such as wetlands; the
production of much needed electricity for irrigation water to support agricultural
production; provision of hydro electric power for homes and for industrial
establishments; job creation from undertakings utilizing the water and generated power;
support improved economic conditions and the well being of the people; regulated river
flow and improved utilization of a river including the establishment of a fishery. The list
of indirect impacts is long and varied.
The creation of sub-surface dams that are more environmentally feasible than surface
dams are faced with various limitations including water recovery. High investments
required during construction and maintenance and operation of pumps (be it petrol/diesel,
solar energy or windmills) are often not economically justified given the low land
productivity inherent in most parts of the dry lands of Southern Africa. Hand pumps are
only feasible if recovery is from shallow wells (Msangi, 1996).
Southern Africa has an expansive area that is characterized by aridity, aridity being a
condition of perpetual moisture scarcity. The inhabitants of such areas have devised
mechanisms which enable them eke a living through manipulations of the environment
and available resources. The activities of these areas are predominantly changing due to
adjustments that must be made in response to the prevailing climatic conditions. An
exceptionally wet year may see the cultivation and harvesting of quick maturing crops
that dry years will not. More often than not the survival techniques include livestock
rearing, mostly keeping of small stock such as goats and sheep. Cattle are kept for milk,
export beef and as a source of wealth in the areas that enjoy relatively humid conditions
as opposed to those that are very dry (Msangi, 1996). Due to excessively high
temperatures, the little moisture that may be received in the form of rain gets evaporated
very quickly soon after a downpour (Msangi, 1996). Climatic variability and uncertainty
has led to precarious living conditions for the inhabitants of the arid lands all over the
world, those in Southern Africa high in the list.
Due to the prevailing climatic conditions and high sense of uncertainty, many water
development projects have been undertaken or are proposed to provide more dependable
water supplies for both animal and human requirements as well as industrial water needs.
In others, irrigation water has been provided to overcome inadequate and/or unreliable
rainfall (Chenje & Johnson, 1996). The main justifying reason for such undertakings is
always that water in its natural state is seldom in a position to satisfy the requirements
which include public water supply for domestic use; regulated flow for hydro-electric
power production; adequate supplies for irrigated agriculture development and for
livestock production. Water development projects have been recorded to cause negative
impacts to the receiving environment (Kaduma, 1977). While these have been
investigated and documented worldwide, in Southern Africa, the need still exists to
identify and/or predict them before the proposed projects are implemented and follow up
monitoring after they are in operation (Chenje & Johnson, 1996; Wood, Stedman &
Mang, 2000). Such projects need to be monitored as they may be affected by
environmental factors caused by other land use management activities taking place in the
Water resources of arid areas 268
project areas. Such activities are known to reduce the performance of projects and do
undermine the sustainability of projects implemented to off set the difficult situations
found in the arid lands (Biswas, 1978; Msangi, 1996). All these impacts need to be
identified and/or predicted during the project planning stage so that appropriate
mitigative measures can be considered before the project is implemented.
The inhabitants of the dry lands in Southern Africa have a rich heritage of managing
and living with their environment including water. They have been irrigating their lands
for centuries. The communities inhabiting the dry lands have lived and adapted to the
environmental conditions arising from many years of experience and folklore handed
down generations. This harmonious existence with nature was interrupted and interfered
with during the last a hundred years or so through the introduction of western cultures
and new ways of viewing the environment. The introduction of improved health and
nutritional facilities as well as monetary economy together with the institutional
requirements that go with it, has disrupted and partially changed the lifestyles of these
communities (Msangi, 1996, 1992; Stone, 1991; Ellis & Swift, 1988).
The population of both people and animals has increased rapidly as food aid and
western medicine have increased survival chances and increased fertility rates. Before
this interruption, land, water and vegetation successfully supported the life styles and
economic activities of the dry lands inhabitants. The forces of nature had adequately
checked imbalances between man and nature so that simple social and economic patterns
had developed and had been harmoniously maintained. The installation of schools,
hospitals, central governments and all their branches imposed new requirements on the
communities and therefore the environmental resources. Water being the central and most
scarce resource in these lands has been subjected to various manipulations and new
development approaches geared towards meeting both the communities’ traditional and
new institutional demands.
Due to increasing populations of both people and animals, water demand far exceeds
supply, thus the need to practice wise use, management and conservation of water
resources in the dry lands of Southern Africa. This requires that social attitudes be
reoriented so that communities appreciate the implications of limited supply as opposed
to the ever increasing demands on the scarce water resources, limited groundwater
recharge rates and the need to conserve the resources such as controlling pollution and
recycling, all new concepts to most of the dry land communities in Southern Africa.
Many cases of efforts to conserve, develop and manage the water resources in the dry
lands have been made and are documented in numerous plans and consultant reports
(Msangi, 1992). Few successful cases have been recorded and many failures have been
experienced. The reasons for failure are mainly due to the inappropriateness of the
technology adopted to the existing environmental conditions or, most often, to the wants
and wishes of the local communities. Many times the wrong sector of the community has
been targeted for training. Women and their children who are the ones mainly responsible
for collecting and managing water for domestic use and sometimes tending small stock,
tend to be side lined for the men who are users rather than managers of the resource.
Women should be at the center of any training aimed at improving existing management
technologies or introducing new ones. Sustainability rests on clear understanding of the
people’s social organizations and gender roles and economic patterns in a given
community.
Impacts of water development in arid lands of Southern Africa 269
Economic and other development activities intended to be introduced into the dry
lands should be focused more on the needs and wants of the people bearing in mind
environmental constraints. For example instead of introducing irrigated agriculture, dry
land farming based on indigenous crop varieties should be employed instead of sprinkler
irrigation to grow exotic crops with high water demand. Flood irrigation and other high
water requiring methods of crop growing have rendered useless large tracts of land
through salinization. The high evaporation rates inherent in the dry lands of Southern
Africa do not favor these methods.
Water development projects are affected by the land use in many different ways.
Unregulated land use system such as indiscriminate clearing of tree cover from a
catchment area can lead to reduced water yield and cut short the lifespan of a project.
Cultivation and/or heavy grazing of such a catchment area can lead to soil erosion and
subsequent sedimentation and siltation of a reservoir, intakes and irrigation channels.
Such a system can also lead to increased surface runoff, flooding in the lower reaches and
lowered groundwater in the upper reaches thus jeopardizing a water development project.
Other impacts include reduced water quality from suspended sediment and agricultural
chemicals and residues from farmlands. This will lead to increased costs for water
purification or adverse health conditions for those depending on the water source either
for domestic use or industrial processing. Habitats for fish and other aquatic animals will
be damaged and the economic standing of those dependent on them will be adversely
affected. Furthermore, maintenance costs for structures will increase dramatically if
sediment has to be cleared regularly, unless ofcourse this was foreseen and budgeted for
right from the beginning.
Closely connected to decreased infiltration and reduced water yield due to compaction
is the loss of water sources such as wells and springs. Reduced infiltration leads to
reduction in levels of ground water table that may cause ground subsidence (ground
surface collapse and curving in) due to over pumping; or if close to coastal areas lead to
intrusion of coastal salt water and soil salinization that may reduce crop production
through increased accumulation of harmful salts in soil particularly where irrigated
agriculture depend on wells or boreholes (Christelis & Struckmeier, 2001).
Other types of land use such as urban land use may cause water pollution due to
inadequate water and waste management from dwellings and industries. Pathogens as
well as organic and chemical pollution can lower the water quality necessitating
expensive water treatment to meet set water quality standards. Alternatively, high
concentrations of discharged organic compounds may create excessive demand on
oxygen resources of a body of water during the conversion process to the extent that the
oxygen concentration in the water is reduced and eventually depleted resulting in death of
living organisms including fish. High concentration of organic matter may also raise the
fertility of the water body to the extent that eutrophication occurs leading to life
decimation in the water body (Wood, Stedman & Mang, 2000).
Water resources of arid areas 270
It becomes apparent that water resource development projects produce serious and
definable impacts on a community both socially and economically while other activities
in the vicinity of the water projects affect and influence the performance of the projects.
Feasibility studies for all impacts must be carried out prior to water projects
implementation. Successful water development and resource conservation should always
strive to incorporate environmental considerations during project planning and project
implementation stages. Similarly, integrated catchment management should encompass
the various resource components and associated management practices to achieve stable
systems.
Environmental legislation should make EIA mandatory in all water development
projects in order to ensure sustainability and high quality water supply for industrial,
agricultural and domestic usage.
People centered planning should be adopted where social, economic and
environmental consequences of an undertaking are given deserving emphasis. Therefore
social-economic as well as environmental impacts should be considered alongside the
often-emphasized physical and technical impacts.
What resources are needed for the study? What resources are available?
– Needed expertise; available expertise
– Needed time for baseline and experimental studies
– Remaining time before the project is supposed to begin
– Required funds to conduct the proposed study; available funds.
Baseline conditions: What are the significant features of the ecosystem presently?
– What is the current pattern of fluctuation in popular sizes for important species
(measured over sufficient time to characterize the range of variations)?
– Which species are playing a dominant or critical role in maintaining ecosystem
processes? What is their abundance, distribution and function of behaviour?
– What is the condition (Quality, quantity, dynamics) of physical resources of the
ecosystem?
– What are the major pathways of interaction between ecological components?
– What sources of stress from natural or human-induced sources already exist (fire, air
pollution, grazing etc)? With what intensity and periodicity do these stresses occur?
Predictions: What will be the major effects of the proposed action? What is known
from each of the following?
– Case studies: Extrapolation of effects from similar instances of disruption to the same
or similar ecosystems elsewhere.
– Modelling: Predictions from conceptual or quantitative models of ecosystem
interaction.
– Bioassay and Microcosm Studies: The effects of simulated disturbances on ecosystem
components under controlled conditions.
– Field Perturbation Studies: Response of a portion of the proposed project area to
experimental disturbance.
Water resources of arid areas 272
Estimation of likelihood:
– What is the probability of occurrence of the predicted events?
– How precisely can the magnitude and likelihood of impacts be estimated?
What steps could be taken to mitigate adverse environmental effects of the proposed
project?
– Could parts of the proposal be reduced or eliminated?
– Could expected damage be repaired or rehabilitated?
– Could ongoing management procedures be instituted to reduce damage?
Impacts of water development in arid lands of Southern Africa 273
REFERENCES
Chenje, M. & Johnson P. (eds.) 1996. Water in Southern Africa. SADC, IUCN SARDC, 238
Biswas, A.K. 1978. Environmental Implications of Water Development for Developing Countries,
Water Supply and Management Journal, 2, 283–297
Christelis, G. & Struckmeier, W. (eds) 2001. Groundwater in Namibia: An explanation to the
Hydrological Map, 128
Christiansson, C. 1981. Soil Erosion and Sedimentation in Semi-arid Tanzania: Studies on
Environmental Change and Ecological Imbalance. Uppsala: Scandinavian Institute of African
Studies and Department of Physical Geography, University of Stockholm.
Darkoh, M.B.K. 1989. Combating Desertification in the Southern African Region: An Updated
Regional Assessment. Nairobi: UNEP
Ellis J.E. & Swift, D.M. 1988. Stability of African Pastoral Ecosystems: Alternative Paradigms and
Implications for Development. Journal of Range Management, 41(6), 450–211.
Kaduma, J.D. 1972. Some Development and Economic Aspects of the Mindu Dam Project,
Morogoro: A Background Analysis for Decision Making, M.A thesis, University of Dar es
Salaam, 140
Kaduma, J.D. 1977. Man-made Lakes: Their Social, Economic and Ecological Impacts—The Case
in Tanzania, PhD thesis, University of Dar es Salaam, 400
Kitheka, J.U. 1993. Soil Erosion and Its Impacts on Surface Water Reservoirs: A Case study of
Nguu Tatu Catchment, NE Mombasa District, Kenya, Proc. 4th Land and Water Management
Workshop, Nairobi, Kenya, 309–329
Mubvami, T. 2000. Environmental Impact Assessment as a Policy Tool for Environmental
Management. IUCN-ROSA A Handbook on Approaches to the Environmental Policy Analysis in
Southern Africa. IUCN—The World Conservation Union.
Mutter, T., Topfer, J. & Wichterich, C. 2002. A Comprehensive Study of the Heinrich Boell
Foundation’s projects abroad in Political Ecology and Sustainability. 1st Ed. Heinrich Boell
Foundation.
Msangi, J.P. 1987. Conservation of Water Resources in the Semi-arid Areas of Tanzania. Journal
of Eastern Africa Research and Development. Vol 17, 63–73.
Msangi, J.P. 1991. Sustainability in Exploitation, Development and Management of Hydrological
Resources of Turkana District. Journal of Eastern African Research and Development, 21, 21–
39.
Msangi, J.P. 1992. Social-Cultural and Demographic Factors in Desertification Control in Kenya’s
Arid and Semi-arid Lands. Proc. of the Workshop on Desertification Monitoring, Assessment
and Control. Nairobi: National Environment Secretariat, 21–32.
Msangi, J.P. 1996. Social-Cultural Factors Affecting Non-Adoption of New Water Harvesting
Technology Among the Dryland Communities in East Africa. In Yue-man Yeung (ed) 1996
Global Change and the Commonwealth. Hong Kong Institute of Asia-Pacific Studies, Chinese
University, Hong Kong, 233–253.
Stone, J.C. (ed) 1991. Pastoral Economies in Africa and Longterm Responses to Drought.
Aberdeen: Aberdeen University African Studies Group.
Wood, A. Stedman-Edwards P. & Mang, J. 2000. The Root Causes of Biodiversity Loss. Earthscan,
399.
Water resources of arid areas 274
Westman, W.E. 1985. Ecology, Impact Assessment, and Environmental Planning. John Wiley &
Sons.
Institutional challenges for small towns’ water
supply delivery in Ghana
Kwabena Biritwum Nyarko
Kwame Nkrumah University of Science and Technology, Kumasi, Ghana
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
About 32% of the Ghanaian population do not have access to safe water (WDI, 2002)
making access to safe drinking water a challenge. The situation is even worse in the rural
an small communities where the majority of the population lives. In 1994, water supply
delivery in Ghana has been separated into urban water supply (served by a public utility,
GWCL) and Community Water Supply (CWS) under community ownership and
management (Nyarko, 2000 & CWSA, 2003a) to improve the supply of water to the
people of Ghana in a sustainable manner.
Community water supply consists of rural and small towns’ water supply. The small
towns’ water supply is a piped system, serving communities with inhabitants between
2,000 and 50,000 who are willing to own and manage the water system. As at the end of
2001 there were 254 small towns’ water supply systems under community ownership and
management (CWSA, 2003a). These systems are decentralised and do not enjoy any
cross subsidies and other benefits of economies of scale as the urban water supply.
Water resources of arid areas 276
Consequently, interest in the delivery of water services in small towns has grown rapidly
in recent years due to the peculiar characteristics of the small towns and the number of
inhabitants it serves.
To ensure the sustainability of the small towns’ systems are fundamental concerns,
which makes the study of institutional issues that affect the sector timely. This paper
examines the institutional framework of the small towns’ water supply component of
community water supply with the aim of enhancing the sustainability of the service
delivery.
revenue as a result of action by the DA, e.g., reduced tariff, etc., shall require that the
DA pay the difference in revenue into the WSDB account. The CWSA in
collaboration with the Regional Co-ordinating Council (RCC) shall ensure
compliance.
● The implementation of small towns’ water supply and sanitation projects shall be in
accordance with the regulations of the Environmental Protection Agency (EPA) and
Water Resource Commission (WRC).
The CWSA Guidelines for operations and maintenance (CWSA, 2003c) also stipulates
that:
● communities through tariffs shall undertake all major repairs and replacements. But for
total rehabilitation, cost sharing arrangement and procurement procedures shall be the
same as for new systems.
● Water produced shall meet Ghana Standards Board Drinking Water Quality standards.
The research approach utilised both quantitative and qualitative methods to gain insight
into the institutional issues of the sector. A literature review of small towns’ water sector
project documents was first conducted to get a thorough understanding of the sector.
Literature on institutional issues such as what the institutional framework should offer
was also reviewed. Based on the literature review, the research instruments (interview
guides and questionnaires) were developed for the various stakeholders, in the sector to
identify the main institutional issues and challenges.
Specific institutional analysis tools used were a combination of the Activity
Responsibility Matrix (ARM) and the Strength Weakness Opportunity Treats (SWOT) as
well as National Macro-environment Analysis. The data was collected from field visits
conducted in 20 small towns’ water supply systems, five district assemblies and three
regional CWSA offices. In addition discussions with 20 WSDB Chairmen and 22
Technical Managers who attended a short course at Kwame Nkrumah University of
Science and Technology, Kumasi, Ghana in August 2003 were used to validate the study.
Institutional challenges for small towns’ water supply delivery in Ghana 281
technical and financial) are well kept and that reports are submitted to DAs and CWSA.
In addition, when reports are prepared and sent to the DAs it hardly get comments from
the DAs. Majority of small towns does not perform routine water quality tests as
stipulated by CWSA. This has been attributed to low capacity at the DA level especially
the DWST to perform their function and inadequate resources at the DWSTs disposal for
their duties. The DAs is also expected to play the role of the Water Resources
Commission (WRC) at the district level in the areas of water abstraction rights and
permitting. This aspect is not yet operational at the DAs level.
The Ministry of Works and Housing (MWH) is responsible for policy making in the
water sector. The provision of infrastructure is the responsibility of the DAs, which is
under the Ministry of Local Government and Rural Development (MLGRD). CWSA as a
facilitating agency cannot force the DAs or the community (WSDB) to execute its water
related activities (eg. ensuring the submission of periodic reports, water quality
monitoring and using appropriate water tariffs). This is attributed to the following: lack of
effective accountability mechanisms between MLGRD/DA and CWSA; the location of
Institutional challenges for small towns’ water supply delivery in Ghana 283
CWSA and DAs in different ministries and the fact that the DAs is an authority on its
own.
To improve on the situation the following option are considered important. First,
establish a works department under the DAs structure with responsibility for
infrastructure (including) delivery so as to deal with CWSA. This has been accepted in
principle but its implementation is yet to start. Secondly, establish clear and explicit
accountability mechanism with benchmarks between the RWST and the DAs within the
region would help. The introduction of yard stick competition with incentives for the
DAs in a particular region or even nationwide would be useful to provide a check on the
DAs to enhance performance. It is important for the CWSA regional team to have good
collaboration with the RCC the appropriate institution to supervise the DAs activities to
enhance CWSA monitoring role of the NCWSP.
The roles and responsibilities of the WSDB demand certain skills, such as technical,
financial, managerial etc. The selection criteria initially specified gender and interest
groups representation, without mention of the skills required. The new draft policy dated
July 2003 adds that the membership of WSDBs shall exclude Traditional Authorities and
DAs, but where necessary, they may participate in WSDB meetings as observers
(CWSA, 2003b). This is laudable since there were interference from traditional
authorities. For example in one small town the WSDB chairman was the chief and his
nephew was also the treasurer.
In a number of communities political agitation and social tensions have affected the
membership, tenure and therefore the effectiveness of WSDBs. Some of the examples
are:
Water resources of arid areas 284
● In a small town (Bimbilla), after a meeting with all stakeholders to increase the tariffs
“a youth movement” managed to convince the District Chief executive (DCE) to
reverse the decision.
● In a small town (Bekwai), the District Security Council dissolved the WSDB in
response to a proposed demonstration threat by some community members. However,
the WSDB were re-instated after about six months.
● In a small town (Juaso) the chief requested for money for farming and was granted by
the WSDB. The DA got to know of it and demanded the money back and the WSDB
was dissolved. As at now (over 6 months) the new WSDB do not have access to their
Bank accounts because signatories have not changed.
At a recent training course organised at KNUST for WSDB treasures, majority of the
participants confirmed having illiterates on the WSDB and indicated that it affects
performance. They attribute that to the community sensitisation during project
preparation, which made them understand that the WSDB is the community
representative and that any one elected by the community could do it. As a result some of
the WSDBs members do not understand the issues, and this reflects in the system’s
performance. In cases where members have the required skills, performance has been
exceptional (Arthur, 2002).
The field visits also revealed that, WSDB perceives themselves as owners of the water
supply, which seems to explain why periodic operational reports are not sent regularly to
the DAs and CWSA. The DAs also do not have incentive mechanisms in place to
enhance the WSDB performance. Most or some of the WSDB do not have approved bye-
laws and hence do not have legal recognition.
At the national level the institutional linkages and the accountability mechanisms
between MLGRD/DAs, CWSA, MWH and the Water Resources Commission (WRC) are
weak. This results in ineffective monitoring and management of the small towns’ water
system as well as the water resources. At the district level the DWSTs members seconded
from the other department does not make the DWSTs permanent in the DAs structure
affecting delivery of water services. The policy on cost recovery is not clear especially on
the recovery of capital expenditure such as rehabilitation and major replacement.
Based on the conclusions, the following recommendations are made:
● At the national level there is the need for a closer collaboration between CWSA,
MWH, WRCand MLGRD through the inter-ministerial coordination and at the
regional level DAs, RCC and CWSA (regional office). In addition there is the need to
include accountability mechanisms in the Memorandum of Understanding (MOU)
between the DAs and CWSA.
● The Government of Ghana should speed up the process of establishing the Works
Department within the DA to strengthen the DWST position in the DA.
● CWSA should clarify the policy on cost recovery. Where there is room for subsidies,
the source of the subsidy and criteria for eligibility should be made explicit.
REFERENCES
Arthur, E. 2002. Manpower survey at the district assembly and community levels for small towns’
water supply.. BSc Thesis, Kwame Nkrumah Univ. of Science and Technology, Kumasi,
Ghana.
Asamoah, K. 1998. Ghana: The Community Water and Sanitation Project. Paper presented at the
Community Water Supply and Sanitation Conference at the World Bank, Washington, DC.
CWSA, 2000. The Project Operational Manual (POM) of CWSP-2, CWSA: 10–11.
CWSA, 2001. Small Towns Water supply and Sanitation Policy, CWSA.
CWSA, 2003a. Investment Opportunities in the Community Water and Sanitation sub-sector. A
presentation to an Americo-German Investors in Ghana, CWSA, pp 2, 8.
CWSA, 2003b. Small Towns Water and Sanitation Policy. Community Water and Sanitation
Agency, Ministry of Works and Housing, Government of Ghana.
CWSA, 2003c. Small Towns Water and Sanitation Policy. Operation and Maintenance Guidelines.
Community Water and Sanitation Agency, Ministry of Works and Housing, Government of
Ghana.
GOG, 1993. Act 462, Local Government Act, Ministry of Local Government and Rural
development. Government Printer, Assembly press, Accra.
Water resources of arid areas 286
GOG, 1998. Act 564, Community Water and Sanitation Agency Act, 1998. Government Printer,
Assembly press, Accra.
Jonah, E. 2003. Performance Assessment of Small Towns Water Supply System: The role of
management models an institutional structure, MSc Thesis, Kwame Nkrumah University of
Science and Technology, Kumasi, Ghana.
MWH, 1999. WATER. Comprehensive Development Framework. Ministry of Works and Housing,
Ghana.
Nyarko, K.B. 2000. Ghana Water and sanitation sector: Drivers for water performance. PhD
Proposal, IHE Delft, The Netherlands.
Sarpong Manu, K. 2001. PPIAF/CWSA PSP in Small Towns Water Study. CWSA, Sept 2001.
WDI, 2002. World Development Indicators, http://www.worldbank.org/data/dataquery.html
Socio-economic performance of Sepeteri
irrigation project in Nigeria
O.O.Olubode-Awosola & E.O.Idowu
Department of Agriculture Economics, Obafemi Awolowo University, Ile-
Ife, Osun State, Nigeria
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
ABSTRACT: This study tried to assess the economic, social and financial
viability of irrigation the project and examine the efficiency of resource
use among the project farmers. Data from all the Sepeteri Project’s
farmers during the 2001/2002 seasons were used. Factor share approach
was used to examine the resource use efficiency among the farmers.
Records of the project’s activities from 1995/96 to 2001/2002 seasons
were summarised to some socio-economic performance indices. All the
farmers perceived irrigation fee cheap however, 77% attributed their low
demand to lack of credits. The irrigation service was acceptable to the
farmers with ease of collection. About 67% of farmers do not accept
responsibility of making the project a success. Farmers were not efficient
in resource use. It was concluded that while the irrigation fee is far below
its economic value, it is high enough for the farmers and this prompts
them to prefer rain-fed to irrigated cropping. The project was not
financially viable due partly to insufficient funding and low level of
demand from farmers.
1 INTRODUCTION
Up to 1960s, Nigeria was almost self-sufficient in staple food crops from the relatively
abundant rainfall. However, from the 1970s, the long drought; the resulting recurrent
desert encroachment and the substantial rate of increase in population brought set back
Nigeria’s agriculture. The emergence of these three phenomena necessitated public
investment in formal irrigation. Irrigation involves development of water resources,
conveyance and distribution of water supply at the field coupled with necessary water
management exercises (Chukwuma 1993). River Basin and Rural Development
Authorities (RBRDAs)’ Irrigation projects were established between 1973 and 1979 to
cover every part of the country. They were to carry out a number of functions particularly
the development of irrigation infrastructure in their respective areas of operation.
Consequently, thousands of tons of crops such as tomatoes, groundnut, wheat, cotton,
Water resources of arid areas 288
millet, maize, etc, were grown by small-holder and commercial farmers. However, the
overall performance of the existing irrigation facilities had been on a decline owing to a
combination of technical, socio-economic and institutional factors (Nwa 1993). The
projects are saddled with inadequacy of untimely funding (Akinkoye 2001) hence, not
self-sustained.
Consequently, the 1995–97 Corporate Plan mandated RBRDA to generate funds
internally to cushion the dwindling funds from the budgetary allocations in order to meet
substantial portions of their recurrent costs. Ordinarily, a guided increase in agricultural
commercialization leads to increase in purchase of farm inputs, scale (farm size) and
specialization in farm enterprises and changes in the role and nature of farm labour
inputs. However, this move, if not guided, has the tendency to expose the RBRDAs to
been more concerned with the activities that return highest internally generated revenue
and possibly less concerned with meeting the irrigation needs of the intended
beneficiaries is vital.
To this end, while most research efforts on improving the performance of public
irrigation projects have focused on the structure, technology and environmental issues,
this research focused on social and economic performance of the system as well as the
resulting effect on its sustainability and achievements of its statutory functions in the face
of commercialization and eventual privatization.
The broad objective of this study is the assessment of the socio-economic performance
of Sepeteri Irrigation Project. The specific objectives are to:
i. Examine the socio-economic characteristics of the farmers;
ii. Examine the factors that affect demand for and supply of irrigation services in the
project area;
iii. Assess the economic, social and financial viability of the irrigated cropping and;
iv. Determine the efficiency of resource use among the irrigation project farmers.
2 CONCEPTUAL FRAMEWORK
Farmers are primarily concerned with the profitability of their enterprises at individual
farm levels. Hence, economic performance of irrigation project farmers is based on a
production function and viewed in terms of the efficiency with which farmers combined
irrigated plot with other resources such as labour, fertilizers, etc., in the context of
institutional framework and management practices of irrigation projects. Farmers are
expected to meet economic optimum criteria by adjusting inputs and outputs to relative
prices. According to the concept of marginal productivity, a rational producer allocates
each variable input according to its market price. This concept is supported by the theory
of equilibrium in factor markets under profit maximization, which requires that a factor
input be paid its value of marginal physical product (VMP). If a factor is paid higher than
its VMP, it is over-utilized and if otherwise it is under-utilized (Henderson & Quandt
1980). However, public irrigation projects in developing nations like Nigeria usually do
not meet the conditions for competitive market analysis as explained above because its
outputs are natural resources (i.e. land and water), which are developed for national
economic efficiency and development. According to Schreiner et al. (1989), such a
project is characterized by concepts like natural monopoly, derived demand, etc.
Socio-economic performance of Sepeteri irrigation project in Nigeria 289
Therefore, it is quite possible for project to record negative returns to the agency
management because of high cost of capital, yet farmers are expected to make profit
consistently.
Therefore, rational irrigation policy should ensure that the only sound reason for fixing
irrigation fee is the net additional benefits it offers. These benefits accrue to the region
and the society as a whole. The major impact of these benefits is to be found in land use,
employment, cropping pattern, farm inputs, etc. So, priorities are always given to these
benefits above financial returns accruing to the government from irrigation fees.
3 LITERATURE REVIEW
efficiencies respectively. Both revealed inefficient water use. The report is that size of
irrigated farms and unit cost of water are factors to be considered in the studies of water
use and management. Mandal et al. (1995) in an attempt to examine resource use
efficiency with respect to farm size of Irrigated HYV Boro Rice Cultivation in
Mymensingh District of Bangladesh used factor share approach by estimating a Cob-
Douglas production function. The study revealed that no farm size group allocated
resources efficiently.
4 METHODOLOGY
irrigation service. Index numbers were used to summarize series of figures over years.
The indexes show how much one-year figures differ from another. Usually, a fairly
typical year’s figure is taken as a base year figure and others are compared to the base
year’s figure. The commonest application of index numbers is in comparison of a series
of annual figures. In this study, 1995/96 irrigation season was taken as a base year such
that subsequent figures were compared with 1995/96 figures. This year was chosen as a
base year to examine how fare have the projects been performing in irrigation services
since commercialization move in 1995. This year was assumed to be a logical base year
for evaluating performance of a previously public irrigation project. To observe the
average annual percentage change over the years, the average annual percentage change
was computed as follows:
(1)
where Indexi=base year Index; indexl=last year index and n=number of years over which
the trend is studied (Harper, 1991).
To examine the performance of the irrigation projects, performance indices according
to Bos (1997) were used. The indices used included Fee Collection Performance, Relative
Water Cost, Users Stake in Irrigation System, Financial Self-Sufficiency and Relative
Cropping Profit Indices were used to assess the operational and strategic performances of
the project management agency. The indexes are specified below:
(2)
(3)
(4)
Water resources of arid areas 292
where Actual Income=Total internally generated revenue from irrigation related services
and Total MO+M Expenditure=total expenditure on irrigation related services. Financial
Self-sufficiency Index reveals the financial viability of the project.
(5)
where Irrigation Cost per ha=cost of irrigation service per ha ( ) and Total
production cost per ha=average total cost of irrigated cropping ( ). Relative Water
Cost Index reveals the tendency of farmers abandoning or continuing with irrigated
cropping. It is computed on the average. However, The Relative Water Cost Index is
perceived to be inadequate to measure the tendency of the farmers abandoning or
continuing with irrigated cropping since for some farmers in the developing nations, ends
justify the means i.e. (Alimi, pers. comm.) they consider profit far more than the cost. It
will then be modified to incorporate the ends, profits from irrigated and rain-fed
croppings as specified below:
(6)
To determine the efficiency of resource use by the respondents, Ordinary Least Squares
(OLS) technique was used to estimate parameters of explanatory variables in the
postulated Cobb-Douglass production function. The marginal values of inputs used were
computed indicating the proportion by which value of crop output changed with one per
cent change in the quantity of each input when the quantities of other inputs were kept
constant. The production elasticities of the inputs were added together to obtain the
returns to scale indicating the proportion by which value of crop output changed with one
percent change in the quantities of all the inputs. It is assumed that the value of output
depends on level(s) of input(s) such as land, labour, capital and management used and
that the production function is a one-equation model (Ogunrowora et al. 1979; Omotesho
et al. 1993; Ayanwale 1995). Thus the production function for project farmers was
specified as follows:
Y=f(X1, X2, X3, X4,
u) (7)
where Y=Total value of crop output ; X1=size of irrigated farm plot (ha);
X2=expenses on fertilizer and other agro-chemicals ; X3=number of farm household
members that assisted in farming activities (man-day); X4=amount spent on hired labour
and u=error term.
Crop output (Y) was measured in monetary term because two crops—Vegetable
(Amaranth sp.) and Okro (Abelmoscus esculentus) were grown together without
measuring for sale in standard unit like kilogram. The log-log stochastic production
function was fitted for the respondents’ values of crop outputs as follows:
ln Y=ln α0+α1 ln
X1+α2 ln X2+α3 ln (8)
Socio-economic performance of Sepeteri irrigation project in Nigeria 293
X3+ln α4X4
where VMPi=value of marginal physical product from using additional unit of input Xi;
MPPi=marginal physical product from using additional unit of input Xi and Py=market
price of the output. So the Allocating/Pricing Efficiency Index is given as
(10)
where MICi=marginal input cost of input i. (i=1, 24). a priori, the expected signs of the
explanatory variables are positive.
revealed that the indigenes within the Project area participated well in the project. About
73% of respondents were within age bracket of 31 and 50 years while about 18% are
above 50 years. Also, women scarcely participated in the projects. About 91% of farmers
were male. The reason may be that the community is a typical agrarian community where
men are predominantly engaged in farming and women engage in other economic
activities or assist the male household heads in farming operations.
About 16% had no formal education while none had tertiary education, while 77% had
between primary school and secondary school education. All the Project farmers had
above 5 years of farming experience. In fact, about 37% had above 20 years of farming
experience. This result justifies locating the project in the area to help agricultural
development and also support the tenet that land should be allocated to farmers with
proven commitment to farming as a career. In the same vein, the project farmers had long
years of participation. Above half had been with the project for over 6 years. However,
few (about 11%) of farmers were full-time farmers. Majority, 88.6% were involved in
other economic activities. About 77% acknowledged credit availability as
Socio-economic performance of Sepeteri irrigation project in Nigeria 295
limiting factors of irrigated plot size while 11% perceived irrigation service fee as high
while others expressed engagement in other activities as a constraint.
6 CONCLUSION
In conclusion, there is higher level of participation from the neighbouring villages to the
project. They were mostly resident farmers. Also, they were mostly male with low level
of formal education but they had long years of farming experience and project
participation. Majority was aged between 31 and 50 years and above. Most of them were
part-time farmers and attributed their small level of irrigation participation to lack of
Water resources of arid areas 298
credit facility. Inadequate funding and deteriorating structures hindered the level of
irrigation supply of the project.
The performance indices revealed that the irrigation service is acceptable to the
intended beneficiaries. Higher cost per hectare of irrigated cropping connotes tendency
among the farmers to abandon irrigated cropping for rain-fed cropping. However, the
farmers do not accept responsibility of making the project a success. The project is not
financially self-sufficient partly for insufficient funding for operations and partly low
level of demand from farmers. The farmers were inefficient in resource use. The results
call for reform in management of irrigation system such that privatization efforts should
not tie down peoples’ land unused; Specifically farmers should be encouraged into
coordinated and recognized WUA that incorporate credit lending and efficient marking.
Further research should look at ways to rationalize family labour. In the same vein, on
hired labour other means of bargaining should explored.
REFERENCES
Adegbola, A.A. & Akinbode, I.A. 1986. A review of old and current agricultural development
schemes in Nigeria: Lessons for future programme designs. In Agricultural Development in
Nigeria. Ife Journal of Agriculture special publication, 8:1–34.
Akinkoye, O. 2001. An overview of organization and management of public sector irrigation
schemes. Paper presented at the National Workshop on Participatory Irrigation Management
organized by National Agricultural Extension and Research Liaison Services (NAERLS),
Ahmadu Bellow University, Zaria in collaboration with the Department of Irrigation and
Drainage, Federal Ministry of Water Resources, Abuja. 26–30 March 2001:12pp.
Alimi, T. (Personal communication).
Ayanwale, A.B. 1995. Resource use efficiency in cassava processing in Oyo North Area of Oyo
state, Nigeria. Ife Journal of Agriculture 16, 17:123–135.
Bos, M.G. 1997. Performance Indicators for irrigation and drainage. Irrigation Drainage Systems
11(2): 119–137.
Chukwuma, G.O. 1993. Some Considerations in Developing Irrigation Research Priorities for
Nigeira. Proc. National seminar on Irrigation Research Priorities for Nigeria held at the
University of Ilorin, Nig. 20–23 April 1993:65–71.
Henderson, J.M. & Quandt, R.E. 1980. Microeconomic Theory: A mathematical Approach 3rd ed.,
McGraw-Hill Kogansha Ltd. Japan, 420pp.
Herpar, W.M. 1991. Statistics, London, Pitman Publishing: 501pp.
Kolawole, A. 1988. RBRDAs and vulnerability to hunger in Nigeria, the case of the South Chad
Irrigation Project. Food Policy 13(4):389–396.
Krishna, R. 1963. Farm Supply Responses in India-Pakistan: A case study in the Punjab Region:
Economic Journal, Sept, 1963.
Kwanashie, M.A., Togun, A., Ajobo, O. & Ingawa, S.B. 2000. Nigeria Water Resources
Management Strategies—Economic and Financing. Technical Report, 16pp.
Mandal, K.C., Sabur, S.A. & Molla, A.R. 1995. Resource use efficiency of irrigated HYV boro rice
cultivation by difference farm size groups and its impact on employment and distribution of
income in DTWII project area of Mymensingh Bangladesh J. Agric. Econs, 8(1):71–87.
Moore, M.R., Gollehon, N.R. & Carey, M.B. 1994. Multi crop production decisions in western
irrigated agriculture: the role of water price, American, Journal of Agricultural Economics,
76(4):859–874.
NCAER (National Council of Applied Economics Research) New Delhi 1959. Criteria for fixation
of water rents and selection of irrigation projects, London, Asian Publishing House.
Socio-economic performance of Sepeteri irrigation project in Nigeria 299
Nwa, E.U. 1993. Irrigation Research Priorities for Nigeria. Proc. National Seminar held at the
University of Ilorin 20–23 April, 1993, Ilorin, Nigeria, Nwa, EU, Pradhan, P. (eds) IIMI, 104pp.
Ogunfowora, O., Esang, S.M. & Olayide, E.O. 1979. Resource productivity in traditional
agriculture: a case study of four agricultural divisions in Kwara State of Nigeria. Journal of
Rural Economics and Development 9(2):119–131.
Omezzine, A. & Zaibet, L. 1998. Management of modern irrigation systems in Oman: allocative vs.
irrigation efficiency. Agricultural Water management, 37(2):99–107.
Omotesho, O.A., Olufe, J. & Oladeji, S.O. 1993. Resource productivity in food crop production in
some selected villages of Oyi Local Government Area, Kwara State, Nigeria. Ife Journal of
Agriculture 14(15): 90–97.
O-ORBRDA 1998. Federal Republic of Nigeria, Ogun-Oshun River Basin and Rural Development
Authority 1997 Annual Report, January, 1999. 49pp.
Schreiner, D.F., Badger, D.D., Welsh, M.P. & Suprato, A. 1989. Policy Applications in Natural
Resource Projects. In Agricultural Policy Analysis Tools for Economic Development (ed)
L.Tweeten, London, Westview Press, 279–321.
Theme D:
Application of geophysical,
GIS, and remote sensing
techniques
Mapping vegetation for upscaling
transpiration using high-resolution optical
satellite and aircraft images in Serowe,
Botswana
Y.A.Hussin1, D.C.Chavarro1, M.Lubczynski1 & O.Obakeng1,2
1
International Institute for Geoinformation Science and Earth
Observation (ITC), Enschede, The Netherlands
2
Geological Survey of Botswana, Lobatse, Botswana
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
As part of the water cycle in the nature or what is well know as the hydrological cycle,
surface water is heated by solar radiation and thus evaporated to the atmosphere.
However, water in plant is emitted through leaves by a process called transpiration.
Vegetation cover is a major component of the hydrological cycle. It has influence on the
hydrology of both ground and surface water and on soils. The illogical use or abuse of
natural vegetation can have a major effect and consequently changes the hydrological
cycle and produce adverse effects. These effects can be very dramatic in arid and semi-
arid regions. In these regions the climate is very sever with extremely hot temperatures
up to 50°C and high evaporation and low annual rainfall. In Botswana, just like any other
semi-arid country, water is scarce and surface water is available only for short periods
after the rainy season. The main water supply is groundwater. Vegetation in Botswana, as
Mapping vegetation for upscaling transpiration 303
Forests cover large areas of the global land surface. For many developing countries, it
represents an important income source for their economies. Due to over exploitation,
forests are currently under constant risk. The protection of forests from disasters (e.g.
fire, disease, erosion, deforestation, over grazing) over extensive area is difficult without
having any information such as condition, area, species, age classes and volume. With
these types of information, it is possible to make a proper management of the forest by
identifying and selecting the appropriate area for different management purposes, such
as, harvesting, protection, etc.
Having all these information collected, there is a need to store the referred information
properly, for better and comprehensive use. For this reason, forest maps play an
important role in organizing gathered information for further strategies and policies
determination in order to make the best use of forest.
Different approaches for mapping, like land survey, aerial photographs and satellite
imagery can be used depending on the level of detail required and the extension of the
area under study. For wide areas, satellite imagery has been shown effective for forest
classification and consequently mapping.
It is recognized that different satellite or airborne imagery can give different results in
terms of information extraction. These different results relate to differences in spatial and
spectral resolution. Vegetation mapping involves the evaluation of the existing data and
information, collecting field data or ground truth, analysing the data and finally
developing the vegetation map and validate it (USGS, 1994). Riquene (2002) have
studied the vegetation condition of the current research area using Landsat-TM images
and ASTER Optical scanner sensors. The study concluded that ASTER images resulted
in better vegetation map than TM because of its higher spatial resolution of the 15 meter
than TM of 30 meter. Further more the results showed that ASTER data gave more
Water resources of arid areas 304
reliable vegetation maps than Landsat TM data. Mapanda (2003), following the steps of
Riquene research, who used high spatial resolution (4 meters) multi-spectral satellite
images of IKONOS in comparison with ASTER images of 15 meter spatial resolution,
have concluded that IKONOS images gave much better results in mapping the vegetation
cover of Serowe, Botswana.
This research is going one step further in using higher spatial resolution of multi-
spectral images. These images were acquired using Tetracam multi-spectral digital
camera. The camera was mounted on a small aircraft and has collected images in 30cm,
60cm and 100cm spatial resolution. This paper is presenting the first preliminary results
of the use of the Tetracam airborne multi-spectral digital images for mapping tree-bush-
shrub-grass-savannas and woodlands vegetation cover of part of Serowe, Botswana.
These images will hopefully be used for up-scaling transpiration of the vegetation in this
area in conjunction with IKONOS MS data.
3 STUDY AREA
The study area is located in the Central District, about 275km NE of Gaborone the capital
of Botswana. Topography is gentle, which varies from 1060 meters above sea level to
approximately 1240. It is characterized to be lower in the east and southeast of the region,
and the highest location in the vicinity of the escarpment edge. From these ones the
average slope is 5% and it gradually decrease to less than 1% towards the east and
southeast.
Soils units, which can be found in that region, are related to arenosols, regosols,
lixisols, luvisols and vertisols. Arenosols are the most common soil units in the study
area. It has low moisture retention capacity than the other soil units.
Climate is a semi-arid with a mean annual rainfall of 447mm. Rainfall occurs mainly
in the summer fallowed by a dry winter season. Summer season stretches from October to
April and the winter begins in May to September (Tyson, 1986) (Obakeng, 2000).
Main vegetation type is thought that belong to the Northern Kalahari Tree and Bush
Savanna. Trees are mostly of Acacia species, which are characterized by the marked
tendency to occur in cluster, and are normally accompanied by a variety of grass species
such as Ariatida and Eragrotis. Vegetation communities are determined by location on
either sandveld or hardveld areas. Dense vegetation is found within and along river
courses. This suggests that the vegetation density is governed by the availability of water,
which may be partly controlled by topography and geomorphology (Obakeng, 2000).
The airborne multi-spectral data was collected using TETRACAM multi-spectral digital
camera, which collects its data in three spectral bands namely red, green and near
infrared. The data is collected in a rectangular frame of 1280×1024 pixels. The size of the
pixels (e.g. ground resolution) would depend then on the altitude of the aircraft above the
ground. The camera would saved the image in DCA format (Digital Camera Format),
which is a compressed file that can be un-compressed and transferred to Bitmap format
Mapping vegetation for upscaling transpiration 305
that can be imported to any image processing software. The Airborne data was collected
in three different spatial resolutions 30, 60 and 100cm.
An area of 10×10km was selected as a study site. It is located in the Hardveld part of
the Serowe terrain, on which two multi-spectral IKONOS satellite scenes of November
2001 and February 2002 were collected. These images are collected with 4 spectral bands
(blue, green, red, and NIR). The spatial resolution of these data sets is 4 meters. A MS
image, so called panchromatic sharpen, was available too. The Pan-Sharpen image is a
MS image fused with the Panchromatic image of IKONOS, which has 1 meter spatial
resolution. Thus the spatial resolution of the new MS image will be improved to 1 meter.
For this study area, two aerial surveys were implemented to collect the multi-spectral
digital camera data. The first aerial survey was done in November 2003 and the second
one was done in February 2004. These surveys were designed and implemented using
Aerial-Photography types of survey. The survey divides the area into flight lines. Within
each flight line, images were collected with a front overlap of 20% and a side overlap
between flight lines of 20% too. The following data where collected:
1. 30cm spatial resolution: 39 flight lines with a total of 910 images
2. 60cm spatial resolution: 21 flight lines with a total of 333 images
3. One meter spatial resolution: 14 flight lines with a total of 187 images.
A qualitative approach was used in the analysis of airborne multi-spectral digital images,
which mainly involves visual interpretation, spectral signature measurements, spatial
features measurements and comparisons of different spectral and spatial data resolution.
The same approach was used with the MS IKONOS satellite data to be compared to the
airborne data.
The results presented in this paper are the first preliminary findings of some exploratory
analysis of airborne MS digital images. Supervised classification and accuracy
assessment was not done because an organized fieldwork to collect ground truth was not
done yet in the study area. The authors are planning for one in early May of this year
(2004).
A spectral signature analysis of the 30cm, 60cm and one-meter spatial resolution
images of the MS digital Tetracam airborne camera using the digital interpretation of the
false color composite, unsupervised classification, and Normalized Difference Vegetation
Index (NDVI) (Figures 1–6) showed that the sample used has 3 different spectral classes
which referred to 3 different species of the bushy vegetation in the area. It also showed
two high contrasted spectral classes, which refer to a soil and a grass classes the selected
sample shown in the mentioned figures.
A comparison of the above findings with the same signature analysis to the MS
IKONOS image of improved one-meter spatial resolution showed similar results of 3
spectral classes of the bushy vegetation and two other high contrasted classes
representing the soil and grass classes.
Water resources of arid areas 306
reflectance from the canopy of the bush. While when using higher resolution (e.g. 30 and
60cm) the representation or the appearance of a small bush will be possible.
Therefore, as the spatial resolution increase the spatial information extracted about an
object on the images (e.g. a bush or a tree) will consequently increase. The information
will include the surrounding area (e.g. soil or grass) of the tree or the bush targeted. This
means that using higher spatial resolution we can define the size and shape of any tree or
bush much accurate than using lower resolution. The size of the crown of a bush or tree is
effecting the estimation of the transpiration of that bush or tree. Consequently this will
effect the process of up-scaling transpiration from the area in general.
As far as the spatial resolution is concern, a general comparison of IKONOS images
and the airborne MS image showed that the information extracted from the 30cm and
60cm spatial resolution images of the airborne MS images is much better than the one-
meter resolution of IKONOS image.
6 CONCLUSIONS
ACKNOWLEDGMENT
This research work was partly supported by the internal research fund of GWFLUX
Project at ITC. However, Botswana Geological Survey (BGS) has offered the main
financial support of the aerial survey missions, fieldwork logistics and transportation. The
authors appreciate and acknowledge the support of Botswana Geological Survey.
REFERENCES
Le Maitre, D.C., Scott, D.F. & Colvin, C. 2000. Information on interactions between Groundwater
and Vegetation relevant to South African Conditions: A review. Groundwater: Past
Achievements and Future Challenges, Silili et al. (eds). Balkema, Rotterdam, 959–962.
Mapanda, W. 2003. Scaling-up and Mapping Transpiration Using Remote Sensing and GIS: A Tool
for Water and Forest Management. Unpublished MSc, ITC—International Institute for
Geoinformation Science and Earth Observation, Enschede.
Obakeng, O.T. 2000. Groundwater recharge and vulnerability: A case study at the margins of the
south-east Central Kalahari Sub-basin, Serowe region, Botswana. Unpublished MSc, ITC—
International Institute for Geoinformation Science and Earth Observation, Enschede.
Water resources of arid areas 312
Riquene, A.H. 2002. Vegetation mapping in Arid Zones: A multi-sensor analysis, the relationship
between Vegetation Distribution and Environmental Factors: A case study in Serowe,
Botswana. Unpublished MSc, ITC—International Institute for Geoinformation Science and
Earth Observation, Enschede.
Tyson, P.D. 1986. Climatic Change & Variability in Southern Africa. Cape Town, South Africa:
Oxford University Press.
USGS, 1994, 19 July 2001. Field Methods for Vegetation Mapping. USGS-NPS. Available:
http://%20biology.usgs.gov/npsveg/fieldmethods/ [2001, 24 August 2001].
Gravity study on groundwater structure in
Central Butana (Sudan)
K.M.Kheiralla
TU Bergakademis, Freiberg, Germany
A.E.Ibrahim
El Neelain, University, Sudan
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
anomaly (Around Jebel Mundara) lies partially in the Basement rocks. Therefore this
research aims to verify the existence of these anomalies and clearly define their
extension. In addition, the study aims at determining the type and the dimensions of the
rocks causing these anomalies.
The study area lies between the Blue Nile River and River Atbara and occupies the
Central Butana area, It is bounded by latitudes 14°60′N and 15°80′N and longitudes
34°00′E and 35°20′E. The study area covers about 10,000km2 (Fig. 1).
The study area is generally flat, with a gentle slope to the North. The general altitude
of the plain is about 500m above mean sea level (m.s.l). The flat monotony of the plain is
occasionally broken by some protruding low to moderately high hills or hill chains,
which hardly exceed 200m above the ground surface.
The climate of the area is arid to semi arid zone of Sudan, characterized by a short
duration of a rainy session in summer (July–September), and along dry season for the rest
of the year. The average annual rainfall is about 200mm. Average annual temperature
over the Butana area is around 40°C in summer (March–October) and 25°C in winter
(November–February).
A number of isolated out crops of Precambrian Basement complex rocks scatter over the
central plains of Sudan. In the Butana region, several hills occur and make prominent
landmarks within the predominantly flat clay plains. The solid geology of the Butana
plain is rather concealed under a veneer of variable thickness of superficial clays
covering over 70% of the Butana area. Several metasedimentary sub-parallel belts extend
for 10–25km in the form of low to moderately elevated ridges surrounded by Butana clay
cover (Iskander et al., 1993). The generalized geological column for Butana region can be
summarized as shown in Table 1.
The majority of the Basement rocks are concealed under the cover of the Butana clay
plain. As mentioned, the structural domain in the area is characterized by northeast
trending lineaments (Fig. 2). Exposed structures in the metasediments display complex
shearing/faulting and tight folding with dipping axial planes where the axes generally
have NE-SW trends and with SW mergence Iskander et al. (1993).
3 GRAVITY METHODS
3.1 Introduction
The gravimeter used in this survey is a Lacoste & Romberg gravimeter, model D108,
which has calibration constant of 1.0863mGal/div and 200mGal measuring range. A total
of 200 gravity readings were measured in the study area following a loop survey, with a
spacing of 2km between stations (Fig. 3). An area of approximately 2412km2 has been
covered. In addition, about 275 gravity points (approximately 2475km2) acquired, by Sun
Oil Company were compiled, in the study.
Water resources of arid areas 316
Elevation of the gravity station was determined by Global Positioning System (Garmin II
GPS Model 1999), with ±10m accuracy. The Global Positioning System, (GPS) device
was used to determine the altitudes of the gravity stations and also for navigation. Control
was provided by the available altitude benchmarks in the area and by elevation contour
maps of scale 1:100,000, determined by the Survey Department in Khartoum. In addition
the altitudes of the previous data (Sun Oil data), were determined by micro barometric
altimeter.
In gravity measurements effects are produced by sources not directly related to the
geological objectives or interest of the study, therefore certain reductions or corrections
are necessary to remove these effects. Such corrections include drift, tidal, latitude and
elevation correction.
This compensates for the earth tides, generated by the complex gravitational
interaction on the Earth by the Sun and Moon. These effects are often calculated from
table published by Geophysical Journal worldwide. In this work it was done by
“GEOSOFT” computer program.
(0.4) is consider as the optimum grid cell size suitable for production of the Bouguer
gravity map.
Contouring of the observed gravity data was performed by “CONTOUR” program of
the GEOSOFT package, whose basic function is to thread contour lines through constant
levels, defined in a gridded GEOSOFT data file (GEOSOFT manual, 1989).
Where, G is the observed gravity, R is the residual field, Z is the regional field, a, b and c
are constants. The condition for the above equation is that
∑R2=min
These are called residual, but do not posses any relationship with local anomalous mass
(Paul, 1967). The only physical significance of this residual is their proportionality to the
second vertical derivative value; hence their zero contours coincide with the zero
contours of the second derivatives (Nettleton, 1976).
Lately dependable method have been introduced by Paul (1967), for computing the
second vertical derivatives, consequently the residual determined by the previous method
loose much of their significance.
Hence,
Where, Z0 is the average regional at the canter of a particular grid, is the average of
the residuals around a circle of radius, “n” on substituting,
Gravity study on groundwater structure in Central Butana (Sudan) 323
Therefore, for the second vertical derivative only the equation is given as
This equation clearly shows that, the regional field constitutes nothing to second vertical
derivative value, and thus this derivative represents entirely the residual value only.
A profile of second vertical derivative of gravity in a direction outward from the
center of a negative anomaly usually shows an outer maximum value and an inner
minimum value Bott (1965). The source of the anomaly may be determined by the
ratio r,
Figure (6) shows that the gravity lows over central part the south of Jebel-Mundara
Bouguer anomaly map has an amplitude of about −50mGal, a rounded-shaped strong
anomaly and of general contours strike which indicates that the study area is narrowly
oriented in a NW-SE structural domain. This anomaly is connected to another gravity low
over J.Qeili, which extends NW beyond the border of the study area.
The abrupt change in the trend direction of the anomalies from northwest (J.Mundara
and Qeili), to northeast (Wad Burwa), may be due to the existence of a structural path
that might have facilitated the emplacement of low-density bodies whose effects are
expressed by the low gravity anomalies. Figure (8) shows a fault or lithological contact as
exhibited by the second vertical derivative profile.
some attention has been drawn to the densities and density contrast between the
representative rocks in the study area. In fact the density contrast between the rocks are
the primary cause of the measured gravity effect.
Densities of main rocks composing Butana (Igneous complex) had been measured by
Ahmed (1968) and range from 2.57 to 2.68gm/cc. A value of 2.65gm/cc is considered as
Water resources of arid areas 326
the average for the whole complex. Densities of the country rocks vary from 2.65 to
2.98gm/cc, and a value of 2.81gm/cc is considered to be the average.
Generally densities of igneous rocks, increase with decrease of silica content or in
other words, it follows the acidity line regardless of the rock being plutonic or volcanic
(Table 2). On the other hand densities of the metamorphic rock increase with the degree
of metamorphism.
account for this low (−ve) gravity anomaly a model representing a granitic intrusion into
the green schist with a density contrast of −0.16gm/cc has been simulated as shown in
Figure (9a). The model (Fig. 9a) shows intrusive granitic mass of a thickness of about
2km into the green schist, and bounded by several step faults, especially at SE side.
4 CONCLUSION
The interpreted gravity data in this research were measured to confirm the existence of
low gravity anomalies in the area that is referred to as Wad Burwa anomaly (Ibrahim,
1993) or otherwise.
Filtering of the gravity data comprises the second vertical derivatives of the gravity
anomaly. The techniques have proven to be effective in revealing local features more
clearly than their respective potential fields. The vertical derivative enhanced and
resolved the regional-residual anomalies more clearly. on the other hand have delineated
density boundaries (lithological boundary).
Modeling of the anomalous field was performed by gravity inversion program
(Cooper, 1991), which simulate two-dimensional geological model of irregular geometry,
mostly representing the mode of occurrence of the granitic into the country rocks of the
Butana region. The surface outcrop of granite controls the shape of the uppermost part of
the model, while the −ve density contrasts between the granitic rock and the other green
schist account for the −ve Gravity lows in the study area, which are generally known to
be shallow Basement Complex terrain.
This gravity low is largely attributed to the occurrence of low-density rocks (granitic
intrusion) into the high-density rocks (green schist) of the Butana region. Thus for it is
not unusual gravity measurements in such areas introduced by granitic into country rock
of higher densities, to reflect apparently anomalies of typical sedimentary basins.
However such ambiguity can be resolved by computation of the second derivative (Bott,
1965). By verifiable of their occurrence and contact with host rock, granitic bodies slope
outwards, thus display (−ve) second derivative gradient as shown in section 3.7.2.1 of
this thesis.
On the other hand the gravity high zones un-ambiguity coincide with the areas of
known shallow Basement Complex. As has been revealed by the quantitative
interpretation, the granitic bodies are bounded by near-surface (shallow) step faults. The
appearance of these faults on the ground surface is completely masked by the Butana clay
Water resources of arid areas 328
soil and probably they can be exposed by intercepting deep cut water courses (Wadi or
Khor). In such conditions the faults can channel water, from surface runoff, to facilitate
occurrence of groundwater storage in the basement (granitic rocks) in the Butana area.
Filed work for ground water trotting and checking of interpreted data is crucial to up
grade the quality of decision. Concurrently geophysical and geological investigation
could be carried out in quest to acquire more precise data in the study area (Wadi and
flood delta, etc.).
The study area of the Butana region has a lack of water supply for both population and
livestock uses, because Basement Complex, which is outcropping, or of shallow depth,
dominates it. However, further more detailed gravity work in addition to the seismic
survey may lead to reveal saturated depression or fracture zones. The presence of water
supply in these zones can be expected.
REFERENCES
Abd el Rahman, E.M., Yehia, A.Y. & Amin, Y.A. 1983. Methods of determination of the proper
regional gravity from Bouguer anomaly profile. E.G S. Proc. of 2nd Ann. M.
Agocs, W.B. 1951. Least squares residual anomaly determination. Geophysics, 16:686–696.
Ahmed, F. 1968. The geology of the Jebel Qeili, Butana and Jebel Sileitaat-Es-Sufr igneous
complex, Nile valley, Central Sudan. Unpublished M.Sc. thesis, Univ. Khartoum.
Ahmed, F. & Ayed, M.A. 1996. Applied geophysical and satellite imagery techniques, for ground
water studies in Central Butana area; ADS report, 25pp, 10–15.
Bannister, A. & Raymond, S. 1989. Surveying Catalog, Singapore, Longman Scientific and
Technical.
Bott, M.H.P. 1965. A geophysical study of the granite problem. Quart. Journ. Geol Soc. London,
112(445): 45–62.
Cooper Ltd. 1991. Program “Grav 2dc”, written by G.R.J. Dep. Geophysics, Witwatersrand, South
Africa.
Dobrin, M.B. 1976. Introduction to Geophysical Prospecting. Mc Graw-Hill, New York.
El kins, T.A. 1951. The second derivative method of gravity interpretation. Geophysics, 16:29–50.
Fajklewicz, Z. 1959. The use of Cracovian Computation in estimating regional gravity. Geophysics,
15:&QJ;465–478.
GEOSOFT reference manual, 1989. Software for earth sciences. GEOSOFT INC, Toronto,
Canada.
Griffin, W.R. 1949. Residual gravity in theory and practice. Geophysics, 14:39–56.
Ibrahim, A.E. 1993. Interpretation of gravity and magnetic data from the Central Africa rift system,
Sudan. Unpublished. Ph.D. Thesis Univ. Leeds, 209pp.
Ibrahim, A.E., Ebinger, C.J. & Fairhead, J.D. 1996. Lithospheric extension NW of the Central
Africa Shear Zone (CASZ) in Sudan from potential field studies. Tectonophysics, 255:70–97.
Iskander, W., Ahmed, A.A., Mokhtar, A. & Fadle, A.S. 1993. Appraisal of mineral and water
resources of central Butana, Eastern region-Sudan ADS report 85pp.
Kearey, P. & Brooks, M. 1988. An Introduction to Geophysical Exploration. Dep. Geol. Univ.
Bristol., 296 pp, ch-6, 138–169.
Nettelton, L.L. 1976. Gravity and magnetic in Oil exploration. Mc Graw Hill, New York, 464p,
138–169.
Paul, M.K. 1967. A method of computing residual anomalies from Bouguer gravity map by
applying relaxation technique. Geophysics, 32:708–719.
Seigal, H.O. 1995. High precision gravity guides. Canada, Ontario, L4K 1B5:120pp.
Sun Oil Company, 1984. Nile blocks gravity survey. Final report, Unpublished.
Gravity study on groundwater structure in Central Butana (Sudan) 329
Talwani, M.J., Worzel, L. & Landisman, M. 1959. Rapid gravity computations for Two-dimension
Bodies with application to the Mendocio submarine fracture zone. J. Geophys. Res., 64:49–59.
Remote sensing and electrical resistivity
studies on groundwater structure zones in
Central Butana (Sudan)
K.M.Kheiralla
TU Bergakademie Freiberg, German, Gustav-Zeuner-Str, Freiberg,
Germany
A.E.Ibrahim
El Neelain University, Sudan
Water Resources of Arid Areas—Stephenson, Shemang & Chaoka
(eds),
© 2004 Taylor & Francis Group, London, ISBN 04 1535 913 9
1 INTRODUCTION
Groundwater has become an important source of water and has played an importance role
in developing industry, agriculture, livestock and domestic purpose. The groundwater
condition in a crystalline rock terrain is multivariate because of the heterogeneity of the
aquifer, due to the varying composition, compaction, and degree of weathering and
density of fracturing. As a result, exploration of groundwater in a crystalline rock terrain
has proved to be a complex phenomenon. However, the presence of a vast crystalline
rock terrain cannot be neglected as an unfavorable zone.
The application of Remote sensing techniques and surface geo-electrical methods is
highly helpful for groundwater exploration lies in delineating potential zones of
groundwater from a large area. Generally the occurrence of groundwater in crystalline
Remote sensing and electrical resistivity studies on groundwater structure zones 331
rock terrain is associated with the geological structural features like lineaments,
fractures/fissure, fault zones.
The study area lies between the Blue Nile River and River Atbara, It is bounded by
latitudes 14° 30′N and 16°00′N and longitudes 33°30′E and 35°30′E. The study area
covers about 20,000km2 (Fig. 1). The study area is generally flat, with a gentle slope to
the North. The general altitude of the plain is about 500m above mean sea level (m.s.l).
The flat monotony of the plain is occasionally broken by some protruding low to
moderately high hills or hill chains, which hardly exceed 200 meters above the ground
surface (Kheiralla, 2001).
According to the 1993 census in Sudan the total population of the study area is
roughly estimated at 30,000 persons, Livestock rising is the major activity of 70% of the
Butana population. The ecological conditions as well as the long experience of the
inhabitants turn pastoralism as the most worthwhile occupation. The animal population
within the area is roughly estimated as some 35,000 heads, composed of about 30,000
sheep and goats, 5,000 cattle and camel (Abd el Ati, 1993). Human and animal
populations in the Butana area receive their water
Central Butana region, as they can store and yield reasonable quantities of water from
their joints and fractures. The alluvial deposits are none water bearing formation,
especially close to the blind deltas of the main Wadis (valley) (Ahmed & Ayed, 1996).
However, groundwater occurs in sand alluvial fans are expected to be found under thick
cover of the Butana clayey.
The surface runoff soon evaporates and/or infiltrates. The crystalline rock terrines,
underlying the Butana area is assumed to be groundwater devoid. Under specific
geological and hydrogeological condition, the crystalline rock terrines can store and yield
appreciable quantities of groundwater variable quality. The weathered and/or fractured
crystalline rocks underlying the drainage system from local aquifer zones, in the Butana
area the hydro-geological significance of the sandy alluvial deposits is that they act as a
membrane through which surface flow can infiltrate to recharge the underlying fractured
zones.
The study aims to investigate the extent of the influence of the drainage and fractured
by the use of Remote sensing and applying geo-electrical methods to delineate the
general hydro-geological aspects of the sediments overlying the crystalline rocks for
exploration of groundwater. In order to understand the significance of the fracture
pattern, geological, hydro-geological, drainage system and lineaments map have been
prepared with the help of Land sat TM imagery. An isoresistvity map is prepared by
conventional survey of equal apparent resistivity (AB/2=60) and then comparing a
lineament map and drainage system map to identify the extent of correlation.
A number of isolated out crops of Pre-Cambrian Basement complex rocks scatter over
the central plains of Sudan (Fig. 2). In the Butana region, several hills occur and make
prominent landmarks
Remote sensing and electrical resistivity studies on groundwater structure zones 333
within the predominantly flat clay plains. The solid geology of the Butana plain is rather
concealed under a veneer of variable thickness of superficial clays covering over 70% of
the Butana area. Several metasedimentary sub-parallel belts extend for 10–25km in the
form of low to moderately elevated ridges surrounded by Butana clay cover (Iskander et
al., 1993). The generalized geological column for Butana region can be summarized as
shown in Table 1.
The majority of the Basement rocks are concealed under the cover of the Butana clay
plain. As mentioned, the structural domain in the area is characterized by northeast
Water resources of arid areas 334
NE vergence. In this search three phases of deformation recognized has reported in the
Butana region (Fig. 2) these are:
● F1—Regional foliation, lineation, and shear zones.
● F2—Planar and linear structures, upright folds and faults bricca.
● E-W faulting and fracture cleavage.
The tectonic events that terminated by continental collision at the end of the Pan-African
developed or reactivated a conjugate set of strike-slip faults and shear zones in both the
Nubian and Arabian shield. The fault/shear zones have two main trends NW-SE (Najd
trend) and the ENE-WSW (Central African Lineament). In the central Butana regions,
major faults and shear
Table 2. Sites proposed for drilling of boreholes.
Well Locality S.W.L Expected Aquifer Apparent
No. (m) yield type resistivity
(103m3/yr) at depth
of 60m
(Ωm)
1 Es 45 90 FBC 125
Subagh
2 Es 39 80 FBC 90
Subagh
(El
Buqaa)
3 Es Sada 25 50 ALL/WBC 60
4 Abu 27 40 WBC 55
Gerad
5 El Fuel 36 70 FBC 115
6 El Bresi 33 65 ALL/WBC 70
7 El Edeid 32 50 FBC 85
El
Tawill
8 El Edeid 24 60 FBC 100
El
Hamur
FBC—Fractured Basement Complex; ALL—Alluvium;
WBC—Weathered Basement Complex.
zones display apparently complicated sets of fractures generally follow the regional
foliation/ schistosity trends in both the metasediments and the underlying Basement
rocks.
The structural domain in the north and southeast of Es Subagh area is NNE to NE
(Fig. 3). Subordinate N-W and E-W trending faults affect mainly the northwestern part of
El Butana (Iskander et al., 1993). The prominent NE trending sets of faults are mainly
strike-slip with a dextral sense of movement. Those discontinuities extend up to 70km as
attested by the linear-controlled drainage system. Some faults sheared the
metasedimentary rock assemblage to significant proportions creating 2–1.5m wide zone
Water resources of arid areas 336
of shearing or brecciation (e.g. Jebel El Rabbda), thrusting the evident in Jebel El Tawill
with NE trends and E or SE dip direction. This orientation lineament direction of
ophiolite transport. Very similar trends have been suggested for the ophiolite and
allochthonous sheets transport in the Engassana Hills (Vail and Duggue, 1986).
This deformation created an important fracture, which have provided Es Subagh area
in Central Butana with fresh water. To the south, J.El Tawill ultramafic belt has been
affected by two boundary thrusts (Fig. 2) 2km apart, resulting into variable degrees of
shearing on the eastern and western flanks. Similar sub-basin has been created and
provides potable water for the inhabitants of the area.
The area has complex hydro-geological conditions owing to complexity in the origin
of the rock units encountered. The compact pre-Cambrian suites of rocks are poor
aquifers. Groundwater occurs in confined conditions in these rocks due to the
development of secondary porosity such as fracturing. Fractured crystalline rocks are less
permeable at greater depth because stress variations that cause fractures are larger and,
over geological time, occur more frequency near the ground surface. Fractures tend to
close at depth because of vertical and lateral stresses imposed by overburden loads and
horizontal stresses of tectonic origin. Apparently these basic conditions, which control
ground water occurrence in crystalline rocks, apply to a large extent to the Butana area.
Groundwater generally occurs in the upper weathered/or fractured zone, which may
extend down to 70m depth as indicated by the lithological logs of Es Subagh two
boreholes. The sheared rocks, which form the bulk of the aquifer, are composed of acid
gneisses, quartzite, marbles and granites. The brecciate marbles in the Butana normally
form the good aquifers. The formation easily dissolvable by moving waters, and thus
forms wide fissures and cavities, which facilitated groundwater storage and transmission
(Iskander et al., 1993). The best aquifer zones comprise the marble cavities along Khor
Abu Gimbil, Es Sufeiya, fractured Basement rocks of Adeid El Tawill, Es Subagh, Reira,
and El Hagar, and the alluvial deposits along Wadi Abu Grad and Wadi Abu Matariq.
Depth to ground water level in the study area varies from 20 to 25m at Reira, between
15 to 50m at Es Subagh, from 20 to 40m below the ground surface at Es Sufeiya, 10 to
20m at El Tawill, and much deeper ground water level (60m below the ground surface) at
El Hagar (Table 2). Groundwater in the fracture Basement Complex occurs under free
water table conditions and at depths ranging from few meters to over 60 meters below the
ground surface. The water table attains its highest level during and shortly after rainy
season and drops to its lowest level immediately before the next rainy season. The
amplitude of the seasonal or annual fluctuation depends to a great extent on the balance
between the recharge and discharge of water to and from the aquifer. Generally
productivities of the wells directly correlate the intensity of rainfall in the study area.
Noticeably in dry years, the water levels drops to the bottom of the wells or completely
dry out. Apparently Groundwater movement in crystalline rocks is non-committal to flow
direction, because fracture trends tend to variation that cause movement are largely
expected variable to flow direction. Basement outcrops from the main catchments area
and ground water movement coincides with the drainage systems controlled structurally.
Groundwater moves away from the surface water divide and generally in the eastern and
NE direction, ground water moves to joint Atbara River.
Eight boreholes have been drilled along watercourses or fractured aquifers at variable
depths (Table 2). These include four at Es Subagh, Qeili, Abu Gimbil, Husheib and Umm
Remote sensing and electrical resistivity studies on groundwater structure zones 337
Sarha villages. Only five of the eight boreholes are successful with the total out put
hardly exceeding 10m3 per hour, per well. Presently only a few are in use. Two hundreds
and fifteen hand-dug wells were excavated into different aquifers (e.g. Fractured,
weathered Basement, alluvial or fan deposits) to depths ranging between 20–55m
(Kheiralla, 2001). Their static water level varies between 15 and 40m, though wells
drilled into the alluvium aquifers are relatively shallower not exceeding about 5m deep.
The static water level in these hand-dug wells fluctuates between 3–5m per year, which
may be considered as a good sign of recharge. The total out put of the hand-dug wells is
estimated to be about 2869*102m3/year. Chemical analysis of samples collected from
some hand-dug wells indicates fair to good water quality suitable for human as well as
animal consumption. Based on the above estimates, the grand total yield of the existing
water sources in the Central Butana is around 33*104m3/year. This amount does not
exceed 2/5th(40%) of the actual demand. These indicate an actuate water supply
shortage.
3 LINEAMENTS
One of the objectives of this work is to delineate the lineaments in more details using the
geoelectrical method and Remote sensing technique. Waters (1990) suggests that there
are two stages involved in hydro-geological investigations based on remotely sensed
images: first, the identification of photo lineaments representing crustal fracturing; and
second, the interpretation of these features with respect to their significance in terms of
potential groundwater flow. Thus, lineaments visible on the land sat TM images my be
expressed by: 1) geomorphologic features such as valleys, straight drainage channel
segments, linear scarp faces, or pronounced breaks in the crystalline rock mass, 2) tonal
differences at the boundaries of contrasting lithological units.
The digitalization of lineaments was carried out through visual analysis at the screen
of land sat TM and linear structure features such as faults and fractures were studied in
the field. Faults can be distinguished from the fracture by the observations of the
slickenside. The Rose diagram (Fig. 3). Constructed from the lineaments map shows the
structural domain is NNE to NE trend. Main trend coincide with the Central Africa
lineaments with an average direction between 5° and 75°, but most of the long and high
frequency lineaments are clustered around 90°, while in the NW direction the NE
trending sets of faults are mainly strike-slip with dextral sense of movement.
Tensional faults, that is those parallel to the direction of the tectonic stress or
orthogonal to the direction of crustal extension, my be believed open and some what
wider than compressive/shear faults, which are orthogonal or inclined with respect to the
direction of tectonic stress and consequently tend to be tighter. Thus, it should be much
easier to recognize tensional faults in a land sat than shear faults and this should be
reflected in the lineaments frequency histogram. These preferred orientation of deep-
seated fractures are responsible for the groundwater potential zones in the study area.
Water resources of arid areas 338
seated fractures are responsible for the formation of groundwater potential zones in the
Butana region, where the density of lineaments is found to be between 0.5 and 2 (Fig. 4).
These zones may also have the continuity of the lineaments extending from high to low
altitudes, which may be buried under transported deposits (20m). This is in conformity
with the well yields of the wells. In this region, lineaments are the most significant
predictors of groundwater occurrence and general geological structures are less
significant.
4 DRAINAGE
A drainage map was prepared with help of land sat TM data (Fig. 5). The drainage
system, which develops in an area, is strictly dependent on the slope, the nature and
attitude of bedrock and on the regional and local fracture pattern. Drainage is studied
according to its pattern type and its texture (Way, 1973). Whilst the first parameter is
associated to the nature and structure of the substratum, the second is related to rock/soil
permeability. Actually, the less a rock is permeable, the less the infiltration of rainfall,
which conversely tends to be concentrated in surface runoff. This gives origin to a well-
developed and fine drainage system.
The low hills of the Butana are mostly composed of the Precambrian Basement
Complex rocks (Ahmed, 1968), these hills and hill chains are arranged to form a
disrupted low regional ridge, which acts as a flat watershed dividing the Butana drainage
system to the Blue Nile River in the
Water resources of arid areas 340
The drainage system was classified as first order and second order based on their role
in groundwater storage. The first order drainage pattern represent fractures or faults
controlling a large part of the study area, affecting a deeper portion of the bedrock and
thus can be play an
The resistivity values of rocks vary depending upon the presence of secondary porosity
such as weathered, fractured and joints. Groundwater prospecting is often combined with
geo-electrical measurements. Vertical Electrical Sounding (VES) are executed to detect
variation resistivity transition with depth. A total of fifty five (55) Vertical Electrical
Sounding measurements utilizing Schlumberger array used in the present study. The
objectives of the resistivity survey in the study area are to determine the lithology,
weathered, fractured pattern, depth to the basement rock and resistivity variation. Vertical
Electrical Soundings were taken at two interest areas, these area are exempted from
Remote sensing and electrical resistivity studies on groundwater structure zones 343
7 CONCLUSIONS
● The study of land sat TM images identified a lineament trending NE-SW direction and
drainage pattern present in the area.
● Moderate to good yield (40–65*103m/yr) are tapping from weathered zones, good yield
(70–90*103m3/yr) are tapping from fracture zones.
● The area as covered with high alluvial and more fractured zones are providing copious
amounts of groundwater.
● Range of resistivities and Expected Yield of different zones are presented in table (2).
● The comparatively high density obtained by lineaments concentration/and or drainage
system indicated the presence of groundwater potential zones.
REFERENCES
Abd el Ati, H.A. 1993. A base line survey Report on Central Butana. ADS project area-UNDP-
Khartoum.
Ahmed, F. 1968. The geology of the Jebel Qeili, Butana and Jebel Sileitaat-Es-Sufr igneous
complex, Nile valley, Central Sudan. Unpublished M.Sc. thesis, Univ. Khartoum.
Ahmed, F. & Ayed, M.A. 1996. Applied geophysical and satellite imagery techniques, for ground
water studies in Central Butana area; ADS report, 25pp, 10–15.
Iskander, W. Ahmed, A A. Mokhtar, A. & Fadle, A.S. 1993. Appraisal of mineral and water
resources of central Butana, Eastern region-Sudan ADS report 85pp.
Kheiralla, K.M. 2001. Geophysical study on groundwater structure at two localities in Central
Butana, Central Sudan. Unpublished M.Sc. thesis, Univ. El Neelain.
Waters, P. 1990. Methodology of lineament analysis for hydro-geological investigation. In Satellite
Remote Sensing for Hydrology and Water Management. E.C.Barret, Power, C.H. & Micallef, A.
eds., New York, Gordon & Breach: 1–23.
Way, D.S. 1973. Terrain analysis, a guide to site selection using aerial photographic
interpretation, Stroudsburg, Dowden, Hutchinson, Ross Inc.
Vail, J.R. & Duggue, J.P. 1986. Bibliography of geological sciences for the Republic of the Sudan.
1837–1985, Center Int. Formation Echanges Geol. Paris, Spec. Publ.
Monitoring and modeling of fluxes on
Kalahari—setup and strategy of the Kalahari
Monitoring project Serowe study case,
Botswana
M.W.Lubczynski1 & O.Obakeng1,2
1
The International Institute for Geoinformation Science and Earth
Observation (ITC), Enschede, The Netherlands
2
Geological Survey, Lobatse, Botswana
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
ABSTRACT: The ongoing discussion about the presence and the rates of
recharge in Botswana, which constrains groundwater sustainability in the
country scale, has led to the initiation of a new recharge project in
Botswana called Kalahari Monitoring Project. In contrast to previous
attempts this project focuses on temporal flux monitoring by using
automated data acquisition systems (ADAS). The framework of that
project is discussed on the base of the example of the Serowe study area,
located at the eastern fringe of Kalahari, where an extensive monitoring
network was installed to provide data for spatio-temporal flux assessment.
This network allows measurements of saturated, unsaturated and surface
zone fluxes. It consists of groundwater table fluctuation monitoring in 21
piezometers, soil moisture and soil suction pressure monitoring in 7
identical profiles comprising measurements at 0.5, 2, 4, 6, 8m b.g.s., one
deep suction pressure profile down to 76m (sensors at 15 different depth
levels), transpiration monitoring using 51 sap velocity thermal dissipation
probes installed at 9 ADAS locations and monitoring of climatic variables
for potential and actual evapotranspiration in 10 towers scattered over the
study area. This data is either interpreted directly (rainfall, transpiration)
or used in 1-D models to calibrate surface and subsurface fluxes such as
evapotranspiration, groundwater evapotranspiration and recharge. For
spatial data assessment the remote sensing (RS) method is proposed. The
evapotranspiration is obtained with RS solution of energy balance,
transpiration by RS upscaling of the sap flow measurements and recharge
by RS and GIS modeling. The final integration of spatial and temporal
data for spatio-temporal flux assessment is carried out by transient
groundwater model calibration with spatio-temporally variable recharge
and groundwater evapotranspiration. The aspect of partitioning of tree
Monitoring and modeling of fluxes on Kalahari 347
1 INTRODUCTION
mass balance measurements or from 1-D recharge modeling, but the method of spatial
representation of such data is not well defined. It is an ongoing dispute on what are the
best and the most efficient schemas to present point flux information spatially, using
interpolation, extrapolation by GIS modeling, stochastic modeling or discrete
groundwater modeling?
Thus in 2001a complex Kalahari Monitoring research project was established in
Botswana focusing on monitoring and modeling of surface and subsurface fluxes. The
main objective of this project is spatio-temporal assessment of subsurface fluxes for
better management of groundwater resources in Botswana. As study areas three
hydrologically different semi-arid locations were selected, Maun area, Localane-Ncojane
area and, Serowe area. Maun area represents relatively wet conditions of the Delta
Okavango with very shallow groundwater table of only few meters below the surface.
Localane-Ncojane area represents the western, driest part of Botswana Kalahari with very
deep groundwater table on average 100–150m b.g.s. The Serowe area of ~2500km2 on
which this study focuses (Figure 1), is currently the most instrumented and the most
intensively investigated research area of the Kalahari Monitoring project.
The Serowe study area was selected as target area of the Kalahari Monitoring project
following the previous research in the same area delivered by SGAB (1988) and WCS
(2000). The latter one included also the numerical groundwater model, after which the
present study area boundaries were assumed (Figure 1).
The study area consists of two contrasting parts, Kalahari sandveld and hardveld,
which have different natural and hydrological conditions. The western sandveld part is
elevated as compared to the hardveld along the prominent escarpment feature. This
elevation is due to the 60–100m eolian Kalahari sand cover on the western, sandveld part,
overlying solid rocks such as Stormberg basalts and Ntane sandstones which in the
eastern part outcrop or are covered by thin, 0–5m Kalahari sand cover (Figure 1). The
sandveld part slopes gently to the west, is fairly flat and featureless without prominent
drainage lines. In contrast, the hardveld part slopes steeper and there is a drainage system
of the intermittent streams, discharging water mainly after the heavy showers. The
majority of the villages in the study area are concentrated along the eastern edge of
Monitoring and modeling of fluxes on Kalahari 349
The automated data acquisition system (ADAS), is a combination of sensors or just only
one sensor installed in the field and operated by a multiple or single channel logger
managing the performance of the sensors. ADAS are very useful in hydrology because
they provide high temporal data resolution so they are well applicable in setting and
calibration of transient models. The simplest example of ADAS is a combination of one
sensor with one logger such as e.g. discussed below automated groundwater table
recorder (AGTR). More sophisticated version is a multi-sensor ADAS operated by one,
multi-channel logger. Such systems composed of various combinations of electronic
sensors are usually mounted as towers on the masts (Lubczynski 2000) and can focus on
monitoring of above-surface, surface, unsaturated zone and saturated zone temporal
variability. The appropriate selection of the sensors and the programming of the loggers
Water resources of arid areas 350
depend on the objectives of the monitoring and both are critical for the success of the
hydrological investigation program determining also the cost-effectiveness of that
program.
In the study area there are eleven multi-sensor ADAS towers named GS00 to GS10.
Two of them are 18m high, GS00 on hardveld and GS10 on sandveld, one on sandveld,
GS09 is 10m high and eight on sandveld GS01–GS08 are 2m high (Figure 1). The
concentration of most of the ADAS towers on the sandveld illustrates current research
focus of the project. The ten towers GS00–GS08 and GS10 are installed permanently
whereas the GS09 is a mobile, retractable mast tower, which if not in mobile campaign
(Obakeng & Lubczynski 2004 in the same issue) then it is temporally located as indicated
in Figure 1. The multi-sensor ADAS towers provide input for assessment of rainfall,
potential evapotranspiration, actual evapotranspiration, transpiration, unsaturated zone
moisture and suction pressure (Lubczynski 2000).
Groundwater fluxes such as recharge and groundwater evapotranspiration vary not only
temporally but also spatially. The spatial distribution of groundwater recharge was first
evaluated in the study area by groundwater modeling (Lubczynski 2000, WCS 2000).
Later two series of chloride data for recharge assessment were collected from the wells,
all over the study area, first by Obakeng (2000) and next by Magombedze (2002) and
assessed spatially by interpolation and also by extrapolation applying integrated GIS
modeling. The summary of those approaches is presented by Magombedze et al. (2004)
in the same conference issue.
An assessment of groundwater evapotranspiration (Eg) is a very difficult issue. A first
attempt to determine Eg spatially in the study area was made by Lubczynski (2000). For
that purpose he applied groundwater modeling in which Eg was considered as state
variable with spatial distribution derived from RS solution of energy balance
(Timmermans & Meijerink 2000). Certainly this was not the ideal procedure since Eg
fluxes were small and likely comparable with the eventual error of the calibrated model.
The recent attempts in defining Eg, lead through the determination of its tree transpiration
(T) component applying sap flow measurements. The methodology of sap flow
measurements on Kalahari and plot level upscaling is discussed in Fregoso (2002), in
Mapanda (2003) and is finally summarized in Lubczynski et al. (2004) in the same
conference issue. The RS upscaling of sap flow measurements for the 10×10km
experimental area covered by multispectral IKONOS image (Figure 1), was attempted by
Fregoso (2002), by Mapanda (2003) and by Keeletsang (2004). A similar attempt for the
same study area but using multi-band TETRACAM digital camera built on the aircraft is
described in Hussin et al. (2004a) whereas the multi-band aerial-photography aircraft
mission itself is described in Hussin et al. (2004b), both in the same conference issue.
Due to the difficulties in classification of tree species, closely related to the large
biodiversity, on Kalahari the RS upscaling protocol is still being improved.
The transpiration mapping by RS upscaling of sap flow measurements unfortunately
does not provide the estimation of the demanded in groundwater management Eg but
provides T. Equalizing the two is only possible if two critical assumptions are fulfilled,
first, that considering large depth of groundwater table in the study area, the Cg is
negligible or definable and the second, that the Tg component of T, can be separated from
unsaturated zone root water uptake (Tu). The first assumption will be tested by setting up
1-D saturated-unsaturated models (see below) for each unsaturated moisture and suction
pressure monitoring profile available in the study area. The action with regard to the
Monitoring and modeling of fluxes on Kalahari 353
applied parameters and as explained by Lubczynski and Gurwin (2004) for Sardon
granite catchment in Spain, they can provide not only the prediction scenarios but also
accurate information on where, when and at which rates fluxes such as recharge and
groundwater evapotranspiration occurred in the analyzed area. The availability of
historical record of rainfall in Serowe village starting in 1922, allows to run backward
model scenario, which will finally allow to provide the demanded in Botswana long-term
temporal characteristic of recharge on Kalahari.
6 CONCLUSIONS
The acquisition of temporal data with ADAS provides unique opportunity for direct
temporal measurement of various flux processes such as rainfall and transpiration. Other
processes such as recharge and evapotranspiration cannot be measured directly but have
to be modeled. ADAS provides full data acquisition support for such models.
The integration of temporal data from ADAS with the spatial data extrapolated with
GIS and RS techniques in numerical models provides the opportunity of model
calibration with spatio-temporally variable fluxes.
In semi-arid and arid climates only models calibrated with spatio-temporally variable
fluxes can provide a reliable system parameterization, reliable spatio-temporal flux
regimes and reliable flux rates. This means, that such models provide the optimal tool for
groundwater management.
ACKNOWLEDGEMENTS
REFERENCES
Arnold, J.G., Williams, J.R., Srinivasan, R., King, K.W. & Griggs, R.H. 1994. SWAT (Soil and
Water Assessment Tool) user’s manual. USDA, Agricultural Research Service, Grassland, Soil
and Water Research Laboratory, Temple, TX.
DHI—Danish Hydraulic Institute. 1993. MIKE SHE water movement-user’s guide and technical
manual, ed.1.0 DHI, Denmark, pp. 81.
De Vries, J.J., Selaolo, E.T. & Beekman, H.E. 2000. Groundwater recharge in the Kalahari, with
reference to paleo-hydrologic conditions. Journal of Hydrology 238, 110–123.
Doherty, J. 2000. PEST—Model-Independent parameter estimation. User’s manual. Watermark
Computing, Australia.
Fregoso, A. 2002. Dry-season transpiration of savanna vegetation. Assessment of tree
transpiration and its spatial distribution in Serowe, Botswana. MSc thesis, Library of ITC—
International Institute for Geoinformation Science and Earth Observation, Enschede, The
Netherlands.
Monitoring and modeling of fluxes on Kalahari 355
Fregoso, A., Chavarro, A. & Lubczynski, M.W. 2004. Sap flow measurements of tree transpiration
on Kalahari, Serowe study case, Botswana. Proc. WRASRA conf. Gaborone 3–7 August, 2004,
Rotterdam, Balkema.
Granier, A. 1987. Evaluation of transpiration in Douglas-fir stand by means of sap flow
measurements. Tree Physiology 3:309–320.
Haase P., Pugnaire, F.I., Fernandez, E.M., Puigdefabregas, J., Clark, S.C. & Incoll, L.D. 1996. An
investigation of rooting depth of the semiarid shrub Retama sphaerocarpa (L.) Boiss. By
labeling of groundwater with a chemical tracer. Journal of Hydrology 177:23–31.
Hussin, Y.A., Chavarro, D. Lubczynski, M.W. & Obakeng O. 2004a. Mapping vegetation for up-
scaling evapo-transpiration using high-resolution optical satellite and aircraft images in Serowe,
Botswana. Proc. WRASRA conf. Gaborone 3–7 August 2004, Rotterdam, Balkema.
Hussin, Y.A., Lubczynski, M.W. & Obakeng, O. 2004b. Designing and implementing an aircraft
survey mission using high-resolution digital multi-spectral camera for vegetation mapping for
up-scaling evapotranspiration of Serowe, Botswana. Proc. WRASRA conf. Gaborone 3–7 August
2004, Rotterdam, Balkema.
Keeletsang, M. 2004. Assessment of dry season transpiration using IKONOS images, Serowe case
study, Botswana. MSc thesis, Library of ITC—International Institute for Geoinformation
Science and Earth Observation, Enschede, The Netherlands.
Le Maitre, D.C., Scott, D.F. & Colvin, C. 2000. Information on interactions between groundwater
and vegetation relevant to South African conditions: A review. In: Past Achievements and
Future Challenges. Balkema, ISBN 9058091597, Rotterdam, 959–961.
Lubczynski, M.W., 2000. Ground water evapotranspiration—underestimated component of
groundwater balance in a semi-arid environment—Serowe case Botswana. In: Past
Achievements and Future Challenges. Balkema, ISBN 9058091597, Rotterdam, 199–204.
Lubczynski, M.W. & Gurwin, J. 2004. Integration of various data sources for transient groundwater
modeling—Sardon study case, Spain. Journal of Hydrology—in revision.
Lubczynski, M.W., Fregoso, A., Mapanda, W., Ziwa, C, Keeletsang, M., Chavarro, D.C. &
Obakeng O. 2004. Dry season Kalahari sap flow measurements for tree transpiration mapping—
Serowe study case, Botswana. Proc. WRASRA conf, Gaborone, 3–7 August 2004, Rotterdam,
Balkema.
Magombedze, L.M. 2002. Spatial and temporal variability of groundwater fluxes in a semi-arid
environment—Serowe, Botswana. MSc thesis, Library of ITC—International Institute for
Geoinformation Science and Earth Observation, Enschede, The Netherlands.
Magombedze, L.M., Frengstad, B. & Lubczynski, M.W. 2004. Spatial variation of groundwater
recharge in a semi-arid environment—Serowe, Botswana. Proc. WRASRA conf, Gaborone 3–7
August 2004, Rotterdam, Rotterdam.
Mapanda, W. 2003. Scaling-up tree transpiration of eastern Kalahari sandveldof Botswana using
remote sensing and geographical information system.
McDonald, M.D. & Harbaugh A.W. 1996. A modular three-dimensional finite difference
groundwater flow model. Washington, D.C., U.S. Geological Survey.
Obakeng, O.T., 2000. Groundwater recharge and vulnerability: A case study at the margins of the
south-east Central Kalahari Sub-basin, Serowe region, Botswana. MSc thesis, Library of ITC—
International Institute for Geoinformation Science and Earth Observation, Enschede, The
Netherlands.
Obakeng, O.T. & Lubczynski, M.W. 2004. Monitoring of evapotranspiration on Kalahari, Serowe
case study, Botswana. Proc. WRASRA conf. Gaborone 3–7 August 2004, Rotterdam, Balkema.
SGAB, 1988. Serowe groundwater resources evaluation. Swedish Geological Co report for
Department Geological Survey, Botswana.
Simunek, J., Sejna M. & van Genuchten M.T. 1998. The HYDRUS-1D software package for
simulating the one-dimensional movement of water, heat and multiple solutes in variably
saturated media, version 2.0, IGWMC—TPS—70, 2002pp., Colorado School of Mines, Golden
Colorado.
Water resources of arid areas 356
Sophocleous, M. & Perkins, S.P. 2000. Methodology and application of combined watershed and
groundwater models in Kansas. Journal of Hydrology 236:185–201.
Timmermans, W. & Meijerink, A., 2000. Remotely sensed actual evapotranspiration; implications
for groundwater management in Botswana. In: JAG: International Journal of Applied Earth
Observation and Geoinformation, 1(1999)3/4, 222–233.
Van der Lee, J. & Gehrels, J., 1990. Modelling aquifer recharge—Introduction to the Lumped
Parameter Model EARTH. Free University of Amsterdam, The Netherlands.
WCS—Wellfield Consulting Services, 2000. Serowe wellfield extension project. DWA,
Department of Water Affairs, Report No. TB10/3/10/95–96.
Geoelectrical investigation for aquifer
delineation in the semi-arid Chad Basin,
Nigeria
A.Iliya1 & E.M.Shemang2
1
Rural Water Supply and Sanitation Agency (RUWASA) Damaturu,
Nigeria
2
Department of Geology, University of Botswana, Gaborone, Botswana
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
The rapid increase in urbanization as well as industrial and agricultural expansion has
focused attention upon the diminishing volume of available groundwater in most major
urban centres within the sahelian zones of West Africa. The study area, which falls within
the Yobe portion of the Chad Basin (Fig. 1) comprises major towns as Damaturu,
Potiskum, Nguru, Gashua, and Geidam, etc. whose combined population is about two
million. As a young state, rapid expansion in industrialization, socioeconomic growth,
etc. is expected and hence the need to address the problem of groundwater resources
evaluation and management. This groundwater development stage, however, can be
viewed as a sequential process consisting of exploration, evaluation and exploitation. The
Water resources of arid areas 358
exploration stage in which surface and subsurface geological and geophysical techniques
are utilized to search for suitable aquifers involve the use of electrical resistivity (VES)
survey and borehole data. The evaluation stage, however, comprises the measurement of
hydrogeological parameters, calculation of aquifer yields (Transmissivity, Storativity,) as
well as hydrogeochemical analysis of water samples collected from boreholes in the
study area.
The exploitation or management stage includes the consideration of optimal
development strategies and assessment of the interactions between groundwater
exploitation and regional hydrological system.
The present study is therefore, aimed at delineating the aquifer system (s) through the
use of surface electrical resistivity techniques thereby pin pointing possible productive
zones, correlate aquifer hydraulic properties with those obtained from VES and hydro
geochemical data. It is also aimed at forecasting the future water requirements/utilization
of the study area. To achieve these objectives therefore, forty two (42) vertical electrical
soundings (VES) using the Schlumberger configuration with a maximum total current
electrode separation of 1000m was carried out in the study area. Some of these have
already been confirmed through drilling of boreholes.
2 REVIEW OF GEOLOGY
The study area located between Longitude 11°N and 13°N and Latitude 11°E and 13°E
(Fig.1) is composed of the Chad Formation outcrops. In general, it consists of successions
of sands, clays, sandy clays and silts with interbedded lenses and layers of sands and
gravels of various levels.
The deposits are generally of lacustrine origin or were formed during periods when
rivers had very low discharges because of climatic and geomorphological conditions. The
beds dip gently towards the centre of the basin not only because of their original attitude
but also mild regional tectonic movements which have affected the basin in recent times.
The Chad Formation may reach a thickness of 600–700m in the central part of the basin
(Offodile, 1992) but thins out rapidly towards the edges. Such a very abrupt reduction in
thickness of sediments near the margins of the Plio-quarternary lake basin could well be
the result of step faulting in the basement rocks of the char depression. This is illustrated
in the lithological data from boreholes (Fig. 2). The products of such activity are
sometimes found at the base of the Chad Formation, as in the case of granitic rocks
encountered in boreholes in Goniri and environs.
These may well be associated with the faults bounding the Chad basin and may to
some extent be contemporaneous with the deposition of the Chad Formation. Data
collected so far indicate that the lithostratigraphic sequence in the study area consists of
sandy clay alteration of the Chad Formation that very probably lies directly on the
Basement Complex rocks.
3 HYDROGEOLOGY
The Chad Basin is described as the largest area of inland drainage in Africa and occupies
parts of Nigeria, Central African Republic and Cameroon. The Nigerian sector of the
basin slopes gently towards the Lake Chad which is the main geographical feature.
Water resources of arid areas 360
Paleocene respectively. Surface water in streams appears seasonally usually from August
to November. For the rest of the year, the streams are dry and the only source of water is
groundwater. This is contained in the three aquifer systems designated as Upper, Middle
and Lower Aquifers (Kogbe et al., 1992) especially in Maiduguri.
While the Lower is considerably deep (over 500m) are tapped by few boreholes, the
Middle and Upper aquifers are on average depth of some 250m and 40m respectively are
obviously over exploited and on many cases have dried-up. This had already focused
attention on the possibility of a perchy aquifer for the Upper aquifer in the Chad basin
(Kogbe et al., 1992). It is worth while mentioning, however, that the above multi-aquifer
systems do not extend throughout the Chad Basin. In Damaturu area, which lies on the
edge of the Chad Formation lake basin, the hydrogeological situation may be summarized
as follows:
a) Total thickness of the Chad Formation is about 130–170m
b) Marked discontinuity of water-bearing levels.
c) Vertical and lateral changes in their hydraulic properties of water-bearing levels.
d) Presence of perchy aquifers where impervious layer levels occur in the upper part of
the formation.
Forty two (42) vertical electrical soundings (VES) using the Schlumberger array
configuration with a minimum and maximum current electrode separation of 320 and
1000m respectively. The equipment used was the ABEM SAS 300B Terrameter.
Sounding was carried out with aim of selecting sites for water supplies to villages and
points were therefore located in and around villages. The VES data was first interpreted
using the conventional curve matching techniques and later using the IPI2WIN software.
Figure 3 shows an interpretation of VES 1 sounding carried out in the area of study.
An Isoresistivity map of the third layer was constructed and the results show that the area
can be separated two zones, the Western and Eastern zones. The boundary between these
two zones probably suggest the contact between two lithological units in the area (the
Kerri-Kerri formation and the Chad formation)
REFERENCES
Bunu, Z.M. and Iliya, A.G. (1992) Understanding the Rainfall Pattern of a Semi-Arid Region: A
case study of Maiduguri. Paper presented at the Fifth National Conference of the Nigerian
Association of Hydrogeologists, Shiroro Hotel, Minna, Nigeria.
Carter, J.D. Barber, W. and Tait, E.A. (1963) Geology of Adamawa, Bauchi and Bornu Provinces
in Northeastern Niogeria. Bull. Geol Surv Nigeria 30, 1–108.
Cratchley, C.R. (1960) Geophysical Survey of the Southwest Part of the Chad Basin, C.C.T.A.
Publication No. 13.
Kogbe, C.A. Schoeneich, K. and Ebah, E.I. (1992) Hydrogeological Framework of Maiduguri
Metropolis in the Chad Basin, NE, Nigeria. Paper presented at the fifth Conference of the
Nigerian Association of Hydrogeologists, Shiroro Hotel, Minna, Nigeria.
Matheis, G. (1965) Short Review of the Geology of the Chad Basin in Nigeria. Journal of Mining
and Geology, 289–294.
Offodile, M.E. (1992) An Approach to Groundwater Study and Development in Nigeria, Mecon
Services Ltd. 300pp.
Monitoring of evapotranspiration on Kalahari,
Serowe case study, Botswana
O.Obakeng1,2 & M.W.Lubczynski2
1
Geological Survey of Botswana
2
The International Institute for Geoinformation Science and Earth
Observation (ITC), Enschede, Netherlands
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
As 70% of the precipitation depth may evaporate annually in semi-arid climates, careful
consideration should be given to the determination of actual evapotranspiration, as well
as potential evapotranspiration. Estimates of potential evapotranspiration in semi arid
climates are an order of magnitude greater than the rainfall depth. In the Serowe study
area the annual potential evapotranspiration amounts to 1350–1450mm (Choudhury,
1997) and the mean annual rainfall is 447mm. Consequently, the actual
Monitoring of evapotranspiration on Kalahari, Serowe case study, Botswana 365
evapotranspiration rates are much smaller than the potential rates because of the limited
amount of water stored in the topsoil.
Evapotranspiration (ET) plays an important role in a groundwater balance in semi arid
climates, as demonstrated by Lubczynski (2000). In a general sense the groundwater
balance equation can be written in the following form
Qin+R−Qout−Eg−A±∆S=0
(1)
The Serowe area has in general, a gentle topography, which varies from ≈1020m.a.s.l. to
≈1240m.a.s.l (Fig. 1).
A major geomorphic feature within Serowe area is the escarpment, which forms part
of a geologically recent axis of uplift known as the Zimbabwe-Kalahari axis (Smith,
1984). It represents the eastern limit of the Kalahari sandveld. All rivers are ephemeral,
and flow occurs only during exceptionally high rainfall events of the annual wet season.
Otherwise, they are dry for most of the year, with groundwater levels often situated at
shallow depths (4–6m) beneath riverbeds. The surface topography is lower in the E and
SE of the region, higher in the western Kalahari plateau and the highest on the NW side
of the escarpment edge, which is a prominent topographic feature in this area. SE from
the escarpment, the average slope is 5%, and it gradually decreases to less than 1%
towards the E and SE. Rock outcrops are found mainly at the escarpment and along river
valleys below the escarpment. Elsewhere Kalahari sands and superficial deposits overlie
rocks. The soil types found in the study area are arenosols, regosols, lixisols, luvisols and
vertisols. Arenosols are by far the most common soil unit covering most of the Serowe
area to the west, north, south and the extreme east.
The climate of Serowe study area, like in other parts of Botswana, is characterized as
semi-arid, with a mean annual rainfall of 447mm (SGS, 1988). Rainfall is seasonal, with
the highest intensity in summer followed by a dry winter period. The summer stretches
Water resources of arid areas 366
from October to April whereas the winter begins in May and ends in August (Bhalotra,
1987).
The main type of vegetation in the study area is thought to belong to the Northern
Kalahari Tree and Bush Savanna (Vossen, 1989; Nash, 1992), despite the existence of
significant spatial
GS00—the most versatile, 18m high, galaxy ADAS tower, installed on the hardveld in
September 2001 (Fig. 1). It is equipped with one net radiometer CNR1 installed at the top
of the tower construction, 3 anemometers and 3 relative humidity & temperature (RH/T)
sensors all installed at 2m, 13m, and 18m heights. In addition there are also: soil heat flux
plates buried at a depth of 2cm, two soil temperature sensors buried at depths of 2cm and
15cm, a tipping bucket rain gauge raised to a height of 1.2m above the ground and 6 sap
flow sensors of Granier type measuring velocity of water transport in tree stems. The data
acquisition and the data storage with 0.5h resolution is managed by the Delta-T logger.
GS01 to GS07—there are seven of the same, 2m high ADAS towers in the study area
(Fig. 1). Each of these towers is equipped with one RH/T sensors attached to a mast at a
height of 2m above the ground surface, a tipping bucket rain gauge raised to a height of
1.2m above the ground, 6 sap flow sensors, one soil moisture and one soil suction
pressure profile with sensors at 0.25m, 2m, 4m, 6m, 8m. The Skye DataHog2 logger logs
the data at half hour intervals in all seven towers.
GS08—this tower was installed as a backup of the tower GS00 in case of its failure
(Fig. 1). It consists of anemometer, RH/T sensors and radiometer CM3 for measuring
incoming short-wave radiation all mounted at the height of 2m above the ground surface.
Other instruments include two soil temperature sensors buried at depths of 2cm and 15cm
below the ground surface and a tipping bucket rain gauge raised to a height of 1.2m
above the ground and 6 sap flow sensors. The Skye DataHog2 logger logs the data at half
an hour interval.
GS09—this is a mobile, retractable 10m high tower, equipped with two anemometers,
two RH/T sensors, installed at 2m and 10m heights, a pair of soil temperature sensors
buried at depths of 2cm and 15cm in the soil and 6 sap flow sensors. During field
campaigns GS09 is moved between stations GS01-GS08 every ten-days, otherwise it is
fixed at its semi-permanent location (Fig. 1). The Skye DataHog2 logger logs the data at
half an hour interval.
Many methods exist for estimating actual evapotranspiration (e.g. Bastiaanssen, 1995)
and potential evapotranspiration (e.g. Hargreaves and Samani, 1985) using
micrometeorological measurements. In this study actual evapotranspiration (AET) was
computed from the energy balance equation in which soil heat flux (G) and the net
radiation (Rn) were considered as known (measured) and the sensible heat flux (H) was
calculated using the temperature profile method (Holtslag and Ulden, 1983). The
potential evapotranspiration was calculated with the FAO Penman-Monteith formula
(Allen et al., 1998).
where Cp (kJ kg−1 °C−1) is the specific heat capacity of air taken as 1013 kJ kg−1 °C−1, ρ
(kg m−3) is air density.
The effect of the modification of forced convection by temperature gradients on
momentum and heat (and water vapor) transfer can be corrected by dimensionless
parameters. One of the widely used stability parameter is known as the Monin-Obukhov
correction factor. The Monin-Obukhov length L (m) is given by
(3)
where k is von Karman’s constant (0.41), g (ms−2) is acceleration due to gravity and T
(°C) is mean air temperature. When L is greater than 0 stable atmospheric conditions
exist and when L is less than 0 unstable atmospheric conditions prevail otherwise the
conditions are neutral.
A simplified method for determination of momentum flux and sensible heat flux (H)
which requires wind speed (uz) (ms−1) at level z(m), a surface roughness length (z0) and a
temperature difference ∆θ (K) between two heights z1(m) and z2(m) in the atmospheric
surface layer as input is provided by Holtslag and Ulden (1983). In this method the
integrated flux-profile relations of Dyer and Hicks, 1970 are used to calculate u* and θ*
from the aforementioned parameters according to
(4)
(5)
The integrated stability correction function for heat transfer (ψh) and momentum transfer
(ψm) for unstable conditions (L<0) can be estimated from Equations 6 and 7 respectively
(6)
(7)
where x is given by
(8)
(9)
The sensible heat flux (H) can be calculated from Equations (2)–(9), starting with a first
guess for L (Monin-Obukhov stability parameter). With L=−5, then u* and θ* are
calculated from Equations (4)–(5). Using Equation 3, L is calculated by using the
estimated values of u* and θ*. The new value of L is substituted into Equations (4)–(5),
primarily to get improved values for u* and θ*. This usually takes about 5 iterations, until
the value of L do not change significantly (<5%). Then H is calculated with Equation 2.
This scheme is referred hereafter as temperature profile method (T-profile). The surface
roughness length (z0) was estimated from vegetation height, leaf area index and other data
according to Raupach (1994).
This method however, will not work when no temperature differences are observed
between two measurement heights, a situation that was occasionally encountered in the
present research work. Once H is known then the actual evapotranspiration can be
calculated from the energy balance equation by applying as input, also soil heat flux and
the net radiation, both directly measured in the study area.
(10)
where λE (MJ m−2day−1) is latent heat flux (evapotranspiration), Rn (MJ m−2 day−1) is net
radiation, G (MJ m−2 day−1) is soil heat flux, γ (kPa °C−1) is the psychrometric constant,
Cp (kJ kg−1 °C−1) is the specific heat capacity of air taken as 1013 kJ kg−1 °C−1, ∆(kPa
°C−1) is the rate of change of the saturation vapor pressure with temperature, es (kPa) is
the saturation vapor pressure, ρa is mean air density at constant pressure, ea (kPa) is the
actual vapour pressure, rs (s m−1) and ra (s m−1) are surface and aerodynamic resistances
respectively.
Not only actual evapotranspiration but also FAO potential evapotranspiration (Allen et
al., 1998) derived from the Penman-Monteith formula (Equation 10) is used as a standard
in hydrology. PET represents water demand (stress) of the hydrological system being also
the upper limit of evapotranspiration (E). The FAO formula is expressed
Water resources of arid areas 370
(11)
where PET (mm/day) is potential evapotranspiration, u2 (m s−1) and Ta (°C) are wind
speed and mean daily air temperature at 2m respectively and other notations are as
described earlier. Basic assumptions in the formulation of Equation 11 is that the surface
resistance (rs)=70 (s m−1) and aerodynamic resistance (ra)=208/u2 (s m−1).
The PET Penman-Monteith formula (unlike other potential evapotranspiration
methods takes into account most parameters that affect evapotranspiration. Most of the
parameters necessary to calculate PET in the study area according to Equation 11 were
either available or could be defined by regression analysis. This allowed assessment of
PET at GS01–GS08 ADAS locations. Similar assessment of E as per Equation 10 is by
far more difficult because of ra and rs parameters.
ra determines the transfer of heat, momentum and water vapour from an evaporating
surface into the air above the vegetation canopy and is inversely proportional to wind
speed and changes with height covering the ground (Maidment, 1993). The ra is
expressed as
(12)
where ra (s m−1) is aerodynamic resistance, d (m) is the zero plane displacement height, uz
(m s−1) is wind speed at a measurement height z (m), zoh (m) is the surface roughness
length for heat transfer and water vapor, which is approximated as 10% of zom, where zom
(m) is defined as the roughness length for momentum transfer. d and zom can be estimated
from other parameters following Raupach (1994).
Several attempts are made in the literature to evaluate rs by means of empirical rules
(e.g. O’Toole and Real, 1986). One such an attempt is the so-called Jarvis type models
(Jarvis, 1976: Stewart, 1988: De Rooy & Holtslag, 1999), in which stomatal (canopy)
resistance is expressed as a minimum rs multiplied by a series of independent stress
functions combined in a multiplicative way, through which each function is representing
the influence of each factor. The main weakness of Jarvis type models is the assumption
that environmental factors operate independently (Monteith, 1995). Another way in
which rs can be estimated is through the inversion of the Penman-Monteith equation
(Equation 13), in which the actual evapotranspiration is considered as known input
parameter (Gash & Stewart, 1975), obtained by other methods (e.g. the temperature
profile and Bowen ratio approaches).
(13)
The surface resistance can also be estimated by substitution of stand transpiration (Ts)
derived from sap flow measurements in place of λE in the inverted Penman-Monteith
equation (Equation 13). This procedure is however practically valid only for dry season
estimates of Ts when the assumption is E=Ts can be made.
As mentioned, in the study area there are a number of ADAS towers monitoring various
hydrological variables. The most important with regard to evapotranspiration are:
radiation, temperature, relative humidity and wind speed.
comparative analysis of evapotranspiration records in GS00 and GS09 shows also that in
most cases (with a few exceptions) the daily actual evapotranspiration rates were higher
at GS00 (hardveld area) than at GS09 (Kalahari sandveld area). This most likely must
have resulted from the larger groundwater evapotranspiration i.e. groundwater root
extraction and upward convection-diffusion of groundwater (lubczynski, 2000) at the
hardveld area where groundwater table was much shallower (often <10m) than in the
Kalahari sandveld area where groundwater was generally deep in order of 70m b. g. s.
Under thick sandveld unsaturated zone, covered by extensive savanna vegetation, the
chances of groundwater evapotranspiration, if present are lower and if so arise solely
from deep tree root extraction such as e.g. of Boscia albitrunca.
Table 2. Minimum and maximum potential
evapotranspiration rates for GS01-GS08 situated on
Kalahari.
2002–03 dry season 2001–03 wet season
daily potential daily potential
evapotranspiration evapotranspiration
(mm/day) (mm/day)
Monitoring Minimum Maximum Minimum Maximum
tower rate rate rate rate
GS01 0.08 6.20 0.76 7.43
GS02 0.01 3.85 0.38 7.59
GS03 0.35 6.38 1.04 7.90
GS04 0.24 6.92 0.90 7.72
GS05 0.55 6.19 1.21 8.11
GS06 0.32 6.16 1.14 8.01
GS07 0.05 5.33 0.78 7.40
GS08 0.03 4.62 0.70 6.85
Monitoring of evapotranspiration on Kalahari, Serowe case study, Botswana 377
7 CONCLUSIONS
The actual evapotranspiration rates obtained by solving the energy balance equation for
GS09 site situated on Kalahari range from 0.01–2.09mm/day, whilst actual
evapotranspiration rates for the hardveld area (GS00) range from 0.01–3.74mm/day.
Actual evapotranspiration rates were generally high at GS00 than on the Kalahari situated
GS09 site. PET is largely spatially variable in the study area; it ranges from 0.01–
8.11mm/day. This variability depends largely on the widely unavailable wind speed
characteristics. However, the correlations between half-hourly wind speed measurements
on Kalahari sandveld showed satisfactory correlation range of 0.63–0.87. This allowed
for reasonably accurate extrapolation for calculation of Penman-Monteith PET in seven
(GS01–07) ADAS locations. The applicability of such extrapolation with regard to actual
evapotranspiration is still being tested mainly because of the difficulties in estimates of rs.
The inversion of Penman-Monteith formula with actual evapotranspiration considered as
dependent variable allowed the calculation of rs for GS00 and GS09 locations. The
comparison of rs and ra derived directly from Equation 12 for the Kalahari and hardveld
areas indicate that aerodynamic resistances were generally higher on the hardveld than on
the sandveld site whereas the surface resistances were indicating an opposite trend.
ACKNOWLEDGEMENTS
This study was conducted as part of a project entitled Kalahari Research Programme that
was cooperatively funded by Botswana Geological Survey and ITC, in The Netherlands.
We appreciate the field assistance of Mr. Ramotsoko and his technicians in collecting
data for this study.
REFERENCES
Allen, R.G., Pereira, L.S., Raes, D. & Smith, M. 1998. Crop evapotranspiration, guidelines for
computing crop water requirements, Food and Agriculture Organization of the United Nations.
FAO Irrigation and Drainage paper 56: Italy.
Bastiaanssen, W.G.M. 1995. Regionalisation of surface flux densities and moisture indicators in
composite terrain, a remote sensing approach under clear skies conditions in Mediterranean
climates. PhD thesis, Wageningen Agricultural University: The Netherlands.
Bhalotra, Y.P.R. 1987. Climate of Botswana, Part 2: Elements of Climate. Department of
Meteorological Services, Ministry of Works, Transport and Communications: Botswana.
Choudhury, B.J. 1997. Global pattern of potential evapotranspiration calculated from the Penman-
Monteith equation, using satellite and assimilated data. Remote sensing of the Environment
61:64–81.
De Rooy, W.C. & Holtslag, A.A.M. 1999. Estimation of surface radiation and energy flux densities
from single level weather data. Journal of Applied Meteorology 38:526–540.
Ecosurv Botswana 1998. Vegetation mapping and ground truthing for radar imagery. Ecosurv, 30
September 1998, Gaborone: Botswana.
Water resources of arid areas 380
Gash, J.H.C. & Stewart, J.B. 1975. The average surface resistance of a pine forest derived from
Bowen ratio measurements. Boundary-Layer Meteorology 8:453–464.
Hargreaves, G.H. & Samani, Z.A. 1985. Reference crop evapotranspiration from temperature.
Applied Engineering in Agriculture 1(2):96–99.
Hernandez, A.R. 2002. Mapping of woody vegetation in arid zones-a multi-sensor analysis, a case
study in the Serowe area, Botswana. MSc. Thesis. ITC—International Institute for
Geoinformation Science and Earth Observation: The Netherlands.
Holtslag, A.A.M. & Van Ulden, A.P. 1983. A simple scheme for daytime estimates of the surface
fluxes from routine weather data. Journal of Climate and Applied meteorology 22(4):517–529.
Jarvis, P.G. 1976. The interpretation of the variations in leaf water potential and stomatal
conductance found in canopies in the field. Philosophical Transactions of the Royal Society of
London. Ser. B 273:593–610.
Lubczynski, M.W. 2000. Groundwater evapotranspiration-Underestimated component of the
groundwater balance in a semi arid environment-Serowe case, Botswana. In Sililo et al. (eds),
Groundwater, past and achievements and future challenges: 959–963. Rotterdam: Balkema.
Maidment, D.R.1993. Handbook of Hydrology, McGraw-Hill: Texas.
Monteith, J.L. 1995. Accommodation between transpiring vegetation and the convective boundary
layer. Journal of Hydrology 166:251–263.
Nash, D. 1992. The development and environmental significance of the dry valley systems in the
Kalahari. central Southern Africa. PhD thesis. University of Sheffield: United Kingdom.
O’Toole, J.C. & Real, J.G. 1986. Estimation of aerodynamic and crop resistances from canopy
temperature. Agronomy Journal 78:305–310.
Raupach, M.R. 1994. Simplified expressions for vegetation roughness length and zero plane
displacement as functions of canopy height and area index. Boundary-Layer Meteorology
71:211–216.
Smith, R.A. 1984. The Lithostratigraphy of the Karoo Supergroup in Botswana. Bulletin 26,
Department of Geological Survey: Botswana.
Stewart, J.B. 1988. Modelling surface conductance of pine forest. Agricultural and Forest
Meteorology 43:19–35.
Swedish Geological Survey (SGS) 1988. Serowe Groundwater Resources Evaluation Project,
Final report, Ministry of Mineral Resources and Water Affairs. Department of Geological
Survey: Botswana.
Vossen, P. 1989. An agrometeorological contribution to quantitative and qualitative rainy season
quality monitoring in Botswana. PhD thesis. State University of Ghent: Belgium.
Electro-seismic survey system
S.R.Dennis, M.du Preez & G.J.van Tonder
Institute for Groundwater Studies, University of the Free State, South
Africa
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
Geophysical techniques have been employed for many years to locate groundwater in
South Africa. Magnetic airborne surveys are very useful in the structural mapping of an
area, particularly to determine the dominant direction of tectonic movement, and the
location of features such as faults, dykes and fracture zones. The majority of geophysical
surveys (magnetic, electromagnetic, electric and gravitational) can yield valuable
information on the global geometry of an aquifer. The results are also often ambiguous
especially in Karoo aquifers with its numerous layers of mudstone and siltstone. In many
instances it is necessary to use a combination of these techniques to overcome these
ambiguities. High resolution radio and seismic tomography are two methods that have
shown some promise for groundwater investigations in the Karoo formations.
The majority of the geophysical methods focus on obtaining information concerning
the rock matrix and subsurface structures. The electro-seismic effect on the other hand is
a direct result of relative movement of an electrolytic fluid with respect to the rock
matrix. Thus the presence of water in the porous rock media is directly responsible for the
generation of electro-seismic signals. This has huge cost saving implications due to the
fact that the presence of water could be determined before any drilling has taken place.
This paper discusses the theory surrounding the electro-seismic effect and finally two
case studies are presented to illustrate the interpretation of the electro-seismic survey
results.
Water resources of arid areas 382
2 ELECTRO-SEISMIC EFFECT
2.1 Background
The electro-seismic effect describes the conversion from seismic to electromagnetic (EM)
energy. Several mechanisms are likely to generate couplings between seismic and EM
energy in the subsurface (Garambois & Dietrich, 2002). The main effects of interest to
geophysicists are electrokinetic and piezoelectric phenomena and variations in electrical
resistivity.
The macroscopic governing equations were derived from first principals by Pride
(1994) which coupled Biot’s theory and Maxwell equations via flux/force transport
equations. In this theory the coupling mechanism is explained by electrokinetic effects
taking place at pore level.
3 SYSTEM DESCRIPTION
The Electro-Seismic Survey (ESS) system consists of a probe, a trigger and base units.
The probe unit is a low noise amplifier connected to a 16-bit A/D converter. A horizontal
dipole antenna is connected to the input of the amplifier. The seismic source used is a
hammer and plate. The hammer is connected to the trigger unit. An inertia switch is fitted
to the hammer and acts as the physical trigger. The base unit is connected to a laptop via
the serial port. Custom software has been developed for data acquisition and processing.
The probe and trigger units interface to the base unit via a wireless link to reduce external
noise coupling into the system. The recorded data set represents 200ms which translates
roughly to 200m at the slowest seismic velocity for subsurface media. Practical results
indicates an average maximum penetration depth of 150m when using the hammer and
plate seismic source.
4 METHODOLOGY
Two methods are used to analyse recorded data. These methods are discussed in the
sections below.
Problems experienced were the instability of the model for certain inputs and the fact
that there was little correlation between the simulated and recorded data.
5 CASE STUDIES
In this section two case studies are presented. The first is to illustrate that the principle of
electro-seismic surveys does work and the second case study illustrate the current
capability of the system used.
Currently the analysis technique used is not capable of indicating the main water strike as
related to possible yield estimation.
Now consider the layer plot on the right hand side of Figure 4. Before drilling starts
the layer interfaces can be visualized from the data set obtained but the layer types can
not be identified until after drilling has taken place and a geological log is available.
The main water strike was found at roughly 47m below surface on a dolerite-
sandstone interface and a blow yield of 4L/s was measured. Very good correlation exists
between the borehole log and the layer plot. Only a few layer types are shown on the
layer plot for illustration purposes.
6 SYSTEM LIMITATIONS
The system is very susceptible to power line noise which makes the data analysis difficult
due to the fact that the 50 to 60Hz noise and the associated harmonics fall within the
bandwidth of the electro-seismic response. Various filtering schemes (notch filters and
sinusoidal subtraction) have been tested with limited success because valuable data gets
lost during the filtering process. Data stacking has provided the best results to date of all
schemes tested. For very noisy data 50 point stacking gives good resolution otherwise 10
point stacking is used in the surveys.
Electro-seismic survey system 387
The analysis technique employed does not account for the attenuation of the seismic
signal with depth and this could lead to the misinterpretation of the data because the
strongest signal response does not necessarily indicate the main water strike.
The current analysis technique does not lend itself to yield estimation and further
research on this topic needs to be done to determine if it is feasible or not.
Accurate depth of features is dependant on the seismic velocity used to do depth
scaling. In most instances this information is not available before hand although good
results have been obtained in areas where the average seismic velocity has been
determined from existing borehole logs.
7 CONCLUSIONS
From the results it is clear that the electro-seismic survey does provide accurate results if
the correct seismic velocity is used for depth scaling. The electro-seismic survey provides
valuable information regarding possible water strike positions and physical layering of
the media but no successful yield estimation could be done to date. Results indicate a
localized response from the electro-seismic effect which is in contradiction to articles
written regarding this subject. Further research is needed to fully understand the electro-
seismic effect and the application thereof.
REFERENCES
1 INTRODUCTION
especially during times of drought. The borehole broke down forcing the pastoralists to
move with their livestock to Lonkewan area approximately 10km south during the
drought periods where an operational borehole exists. Due to the influx and concentration
of large numbers of livestock in Lonkewan area, land degradation and pasture depletion
was inevitable. The soils in Lonkewan area were exposed to erosion by wind and surface
run-off. In 1998, drought preparedness intervention recovery programme (D.P.I.R.P), a
non-governmental organization in charge of poverty alleviation for the Samburu pastoral
community, commissioned a contractor to rehabilitate the borehole in Lolmolok area.
This was not successful since metallic objects in the borehole rendered it impossible to
rehabilitate. A replacement borehole was drilled about 1.2km north of the old one down
to a depth of 140m. However, no water was struck up to this depth.
2 GEOLOGY
The investigated area is covered by black cotton soil developed from the weathering of
underlying basalt and/or phonolites. The basalt in the area is underlain by Losiolo and
Rumuruti phonolitic groups, Figure 1 (Hackman 1988). Five flows of the Losiolo
phonolite sequence are recognized on the Loroki plateau. The phonolites are typically
black in colour and fine grained. On weathering, the phonolites attain a purplish grey
colour. However, the uppermost flow is tough, brownish grey and fine grained. Parts of
the flow show alignment of alkali-feldspar phenocrysts and occasionally biotite flakes. It
is pitted where small nepheline phenocrysts have weathered out. Clastic bases and
scoriaceous tops to the flow may be up to 5m thick (Hackman 1988). The individual
thicknesses of these flows have not been established. However, on the eastern shoulder of
the rift valley, the thickness is estimated to be 600m. Metamorphic basement system
rocks are likely to be encountered beyond this depth.
Borehole site investigations in volcanic rocks of Lolmolok area, Samburu district, Kenya 391
The structures within the area deduced from aerial photo interpretation are faults and
fractures which occur both in the basement system rocks and in the volcanics. The
general trend of the major lineaments is roughly north-south. These lineaments are easily
discerned on both aerial photographs and Landsat images at higher altitudes, but are less
discernible in the lower regions where they are obscured by topsoils developed from the
weathering of basalts and phonolites. From the aerial photography interpretation, it can
be concluded that the river channels are structurally controlled as is manifested by the
sudden angular changes in the channel courses (Figure 2).
The investigated area is marked by a medium potential for groundwater abstraction since
it is exclusively volcanic, with a rainfall of about 600mm per annum, which is higher
when compared to that of adjacent areas. This supports the vegetation on the Loroki
plateau. The plateau is marked by three rainy seasons which comprise the monsoon
controlled long rains (March–May) and the short rains (October–November). In addition,
humid air streams with their origin in the west cause a third rainy period in July–August,
known as the continental rains. The continental rains are of great importance as they
provide an extended growing period for crops and pasture development (Flury 1987). The
most distinct dry periods occur from December to February and during the month of
September. Assuming a conservative effective rainfall of 1% and a catchment area of
1480km2 over the Loroki plateau, the available recharge is estimated to be
8.9×106m3/year (Water
Water resources of arid areas 392
rainfall over the Loroki plateau can theoretically sustain approximately 100 boreholes
(WRAP 1991).
The Loroki plateau is drained by two main intermittent streams; Amaya in the further
west of the plateau and Enkare Narok (Figure 2) in the east. A network of seasonal
streams (laggas) collect surface water on the western part of the plateau and drain into
Amaya river. The laggas on the eastern part of the plateau drain into Enkare Narok which
in turn drains into Ewaso Nyiro river. The main NNE-SSW watershed along the central
part of the Loroki plateau divides streams flowing westwards into the Rift valley from the
major eastward flowing river systems. Ewaso Nyiro, Ewaso Narok and Ol Keju Losera
are the largest perennial rivers on the eastern side of the water divide (Figure 2).
Groundwater within the area is located within old land surfaces between lava flows,
fractures, fault zones and contact zones of various rocks. The fresh and massive
phonolites do not contain significant amounts of water. To the east of the plateau, lavas
wedge out against the basement rocks whereas the rift valley shoulder forms the western
boundary. Since the geology, physiography and drainage patterns are uniform over a
large area, a regional aquifer system is expected. There are extensive aquifers of old land
surfaces which are interconnected by fractures and faults. Although the boreholes drilled
in the area are few and their distribution insufficient to ascertain the regionality of the
aquifer system, evidence from similar areas in Laikipia and Meru shows the existence of
a regional aquifer system in the volcanic sheets (Ground water survey 1998,
unpublished).
4 SITE INVESTIGATION
The investigated area is shown in Figure 3. The main objective of borehole site
investigation is to determine the optimum drilling site and reasonable depth of drilling
which should yield sufficient
Borehole site investigations in volcanic rocks of Lolmolok area, Samburu district, Kenya 395
● Processing and interpretation of the data acquired in the field which led to selection of
suitable drill sites, indication of potential yield and depth of aquifers and general
prospects for groundwater development in Lolmolok area.
The VLF-EM survey was executed in order to evaluate the pattern and form of the
underlying geology. Qualitative evaluation of the data enabled the identification of
structural anomalies (fault zones, fractures, lithological contacts) and thicknesses of the
weathered zones. The identified structural anomalies were targeted for detailed
investigations using the VES method. The VES method was used mainly to identify
thickness of the weathered formation and indirectly identify potential water bearing zones
(Beeson and Jones, 1988). VLF-EM profiling was carried out with an ABEM WADI
VLF instrument while ABEM SAS 3000B terrameter was used for VES by application of
Schlumberger array. A total of six VLF-EM profiles and nine vertical electrical
soundings were carried out. Six VES measurements were executed on high conductivity
anomalies identified from the VLF-EM profiles. Two soundings were executed next to
the abadoned borehole (C-2847) and at the dry replacement borehole about 1.2km north
of C-2847. The other sounding (VES 9) was carried out within the study area at a random
location with no apparent conductivity anomalies or borehole for comparison purposes.
Figure 3 shows the locations of VLF-EM and VES traverses in the study area.
The geophysical site investigations in Lolmolok area have been backed by resistivity
data, vertical electrical sounding curves and very low frequency electromagnetic profiles.
The VES data has been processed using Schlumb software which is based on the inverse
filter coefficients of Ghosh (1971). The results of the modeled VES curves are presented
in Figures 4, 5, 6, 7, 8 and 9 below. The VLF-EM profiles carried out along the VES sites
3, 4, 5 and 6 are shown in Figures 10, 11 and 12. VLF-EM profiles along traverses 4, 5
and 6 did not indicate the presence of any subsurface conductive zones and are not
presented in this paper.
6 RECOMMENDATIONS
On the basis of the control sounding carried out at the abadoned borehole (C-2847), four
vertical electrical sounding sites were recommended for drilling within the area initially
set aside by the local community. These include VES sites 6, 5, 4 and 3 (Figure 3) in that
order of priority. Groundwater is expected to occur within the weathered formations,
faults and fracture zones at depths ranging between 30m and 170m below ground level.
From the results and recommendations of this survey, two boreholes herein referred to
as C-2000A and C-2000B were drilled at VES sites 6 and 4 respectively. Borehole C-
2000A was drilled
Borehole site investigations in volcanic rocks of Lolmolok area, Samburu district, Kenya 397
Figure 4. Hydrogeological
interpretation of VES 3.
Water resources of arid areas 398
Figure 5. Hydrogeological
interpretation of VES 4.
Borehole site investigations in volcanic rocks of Lolmolok area, Samburu district, Kenya 399
Figure 6. Hydrogeological
interpretation of VES 5.
Water resources of arid areas 400
Figure 7. Hydrogeological
interpretation of VES 6.
Borehole site investigations in volcanic rocks of Lolmolok area, Samburu district, Kenya 401
Figure 8. Hydrogeological
interpretation of VES 7 next to
borehole C-2847.
Water resources of arid areas 402
Figure 9. Hydrogeological
interpretation of VES 9.
Borehole site investigations in volcanic rocks of Lolmolok area, Samburu district, Kenya 403
7 CONCLUSION
A geological and hydrogeological assessment of Lolmolok area has been carried out
using geophysical investigations. Aerial photo interpretation aided in the identification of
lineaments and other lateral anomalies on the surface. Geophysical exploration
techniques aided in the location of lateral variations and evaluation of thicknesses and
resistivities of rock formations. The results of VES 9 show the effectiveness of combined
geophysical survey techniques in borehole site investigations. In Loroki plateau, aquifers
associated with phonolites have the potential of yielding up to 5m3/hr. Fractured and
contact zone aquifers have the potential of yielding 10m3/hr. Recharge estimated to be
1% of rainfall over the Loroki plateau can sustain up to 100 boreholes (Q=4m3/hr) and
the safe drilling depth is 200m below ground level. Water occurs at various depths
between 30m and 170m below ground level.
REFERENCES
Anyumba, J., Van Dongen, P. & Nzomo, J. 1993. Borehole site investigations in fractured hard
rock aquifers in Gachoka division, Embu district, Kenya. Proc. 5th conference on the Geology
of Kenya, Geol. Soc. Kenya: 116–121.
Beeson, S. & Jones, C.R.C. 1988. The combined EMT/VES geophysical method for siting
boreholes. Groundwater, 26(1):54–63.
Flury, M. 1987. Rainfed agriculture in the central division, Laikipia district, Kenya. Univ. of
Berne, Switzerland.
Borehole site investigations in volcanic rocks of Lolmolok area, Samburu district, Kenya 405
Ghosh, D.P. 1971. Inverse filter coefficients for the computation of apparent resistivity standard
curves for a horizontally stratified earth. Geophys. Prospect., 19:769–775.
Hackman, B.D. 1988. Geology of the Baringo-Laikipia area. Ministry of Environment and Natural
Resources, Mines and Geology dept., 97:1–25.
Mulwa, J.K. 2001. Geological and structural set up of Kiserian-Matasia area and its influence on
groundwater distribution and flow. M.Sc. Thesis, Univ. of Nairobi, Kenya.
Water Resources Assessment Project (WRAP), (1991). Water resources assessment study in
Samburu district; District water development study 1993–2013, part 1, water demand and water
resources. Water Resources Assessment Division, Nairobi, Kenya; TNO-DGV Institute of
Applied Geoscience, Delft, The Netherlands.
Groundwater evaluation in a complex
hydrogeological environment—a GIS based
approach
B.Mudzingwa, J.L Farr, R.Gumiremhete & T.Kellner
Wellfield Consulting Services, Botswana
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
The project location is shown in Figure 1. The project’s aim was to establish groundwater
potential of all aquifers in the study area and to address the key issue of providing
adequate quantities of suitable quality water to support both livestock and the associated
local communities. The area is underlain by sedimentary strata of the Karoo Supergroup
to the north and Proterozoic metasediments (quartzites, shales) of the Transvaal and
Waterberg Supergroups in the south and central parts of the area. Intrusive rocks of the
Molopo Farms Complex (mafic and ultramafic rocks) underlie the southeastern portion of
the area (Fig. 2). Other intrusives found within the project area are dolerites, granites and
Groundwater evaluation in a complex hydrogeological environment 407
Satellite imagery (Landsat ETM+) and aerial photographs were interpreted and then
integrated with regional geophysical data to develop a tectonic model and locate potential
borehole targets.
Geophysical work involved interpretation of regional geophysical data sets (gravity,
aeromagnetics and seismics) and subsequent detailed ground geophysical surveys over
selected zones.
Water resources of arid areas 408
Within the Proterozoic quartzite aquifer, hydrochemistry and recharge mechanisms are
largely controlled by the distribution and interconnectivity of the fracture systems,
structures and Kalahari thickness. Water type is dominantly Na-Cl/Na-Cl-SO4 down the
hydraulic gradient.
Temporal variations in rainfall, recharge rates and groundwater abstraction regimes
are likely to induce temporal changes in water type and quality as seen in Borehole BH
1102 in Figure 3.
All water types have been identified in the project area. Group I—Ca-Mg-HCO3,
Group II—Ca-Mg-Na-HCO3-Cl, Group III—Na-Cl-HCO3, Group IV—Na-Cl and Group
V—Na-Cl-SO4/ Na-SO4-Cl
Group I, II and III are common in areas with thin Kalahari cover, Group II and III are
regarded as a mixture of Groups I and IV. Group V water type is dominant in Karoo to
the north and within the Kalahari occurring along the Molopo River.
Ground water level and rainfall monitoring data was not sufficient to accurately
determine recharge rates as the monitoring period was limited (1 year).
Both newly acquired and historical isotope data from boreholes sampled within the
project area were used for isotope analysis for 18O, 13C, 14C, 2H (deuterium), 3H (tritium)
and N.
The relationship between 18O and deuterium in the study area follows the pattern that
has been seen in other parts of Botswana. The two isotopes fall along an evaporation line
with slope 5.3
Water resources of arid areas 410
show pre-1950 recharged water. 14C ranges from a young 101pmc to values of 2 and
4pmc, the latter certainly representing fossil water. Where both 14C and tritium have been
analysed, the high values for both isotopes are thought to represent recharge. The Molopo
valley exhibit different isotopic signatures from other samples within the project
suggesting different mechanisms of recharge.
Mean Residence Time (MRT) were semi-quantitatively estimated depending on a
combination of 14C, 13C and tritium contents based on the following criteria (Talma &
Tredoux, 2003):
MRT<200 years: 14C>70 pmc and MRT between 1000 and 15000 years: 14C
tritium >1 TU. between 20 and 50pmc and tritium <1 TU,
MRT<1000 years: 14C between 50
and 70 pmc and tritium <1 TU, Fossil water>15000 years: 14C<5pmc.
4 GIS APPLICATIONS
(1)
where is the weighted score of an area (polygon or pixel), Sij is the score of the j-th
class of the i-th map and Wi is the weight of the i-th input map.
possible to delineate a general trend in their spatial distribution. A borehole yield map is
thus useful in showing areas with a high probability of Groundwater Availability.
High groundwater recharge potential areas have higher Groundwater Availability as a
result of regular groundwater replenishment.
The groundwater quality map was classified with much emphasis given to the
distribution of TDS of a certain limit with respect to availability of groundwater suitable
for livestock with a cut-off value of 10,000mg/l, with values above this cut-off figure
designating zones of non-available groundwater.
The Karoo Aquifer had the highest groundwater potential, 12% of the data indicating
yields greater than 20m3/hr, with the Kalahari and Proterozoic Quartzite Aquifers having
values of 9 and 1% respectively.
A Groundwater Availability map was thus produced using the algorithm shown in
equation 2 below.
(2)
where Ywght, Gwqwght, Geowght and Rechwght represent the yield, groundwater quality,
geology and recharge weighted maps and Wy, WGwq, Wgeo and WRech represent the
weights of each of the maps respectively.
After statistical evaluation of the map histogram, the map was then sliced into 3
classes of high, medium and low groundwater availability as shown in Figure 5.
Highest potential is found within the Karoo Aquifer to the north, the Kalahari Aquifer
to the south and along the ENE-WSW ridge which runs through the centre of the project
area. These zones are associated with high borehole yields, high recharge and relatively
moderate TDS. The lowest groundwater availability is found in areas where borehole
yields are very low to dry, TDS is high (generally >10,000mg/l) and recharge potential is
low. This area coincides with thick Kalahari zones; illustrating that Kalahari thickness
has a significant influence on groundwater availability.
(3)
Water resources of arid areas 418
Dwlwght, Dwswght, Obtwght and Afcwght represent the weighted depth to water level, depth to
water strike, overburden thickness and aquifer formation competence maps respectively.
WDwl, WDws, WObt and WAfc represent weights of each of the four maps.
The Groundwater Development Map is shown in Figure 6. Highly favourable to
favourable conditions are found along the ENE/WSW ridge associated with thin Kalahari
cover. The Karoo Aquifer to the north and the Kalahari Aquifer to the south are
characterised by unfavourable groundwater development conditions due to poor aquifer
formation competence, deep water strikes and deep water levels. Unfavourable areas
within the Proterozoic Quartzite Aquifer coincide with thickest Kalahari cover.
Most favourable livestock water prospectivity zones occur in areas characterised by good
groundwater quality, high potential recharge and thin Kalahari cover making
groundwater development most favourable.
5 CONCLUSIONS
REFERENCES
Bonham-Carter, G.F. 1996. Geographic information systems for geoscientists, Modelling with GIS,
Computer Methods, Geosciences, Vol. 13, Canada, Pergamon Press: 398pp.
Groundwater evaluation in a complex hydrogeological environment 419
Carney, J.N. & Aldiss, D.T. & Lock, N.P. 1994. The Geology of Botswana, Bulletin Geological
Survey Botswana. pp37, 113.
Department of Geological Survey, 1999. The National Geological Map of the Republic of
Botswana, Scale 1:1,500000, First Digital Draft Copy.
Department of Geological Survey, 2003. Werda-Mabutsane-Sekoma TGLP Groundwater Survey.
Final Report volume 1 by Wellfield Consulting Services.
Application of 2-D resistivity imaging
combined with time domain electromagnetic
survey to map shallower aquifers in Kunyere
valley, northwest Botswana
E.M.Shemang1, H.Kumar2 & J.Ntsatsi3
1
Department of Geology, University of Botswana, Botswana
2
Water Resources Consultants Ltd, Gaborone, Botswana
3
Department of Water Affairs, Gaborone, Botswana
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
The area of study (Kunyere valley) is located in the North western part of Botswana (see
Figure 1).
Geologically, the area is underlain by sediments of the Kalahari beds which consists of
unconsolidated sands, whose grain sizes range from very fine to medium grained
Application of 2-D resistivity imaging 421
(occasionally coarse grain); clays, silts, calcretes, mudstones, siltstones and sandstones
(DWA, 1997).
The Kunyere valley system comprised of three tributary valleys (Marophe, Xudum
and Matsibe) and the Kunyere Fault. Hydrogeologically, the valley system comprised of
multi-layered fresh water aquifer systems with intervening semi-confining units. Shallow
unconfined and/or semi-confined units comprise the uppermost freshwater aquifer system
which are underlain by an upper semi-confining unit. The lower semi-confined systems
contain fresh water and the fresh/ brackish groundwater interface appears to be within the
fine-medium sands that comprise these lower semi-confined aquifer systems. Figure 2 is
a geologic cross section for the Kunyere Valley oriented southwest to northeast.
A two-dimensional (2-D) Electrical Resistivity Imaging (ERI) survey was conducted
with the objective of assessing if ERI cross-sections would provide better
definition/resolution of aquifer geometry than TEM interpreted sections that were derived
from data obtained at discrete locations.
Two profiles were surveyed near boreholes BH8255/57 and BH9712/13 in the
Kunyere Exploration Area (Figure 1). The first line (Line-A) runs through two
exploration boreholes, BH8255 and BH8257, and stretches for a distance of 1km. The
2 DATA ACQUISITION
TEM soundings in the area were conducted using a Geonics™ time-domain EM unit
comprised of a PROTEM-D receiver and a TEM-47 transmitter. Data was acquired at
base frequencies of 25Hz, 62.5Hz, and 237.5Hz using a square transmitter loop, with a
high frequency receiver coil (area 31.4m2) located at the centre of the loop.
The 2-D ERI data was acquired using an ABEM™ Lund Imaging system using a
400m long multi-core cable with 5m electrode spacing in the Wenner Array. Data was
acquired using continuous, roll-along, vertical electrical sounding method described by
Dahlin and Bernstone (1997). A microcomputer, together with an electronic switching
unit, was used to automatically select the relevant four electrodes for each measurement.
To extend the profile line the roll-along method was employed. This involves moving the
cable past one end of the profile line, by several electrode spacings, after completing a
sequence of measurements.
To plot the 2-D ERI data, a resistivity pseudo-section contouring method is used. In
this method, the horizontal location of the plotting point is placed at the mid-point of the
Application of 2-D resistivity imaging 423
set of electrodes used to take that measurement. The vertical location of the plotting point
is placed at a distance that is proportional to the separation between electrodes.
One-dimensional inversions of the TDEM sounding were carried out using Interpex™
Temix-XL software. The inversion results were then utilized to draw geoelectric cross-
sections along each profile line. An interpretation of each section, in terms of different
hydrogeological units was carried out by grouping the various resistivity ranges. A broad
classification of different hydrogeological units, based on the TEM resistivity, is given in
Table 1.
Inversion of the 2-D CVES field data was carried out using Geotomo™,s
“RES2DINV” software programme. The inversion routine used by this program is based
on the smoothness-constrained least-squares method (De Groot-Hedlin and Constable
1990; Sasaki 1992 and Dahlin, 1996). The output of the inversion routine provide a
pseudo-section of the apparent resistivity and more significantly a true 2-D resistivity
cross-section from which all array geometry factor has been removed along the profile
line. Resistivity variations then reflect true subsurface variations in lithology and water
quality.
Table 1. Resistivity ranges of different
hydrogeological units.
Formation Resitivity
(Ωm)
Dry surficial sand (top 10m) 20–5000
Saturated Sand (Fresh water, TDS 10–70
<1500mg/L)
Saturated Sand (Brackish to saline <10
water, TDS >2000mg/L)
Clayey or Silty Sand within the <3
freshwater zone
Inversion of the TEM soundings provided estimates of the depth to the base of fresh
water aquifers. The results show that the depth to the base of the fresh groundwater
aquifers varies considerably in the area. In some localized areas, the base of the aquifer
occurs up to 100mbgl.
Two-dimensional resistivity imaging sections along Line-A and Line-B with borehole
locations are shown in Figures 3 and 4 respectively. These figures also show the location
of boreholes crossed by these profile lines. With the exception of monitoring borehole
BH9712, all the boreholes are exploration boreholes that have been pump tested. The
tested yield and formation water EC of each borehole is also shown on these figures. The
numbers on the left side of
Water resources of arid areas 424
35m, the ERI sections are unable to resolve different lithological units or the base of fresh
water aquifer, which might be due to insufficient formation resistivity contrasts. The
TEM sounding data also does not resolve different lithological units, however, it does
resolve the base of the fresh water aquifer to be at 68m.
Another feature of interest in this ERI section are the two low resistivity zones that
occur between stations 0–200m and between stations 280–400m. These zones are
separated by vertical boundaries and represent areas with dominantly clay and saline
water. Boreholes BH8255/57 are located near the boundary of a clayey formation/saline
water zone. This example shows advantage of 2-D ERI over TEM soundings in
delineating lateral sub-vertical lithological boundaries cost-effectively. Such features can
not be picked up by conducting TEM soundings at discrete locations and are pertinent for
mapping the aquifer geometry.
The results show that while widely spaced TEM soundings are adequate for broad
delineation of the aquifer there are significant resistivity versus depth variations between
these stations which require a much closer station spacing to ensure their adequate
mapping.
Section Line-B (Figure 4) shows a variably thick (7–35m) resistive layer (>400Ωm)
overlying low resistivity (30–193Ωm) formations. The top layer is interpreted as
dominantly sandy formation that overlies low resistive clayey sand.
5 SUMMARY
REFERENCES
Dahlin, T., and Bernstone, C., 1997, A roll-along technique for 3D resistivity data acquisition with
multielectrode arrays; Proc. Symposium on the Application of Geophysics to Engineering and
Environmental Problems, Reno, Nevada, 927–935.
Dahlin, T., 1996, 2D resistivity surveying for environmental and engineering applications: First
break, 14, 275–284.
deGroot-Hedlin, C., and Constable, S., 1990, Occam’s inversion to generate smooth, two-
dimensional models from magnetotelluric data: Geophysics, 55, 1613–1624.
Water resources of arid areas 426
DWA, 1997, Maun Groundwater Development Project. Phase 1 Exploration and Resources
Assessment. Final Report. pp 134.
DWA, 2003, Maun Groundwater Development Project. Phase 2 Resources Assessment and well
field development. Main report. pp 80.
Loke, M.H., and Barker, R.D., 1996, Rapid least-squares inversion of apparent resistivity
pseudosections by quasi-Newton method: Geophysical Prospecting, 44, 131–152.
Sasaki, Y 1992, Resolution of resisivity tomography inferred from numerical simulation:
Geophysical Prospecting, 40, 453–464.
Theme E:
Climate change and its impact
Hydraulic studies in the design of sand dams
A.S.Nzaba, H.O.Farah & T.C.Sharma
Moi University, Eldoret, Kenya
C.W.M.Sitters
Faculty of Civil Engineering and Geosciences, Delft, The Netherlands
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
A sand dam or subsurface dam is a weir wall placed across a watercourse behind which
successive floods deposit layers of sediment. The sediment absorbs substantial amount of
floodwater which can later be drained for domestic or livestock use. In Kenya the use of
sand dams has been proposed as a viable option for water supply in arid and semi arid
area with seasonal water course and plenty of sediment flowing within the channels. Sand
dams have been utilized in Namibia, Indian, Germany and the semi arid regions of
Arizona and California of U.S.A (Barrow, 1987). The advantage of sand dams is that
evaporation is restricted and contamination of water from the surface is minimal.
A sand dam is built in suitable channels where conditions to create a water tight
subsurface dam exist. Such conditions exist in rocky or stony catchments. The amount of
Hydraulic studies in the design of sand dams 431
water stored or drained depends on the characteristics of the trapped sediments. The
coarser the sediments, the higher the storage capacity and yield of the sand dam. In areas
where significant quantities of fine sediments may be transported during floods, certain
precautions may be taken during construction of the weir wall to ensure that only the
coarse fraction of the sediments is retained in the dam. The weir is raised in stages to
prevent fine sediments from settling in the dam basin. The height of each is carefully
controlled and each new stage is constructed only after the previous stage has effectively
filled up with sediments. Some of the problems that hamper the effective utilization of
sand dams is low capacity of water storage due to low porosity of sediments and low
quantity of extractable water due to low specific yield of the sediments.
The first objective of this study was to investigate the relationship between the
porosity and specific yield of the sediments and step height increment used during
construction of sand dams. The second objective is to investigate ways of trapping
sediments with high water storage and high specific yield.
(2)
(3)
Water resources of arid areas 432
(4)
(5)
where Mc is the Strickler coefficient scale. The following assumptions are made:
(a) The influence of the roughness of channel walls are not taken into account.
(b) Relative intensity of turbulence is the same on the model and the prototype.
(c) Shapes of the grain distribution curve for sediments in prototype and model are
identical.
(d) The shapes of the sediment grains in the model and in the prototype are identical.
From the above equations it is possible to determine all basic scales for the model studies
and from these scales other quantities such as geometric (area and volume), kinematic
(time, velocity and acceleration) and dynamic (pressure, energy and power) can be
determined.
The ratios of width, specific weight of bed load, decisive diameter of bed load mixture
and the Strickler coefficient obtained from corresponding prototype and model properties
are as follows:
MB=163.6, Mγ′s =1.0, MD=1.0 and
Mc=1.9.
The following set of linear equations are obtained from the above ratios:
−2.21=1.51og MD−Log MQ
=Log MD−log ML−Log Ms
0.56=Log ML−1.331og MD
0.02=Log Ms+log MD
0.00=1.5log MD+1.5log MS−log Mq′s
The following values for water discharge, length, depth and specific discharge of bed
load scales result:
MQ=3224.83370, ML=51.4450,
MD=7.3396, MD=0.1427, Mqs′=1.0715
(6)
where V is total volume of soil mass, Vv is the volume of voids, Va is the volume of air
and Vw is the volume of water. Eqn.6 can be re-written as
(7)
where ρB is the bulk density, ρs is the density of the solid particle and Msl is the mass of
solids. In determining η, V was obtained by core cutter method. A miniature core made of
plastic cylinder of specific height and diameter was pressed down into the sediment layer
until it was filled with sample material. The ends of the core were the trimmed flat to the
ends of the cutter by a straight cutting edge. The volume of the soil collected in the
undisturbed state V, was the volume of the core. The sample collected in the core was
oven dried at temperature of between 105°C and 110°C for 24 hours and weighed. This
gave the mass of solids (Msl).
To determine the specific gravity of the solids ρs, complete density bottles were dried
at 105 to 110°C and cooled in a desiccator and then weighed (m1). A 5 to 10g sample of
the solid which pass through a 2mm sieve was oven dried and put in the density bottle
directly from the desiccator. The bottle and its contents were then weighed (m2).
Sufficiently air free distillated water was added so that the soil in the bottle was just
covered with water. The bottle was then placed in a desiccator to eliminate air
completely. More air free distillated water was then added until the bottle was full and the
Water resources of arid areas 434
contents and the bottle are then weighed (m3). Finally the bottle was emptied and cleaned
and filled up with air free distilled water. The bottle is weighed (m4). The specified
gravity of the soil particles is given by
(9)
(10)
where d10, d30 and d60 are grain sizes for which 10, 30, and 60% of grain particles are
smaller than these sizes respectively. A uniformly graded soil will have its Cu less than
3.0 and well graded soil will have Cu greater than 6 and Cc more than 1.0 but less than
3.0.
after the last height increment of the weir. Sampling was done at the inlet, middle and
outlet of the channel.
significance. The letters O, M and I represent the outlet, middle and inlet sections of the
channel respectively. Figure 2 shows that step height increments affect the sizing of
sediments at the outlet and middle sections. The effect of moderately high step increment
of about 10mm to about 30mm (73 to 220mm on the prototype) have more desirable
effect as bigger grain sizes are obtained. However the step height increment has no effect
on the coefficient of uniformity and curvature.
The effects of step height increment on Sr, Sy and η are shown in Table 1. The results
show that step height increment has no effect on Sr, Sy and η.
and more importantly they give highest storage per unit volume (specific capacity). The
discharged should at least be of the order of 65l/s per m width. The upper limit to these
values could not be known because of the scale constraints for the model and finally a
solid weir is better than a meshed or perforated weir.
3 CONCLUSIONS
The sand dams simulation experiments, based on the notions of hydraulic modeling
where a geometrically distorted existing laboratory flume model was used, revealed a
number of trends in order to achieve the best sediments in terms of storage and yield
properties. The following conclusions can be made for dam construction:
● That the step height increment use for design of sand dams should be at least 185mm.
A much higher step height increment of up to 350mm is desirable and more
economical
● The weir used across the water channel for trapping sediment should be of solid type
● The initial channel slope should be high or at least 0.9%
● The flow rate during sediment deposition should be at least 65l/s per m width of the
channel.
REFERENCES
Armfield, 1993. Instruction manual: Sediment transport channel. England, Armfields Ltd.
Barrow, C. 1987. Water resources and agricultural development in the tropics. United Kingdom,
Longman Scientific and Technical Publishers.
Hydraulic studies in the design of sand dams 441
British Standards Institution, 1995. BS 1377: Methods of test for soils in civil engineering
purposes.
Burger, S.W. & Beaumont, R.D. 1967. Sand storage dams for water conservation. CSIR report R.
Meg 329.
Kibiiy, J.K. 2001. Runoff harvesting using sand dams in ASAL areas of west Pokot, Kenya: some
design aspects. PhD thesis, Moi University, Kenya.
Linsley, R.K., Kohler, M.A. & Paulhus, J.L.H. 1982. Hydrology for Engineers. New York,
McGraw-Hill.
Mead, R. 1994. The design of experiments: Statistical principles for practical applications. New
York, Cambridge University Press.
Novak, P. & Cabelka, J. 1981. Models in hydraulic engineering—Physical principles and design
applications. London, Pitman.
Nzaba, S. 2001. Hydraulic similitude studies in the design of sand dams. Msc thesis. Moi
University, Kenya
Thomas, D.B. 1999. Where there is no water—A story of community water development and sand
dams in Kitui district, Kenya. Nairobi, Majestic Printing Works Ltd.
Designing and implementing an aircraft
survey mission using high-resolution digital
multi-spectral camera for vegetation mapping
for upscaling transpiration of Serowe,
Botswana
Y.A.Hussin1, M.W.Lubczynski1, O.Obakeng1,2 & D.C.Chavarro1
1
International Institute for Geoinformation Science and Earth
Observation (ITC), Enschede, The Netherlands
2
Geological Survey of Botswana, Lobatse, Botswana
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
The existence and presents of people, plants, and animals are very much related to water
availability. In arid and semi-arid regions the main source of water is groundwater. It is
well known that vegetation cover has an influence on the hydrological cycle through
interception, infiltration through to ground and transpiration of water. Thus it influences
the amount and quality of ground water. Therefore, vegetation plays an important rule in
hydrological cycle. Mapping vegetation cover especially in semi-arid region like
Botswana is very important and can help as a good indicator of ground water since water
is scarce in this country. The growing human population leads to increasing demand for
water, both in terms of quantity and quality, especially for drinking water. Therefore,
water for people and for the environment is an important issue to be considered in the
framework of sustainable use of the resources.
Designing and implementing an aircraft survey mission 443
Having in mind all what has been said in the above, different types or species of
vegetation would deals differently with the issue of influencing the hydrological cycle by
intercepting, infiltrating through to ground and transpiring of water.
The primary goal of the vegetation mapping effort is to classify vegetation to different
ecosystem, communities and species. Vegetation mapping has been developed because of
the demand for obtaining more detailed map about species habitat distribution,
community mapping and in general its resources. In this process the availability of
suitable sensors capable of recording images in a suitable spectral and spatial resolution
is an important tool in the analysis of vegetation mapping.
The process of vegetation mapping can be greatly aided by interpretation of air and
satellite-based images. These images (often captured in multiple spectral bands) can be
used to delineate a variety of vegetation types based upon species and/or structure.
Further, visual interpretation or digital processing of multi-spectral data can help
delineating certain features of interest of vegetation (e.g., vegetation types, species,
canopy size and structure, density, biomass, leave area coverage), thus aiding to the
mapping process.
A flight campaign was planned to collect very high spatial resolution (e.g. 30, 60 and
100cm) aircraft images of an area in Serowe, Kalahari, Botswana using the multi-spectral
digital camera TETRACAM for transpiration mapping by up scaling of sap flow
measurements.
The objective of this paper was to design and implement an aircraft survey mission
using digital multi-spectral camera for vegetation mapping for upscaling transpiration of
part of Serowe, Botswana.
2 STUDY AREA
The study area is located in the Central District, about 275km NE of Gaborone the capital
of Botswana. Topography is gentle, which varies from 1060 meters above sea level to
approximately 1240. It is characterized to be lower in the east and southeast of the region,
and the highest location in the vicinity of the escarpment edge. From these ones the
average slope is 5% and it gradually decrease to less than 1% towards the east and
southeast.
Soils units, which can be found in that region, are related to arenosols, regosols,
lixisols, luvisols and vertisols. Arenosols are the most common soil units in the study
area. It has low moisture retention capacity than the other soil units.
Climate is a semi-arid with a mean annual rainfall of 447mm. Rainfall occurs mainly
in the summer followed by a dry winter season. Summer season stretches from October
to April and the winter begins in May to September (Tyson, 1986; Obakeng, 2000).
Main vegetation type is thought that belong to the Northern Kalahari Tree and Bush
Savanna. Trees are mostly of Acacia specie, which are characterized by the marked
tendency to occur in cluster, and are normally accompanied by a variety of grass species
such as Ariatida and Eragrotis. Vegetation communities are determined by location on
either sandveld or hardveld areas. Dense vegetation is found within and along river
courses. This suggests that the vegetation density is governed by the availability of water,
which may be partly controlled by topography and geomorphology (Obakeng, 2000).
Water resources of arid areas 444
The airborne multi-spectral data was collected using TETRACAM multi-spectral digital
camera, which collects its data in three spectral bands namely red, green and near
infrared. The data is collected in a rectangular frame (e.g. image) of 1280×1024 pixels.
The size of the pixel (e.g. ground resolution) is depending on the altitude of the aircraft
above the ground. Therefore, the camera was mounted on a small aircraft (e.g. Cessna
210) (Figures 1 and 2). The camera saves the image in DCA format (Digital Camera
Format), which is a compressed file that can be
(Table 1). The image data is stored in what is known as the flash cards of 256MGB. Then
the
ACKNOWLEDGMENT
This research work was partly supported by the internal research fund of GWFLUX
Project at ITC. However, Botswana Geological Survey (BGS) has offered the main
financial support of the aerial survey missions, fieldwork logistics and transportation. The
authors appreciate and acknowledge the support of Botswana Geological Survey.
REFERENCES
Obakeng, O.T. 2000. Groundwater recharge and vulnerability: A case study at the margins of the
south-east Central Kalahari Sub-basin, Serowe region, Botswana. Unpublished MSc, ITC—
International Institute for Geoinformation Science and Earth Observation, Enschede.
Tyson, P.D. 1986. Climatic Change & Variability in Southern Africa. Cape Town, South Africa:
Oxford University Press.
Relevance of groundwater interaction with
surface water to the eco-hydrology of semi-
arid regions
John Y.Diiwu
Alberta Research Council Inc., Vegreville, Canada
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
Groundwater systems are not isolated from surface water systems, but are in continuous
dynamic interaction at local, intermediate and regional scales. The degree of the
interaction between groundwater and surface water depends on physiographic and
climatic conditions. Irrespective of the degree of the interaction between the two systems,
development and/or contamination of one ultimately affects the other, and hence the
entire ecosystem (Lamontagne et al., 2003). An understanding of the basic principles of
Relevance of groundwater interaction 451
the interactions is therefore needed for effective management of water resources. This is
even more imperative for semi-arid regions where water resource systems are highly
vulnerable due to climate change and anthropogenic activities.
Interest in the relationship of groundwater with streams, lakes, wetlands and estuaries
increased in recent years due to concerns about acid rain, eutrophication, and the
disappearance of coastal ecosystems as a result of development (Winter, 1995). In the last
two decades, attention has been focused on exchanges between near-channel and in-
channel water, which are necessary for evaluating the ecological structure of streams and
for designing stream restoration and riparian zone management programmes. The need
for a holistic approach to environmental protection has heightened the attention of
ecologists, geoscientists, and watershed managers to groundwater interaction with surface
water.
The partitioning of precipitation into surface runoff, infiltration and potential recharge/
discharge is highly variable in space and time in semi-arid regions. Understanding the
spatial and temporal variability of these processes at a range of scales improves our
ability to quantify and manage the available water resources. The recharge/discharge
component, which links groundwater and surface water systems has received renewed
attention in the last few decades. This paper is therefore intended to review fundamental
ecological and hydrological concepts useful for understanding groundwater interaction
with surface water, discuss the relevance of the interaction to the ecology of semi-arid
regions, and provide information for further studies of this important pathway between
groundwater and surface water systems.
Surface and subsurface water interactions occur by subsurface lateral flow through the
unsaturated soil and by infiltration into or exfiltration from the saturated zones. Also, in
the case of karst or fractured terrain, interactions occur through flow in fracture or
solution channels. In general, subsurface flow through porous media is sluggish. The
mechanisms by which subsurface flow enters streams quickly enough to contribute to
streamflow responses to individual rainstorm and snowmelt inputs are discussed in the
literature (Beven, 1989; Winter, 1995 and Lamontagne et al., 2003). In particular, Beven
(1989) identifies four mechanisms that account for fast subsurface contributions to the
storm hydrograph: translatory flow, macropore flow, groundwater ridging, and return
flow.
Translatory flow, also known as plug flow or piston flow (Hewlett and Hibbert, 1967),
is easily observed by allowing a soil column to drain to field capacity and then slowly
adding a unit of water at the top. It would be observed that some water flows from the
bottom immediately, but this is not the same water that was added at the top. Macropore
flow is fast flow through larger noncapillary soil pores, resulting in rapid subsurface
responses to storm events (Beven and Germann, 1982). Groundwater ridging describes
the large and rapid increases in hydraulic head in groundwater during storm events
(Sklash and Farvolden, 1979). As a result, an increase occurs in the net hydraulic gradient
toward the stream and/or the size of the seepage face, thus enhancing fluxes to the
Water resources of arid areas 452
stream. The streamflow contribution induced thereby may greatly exceed the quantity of
water input that induced it. Return flow is the discharge of subsurface water to the
surface. This may result if the water table and capillary fringe are close to the soil
surface, such that small amounts of applied water are necessary to saturate the soil
surface completely (Dunne and Black, 1970). The response of any particular watershed
may be dominated by a single mechanism or by a combination of mechanisms, depending
on the magnitude of the storm event, the antecedent soil moisture conditions in the
watershed, and/or the heterogeneity in soil hydraulic properties in the watershed (Sklash
and Farvolden, 1979).
Large scale exchange of groundwater with surface water is controlled by the distribution
and magnitude of hydraulic conductivities, both within the channel and the associated
alluvial plain sediments; the relation of stream stage to the adjacent groundwater level;
and the geometry and position of the stream channel within the alluvial plain (Woessner,
2000). The direction of the exchange processes varies with hydraulic head, whereas flow
depends on sediment hydraulic conductivity. Storm events and seasonal patterns alter the
hydraulic head and thereby induce changes in flow direction. Two net directions of flow
are: the influent condition where surface water contributes to subsurface flow (losing
stream, Figure 1), and the effluent condition where groundwater drains into the stream
(gaining stream, Figure 2). On one hand, variable flow regimes could alter the hydraulic
conductivity of the sediment via erosion and deposition processes and thus affect the
intensity of groundwater interaction with surface water (Brunke and Gonser, 1997).
During periods of low precipitation, baseflow in many streams constitutes the discharge.
On the other hand, under conditions of high precipitation surface runoff and interflow
gradually increase, resulting in higher hydraulic pressures in the lower stream reaches,
which cause the river to change from effluent to influent condition, infiltrating its banks
and recharging the aquifer. Thus,
5 QUANTITATIVE ANALYSIS
The flow of water on the surface, and in the unsaturated and saturated zones is driven by
gradients from high to low potentials. The hydraulic connection between the stream and
groundwater may be direct, as shown in Figures 1 and 2 above. On the other hand, it may
be disconnected by an intervening unsaturated zone, with streams losing water by
Water resources of arid areas 456
seepage through a streambed down to a deep water table, as shown in Figure 3 above.
The degree of connection can change over different reaches within any one stream and
from time to time over the same reach.
For hydraulically connected stream-aquifer systems, the resulting exchange flow is a
function of the difference between the river stage and aquifer head. A simple approach to
estimate flow is to consider the flow between the river and the aquifer to be controlled by
the same mechanism as leakage through a semi-impervious stratum in one dimension
(Rushton and Tomlinson, 1979). This mechanism, based on Darcy’s law, where flow is a
direct function of the hydraulic conductivity and head difference, can be expressed as:
q=k∆h
(1)
Where, ∆h=ha−hr (ha is aquifer head, and hr is river head/stage); q is flow between the
river and the aquifer (positive for baseflow for gaining streams, and negative for river
discharge for losing streams); and k is a constant representing the streambed leakage
coefficient (hydraulic conductivity of the semi-impervious streambed stratum divided by
its thickness). Equation (1) can be used to represent both baseflow and river discharge,
even though in practice, the mechanisms representing the two processes can be different.
At times of high recharge, the leakage calculated by the linear relationship in Equation
(1) is much greater than would occur in practice and takes no account of water as its
volume increases. For such increased resistance to flow Rushton and Tomlinson (1979)
propose a nonlinear relationship of the form:
q=k1[1−exp(−k2∆h)]
(2)
Where k1 and k2 are constants. In cases where the suggestion of a maximum flow rate is
not acceptable, Rushton and Tomlinson (1996) propose a combination of linear and
nonlinear relationships of the form:
q=k1∆h+k2
[1−exp(−k3∆h)] (3)
6 ECOLOGICAL SIGNIFICANCE
In semi-arid regions where intense runoff occurs in a relatively short periods of time,
closed topographic depressions of varying sizes are filled by runoff to form ephemeral
ponds or wetlands. Playas in arid and semi-arid regions are some examples of such
Relevance of groundwater interaction 457
ephemeral ponds (Gordon et al., 1992; Brunke and Gonser, 1997). As the water level in a
pond occupying a depression rises in response to input from overland flow and
streamflow, water flows from the pond to groundwater where the adjacent groundwater
level is lower than the pond. The period of standing water in the depression affects the
species richness of aquatic invertebrates, amphibians, and their predators. From a study
of 22 wetlands Snodgrass et al. (2000) found that amphibian species richness increased
with increase in the duration of standing water in the wetlands, but found no significant
relationship between species richness and wetland size. In semi-arid regions, intense
runoff coupled with high evapotranspiration produces wetlands with intermediate
duration of interaction with groundwater. This is crucial for biodiversity because such
wetlands maintain high productivity by periodic drying, which results in routine recycling
of organic materials and nutrients (Gordon et al., 1992; Snodgrass et al., 2000).
The hyporheic zone, as shown in Figure 4 above, is a mixture of surface water and
groundwater, and so has physical and chemical characteristics considerably different
from stream water. The zone is therefore an ecotone between the surface environment
characterized by light, high dissolved oxygen, and temperature fluctuation and the
groundwater environment characterized by darkness, less oxygen, and stable temperature
(Gilbert et al., 1994). Invertebrates living in the hyporheic zone exploit the groundwater
environment to varying degrees. Some species spend their entire life cycle in the
hyporheic zone, while others spend their egg and laval stage in the zone, and then move
to the surface environment to spend their adult life. A third category of species use the
hyporheic zone only to seek protection from unfavorable situations (Gilbert et al., 1994).
The food web of the hyporheic zone is fueled by the heterotrophic microbial communities
which depend on dissolved oxygen provided by surface water exchange, particulate
organic carbon, and dissolved organic carbon in nutrient-rich groundwater. The microbes
provide food for grazers, which in turn provide food for invertebrate predators. Dissolved
organic carbon stored in the hyporheic zone can serve as a food source when it is not
readily available in surface water, and therefore has a crucial influence on the metabolism
of the fluvial ecosystems (Brunke and Gonser, 1997).
The hyporheic zone provides a number of ecologically important services. When
surface water recharges groundwater, there is opportunity for organic pollutants and
detritus to become trapped in the sediment. The bacteria may then catalyze reactions that
could change the chemicals into less toxic forms or into available nutrients. For instance,
in contaminated aquifers many bacterial micro-organisms residing in groundwater and
sediment interstices can aid in groundwater remediation by degradation and
denitrification (Gilbert et al., 1994). During floods, excess water that enters bank storage
may percolate to recharge groundwater or may re-emerge at a different location in the
watershed and at a different time. These diversions allow the onslaught of water into
streams to be delayed by days, weeks, or even months and thus mitigates the effects of
flood flows (Winter, 1995; Brunke and Gonser, 1997). The interaction of groundwater
with surface water within the hyporheic zone also has a thermal service. Since
groundwater temperatures remain relatively constant, the water that discharges tends to
be cooler than surface water in semi-arid regions. The hyporheic zone therefore serves as
a thermal refuge for fish and other aquatic species in semi-arid regions. The zone also
serves as a habitat for micro-organisms, macro-invertebrates, fish and wildlife; provides
flow augmentation; refugia for endangered aquatic species under conditions of increased
Water resources of arid areas 458
fragmentation and degradation of aquatic habitat; and food source for fish in surface
water ecosystems and organic matter for microbial activity in groundwater ecosystems
(Winter, 1995). Surface water moving into groundwater is one of the ways in which
microorganisms may colonize groundwater environments. The presence or absence of
certain groundwater species may indicate the location of groundwater-surface water
interaction zones and a decline in the diversity of groundwater species may indicate a
decline in water quality (Gilbert et al., 1994). Groundwater invertebrates and micro-
organisms are an important food source for fish, and so the interaction of groundwater
with surface water, which determines the availability of such organisms, has the potential
to affect the viability of native fish populations (Gilbert et al., 1994).
Valley bottoms in semi-arid regions often serve as desirable areas of grazing and
agriculture because of continuous availability of soilwater in the unsaturated zone and
hence green pasture throughout the year. While these areas have the ability to introduce
the cooling effects of groundwater to surface water and continuously make soilwater
available in the unsaturated zone, they are also easily degraded by mismanagement.
Grazing and agriculture may cause accelerated erosion and soil compaction in the valley
bottoms, thus causing the permanent loss of such vital components of the ecosystem in
semi-arid regions (Gilbert et al., 1994).
In semi-arid regions, crop production requires consumptive use of large quantities of
water. Water, which is already scarce must be shared among several consumptive as well
as non-consumptive uses. Consequently, society faces serious water management
problems. The decline of groundwater levels due to over-pumping ultimately results in
reduced baseflow, which would have discharged into surface water to sustain aquatic life
during periods of low flow. At sufficiently large pumping rates, these declines induce
flow out of the body of surface water into the aquifer, and this leads to streamflow
depletion. As discussed in a previous section, groundwater-surface water interactions are
also important in situations of groundwater contamination by polluted surface water, and
in situations of degradation of surface water by discharge of saline or other low quality
groundwater. An understanding of groundwater-surface water interaction in semi-arid
regions is therefore important for the sustainable management of water resources in those
regions.
8 RESEARCH NEEDS
discharge and recharge along active channels in varying geomorphic settings needs to be
further investigated (Winter, 1995). Jones and Holmes (1996) concludes that whereas
surface-hyporheic exchanges and water residence times are known to be important
regulators of subsurface biochemical transformations, the manner in which these
parameters vary across streams and under different climatic conditions, such as semi-arid
regions is not yet known.
The effect of heterogeneity on water fluxes in general, and specifically between
groundwater and surface water is still a major challenge. The hydraulic properties of
stream and lake beds control the interactions between groundwater and surface water
systems, but these properties are normally difficult to measure directly. The primary
limitation has so far been the difficulty of spatially defining the hydraulic properties and
heterogeneities of a stream and lake beds. Streambed clogging and stream partial
penetration are factors which are equally important as heterogeneity. All these factors
need to be considered during analytical treatments of groundwater-surface water
interactions (Jones and Holmes, 1996). Moreover, the relative importance of streambed
clogging, stream partial penetration and heterogeneity under semi-arid conditions needs
to be further investigated.
At the current state of research, most techniques and models developed for
groundwater-surface water interactions were based on information from humid regions
(Winter, 1995). There is therefore a need to revise such techniques and models utilizing
both in-situ and remote sensing observations from semi-arid regions. These techniques
also need to be coupled with Geographic Information Systems (GIS) technology and
statistical analysis to study groundwater-surface water interactions in semi-arid regions in
a multidisciplinary and multiscale approach.
9 CONCLUSIONS
The realization that hydrological and ecological settings are inter-related has prompted
the coining of the term “ecohydrology” to describe this inter-relationship (Wassen and
Grootjans, 1996). Baird and Wilby (1999) provide several examples from a range of
environments on how exchange between groundwater and surface water affects interface
ecology, and how biological communities affect groundwater-surface water exchanges.
Several studies investigating the advantages of the inter-relationship have also been
reported in the literature. However, there are still many gaps in our understanding of the
processes involved in groundwater-surface water interactions, and the environmental
implications of the exchanges. The boundaries between hydrological and ecological
research are gradually disappearing, yet a need remains for closer collaboration between
these traditionally distinct disciplines, and among researchers working in different
climatic regions so that research results may be pooled and applied to the benefit of the
global environment, such as for the sustainable management and utilization of natural
resources.
Water resources of arid areas 460
REFERENCES
Baird, A.J. & Wilby, R.L. 1999. Eco-Hydrology: Plants and Water in Terrestrial and Aquatic
Environments. New York, Routledge Press.
Baxter, C.V. & Hauer, F.R. 2000. Geomorphology, hyporheic exchange and selection of spawning
habitat by bull trout (Salvelinus confluentus). Canadian J. Fisheries and Aquatic Sciences
57:1470–1481.
Beven, K.J. 1989. Interflow. In: H.J.Morel-Seytoux (ed) Unsaturated flow in hydrologic modeling:
theory and practice, Kluwer, Dordrecht: 191–219.
Beven, K.J. & Germann, P.F. 1982. Macropores and water flow in soils. Water Resour Res,
18:1311–1325.
Brunke, M. & Gonser, T. 1997. The ecological significance of exchange processes between rivers
and groundwater. Freshwater Biol. 37:1–33.
Carrere, R. 1996. Pulping the South: Brazil’s pulp and paper plantations. Ecologist, 26:206–214.
Dahm, C.N. & Valett, H.M. 1996. Hyperheic zones, In: F.R. Hauer & G.A. Lamberti. (eds).
Methods in Stream Ecology. San Diego, California, Academic Press: 107–119.
Dunne, T. & Black, R. (1970). An experimental investigation of runoff production in permeable
soils. Water Resour Res 6:478–490.
Gilbert, J., Danielpool, D. & Stanford, J.A. 1994. Groundwater Ecology, San Diego, California,
Academic Press.
Gordon, N.B., McMahon, T.A. & Finlayson, B.L. 1992. Stream Hydrology: An Introduction for
Ecologists, Chichester, Wiley.
Hewlett, J.D. & Hibbert, A.R. 1967. Factors affecting the response of small watersheds to
precipitation in humid areas. In: W.E.Sopper and H.W.Lull (eds) Proc Int. Symp on Forest
Hydrology, Oxford, Pergamon Press: 275–290.
Hoehn, E. 1998. Solute exchange between river water and groundwater in headwater environments.
In: Proc. Headwater ‘98 Conf Hydrology, Water Resources and Ecology in Headwaters,
Meran/Merano, Italy, IAHS Publ 248, Wallingford, 165–171.
Jones, J.B. & Holmes, R.M. 1996. Surface-subsurface interactions in stream ecosystems. Trends
Ecol Evol, 16:239–242.
Lamontagne, S., Herczeg, A.L., Dighton, J.C. & Pritchard, J.L. (2003). Groundwater-surface water
interactions between streams and alluvial aquifers: Results from the Wollombi Brook (NSW)
Study, Part II-Biogeochemical Processes. CSIRO Land and Water Technical Report 42/03.
Lee, D.R. & Cherry, J.A. 1978. “A field exercise on groundwater flow using seepage meters and
minipiezometers”. J. Geol. Educ., 27:6–10.
Rushton, K.R. & Tomlinson, L.M. 1979. Possible mechanisms for leakage between aquifers and
rivers. J. Hydrol 40:49–65.
Sklash, M.G. & Farvolden, R.N. 1979. “The role of groundwater in storm runoff’. J. Hydrol,
43:45–65.
Snodgrass, J.W., Komoroski, M.J., Bryan, A.L.J. & Burger, J. 2000. Relationships among isolated
wetland size, hydroperiod, and amphibian species richness: Implications for wetland
regulations. Conserv Biol. 14: 414–419.
Wassen, M.J. & Grootjans, A.P. 1996. Ecohydrology: An interdisciplinary approach for wetland
management and restoration. Vegetation, 126:1–4.
Winter, T.C. 1995. Recent advances in understanding the interaction of groundwater and surface
water. Rev Geophys (Suppl):985–994.
Woessner, S.M. 2000. Stream and fluvial plain groundwater interactions: rescaling hydrogeologic
thought. Groundwater, 38:423–429.
Impacts of climate change in water resources
planning and management
Alfred Opere
Department of Meteorology, University of Nairobi, Kenya
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
Water is one of our most important natural resources. Without it, there would be no life
on earth. The lifestyle we have become accustomed to depend heavily upon having plenty
of cheap, clean water available as well as an inexpensive, safe way to dispose of it after
use.
The supply of water available for our use is limited by nature. Although there is plenty
of water on earth, it is not always in the right place, at the right time and in the right
quality. Adding to the problem is the increasing evidence that chemical wastes
improperly discarded yesterday are showing up in our water supplies today.
Freshwater is a crucial resource for sustainable development. Considering the current
situation and the multifaceted dimensions of the water crisis, there is acute need for
action—no time to waste:
Water resources of arid areas 462
1. More than 1 billion people worldwide lack access to safe drinking water and close to 3
billion are not provided with adequate sanitation.
2. Ecosystems are harmed as a consequence of pollution or of drying out.
3. People and nations are competing for scarce and finite water resources.
Moreover the problems of variability of precipitation and climate, changing human
settlement and land use changes as well as exploitation of natural resources caused a
considerable increase of the number of catastrophes, such as floods and droughts, over
the past 50 years.
Of the 19 countries around the world currently classified as water-stressed, more are in
Africa than in any other region, and this number is likely to increase, independent of
climate change, as a result of increases in demand resulting from population growth,
degradation of watersheds caused by land use change and siltation of river basins. A
reduction in precipitation projected by some GCMs for the Sahel and southern Africa, if
accompanied by high inter-annual variability, could be detrimental to the hydrological
balance of the continent and disrupt various water-dependent socio-economic activities.
Variable climatic conditions may render the management of water resources more
difficult both within and between countries. A drop in water level in dams and rivers
could adversely affect the quality of water by increasing the concentrations of sewage
waste and industrial effluents, thereby increasing the potential for the outbreak of
diseases and reducing the quality and quantity of fresh water available for domestic use.
Adaptation options include water harvesting, management of water outflow from dams
and more efficient water usage.
Rainfall is the key input variable that activates flow and mass transport in hydrological
systems, and models for simulating and forecasting rainfall in space and time can play an
important role in enhancing our understanding of the hydrological system response, and
in the design and operation of water resource systems.
Today, we face record consumption, uncertain supplies, and growing demands for
protection from flooding and pollution. The health and economic effects of a shortage of
clean water are matters of great concern. Hydrology has evolved as a science in response
to the need to understand the complex water systems of the earth and help solve water
problems. Hydrologists play a vital role in finding solutions to water problems.
There are many pathways the water may take in its continuous cycle of falling as
rainfall or snowfall and returning to the atmosphere. It may be captured for millions of
years in polar ice caps. It may flow to rivers and finally to the sea. It may soak into the
soil to be evaporated directly from the soil surface as it dries or be transpired by growing
plants. It may percolate through the soil to groundwater reservoirs (aquifers) to be stored
or it may flow to wells or springs or back to streams by seepage. The cycle for water may
be short, or it may take millions of years.
People tap the water cycle for their own uses. Water is diverted temporarily from one
part of the cycle by pumping it from the ground or drawing it from a river or lake. It is
used for a variety of activities such as households, businesses and industries; for
irrigation of farms and parklands; and for production of electric power. After use, water is
returned to another part of the cycle: perhaps discharged downstream or allowed to soak
into the ground.
Used water normally is lower in quality, even after treatment, which often poses a
problem for downstream users. The hydrologist studies the fundamental transport
processes to be able to describe the quantity and quality of water as it moves through the
cycle (evaporation, precipitation, streamflow, infiltration, groundwater flow, and other
components).
The engineering hydrologist, or water resources engineer, is involved in the planning,
analysis, design, construction and operation of projects for the control, utilization, and
management of water resources. Water resources problems are also the concern of
meteorologists, oceanographers, geologists, chemists, physicists, biologists, economists,
political scientists, specialists in applied mathematics and computer science, and
engineers in several fields.
Hydrologists help cities by collecting and analyzing the data needed to predict how much
water is available from local supplies and whether it will be sufficient to meet the city’s
projected future needs. To do this, hydrologists study records of climate such as rainfall,
Water resources of arid areas 464
snow-pack depths and river flows that are collected and compiled by hydrologists in
various government agencies.
Managing reservoirs can be quite complex, because they generally serve many
purposes. Reservoirs increase the reliability of local water supplies.
Deciding how much water to release and how much to store depends upon the time of
year, flow predictions for the next several months, and the need of irrigators and cities as
well as downstream water-users that rely on the reservoir. If the reservoir is also used for
recreation or for generation of hydroelectric power, those requirements must be
considered. Decisions must be coordinated with other reservoir managers along the river.
Hydrologists collect the necessary information, enter it into a computer, and run
computer models to predict the results under various operating strategies. On the basis of
these studies, reservoir managers can make the best decision for those involved.
The availability of surface water for swimming, drinking, industrial or other uses
sometimes is restricted because of pollution. Pollution can be merely an unsightly and
inconvenient nuisance, or it can be an invisible, but deadly, threat to the health of people,
plants and animals.
Hydrologists assist public health officials in monitoring public water supplies to
ensure that health standards are met. When pollution is discovered, environmental
engineers work with hydrologists in devising the necessary sampling program. Water
quality in estuaries, streams, rivers and lakes must be monitored, and the health of fish,
plants and wildlife along their stretches surveyed.
Related problem concerns acid rain and its effects on aquatic life, and the behavior of
toxic metals and organic chemicals in aquatic environments. Hydrologic and water
quality mathematical models are developed and used by hydrologists for planning and
management and predicting water quality effects of changed conditions.
It would be difficult to think of any human activity or interest that is not in some way
affected by weather and climate. And so, it should be a matter of considerable concern
that, if we continue to emit carbon dioxide (CO2), methane (CH4), nitrous oxide (N2O)
and other radiatively active trace gases (the “greenhouse gases”) at current or even
significantly reduced rates, the lower portion of the atmosphere will grow warmer. As a
result, patterns of cloudiness, precipitation, humidity, windiness and possibly the
frequency and severity of extreme weather events to which we are now accustomed will
be different—not necessarily or universally worse, but different. If this be so, then food
and water supplies, the health of ecosystems and of humanity will be affected, maybe for
the better, more likely for the worse.
is quite uncertain. Many of the GCMs indicate that some regions will be drier as a result
of global warming. Some GCMs anticipate increased precipitation for other parts of the
world, such as sub-Sahelian Africa. As a consequence, the chronic droughts that have
long afflicted these regions could become a thing of the past.
5 IMPACTS OF CLIMATE
What follows are some possible (but far less certain) impacts of impending climatic
change.
6 CONCLUSION
There is uncertainty with respect to the prediction of climate change at the global level.
Although the uncertainties increase greatly at the regional, national, and local levels, it is
at the national level that the most important decisions would need to be made. Higher
temperatures and decreased precipitation would lead to decreased water-supplies and
increased water demands; they might cause deterioration in the quality of freshwater
bodies, putting strains on the already fragile balance between supply and demand in many
countries. Even where precipitation might increase, there is no guarantee that it would
occur at the time of year when it could be used; in addition, there might be a likelihood of
increased flooding. Any rise in sea level will often cause the intrusion of salt water into
estuaries, small islands and coastal aquifers and the flooding of low-lying coastal areas;
this puts low-lying countries at great risk.
On a more optimistic note, technological innovations to increase water supply, by
evaporation suppression, water reuse and recycling systems, satellite-guided and
computer-controlled irrigation systems, can help mitigate some of the negative impacts of
climate change. Water demand management through appropriate climate factoring and
institutional adaptation will be the key to increasing flexibility of water resource systems
in the face of climatic change.
REFERENCES
Houghton, J. 1997. Global Warming: The Complete Briefing. Cambridge, Cambridge University
Press.
x
Rosenberg, N.J. & Cooper, C.I. 1982. Likely impacts of a likely global warming (CC’82).
Watson, R.T., Zinyowera, M.C., Moss, R.H. & Dokken, D.J. eds. 1997. 1PCC. Special Report on
the Regional Impacts of Climate Change: An Assessment of Vulnerability. Summary for
policymakers.
White, R.M. 1998. “Kyoto and Beyond,” Issues in Science and Technology, 1998.
Wuebbles, D.J. & Rosenberg, N.J. 1998. “The Natural Science of Global Climate Change.”
Chapter 1 in S Rayner and EL Malone, eds. Human Choice and Climate Change, Vol 2.
Resources and Technology Columbus, Battelle Press, OH, 1–143.
Turning a liability into an asset: the case for
South African coalmine waters
B.H.Usher & F.D.I.Hodgson
Institute for Groundwater Studies, University of the Free State,
Bloemfontein, South Africa
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
1.1.4 Geology
The Karoo Supergroup in the Witbank region comprises the Ecca Group and Dwyka
Formation. The total thickness of these sediments ranges from 0–100m.
The Ecca sediments consist predominantly of sandstone, siltstone, shale and coal.
Combinations of these rock types are often found in the form of interbedded siltstone,
mudstone and coarse-grained sandstone. Typically, coarse-grained sandstones are a
characteristic of the sediments in the Witbank Area. Of all the unweathered sediments in
the Ecca, the coal seams often have the highest hydraulic conductivity. Packer testing of
the No. 2 Seam and underlying Dwyka tillite has a hydraulic conductivity.
Five coal seams, numbered from bottom to top as No. 1–5, are present. Only two of
the seams are mineable over most of the area. These are the No. 2 and 4 Seams, which
are usually separated by sediments of a total thickness in the order of 20–30 m. Seams 1
and 5 are, however, mined locally.
Dolerite intrusions in the form of dykes and sills are present within the Ecca Group.
Faults are rare. However, fractures are common in competent rocks such as sandstone
and coal.
In terms of expected sources of water, the table above (Table 1) from Hodgson and
Krantz summarises, the most important information.
The high recharge percentage of around 20% is due to a multitude of factors such as
ponding, areas of spoils exposure, restabilisation cracks and influx into ramp areas.
Several researchers have confirmed these high recharges through decant measurements
(Hodgson, 1999, van Tonder et al., 2003). The high influx is naturally an important
driver on the observed water quality in opencast pits.
In underground mines the following sources of water could be encountered:
– Water encountered in the seam as mining commences. This is fairly low except where
fractures or fissures as they are known within the mining industry are encountered.
– Recharge through the roof lithologies. The magnitude of this varies depending on
mining induced fracturing of the overlying sediments.
– Direct recharge where cracks from the collapse of mining areas, usually due to high
extraction mining, run through to the higher-yielding transmissive aquifers nearer
surface.
– Regional groundwater flow, which will usually flow along the coal horizon, due to its
higher hydraulic conductivity compared to the surrounding sediments.
– Influx through the floor lithologies. This can play an important role in areas where the
floor is transmissive, but where the mining floor is close to the Dwyka such as where
the No. 1 or No. 2 seam are mined, such influxes are negligible.
Table 2. Key mining and storage statistics for
Mpumalanga coalfields.
Future
Future volume
Current area Total Volume coal to Total
mined to be area to Extraction coal be volume Storage
Mining area mined be height Extraction mined mined to be volume
type (ha) (ha) mined (m) rate (Mm3) (Mm3) removed (Mm3)
U/G 5 7842 7050 14892 2.5 0.65 127 115 242 242
seam
U/G 4 13485 2833 16318 3 0.65 263 55 318 255
seam
U/G 2 98550 39695 138245 3 0.65 1922 774 2696 1887
seam
U/G 1 2525 0 2525 2.5 0.6 38 0 38 30
seam
Opencast 13557 14480 28037 3.5 0.9 427 456 883 221
Totals 135959 64058 200017 2777 1400 4177 2635
The water encountered, as mining continues, will subside over time. In bord-and-pillar
areas where mining has been completed, features such as roof bolts, drilled to stabilise
the roof during mining, act as local sinks for water to drain to. Where such a roof bolt
intersects horizontal or vertical fractures, increased influx is experienced. This is
particularly problematic in areas adjacent to water storage compartments, where seals are
Water resources of arid areas 472
installed to accommodate a head in excess of the mining height. As the water level rises,
intersection of the naturally occurring bedding plane fractures occurs. These act as more
transmissive conduits, which allow the water to flow more freely toward the locally
created sink.
(1)
(2) Fe2++1/4 O2+H+=>Fe3++1/2 H2O (rate limiting step)
(3) Fe3++3H2O =>Fe(OH)3 (yellow boy)+3H+
(4)
In the coalfields there are co-existing carbonates such as calcite and dolomite, which can
neutralise the acidity generated. The reaction with calcite is given by:
The results of this are that many of the mine waters are not necessarily acidic, but often
high in dissolved salts. Additionally, the sediments overlying the mines, can be fairly
saline, and particularly in the southern portion of the coalfields, are high in sodium.
Turning a liability into an asset 473
compartments within mines. This results in water flow between mines, or onto the
surface. This flow is referred to as intermine flow (Grobbelaar, 2001). Projections for
future volumes of water to decant from the mines have been made by Grobbelaar et al.
(2000). In total, about 360ML/d will decant from all the mines in combination. On a
catchment basis, it relates to the following (ML/d):
Wilge/Klip Olifants Klein Vaal Komati
Olifants
23 170 45 120 2
water. An additional 40ha center pivot has been in operation since 2002, extracting water
from another section of the mine. According to the farmer, he applied approximately
35mm per week from the mine onto the 40ha crop of maise. The ideal mine-water
management option is already in place for Minnaar Colliery. Excess water is utilised for
irrigation and the mine water is simultaneously flushed. Care should, however, be
exercised not to over utilise the source, as a permanent drop in the water level can result
in an influx of oxygen and subsequent oxidisation of pyrite (Vermeulen, 2003).
surface-holding facility where aeration is possible, is necessary. Such a facility could also
be used for quick release of the water during flood discharge. Very few other chemical
benefits would be forthcoming from mine water mixing, because most of the constituents
are undersaturated in this water.
5 DISCUSSION
In a water-stressed country like South Africa, all water must be regarded as a potential
resource. The proximity of the Mpumalanga coalfields to major water users such as
power stations, industries and the Gauteng province which is the most highly populated
and industrial area of the country, means that this water is a prime resource.
Water can only be used if the quality thereof is fit for its intended use. The water
management focus at the coalmines needs to change from a volume-driven focus to a
quality driven focus. If water qualities can be dealt with in a way that an array of qualities
for different purposes can be provided, the cost-savings or rebates from the waste
discharge costs will be such that the current waste becomes a prime asset for the mine
and the country.
REFERENCES
Hodgson, F.D.I. 1999. Intermine flow between Tavistock and Rietspruit Collieries. Unpublished
report.
Hodgson, F.D.I. & Krantz, R.M. 1998. Groundwater quality deterioration in the Olifants River
Catchment above the Loskop Dam with Specialised investigations in the Witbank Dam Sub-
Catchment. Report to the Water Research Commission by the Institute for Groundwater Studies,
Univ. of the Orange Free State.
Hough, J.J.H. 2003. Evaluation of Management Options for Intermine Flow and associated
Impacts in the Central Witbank Coalfield. Unpublished M.Sc thesis. Univ. of the Free State.
Stumm, W. & Morgan, J.J. 1970. Aquatic Chemistry, 2nd Ed. New York, John Wiley & Sons, Inc.,
New York.
Usher, B.H., 2003. Development and Evaluation of Hydrogeochemical Prediction Techniques for
South African Coalmines. Unpublished Ph.D thesis, Univ. of the Free State.
Van Niekerk, A.M. 2001. Innovative And New Mine Water Treatment Technologies. Proc. Conf.
on Environmentally Responsible Mining in Southern Africa, Sept 2001. Johannesburg, South
Africa.
van Tonder, G.J., Vermeulen, P.D., Kleynhans, J. & Cogho, V. 2003. Prediction Of The Decant
Rate And Sulphate Concentration From Rehabilitated Open Cast Coal Mines In South Africa.
Paper submitted to the 6th International Conference of Acid Mining Drainage. In press.
Vermeulen, P.D. 2003. Investigation of decant water from the underground collieries in
Mpumalanga. Unpublished M.Sc thesis, Univ. of the Free State.
Environmental hydrogeology of the dolomite
aquifer in Ramotswa, Botswana
Michael Staudt
Geological Survey of Finland, Espoo
Horst Vogel
Department of Geological Survey, Botswana
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
This study was carried out as part of a technical co-operation project between the
Botswana Department of Geological Survey (DGS) and the German Federal Institute for
Geosciences and Natural Resources (BGR). The study objective was to establish the
quality of the groundwater resources in Ramotswa and to produce environmental
hydrogeology maps for regional and urban planners.
One vulnerable area is the village of Ramotswa, which experienced one of Botswana’s
worst cases of groundwater pollution during the 1990s. The successful promotion of pit
latrines and the location of Ramotswa on top of Botswana’s most productive dolomite
aquifer meant disaster as human wastewater polluted the shallow aquifer in no time at all.
As a result, the entire wellfield had to be abandoned in favour of surface water from the
dam in the nearby capital city of Gaborone.
In 2001, the Department of Geological Survey (DGS) in Lobatse decided to carry out
an environmental hydrogeology investigation in Ramotswa because the aquifer there is
possibly the most productive in a country where water resources are scarce.
The study results revealed that the most obnoxious groundwater pollution problem
was still due to high nitrate concentrations in several boreholes. The recorded maximum
Water resources of arid areas 480
nitrate concentration was 442mg L−1. The latter is attributed to the continued use of pit
latrines.
The industrial complex in the Ramotswa station area turned out to constitute another
area of concern. High nitrate levels and elevated concentrations of sulfate, sodium, and
chloride were identified. This was indicative of anthropogenic pollution, as were the
observed nickel and possibly the aluminium levels.
The study area includes the settlements of Ramotswa, Ramotswa station (Taung),
Boatle, and surroundings. The size of the study area was approximately 174km2. The
dominant Ngotwane river valley follows mainly N-S trending structures through the
study area. River infiltration into the aquifers occurs beneath the Ngotwane river bed,
which is approximately 2m thick. The overall hydraulic gradient is 1:300 along the
course of the river. Other rivers are the Taung and the Boatle, which are tributaries to the
Ngotwane river. All rivers are ephemeral and belong to the Limpopo river basin. The
Ngotwane river forms the international boundary with the Republic of South Africa to the
east.
The rainfall pattern is seasonal, as is generally the case in Botswana. Mean annual
rainfall amounts to approximately 475mm. In good years annual rainfall may exceed
1000mm, while in poor years it may be as low as 125mm. Rains fall mainly in short, high
intensity events. Occasionally, heavy rainfalls provide the bulk of the annual
precipitation. Rainfall is the primary source, which replenishes water resources in the
study area.
The by far most important form of landuse is livestock grazing. Although the
recommended stocking rates are 3 to 6 livestock units per hectar, the present stocking rate
in the study area is much higher, around 12 to 14 livestock units per hectar. This results
in overgrazing and environmental degradation.
2 HYDROGEOLOGY
There are two primary aquifer systems in the study area, namely the Ramotswa dolomite
formation, which underlies most of the major village of Ramotswa, and the Lephala
formation to the south and south-east. The latter consists of clastic cherts and
conglomerates. In between the two major formations are three lesser formations, namely
the Maholobota formation, which consists of interlayered dolomite with minor chert, the
Magopane formation, which is made up of bedded chert and minor dolomite, and the
Ramotswa shale formation, which comprises siltstone and shales. All five formations
belong to the Transvaal supergroup.
All these aquifers are considered to be in local hydraulic connection via predominately
N-S trending fracture zones. The dominant feature of the system is a marked anisotropy
associated with high density fracturing. In the dolomites, the active groundwater
circulation has favoured local karstification along structural lineaments producing high
transmissivity (T) and storativity (S).
The Ramotswa dolomite aquifer consists of two different karst zones, a shallow and a
deep zone (Institute of Hydrology, 1986). The upper karstic zone has a variable thickness
of 20 to 50m and receives recharge from the river and percolating rainwater. Dolomite
Environmental hydrogeology of the dolomite aquifer in Ramotswa, Botswana 481
solution appears preferentially along fractures. The deeper karstic zone has a thickness of
beween 25 to 50m and recharge is probably from across the border in South Africa.
Within the dolomite aquifer, areas of major linear karst and unfractured dolomite
country rock have to be distinguished. While high yielding wells are commonly found
along the major linear karst, low yielding boreholes are common in the country rock.
During field work the boreholes in this latter zone rapidly depleted at low pumping rates.
Away from the river, dry boreholes were common.
Hence it came as no surprise that the main production wells are located in E-W
direction along the major linear karst in the southern part of the study area. The linear
karst there features a dense fracture pattern with different fracture directions. This
fracturing, in combination with infiltration of river water and the intersection of minor
side valleys in E-W direction, has produced favourable conditions in terms of
permeability.
Borehole logs perfectly highlighted the 3 subdivisions described above, namely the
upper aquifer extending to a depth of about 30m followed by a zone of less water-bearing
fissures between 30–50m depth, and a deeper aquifer of variable thickness between 45
and 100m depth.
The Lephala formation aquifer outcrops in the southern edge of the project area, where
the rocks are faulted against the dolomites. The Lephala aquifer is similar to the
Ramotswa dolomite aquifer but is unaffected by karstification. The formation is
characterized by two fissured zones which are separated by a less fisured zone, a
thickness of the upper zone of 30 to 40m, and a thickness of the lower zone of
approximately 30m.
Yields of the boreholes in the Lephala formation depend on their proximity to the
river, the intersection of the fissured zones, and the extent of the secondary infills.
Recharge may be restricted to surface runoff via fractures or infiltration from the river.
At the onset of the study, a well census was carried out. The aim was to identify all
boreholes in Ramotswa and the surrounding areas. Once located, the site of the boreholes
was determined by means of a hand-held GPS 12XL (http://www.garmin.com/). While
carrying out the well census, the manually produced preliminary borehole location map
was transferred into the ESRI ArcView GIS software environment
(http://www.esri.com/). Similarly, all borehole parameters such as construction details,
lithologies, water levels, water yields, water strikes, usage, coordinates, and owner data,
were entered into the GeODin digital data base (http://www.geodin-system.com/).
In the second phase all environmental hazards to groundwater were identified in the
field. They were mainly anthropogenic and were differentiated as point hazards, such as,
for example, pit latrines, and area hazards, such as, for example, cemeteries. Again, the
location of all hazards was determined by GPS so as to produce a digital environmental
hazard map.
During the third stage of the field survey, groundwater from all accessible boreholes
was sampled for chemical analyses. All in all, a total of 31 boreholes were sampled, 21 of
which were also sampled for trace elements. It was most unfortunate that a number of
Water resources of arid areas 482
extra boreholes could not be sampled either because of vandalism or inaccessibility due
to plentiful rains.
Prior to sampling, the groundwater level was determined with the help of a dipper. At
the same time the height of the casing and the diameter of the borehole were recorded.
Before the pump was lowered into the borehole, the borehole log (whenever available)
was checked in order to determine the most appropriate depth level to place the pump.
Unfortunately, archive records were often incomplete or nonexistent. Because in such
cases one could not know the depth of the water strike, the pump was installed half way
down of the overall borehole depth.
Most of the boreholes were not equipped with a pump. In these cases a submersible
pump (Grundfos MP1) was employed, which was driven by a mobile power generator.
To reach the desired depth, 3-m-long PVC tubes were used and put together. For security
reasons the pump was never placed at very shallow depth; but the drawdown still had to
be observed regularly. Based on the depth of the well and its casing diameter, the volume
of the water column was calculated and a minimum pumping time calculated based on
pumping yield [L/s].
Once pumping began, a series of field parameters [electrical conductivity (EC), water
reaction (pH), water temperature (degrees Celsius), dissolved oxygen concentration (O2),
and total dissolved solids (TDS)] were measured employing field electrodes in a flow cell
(http://www.wtw.com/). Parameter values were checked every 10 minutes, and from time
to time the flow cell was cleaned. As soon as the readings had stabilized for a period of at
least 15 minutes, groundwater samples were taken and filled in 250-ml PVC bottles. The
samples earmarked for the analysis of the trace elements were filtered through 0.45µm
filters and acidified. All samples were immediately put into a cooler box and later stored
in a fridge.
After the groundwater sampling campaign had been completed, a GPS precision
survey was carried out so as to determine the precise elevations of the sampled wells. But
because of the high costs involved only 17 sites were surveyed. The elevations of the
remaining boreholes were extracted from Geotechnical Consulting Services (2000). Since
the difference between the two data sets was only ±0.3m, the accuracy level needed to
obtain groundwater flow directions was adequate.
4 RESULTS
that were above the maximum allowable limit of 400mg L−1. Again, this was indicative
of pollution as were the much higher than acceptable (200mg L−1) chloride [Cl¯]
concentration within the Ramotswa station area, which again suggested anthropogenic
pollution.
Sodium (Na+) did not show elevated levels in the major village. But again three
boreholes in the Ramotswa station area revealed high levels. In all three boreholes the
levels were higher than the maximum allowable level of 400mg L−1(BOS, 2000).
Four boreholes revealed aluminium (Al3+) levels above the maximum allowable level
of 200µg L−1 (BOS, 2000).
Yet, the fact that nitrate levels were still high at the end of 2001 highlights a persistent
pollution problem. Most likely the problem emanated from the still large number of pit
latrines. In 1991, 2432 pit latrines were in use that is 66% of all households used pit
latrines then (Enneco, 1996). Given the fact that the population grew over the last two
decades from approximately 14000 inhabitants in 1981 to 20680 in 2001, the number of
pit latrines has most likely increased further.
Hence, the nitrate pollution problem in Ramotswa will not be solved unless all pit
latrines are replaced with flush toilets that are connected to a properly functioning sewage
system. But in late 2001 not even the new sewage system was in proper working
condition. At several places the new sewers were leaking, ironically because of damages
caused by excavators who were still working on the completion of the system.
Livestock may be another source of nitrate pollution. In places where livestock is
highly concentrated, animal waste may contribute to groundwater pollution. A case in
point was the cattle kraal next to a borehole where the watering site was less than 50m
away from the well house. Here the nitrate concentration was 148mg L−1 in late 2001.
The primary health hazard from drinking water with nitrate-nitrogen (NO3-N) occurs
when nitrate (NO3) is transformed to nitrite (NO2) in the digestive system (Vogel, 2002).
While NO3 is not very toxic, NO2 is toxic. The nitrite oxidizes iron in the hemoglobin of
the red blood cells of humans (and other warm-blooded animals) to produce
methemoglobin. Methemoglobin lacks the ability of hemoglobin to transport oxygen to
body tissues. This creates the condition known as methemoglobinemia (“blue baby
syndrome”), in which red blood cells carry insufficient oxygen to the individual body
cells thus causing the veins and skin to appear blue (“internal suffocation”). This
condition is especially serious in infants.
The results of this study confirmed that groundwater pollution problems still exist in
Ramotswa. The most obnoxious pollution problem was again due to high nitrate
concentrations in several boreholes. The latter was attributed to the continued use of pit
latrines. In fact, at sites where the soil is very thin, and given a mean depth of the pit
latrines of 1.7m (WLPU, 1985), human waste may enter directly into the aquifer. Clearly,
pit latrines in the study area ought to be replaced with facilities connected to the new
sewerage system. But the latter also must be maintained properly.
The industrial complex in the Ramotswa station area turned out to constitute another
area of concern. In this area five boreholes were found that produced contaminated
groundwater. High nitrate levels and elevated concentrations of sulfate, sodium, and
chloride were identified. This was indicative of anthropogenic pollution, as were the
observed nickel and possibly the aluminium levels.
Because of these pollution indicators and also because of the observed elevated but
possibly natural arsenic concentrations along the river Ngotwane, it was decided to re-
sample all boreholes starting in late 2003. In addition, several blocked boreholes, which
could not be sampled in 2001, will be rehabilitated and a very few new boreholes will be
drilled in strategically important places. The objective of this programme is to thoroughly
Environmental hydrogeology of the dolomite aquifer in Ramotswa, Botswana 487
determine groundwater pollution levels in the Ramotswa area, and to contribute towards
the reduction of groundwater pollution.
This will also help alleviate a further matter of concern that became glaringly obvious
while carrying out this study, namely the lack of reliable data for the important
Ramotswa wellfield. For example, it was impossible to establish trend lines of certain
pollutants over time from the data retrieved from the national borehole archive. Only a
very few were possible but even these raised doubt with regards to data reliability.
Last but not least, the groundwater protection zones and previously made
recommendations (WLPU, 1985; Water Surveys, 1994) ought to be strictly observed.
This, and the elimination of the existing pollution problem, may in future allow for the
renewed use of Botswana’s most important aquifer.
REFERENCES
WLPU 1985. Ramotswa Wellfield Pollution Study. Watermeyer, Legge, Piesold & Uhlmann
(WLPU), Final Report, Dept. of Water Affairs (DWA), Gaborone, Botswana.
Investigation of natural enrichment processes
of nitrate in soil and groundwater of semi-arid
regions: case study—Botswana
S.Stadler, M.von Hoyer, W.H.M.Duijnisveld & T.Himmelsbach
Federal Institute for Geosciences and Natural Resources (BGR),
Hannover, Germany
M.Schwiede & J.Böttcher
Department of Soil Ecology, University of Hannover, Germany
H.Hötzl
Department of Applied Geology, University of Karlsruhe, Germany
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
immediate result of the AIDS epidemic which requires bottle-feeding of the infants, often
with nitrate-rich water (Colvin, 1999). Nitrosamines are harmful to humans of any age.
Elevated nitrate concentrations also shorten the lives of cattle, lead to deficient growth
and decreased reproduction (Davidson et al., 1964).
Plenty of research efforts have been conducted towards anthropogenically caused
nitrate accumulation in groundwater, especially relating to spatial nitrate input from
fertilizers. However, not much attention has been paid to natural nitrate enrichment in
groundwaters of semi-arid and arid regions (Burt et al., 1993; Van der Hoek et al., 1998).
Only few studies until now have focused on natural nitrate sources (Edmunds and
Wright, 1979; Edmunds and Gaye, 1997; Barnes, 1992; Aranibar et al., 2003; Walvoord
et al., 2003). In the Kalahari of Botswana natural nitrate accumulation processes appear
to play an important role and were firstly investigated by Heaton, 1983, 1984.
Enrichment processes, however, are not yet fully understood.
To address this topic a research cooperation between the Federal Institute of
Geosciences and Natural Resources (BGR), the Department of Applied Geology of the
University of Karlsruhe, the Institute of Soil Science of the University of Hannover and
the Department of Geological Survey (DGS) of Botswana was established.
In this study we investigate the natural enrichment processes of nitrate in groundwater
of the Ntane-Sandstone Aquifer between Serowe and Orapa in the Kalahari of Botswana.
The aim is to identify the (major) natural input factors and the fate of nitrate in
groundwater. We use a combined approach of soil science and hydrogeology to fully
trace the sources.
1.2 Concept
As the Kalahari region offers much potential for nitrogen input from different
compartments (see Fig. 1), it was necessary to validate and investigate the individual
sources. Dominant nitrogen input was expected from the following: i) Natural vegetation,
especially acacia trees and shrubs, the predominant vegetation in the area, are known for
their N-fixing potential. ii) Cattle, though extensive, cause (over)grazing, displacing
natural mammals, and leave nitrogen from manure especially in Kraals (cattle posts). iii)
Termite mounds, occurring in the area, allow an accumulation of nitrogen in symbiosis
with fungi cultures. All these features, coupled with climatic aspects such as downward
flushing during major rainfall events in combination with a low seepage rate, and the
transport by surface and preferential flow of seeping water in macropores (root channels,
Investigation of natural enrichment processes of nitrate 491
animal burrows), enhance nitrate enrichment in the subsurface. Lithology could provide
further nitrogen, e.g. in calcretes and silcretes of the uppermost formation
nitrate-rich uprising groundwater from the lower lying aquifer (Mosolotsane Fm.) or
other allochthonous processes might play a role.
For those reasons we used a combined approach of hydrochemical, hydroisotopic, and
soil scientific methods. Soil scientific investigations include (i) the quantification of the
soil nitrogen balance at typical sites of the Kalahari, and the evaluation of stocks of
mineralized nitrogen in soils and unsaturated zone with respect to vegetation and cattle
grazing intensity, (ii) the identification of nitrogen dynamics in the root zone, and
investigation of leaching processes and the transport and transformation of ammonia and
nitrate in the unsaturated zone, and (iii) the development of a model for prognoses of
nitrate leaching into groundwater with respect to vegetation and land use (grazing)
intensity. Hydrogeological and hydrochemical investigations include (i) the
determination of hydrochemical characteristics of the groundwater types occurring in the
Serowe-Orapa (Botswana) groundwater system (ii) the determination of groundwater
residence times and recharge sources by a wide range of environmental tracers (iii) the
investigation of input processes leading to the varied spatial distribution of nitrate in the
aquifer, including rock and water analysis, and (iv) the investigation and modeling of the
reactive transport and fate of nitrate under confined conditions.
2 SOIL
increasing distance from the cattle post and thus are also correlated with grazing
intensity.
Also, at the base of termite mounds extremely high concentrations of up to 750mg
NO3-N/kg were measured (Fig. 4). In between these two termite mounds the
concentrations are comparable to the values found under natural vegetation at the test
farm (Fig. 3). Thus cattle posts and termite mounds are potential sources of nitrate
pollution of the groundwater.
To get more insight into the spatial and temporal variability and heterogeneity of
nitrate and ammonium concentrations, appropriate soil profiles and sites were sampled.
Figure 5 shows the results of soil profile sampling below a kraal in October 2002 and
again at the same site in October 2003. The overall picture of the nitrate concentration
profiles on both dates is more or less the same, but spatial heterogeneity masks the
detection of a potential vertical downward movement of the nitrate profile during this
period of one year. Very interesting are also the concentration profiles of two soil profiles
sampled 50m apart at the same date in November 2003 (Fig. 6). The nitrate profiles A
and B look very much the same, which indicates, that the spatial variability seems to be
quite low. However, it is currently not possible to explain the increased nitrate
concentrations found in 8–12m depth. A more thorough analysis of the climatic
conditions during
the last 30 years might bring better insight into the leaching conditions during this period
and might help to explain the measured concentration profiles in Figure 6.
3 GROUNDWATER
depth and according variation of redox parameters, e.g. oxygen and bicarbonate. This
could be a sign for the possibility of autochthonous nitrate reduction (Fig. 9). Due to local
heterogeneities a complex interaction of both options is assumed.
REFERENCES
Aranibar, J.N., Anderson, I.C., Ringrose, S. & Macko, S.A. 2003. Importance of nitrogen fixation
in soil crusts of southern African arid ecosystems: acetylene reduction and stable isotope
studies. J. Arid Env. 54: 345–358
Barnes, C.J., Jacobsen, G. & Smith, G.D. 1992. The origin of high nitrate ground waters in the
Australian arid zone.—J. Hydrol.137:181–197
Burt, T.P., Heathwaite, A.L. et al. 1993. Nitrate: Processes, patterns and management, Chichester,
John Wiley & Sons, Chichester, 444pp.
Colvin, C. (1999): Increased risk of methemoglobinemia as a result of bottle feeding by HIV
positive mothers in South Africa.—IAH Congress, Bratislava.
Davidson, K.L. et al. 1964. Nitrate toxicity in dairy heifers. I. Effects on reproduction, growth,
lactation and vitamin A nutrition.—J. Dairy Sci., 47:1065–1073.
De Vries, J.J., Selaolo, E.T. & Beekman, H.E. 2000. Groundwater recharge in the Kalahari, with
reference to paleo-hydrologic conditions. J. Hydrol. 238:110–123
De Wit, P.V. & Nachtergaele, F.O. 1990. Soil Mapping and Advisory Services Botswana:
Explanatory Note on the Soil Map of the Republic of Botswana (Typifying Pedons and Soil
Analytical Data); Gaborone, December 1990
Edmunds, W.M. & Gaye, C.B. 1997. Naturally high nitrate concentrations in groundwaters from
the Sahel. J. Environ. Qual. 26:1231–1239
Edmunds, W.M. & Wright, E.P. 1979. Groundwater recharge and paleoclimate in the Sirte and
Kufra Basins, Lybia. J. Hydrol 40:210–241
Heaton, T.H.E, Talma, A.S & Vogel, J.C. 1983. Origin and history of nitrate in confined
groundwater in the Western Kalahari, J. Hydrol 62:243–262
Heaton, T.H.E. 1984. Sources of the nitrate in phreatic groundwater in the western Kalahari, J.
Hydrol 67: 249–295.
Lavelle, P. & Spain, A.V. 2001. Soil Ecology. 514 pp. Dordrecht, Kluwer Academic Publishers.
Lubczynski, M. 2000. Groundwater evapotranspiration—Underestimated component of the
groundwater balance in a semi-arid environment—Serowe case, Botswana. In: Sililo et al. (eds):
Groundwater: Past Achievements and Future Challenges. Rotterdam, Balkema: 199–204
Marrett, D.J., Khattak, R.A., Elseewi, & Page, A.L. 1990. Elevated nitrate levels in soils of the
eastern Mojawe desert.—J. Environ. Qual. 19:658–666
Martini, I.P. & Chesworth, W. 1992. Weathering, Soils & Paleosoils—Developments in Earth
Surface Processes2:309–377, Department of Land Resource Science, University of Guelph,
Ont. N1G 2W1, Canada, Elsevier, Amsterdam-London-New York-Tokyo
Mokokwe, K. 1999. Occurrence of groundwater with high nitrate content—Orapa wellfields.
Inception report, DGS, Lobatse.
Mphinyane, W.N. 2001. Influence of livestock grazing within biospheres under free range and
controlled conditions in Botswana. PhD thesis: Department of Plant Production and Soil
Science, Faculty of Natural and Agricultural Sciences, University Pretoria
Perkins, J.S. & Thomas, S.G. 1993. Environment Responses and Sensitivity to Permanent Cattle
Ranching, Semi arid Western Central Botswana In: Landscape Sensitivity: 273–286, edited by
D.S.G.Thomas, & R.J.Allison (eds), John Wiley and Sons Ltd
Schulze, E.D., Gebauer, G., Ziegler, H. & Lange, O.L. 1991. Estimates of nitrogen fixation by trees
on aridity gradient in Namibia. Oecologia 88:451–455, Springer Verlag.
Selaolo, E.T. 1998. Tracer studies and groundwater recharge assessment in the eastern fringe of
the Botswana Kalahari, the Letlhakeng-Botlhapatlou area. Thamaga, Botswana.
Water resources of arid areas 504
ABSTRACT: Climatic and runoff data in semi arid regions of Africa were
used to assess the implications of climatic variability on some water
balance parameters. Result shows the coefficient of determination
between discharge and rainfall as 0.64 to 0.72, soil moisture capacity of
170mm and infiltration of 608.3mm. Soil moisture deficit and surface
runoff are 404.1mm/m2 and 161.6mm/m2 respectively. Responses to
climate change include temperature rise of about 1.5°C, incidence of
droughts, desertification, aridity, decline in rainfall, decline in runoff and
water shortages for plant, animals and man. Water resources management
strategy is vital for sustainable use of the region’s meagre water resources.
Suggested sustainable water resources management policy will include:
effective water pollution control, water re-use, water recycling, water
demand management involving restriction/reduction in water use, control
of distribution losses and the exploitation of new sources of water such as
rain water harvesting and desalination.
1 INTRODUCTION
Semi arid regions of Africa has greatly been affected by climate change and climatic
variations over the past centuries and decades. Responses to climate change in drylands
of Africa include incidence of droughts, desertification, aridity, decline in rainfall,
decline in runoff and water shortages for plant, animal and man. During the late sixties
and early seventies, for example, the southern border region of Sahara desert, the Sahel,
succumbed to prolonged drought, desertification and famine. Climatic variations
resulting in disastrous decline in rainfall over semi arid regions of African continent have
impacted negatively on various components of water resources. Droughts and disastrous
decline in rainfall which are direct consequence of climatic variability have been on the
increase since the 1960’s in drylands of Africa. These phenomena have been attributed to
complex interplay of sea surface temperature anomalies over the tropical Atlantic Ocean
and the attendant latitudinal shifts of circulation systems during the dry years,
Water resources of arid areas 506
2 METHOD OF STUDY
Climatic data (daily and monthly rainfall, temperature and evaporation), runoff records
(discharge and stage), from 1900 to 1999 were collected from different stations in semi
arid regions of Africa and used in the analysis. As part of analysis carried out in the
study, rainfall totals and means for the study period were taken as series. The series were
subjected to a battery of time series analysis procedure including test for variability and
Gaussian low pass filter technique for detecting marked fluctuations. Regression, rainfall-
runoff and water balance models were also constructed. Water balance parameters of
Potential Evapotranspiration (PET), Actual Evapotranspiration (AET), Readily Available
Supply in the Soil (RAS), Soil Moisture Deficit (SMD), Surface Runoff (SR), Infiltration
and Areal Annual Volume were calculated according to water balance estimation method
(Umoh, 1999).
The rainfall regimes for the stations are generally in the form of alternating wet and dry
epochs (Figure 1). The annual rainfall curve for Kaduna (West Africa) indicates three
periods of increased rainfall: 1905–1909; 1925–1929; 1946–1949; and three periods of
marked decrease: 1911–1916; 1931–1938; 1952–1999 (Figure 1). In Botswana (Southern
Africa) two periods of increased rainfall were 1921–1927 and 1930–1939 and four
periods of marked decrease: 1913–1915; 1940–1948, 1951–1958 and 1960–1999. In
between these periods, the pattern of rainfall is oscillatory.
Descriptive statistics for annual rainfall series in the semi arid zone of West Africa for
the study period are shown in Table 1. Izom registered the highest mean annual rainfall of
1513.1mm while Lokoja had the least being 1122.1mm. The standard deviation ranged
between 179.9 for Kontagora and 323.3 for Kachia. It can also be seen from Table 1 that
annual variability in rainfall differ from station to station. Results of analysis of basic
statistical parameters of Kaduna reveals that the mean annual rainfall over 100 years
period is 1356.3mm with standard deviation (SD) of 319mm. Of this total, 1254.1mm or
92.5 percent (with a SD of 244.5mm) fell in the months of May to October. Mean annual
rainfall and rainy season coefficient of variation (CV) are 23.5% and 19.5% respectively.
The low values of annual coefficient of variation for the station suggests that rainfall is
less variable from year to year and that its variability is greater in the individual months.
The decade 1950–1959 was characterized by above normal precipitation over semi
arid regions of Africa, although rainfall deficiencies prevailed over the near equatorial
region. Latter, during the period 1960–1969, this rainfall anomaly pattern dramatically
reversed with rainfall deficit observed for semi arid regions of Africa while the equatorial
region experienced widespread abundance of rainfall. More recently, the pattern has been
one of increased aridity throughout the region. Mean rainfall decreased by 30–45% in the
Sahel between periods 1931–1960 and 1968–1997 and generally 15–25% across the rest
of the semi arid regions of Africa.
Water resources of arid areas 508
peak of AET occurs in December. The difference in maximum AET month within the
three drainage basin can be attributed to the fact that cessation of rainfall in Gbako and
Gurara basins is in November while that of Kaduna is October, thus resulting to one extra
month of rain duration over Kaduna basin. This extra rainfall month consequently
increase the duration of moisture in the soil, to the extent that, in the month of December
the AET is equal to PET, though the moisture availability is relatively low (15mm in
Gbako and 33mm in Gurara).
The only source of recurrent water is rainfall in the region. It supplies water for both
surface (overlandflow and streamflow) and ground water runoff. Rainfall is seasonal in
the region and most rivers dry up during the dry season. There is a close relationship
among rainfall, overlandflow and stream discharge. Drought and desertification disrupt
this closeness since they decrease both surface flow and ground water storage. In the
semi arid region, a lot of rainfall, that usually comes with high intensity, runs off rapidly.
Such conditions impede groundwater recharge and create flash flow on the surface. When
Hydroclimatological approach to sustainable water resources management 511
drought occur in the region, much less amount of water is available for the two processes
because the rainfall is less and a high proportion of the water is evapotranspired.
Effects of climate change include increasing incidents of drought and aridity,
incursion of desert-like conditions and decline in agricultural activity (Olofin, 1985;
UNEP, 1987; Issar, 1998). The implications of climatic variability on water resources is
demonstrated by the coincidence of low discharge, recharge and runoff with periods of
drought. An indication of long term climate changes in the semi arid regions of Africa is
depicted by a 90-year long record of flow in the Zambezi River at Victoria Falls shown in
Figure 2. Average flow over the study period was 1056m3/s. Annual flows were mostly
below the average for 38 years prior to 1945. Lower than average flows was recorded
from 1910 to 1940 and 1980 to 1995. Low discharge and recharge affect the surface and
the ground water of the region adversely. Under such conditions, there is very little water
available on the surface and much less available as groundwater. Most rivers dry up
during the dry season because of seasonality in rainfall.
Climate change has serious impact on water resources of various semi arid African
countries. In the region, the incidence of seasonal flow cessation is on the increase, as
shown by some streams in Angola, Botswana, Namibia, Tanzania and Zimbabwe,
resulting in water shortages for industrial and urban domestic supplies (Magadza, 2000).
Major rivers such as Zambezi, Kafue
reservoirs at a dangerously low level, leading to power rationing and load shedding.
Smaller streams, small dams and wetland areas in the valleys dried up early, reducing
water for human consumption, livestock use and vegetable production.
Compounding the effects of climate change on water resources in semi arid Africa, is
the explosive urban growth that is placed at 4% per year for developing countries.
Provision of clean water in the big cities is a challenging task of development strategists.
It is known that already urban water needs compete with agricultural water requirements.
Groundwater levels are known to be dropping rapidly as a result of demands of nearby
urban centres. When such urban growth with its concomitant water demand is set against
declining water availability due to climate change, one will agree with Nakayama (1998)
that water may be the 21st century’s oil.
Water resources of semi arid regions of Africa are very vulnerable to climate change
through occurrence of droughts and declining rainfall. The most important input that can
be made to cushion the adverse effect of climate change on water resources in the region
is to adopt a water management option that would stand up to the extremes of climatic
variations that is currently plaguing the region. Water conservation is vital for sustainable
use of the region’s meagre water resources. Components of such water conservation
policy will include: effective water pollution control, water re-use, water recycling, water
demand management involving restriction/reduction in water use, control of distribution
losses and the exploitation of new sources of water such as rain water harvesting and
desalination.
5 CONCLUSION
Semi arid regions of Africa has experienced pronounced climatic variations with their
accompanying hydroclimatic events and consequences. Rainfall and runoff have
fluctuated over the years with a declining trend. The decline in rainfall is reflected by
overall decline in annual peak flood and minimum water level. The dominant impact of
climate change is reduction in runoff. Current trends in major basins indicate a decrease
in runoff of about 17% over the past decade. The severity of the dry season increases
during the sequence of months with excessive potential evapotranspiration. Spatial and
temporal climatic variability and alterations in water balance parameters have serious
implications on human activities in water related sectors.
Climate change exerts remarkable effects on river flows, groundwater recharge and
other biophysical components of the water resource base, and demands for that resource.
The consequences, or impacts, of such changes on risk or resource reliability depend not
only on the biophysical changes in stream flow, recharge, sea-level rise and water quality,
but also on the characteristics of the water management system. Possible changes in
water resources and demand will impact on water supply, flood risk, power generation,
navigation, pollution control, recreation, habitats and ecosystems services in the absence
of planned adaptation to climate change. Luckily, some suggestions are proffered for
effective water resource management in the face of increasing effects of climate change.
Hydroclimatological approach to sustainable water resources management 513
REFERENCES
Desanker, P. & Magadza, C. 2001. Africa, In Mc Carthy, J.J. Canziani, O.F., Leary, N.A. Dokken,
D.J. & White, K.S. (eds), Climate Change 2001: Impacts, Adaptation and Vulnerability: 489–
531. U.K. Cambridge University Press.
Hastenrath, S. 1990. A generalized classification of South African summer rain-bearing synoptic
systems. Journal of Climatology 4, 547–560.
Issar, A.S. 1998. Climate, history and water. Work in Progress, United Nations University
15(2):16–17.
Lamb, P.J. & Peppler R.A. 1992. Further case studies of tropical Atlantic surface atmospheric and
oceanic patterns associated with sub-Saharan drought. Journal of Climatology 5:476–488.
Magadza, C.H.D. 2000. Climate change impact and human settlements in Africa: prospects for
adaptation. Environmental Monitoring, 61:193–205.
Nakayama, M. 1998. Water: The 21st Century’s Oil? Work in Progress, United Nations Univ.,
15(2):18–19.
Nicholson, S.E. & Entekhabi, D. 1986. The quasi-periodic behaviour of rainfall variability in
Africa and its relationship to Southern Oscillation. Journal of Climate and Applied Meteorology
34:331–348.
Olofin, E.A. 1985. Climatic constraints to water resource development in the Sudano-Sahelian
Zone of Nigeria. Water International 10(1):29–37.
Skoftland, E. 1995. Water resource management in southern Africa—a vision for the future. Conf.
of SADC Ministers responsible for Water Resources Management, 23–24 Nov, 1995. Pretoria,
South Africa.
Umoh, U.T. 1999. Climatic variability and rainfall-streamflow relationships in parts of the Niger
River Basin of Nigeria. Africa Climatological Research Series Vol. 3:10–22.
United Nation’s Environment Programme (UNEP) 1987. The changing atmosphere. Environmental
Brief No. 1. UNEP.
Impact of cultivation practices on multiple
uses of water in the Alemaya catchment,
eastern Ethiopia
Y.E.Woyessa
School of Civil Engineering and Built Environment, Technikon Free State,
South Africa
A.T.P.Bennie
Department of Soil, Crop and Climate Sciences, University of the Free
State, South Africa
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
ABSTRACT: Soil degradation and water pollution are the most important
environmental issues facing most of the developing world today. Across
Ethiopia, there has been considerable degradation of land, water and
biodiversity resources, which has, and will continue to have major
impacts on rural communities’ economic production and natural
ecosystems. Most of the productive topsoil in the highlands of Ethiopia
has been degraded, resulting in chronic food shortages and persistent
poverty. Serious erosion is estimated to cause soil degradation on some
25% of the highlands of Ethiopia. As a result of upland erosion, siltation
of dams, lakes and reservoirs is becoming a common phenomenon. A
field study was conducted in the catchment of Lake Alemaya, eastern
Ethiopia, with the objective of evaluating the effect of land cultivation
practices on runoff and soil loss, and its possible impact on the siltation of
Lake Alemaya. The result showed that it was possible to reduce runoff
and soil loss from the cultivated agricultural fields and subsequently
reduce sediment load deposited in the Lake by using proper crop residue
management on the agricultural fields.
1 INTRODUCTION
Soil degradation and water pollution are the most important environmental issues facing
most of the developing world today. Soil erosion has long been recognized as a negative
attribute of agriculture and pastoralism. It is evident that erosion is partly a natural
Impact of cultivation practices on multiple uses of water 515
process and cannot be avoided entirely. The objective should be to minimize erosion
within economic limits rather than trying to eliminate it.
It is useful to consider the short and long term impact of erosion. Short-term effects
occur within a season, such as damage to infrastructure, loss of crop area, lower yields
due to less water. Longer-term effects are generally manifested over decades, such as an
accelerated nutrient decline, lower plant available water capacity and siltation of dams,
lakes and low lying wetlands. In general, poor catchment management leads to short term
effects, while the long term impact of poor catchment management results in downward
spiral phenomena, which occurs when longer term degradation effects make management
in the short term difficult. For example, higher runoff leads to lower yield and soil cover,
which increases erosion, thus accelerating the degradation cycle.
The considerable degradation of land, water and biodiversity resources across Ethiopia
will continue to have major impacts on rural communities to produce economically and
to conserve natural ecosystems. Most of the productive topsoil in the highlands of
Ethiopia has been degraded, resulting in poor yields which enhance chronic food
shortages and persistent poverty. Serious soil erosion occur on an estimated 25% of the
area, and some estimates suggest that 4% of the highland is so seriously eroded that it
will not be economically productive again in the foreseeable future (SCRP, 1996). As a
consequence of this land degradation, the production capacity of the soils of the
Ethiopian highlands is calculated to be declining at a rate of 2–3% annually (Hurni, 1993;
cited by Zeleke, 2000). This is a potential threat to national food security if allowed to
continue uncontrolled and every effort should be made to reverse the situation.
Sonneveld & Keyzer (2003) investigated national agricultural revenues under
alternative scenarios of soil conservation, land accessibility and technology. Their results
showed that without soil erosion control, agricultural production will stagnate in future
resulting in distressing food shortages, while rural incomes will drop dramatically below
the poverty line. Even with the adoption of modern technology, soil conservation remains
essential, especially over the long term.
The vicious cycle of poverty and environmental degradation is the result of
interrelationships between the socioeconomic aspects of farmers and land degradation.
The socioeconomic situation of farmers determines their capability to implement
environmentally viable natural resource conservation measures. These include farming
practices and farmers’ attitudes toward rational use of resources. A study conducted by
Daba (2003) in the Hararghe highlands of eastern Ethiopia showed that farmers’
perception about the danger of gully erosion is significantly related to aspects like
severity of water scarcity, location of the farm within the landscape and the literacy level
of farmers. Willingness of farmers to adopt a new or improved soil and water
conservation practices is determined by fertilizer availability as an incentive and the
education of farmers. Daba (2003) concluded that for the Hararghe highland conditions,
design and implementation of soil and water conservation measures should be based on
farmers’ priorities such as addressing the alleviation of the water scarcity problem.
During the early 1980’s the Government of Ethiopia, through the Soil Conservation
Research Project (SCRP), placed a very high priority on stopping land degradation and
restoring natural resources. Some of the key aims of the SCRP were to generate
information on runoff and erosion processes under different agro-climatic conditions and
land use practices for application of proper planning and design of soil and water
Water resources of arid areas 516
conservation practices. It was underlined that such an effort would enhance and promote
sustainable development and protect the environment for future generations. Achieving
these outcomes is essential to revitalize rural communities and maintain sustainable
economic growth, and to ensure that our natural biodiversity is maintained for future
generations. It is believed that this can be achieved through integrated catchment
management.
Catchment management should be based on a partnership between community and
Government. Planning and implementation of natural resource management programs
should create opportunities for community engagement. Management of natural resources
should recognize the linkages between land and water and that the management of one
can impact on the other.
The catchment of Lake Alemaya is located in the Eastern Hararghe Region of eastern
Ethiopia, and is characterized by intensive agriculture and a high population pressure.
The average landholding per family varies between 0.5 to 1.7 hectares, which puts a high
pressure on the utilization of land resources, such as soil and water. The Lake Alemaya
provides the local population with domestic water supply as well as water for livestock
and irrigation purposes. However, due to an increase in population pressure and the
intensive agriculture practiced in the catchment in general and around Lake Alemaya in
particular, the capacity of the Lake Alemaya to store water is decreasing at an alarming
rate in the past few years, threatening the water supply to nearby towns. The main reason
for this situation is reported to be the accumulation of silt in the Lake due to continuous
erosion from the farmlands in the catchment.
The aim of this paper is to present results of an experiment conducted in the Alemaya
catchment, eastern Ethiopia, with the objective of evaluating the effects of cultivation
practices on the catchment hydrology, with especial emphasis on runoff and soil loss
from cultivated fields, and its possible impact on the siltation problem of Lake Alemaya.
The experimental site was located within the Lake Alemaya catchment on the Alemaya
University Campus, which is situated 550km east of Addis Ababa, the capital city of
Ethiopia. The climate is characterized as “Dry Weyna Dega” zone (Hurni, 1986), with an
altitude of 1960 meter above sea level and a mean annual rainfall of about 800mm. The
mean daily minimum and maximum temperatures are 10.1°C and 23.6°C respectively.
The rainfall pattern is bi-modal with a short rainy season from March to May and a long
rainy season during the months of July, August and September. The soil of the study area
is characterized as a Regosol with good internal drainage. The particle size distribution of
the topsoil, determined using the standard Pipette procedure as described by Day (1965),
is clayey with 45.1% clay, 22.2% silt and 32% sand. The mean bulk density of the soil
varies from 1150 to 1280kg m−3.
In this experiment, three tillage practices, namely no-tillage (NT), traditional tillage
(TT), and conventional tillage (CT) (mouldboard ploughing as a primary tillage) each
with four levels of wheat (Triticum aestivum L.) residue cover, namely 0t ha−1, 2t ha−1, 4t
ha−1, and 8t ha−1 as treatments were compared. Traditional tillage is the land preparation
method with a shallow tillage tool called “Maresha” pulled by a pair of oxen. The
Impact of cultivation practices on multiple uses of water 517
experiment was laid out in a split plot design with tillage practices as main plots and
residue cover rates as sub-plots. The size of each plot was 4 meters wide and 6 meters
long. All treatments were replicated twice. Following the tillage treatments the calculated
amounts of wheat residue were spread evenly over each plot.
The experimental plots were laid out on a field with homogenous soil and uniform
slope of 5–6%. A corrugated iron sheet border was placed around each plot to measure
runoff and soil loss. Pieces of corrugated iron sheet were carefully attached to one
another by rivets. The 400mm corrugated iron sheets were inserted 200mm deep into the
soil in order to prevent run-on and runoff to and from the plots. At the down slope side of
each plot a trough made of iron sheet was installed which collected the runoff in a barrel
with a capacity of 0.2m3 (200 liters). During the course of the experiment growth of
weeds within the plots was chemically controlled by spraying an herbicide (glyphosate).
During the second year (summer 2001) of the experiment, the same experimental plots
were used with the same layout and replication. The tillage operation was conducted after
carefully removing the corrugated iron sheet borders and the troughs. After completion of
the tillage operations the metallic borders and troughs were re-installed and fresh wheat
straw was spread on each plot at the required rates.
Runoff amount and sediment mass were collected in a barrel for each plot. The
volume of runoff in each barrel was measured after every runoff producing rainfall event
by inserting a steel tape into the barrel and measuring the depth of water. The depth of
runoff in the barrel was converted into volume. The collected runoff was mixed
thoroughly after which a sample was taken to determine the sediment concentration,
where after the barrel was emptied and cleaned. Rainfall was recorded with an automatic
rain gauge situated in the middle of the experimental area.
In addition to the controlled experiment on runoff plots, visual observation at selected
sites in the Alemaya catchment and informal discussions with small groups of farmers
were conducted in order to collect information on farming practices and visible land
degradation symptoms, such as erosion gullies.
3.1 Runoff
The average total runoff for the three tillage and four rates of residue treatments, during
the main rainfall seasons of both experimental years (2000 and 2001), are given in Figure
1. Figure 1 illustrates that residue rates of 41 ha−1 and 81 ha−1 had effectively eliminated
runoff from all three tillage treatments for both years.
Although there was no statistically significant effect of tillage practice on total runoff,
the average total runoff from the three tillage practices was in the order of NT>TT>CT.
Tillage affected runoff only during the first few storms of the season, where runoff from
no-tillage was significantly higher than from the freshly tilled traditional and
conventional tillage plots, before the impact of raindrops destroyed the roughness created
by tillage operations to form a surface seal or crust. Generally, the effect of no-tillage on
runoff is reported to take a long time to
Water resources of arid areas 518
The residue mass was converted to percentage surface residue cover using a
measurement technique as described by Lang & Mallett (1982). Accordingly the residue
masses of 2, 4 and 8t ha−1 corresponded to 62%, 76% and 92% of residue cover
respectively. The mean total runoff from the three tillage practices were related to the
percentage residue cover for 2000 and 2001 rainfall seasons (Fig. 2). Figure 2 illustrates
the decrease in runoff with an increase in residue cover. The linear decline in total runoff
with increasing residue cover for the first year (Fig. 2a) resulted mainly from the high
runoff measured at the 2t ha−1 or 62% residue cover during a high intensity storm of the
12th of August 2000. When the runoff from the 12th of August storm is not included in
the curve fitting, the function changes to an exponential form similar to the 2001 data
depicted in Figure 2b.
It was observed that for most of the storms a residue rate of 2t ha−1 reduced runoff
significantly compared with the bare plots. At higher intensity storms, such as the one on
the 12th of August
Water resources of arid areas 520
surface cover reduces erosion by reducing the runoff volume through stubble protecting
the soil surface, thus reducing aggregate breakdown and compaction of the soil surface
by raindrop impact (Edwards, 1982; Freebairn et al., 1993).
Impact of cultivation practices on multiple uses of water 521
This was believed to be one of the reasons for the lower soil loss per storm during the
second year.
For six of the eight storms that caused erosion during 2000 and 2001, there were
significant differences in soil loss only between the bare and 2t ha−1 residue rates. Among
the different residue rates there were no significant differences. For the other two storms,
there were significant differences between 0 and 2t ha−1, and between 2 and 4t ha−1
residue rates. From this it can be concluded that a residue rate of 2t ha−1 will be sufficient
to control erosion from this soil, except for the occasional high intensity storms where a
higher residue rate of 4 to 6t ha−1 will be preferred.
The decrease in the total soil loss per season, representing the mean of the tillage
treatments, with increasing percentage residue cover, is presented in Figure 3 for both
seasons. The decrease in soil loss was linear for the first year, with a reduction of 0.053t
ha−1 for every 1% increase in the percentage residue cover. The decline was non-linear
for the second year with a more rapid decline with an increase of residue cover from 0 to
60%. Linear relationships between soil loss and the percentage residue cover have also
been reported by some authors (Singer et al., 1981; Kinnel, 1996; Cruse et al., 2001),
while others have reported non-linear (inverse exponential or polynomial functions)
relationships between the two parameters (Gilley et al., 1986a; Papendick et al., 1990;
Freebairn et al., 1993).
The mean total soil losses averaged over the four rates of residue cover, for the three
tillage treatments, for both years, are given in Figure 4. It shows that the mean seasonal
total soil loss was the highest for NT, followed by CT. Traditional tillage had the lowest
soil loss for both seasons. These differences, however, were not statistically significant.
Despite the higher sediment concentration from CT compared with the other two tillage
practices, the soil loss from CT was lower than from NT. The soil loss from a given area
is the product of the sediment concentration and runoff amount. Despite the lower
sediment concentration, the higher runoff generated from NT resulted in a higher soil loss
than from CT. Lindstrom & Onstad (1984) warned that soil erosion could be a serious
problem on NT when insufficient residue cover is present to reduce the runoff flow
velocity. In another study, Myers & Wagger (1996) reported that residue cover did not
substantially reduce runoff from NT treatments but consistently decreased soil loss.
Overall, the type of tillage practice had little effect on the total soil loss except on the
bare NT plots that had the highest erosion for all the storms. When NT is practiced on
this type of soil, care should be taken that a residue cover of at least 2t ha−1 wheat stubble
or 62% residue cover should be maintained at all times. The beneficial effect of tillage
was generally limited to the first few erosive storms until the tillage induced roughness
and depression storage disappeared. It is possible that two years were not long enough to
detect the differences due to tillage practices. The effect of a tillage practice is reported to
take several years to change certain physical and hydraulic properties of the soil. For
instance, Dickey et al. (1989) reported that between 5 and 6 years is required under no-
tillage for changes in soil physical properties, resulting in higher water intake, to become
measurable. Voorhees & Lindstrom (1984) reported that 3 to 4 years were required
before conservation tillage had a more favorable porosity in the upper 0 to 150mm depth
of the soil.
Impact of cultivation practices on multiple uses of water 523
of irrigation water. Two factories in the city of Harar (Harar Brewery and Hamaressa
Edible Oil Mill) which previously depended on their water supply from Lake Alemaya
since their establishment had to seek for alternative ground water supply 30km away due
to the unreliability of water supply from the Lake.
The whole catchment area drains into Lake Alemaya. Unsustainable intensive
agricultural practices coupled with a high human population pressure are among the
factors that enhance the degradation of land and water in the catchment. The size of Lake
Alemaya is shrinking tremendously both in depth and in surface area. Some of the
original Lake area is now grazing land for the nearby livestock owners. The Lake, which
used to supply the local population with multiple uses of water, is no longer a dependable
source of water. Although there could be other factors contributing to the present
condition of the Lake, it is believed that unacceptably high erosion from agricultural
lands in the catchment is mainly responsible for the siltation and the eventual dry up of
some parts of the Lake. Gullies as deep as 5 meters and more are running into the Lake.
In some cases the density of gullies is alarmingly high with more than 10 gullies running
into the Lake over a distance of just two kilometers.
grown for the export market to neighbouring countries such as Somalia, Djibouti and
some Arab countries such as Yemen. During the dry season when irrigation is very
crucial for growing of this cash crop, one kilogram of good quality chat can sell for up to
400 Ethiopian Birr (1US$=8.5 ET Birr) on the export market, which is a huge income for
chat growing farmers. In the year 1999/2000, Ethiopia earned 618.8 million birr in hard
currency by exporting 15,684 Metric ton of chat (Tefera, 2003). This market incentive
encourages farmers to withdraw more and more water for irrigation from Lake Alemaya.
Farmers in the surroundings of the Lake use several small pumps in series arrangements
to withdraw water from the Lake to irrigate fields as far as to the top of the catchment
boundary.
At the end of every harvest season no crop residue is left on the agricultural fields due
to the fact that the biomass is harvested and the multiple uses of the remaining crop
residue, such as for fuel, construction, animal feed, etc., exposing the bare soil surface to
the erratic and sometimes very aggressive rainfall. It is at this time of the season, before
establishing crops in the field, that most of the erosion occurs.
Extrapolation of the results from this study to a catchment level, with the present land
use practices where insignificant amount of crop residue remains in the fields, supports a
scenario of continuous sediment buildup in the Lake. The total area of the Alemaya
catchment draining into the Lake Alemaya is estimated at 1680 hectares. The average soil
loss from the bare traditional tillage plots at the experimental site was 2.46t ha−1. If we
assume this to be the average annual soil loss for the catchment, the total amount of soil
running into the Lake will be 4133ton. This will be the amount for the main rainfall
season of July to September, whereas additional erosion will also occur during the short
rainy season for the months of March to May with substantial amounts of rainfall. With
an average bulk density of 1260kg m−3 for the soil of the experimental site, the total
volume of soil deposited in the Lake will be approximately 3280m3 per annum. This is
sufficient to reduce the depth of the Lake annually by about 33mm over an area of 10
hectares.
The benefit of crop residue on the surface of the soil in reducing runoff and soil loss
was demonstrated in this experiment. The presence of wheat residue at a rate of 2t ha−1
was sufficient to reduce soil loss to a minimum level. This shows that it is technically
possible to reduce the quantity of sediment deposited in the Lake with proper
management of crop residue on the farm lands in the catchment. However, the high
population pressure and the nature of subsistence farming in the area, which includes
both crop and livestock production (termed as a mixed farming system), puts a lot of
pressure on both land and water resources. Under the present circumstances it appears
that the use of crop residue and other practices for soil and water conservation is not a
priority issue for the small farmers in the area.
4 CONCLUSION
and soil loss from agricultural lands through proper crop residue management by
retaining at least 65 to 70% crop residue cover on the agricultural fields. However, this
objective is difficult to achieve under subsistence farming conditions in the catchment
where the main crop is grown for its biomass yield leaving little residue behind and
where there are also other multiple uses for crop residue, such as for fuel, animal feed and
construction purposes. These are the most pressing and immediate needs of subsistence
farmers in the catchment, whereas erosion control seems by far less important to them.
Therefore, investigation into the causes and possible remedies to land and water
degradation at a catchment level requires a holistic and an integrated approach taking into
account the socioeconomic circumstances of farmers and ensuring active participation of
the community in the process of planning and implementation of an integrated catchment
management approach.
REFERENCES
Cruse, R.M., Roberto, M. & Mize, C.W. 2001. Surface residue effect on erosion of thawing soils.
Soil Sci. Soc. Am.J. 65:178–184.
Daba, S. 2003. An investigation of the physical and socioeconomic determinants of soil erosion in
the Hararghe Highlands, eastern Ethiopia. Land Degradation and Development. 14(1):69–81.
Day, P.R. 1965. Particle fractionation and particle size analysis. In: C.A.Black (ed.) Methods of soil
analysis. Amer. Soc. of Agron., Madison, WI. pp. 545–567.
Dickey, E.E., Eckert, D.J., Larson, W.E., Johnson, R., Mannering, J., Kinsella, J., Wickner, I. &
Cruse, R.M. 1989. To till or no to till during drought. J. Soil Water Conserv., 44:117–120.
Edwards, W.M. 1982. Predicting tillage effects on infiltration. In: P.W.Unger and D.M.Van Doren
(eds), Predicting tillage effects on soil physical properties and processes: ASA special
publication, 44. Am. So. Agron. and SSSA, Madison, WI. pp. 105–115.
Freebairn, D.M. & Wockner, G.H. 1986. A study of soil erosion on vertisols of the eastern Darling
Downs, Queensland. I. Effect of surface conditions on soil movement within contour bay
catchments. Aus. J. Soil Res. 24:135–158.
Freebairn, D.M., Loch, R.J. & Cogle, A.L. 1993. Tillage methods and soil and water conservation
in Australia. Soil Tillage Res. 27:303–325.
Gilley, J.E., Finker, S.C., Spomer, R.G. & Mielke, L.N. 1986. Runoff and erosion as affected by
corn residue: Part I. Total losses. Transactions of the ASAE.Vol. 29(1):157–160.
Hillel, D. 1980. Application of soil physics. Academic Press, New York. 385 p.
Hillel, D. 1971. Soil water: physical principles and processes. Academic press, New York.
Hurni, H. 1986. Guidelines for development agents on soil conservation in Ethiopia. Community
Forests and Soil Conservation Development Department, Addis Ababa, Ethiopia. 100 p.
Karen, R. 1990. Water-drop kinetic energy effect on infiltration in sodium-calcium-magnesium
soils. Soil Sci. Soc. Am. J. 554:983–987.
Kinnell, P.I.A. 1996. Runoff and sheet erosion from tillage trials under artificial rainfall at Harden,
New South Wales. Aust. J. Soil Res. 34:863–877.
Lang, P.M. & Mallett, J.B. 1982. The effects of various methods of primary tillage and seedbed
preparation upon the maintenance of surface residues in a maize monoculture system. Crop
Production. 11:55–58.
Lattanzi, A.R., Meyer, L.D. & Baumgardner, M.F. 1974. Influence of mulch rate and slope
steepness on interrill erosion. Soil Sci. Soc. Am. Proc. 38:946–951.
Lindstorm, M.J. & Onstad, C.A. 1984. Influence of tillage systems on soil physical parameters and
infiltration after planting. J. Soil Water Conserv. 39:64–68.
Impact of cultivation practices on multiple uses of water 527
Meyer, L.D. 1981. How rain intensity affects interrill erosion. Transactions of the ASAE. 24:1472–
1476.
Myers, J.L. & Wagger, M.G. 1996. Runoff and sediment loss from three tillage systems under
simulated rainfall. Soil Tillage Res. 39:115–129.
Papendick, R.I., Parr, J.F. & Meyer, R.E. 1990. Managing crop residues to optimize crop/livestock
production systems for dryland agriculture. Advances in Soil Science. 13:253–272.
Rao, K.P.C., Steenhuis, T.S., Cogle, A.L., Srinivasan, S.T., Yule, D.F. & Smith, G.D. 1998a.
Rainfall infiltration and runoff from an Alfisol in semi-arid tropical India. I. No till systems. Soil
Tillage Res. 48: 51–59.
Rao, K.P.C., Steenhuis, T.S., Cogle, A.L., Srinivasan, S.T., Yule, D.F. & Smith, G.D. 1998b.
Rainfall infiltration and runoff from an Alfisol in semi-arid tropical India. II. Tilled systems.
Soil Tillage Res. 48: 61–69.
Roth, C.H., Meyer, B., Frede, H.G. & Derpsche, R. 1988. Effect of mulch rates and tillage systems
on infiltrability and other soil physical properties of an Oxisol in Parana, Brazil. Soil Tillage
Res., 11:81–91.
Sallaway, M.M., Lawson, D. & Yule, D.F. 1988. Ground cover during fallow from wheat, sorghum
and sunflower stubble under three tillage practices in central Queensland. Soil Tillage Res.
12:347–364.
Singer, M.J., Matsuda, Y. & Black, J. 1981. Effects of mulch rate on soil loss by raindrop splash.
Soil Sci. Soc. Am. J. 45:107–110.
Soil Conservation Research Project, 1996. Data base report (1982–1993). Watershed Development
and Land Use Planning Department, Ministry of Agriculture, Addis Ababa, Ethiopia.
Sonneveld, B.G.J.S. & Keyzer, M.A. 2003. Land under pressure: soil conservation concerns and
opportunities for Ethiopia. Land Degradation and Development. 14(1), 5–23.
Tefera, T.L., Krieten, J.F. & Perret, S. 2003. Market incentives, farmers’ response and policy
dilemma: A case study of chat production in the eastern Ethiopian highlands. Agrekon, Vol.
42(3), 213–227.
Unger, P.W., Stewart, B.A., Parr, J.F. & Sing, R.P. 1991. Crop residue management and tillage
methods for conserving soil and water in semi-arid regions. Soil Tillage Res. 20:219–240.
Voorhees, W.B. & Lindstrom, M.J. 1984. Long-term effects of tillage method on soil tilth
independent of wheel traffic compaction. Soil Sci. Soc. Am. J. 43:152–156.
Geochemical evidence and origin of salinity
in the shallow basinal brine from the
Makgadikgadi Pans Complexes, northeastern
Botswana
L.N.Molwalefhe
Department of Geology, University of Botswana, Gaborone, Botswana
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
Inland salt lakes are common in southern Africa, and many occur in Namibia, Botswana
and South Africa (Lancaster, 1978; Shaw, 1988; Seaman et al., 1991). All inland lakes in
southern Africa are shallow, and most are ephemeral with salinities that are less than
50g/L (Seaman et al., 1991). This study was conducted at Sua Pan, part of the
Makgadikgadi Pans Complexes of northeastern Botswana (Figure 1). The Makgadikgadi
pans are 200km from east to west and 120km from north to south and occupy a
depression that forms the center of a basin of internal drainage, the lowest point of which
is approximately 890m above sea level (masl) located at the northeastern corner of Sua
Pan. There are two major pans, Ntwetwe to the west and Sua to the east, and numerous
small pans, making up the Makgadikgadi Pans Complexes. The investigated site is
located in the northern part of Sua Pan between latitude 20°22′S to 20°27′S and longitude
25°54′E to 26°06′E. The aim of this research is to conduct a focused stable isotope-based
study to investigate the sources of salinity in groundwater brines from Sua Pan. Apart
Geochemical evidence and origin of salinity in the shallow basinal brine 529
from explaining the processes that control the formation of brines in the Makgadikgadi,
this study will also contribute towards bridging of the critical gaps in our knowledge
regarding the causes of salinity over a wide range of geologic environments.
2 GEOLOGY/HYDROGEOLOGY
The geological history of the Makgadikgadi pans is tied to the tectonic and magmatic
events that have affected the area since the Jurassic times. The rocks surrounding the
Makgadikgadi Pans belong to the Kalahari Beds and consist of aeolian sands, sandstones
and various fluviatile and lacustrine deposits. The thickness of the Kalahari deposits
varies from zero in the eastern and southern margins of the pans to over 100m north of
Sua Pan (Gould, 1986).
Similar intrusions that occur in the central parts of Botswana are dated to be coeval
(187 Ma) with the major flood basalt that is widespread in southern Africa (Smith, 1984).
All the rocks
around the pans are highly fractured by post-Karoo faulting and dyke emplacement
(Baillieul, 1979).
Historical water level data (Gould, 1986) showed that groundwater flow is largely
responsive to the topographic slope of the base of the aquifer rather than to changes in the
configuration of the water table. The base of the aquifer is very irregular (Figure 2),
probably owing to some structural control, and dips consistently from the southeast
toward the northwest (Gould, 1986). Pumping tests carried out by Paulsen (1971) defined
the aquifer conditions to be confined to semi-confined.
Groundwater samples were collected from the medium grained sand aquifer using a
network of high capacity boreholes and were analyzed in the laboratory for their
chemical and isotopic content. The main tracers that were analyzed for were the stable
hydrogen and oxygen isotopes and the major element chemistry of the brine.
The water temperature, specific conductance and pH measurements were determined
at the wellhead using appropriate electrodes. Titrimetric determinations of alkalinity were
not accomplished in the field due to the high concentration of and in the
samples. Samples for anion and cation analyses were collected unfiltered in high-density
polyethylene (HDPE) bottles. Samples for the analyses of cations were acidified to pH2
or lower using high purity nitric acid to prevent metals from precipitating during storage.
The samples were refrigerated at 4°C during storage. For the stable isotopes of oxygen
and hydrogen, water samples were collected in 20ml glass vials with inverted cone
closures to minimize headspace and potential evaporation (Gat, 1996).
Geochemical evidence and origin of salinity in the shallow basinal brine 531
Concentrations of anions (Cl−, Br− and ) and cations (Na+, K+, Ca2+ and Mg2+)
were analyzed using a Dionex DX-500 ion chromatograph (Dionex, Sunnyvale, CA,
USA) equipped with a conductivity detector and an AS40 auto sampler. All samples were
diluted 500-fold using ultra-pure de-ionized water to bring the concentrations within the
calibration range of the standards.
The stable oxygen isotope (δ18O) was analysed following the widely used and
established technique of isotopically equilibrating water with pure CO2 in a constant
temperature environment (Epstein and Mayeda, 1953). In this procedure, 2ml of water
sample is equilibrated with 0.5 atmospheres of CO2 for 25 h at 25°C.
In this study, all groundwater samples were concentrated chloride brines and had to be
vacuum distilled at 900°C (Molwalefhe, 2003) before reaction with zinc (Coleman et al.,
1982). Chlorides have been shown to inhibit the metal-water reaction (Yang et al., 1996).
The results for δD and δ18O isotope ratios are reported on the V-SMOW scale. The
analytical precision for δ18O measurements is 0.2‰ and for δD is ±2‰.
Water resources of arid areas 532
before calcite saturation could be reached. This raises the pH to 8 or more. Considering
the low pCO2 of the atmosphere, it is difficult to account for high DIC in the
Makgadikgadi brine using a single-stage mechanism of evaporative concentration of rain
or surface water alone. A combination of occasional flooding of the pans followed by fast
evaporation prior to infiltration, and contacting with a high pCO2 environment in the
Geochemical evidence and origin of salinity in the shallow basinal brine 533
By using the average δ18Oprec value for the Pretoria meteoric water (−3‰), average
δ18OGW of +3‰ (see Figure 4) and assumed humidity of 50%, the calculated amount of
water, f, that remains is 69%. This suggests that about 31% evaporation has taken place
to cause the shift in the brine compositions.
The above calculation does not provide a realistic conclusion that the salinities in the
brine are a result of evaporating surface water by only 31%. If equation 1 is considered,
and a concentration factor of 1000 (f=0.001) is applied to surface waters to produce the
salinities, the isotopic values for the remaining water can be calculated to determine if
isotopic values that could be produced in the brine.
A calculated value of +113‰ is obtained which is much different from the isotopic
value of the brine (+3‰). Therefore, another source might be contributing water that is
isotopically light and more saline. In Figure 5, the meteoric water equation is applied on
the actual brine compositions to provide a reconstruction of the expected isotopic
composition in the brine. Values that are expected because of evaporating rainwater fall
along a line, and indicate heavier compositions compared to the measured values.
Assuming the brine represents evaporated meteoric water, then the isotopic compositions
should range from +17 to +32‰ for δ2H. It is clear from Figure 5 that
5 CONCLUSIONS
The geo-chemical and isotopic studies of the hyper-alkaline Na-Cl brine from the
northern part of Sua Pan in the Makgadikgadi area have indicated that the brine has likely
evolved along various geochemical pathways and underwent more than one stage of in-
Geochemical evidence and origin of salinity in the shallow basinal brine 537
lake modifications via a sequence of surface and subsurface processes. The study
demonstrated the need for a multiple-stage process to account for the high salinities in the
brine. Chemical data also suggest that surface evaporative processes have played an
equally significant role in concentrating salts in the brine while DIC concentrations
showed that the brine has imprints of a magmatic source. This model indicates that the
evolution of the Makgadikgadi brine is partially maintained by underground discharge
and partially by meteoric recharge.
REFERENCES
Baillieul, T.A. 1979: Makgadikgadi Pans Complex of central Botswana. Geological Society of
America Bulletin, Part II, v.90 pp. 289–312.
Campbell, A.R. and Larson, P.B. 1998: Introduction to stable isotope applications in hydrothermal
systems. In: Techniques in hydrothermal ore deposits geology, Society of economic geologists,
Reviews in economic geology v.10, pp. 173–193.
Coleman, M.L., Shepard, T.J., Durham, J.J., Rouse, J.E. and Moore, G.R. 1982: Reduction of water
with zinc for hydrogen isotope analysis. Anal. Chem., 54:993–995.
Coplen, T.B., Herczeg, A.L. and Barnes C. 2000: Isotope engineering—Using stable isotopes of the
water molecule to solve practical problems: In Environmental Tracers in Subsurface Hydrology.
P.Cook and A.L.Herczeg (Eds) Kluwer Academic, Dordrecht, pp. 79–110.
Clark, I.D. and Fritz, P. 1997: Environmental Isotopes in Hydrogeology. Lewis Publishers, New
York, pp. 112–123.
Epstein, S. and Mayeda, T. 1953: Variations of 18O content of waters from natural sources.
Geochim. Cosmochim. Acta, 4:213.
Gat, J.R. 1996: Oxygen and hydrogen in the hydrologic environment. Ann. Rev. Earth Plan. Sci.
24:225–262.
Gonfiantini, R. 1986: Environmental isotopes in lake studies. In: Handbook of environmental
isotope geochemistry, Vol 2, eds. P.Fritz and J.-C.Fontes, Elsevier, Amsterdam. pp. 113–168.
Gould, D. 1986: Brines of Sowa Pan and Adjacent Areas, Botswana. In Mineral Deposits of
Southern Africa Vols I and II, Anhacusser C.R. and Maske S. (Eds) pp. 2289–2299.
Lancaster, I.N. 1978: The pans of the southern Kalahari, Botswana. Geographical Journal 144:80–
98.
Reeves, C.V. 1972: Rifting in the Kalahari? Nature 237:95–96.
White, D.E. 1974: Diverse origins of hydrothermal ore fluids. Economic Geology 69:954–973.
Yang, W., Krouse, H.R. and Spencer, R.J. 1996: Stable isotope analyses of microlitre quantities of
water from inclusions in halite and concentrated brines. Chem. Geol (Isot. Geosc. Sect.)
130:139–145.
Theme F:
Vulnerability and risk
Decision support for optimal water system
planning: a Wadessy case study
A.A.Ilemobade
School of Civil and Environmental Engineering, University of the
Witwatersrand, South Africa
D.Stephenson
Department of Civil Engineering, University of Botswana, Gaborone,
Botswana
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
The design of water distribution systems (WDSs) has received a great deal of attention
because of its importance to industrial growth and water’s crucial role in society for
health, fire-fighting, and quality of life, particularly in light of increased urban
development and water use (Sherali et al 1998:1381). WRNs are essential components of
all WDSs as they convey potable water from source, pump station or storage to the
consumers. The cost of these networks may amount to as much as 60% of the entire
water supply scheme (Sarbu & Borza 1997; Stephenson, 1998:49) and as a result,
operation and maintenance costs may soar higher if networks are ill designed (Ilemobade
Water resources of arid areas 542
& Stephenson 2002:80). WRNs also account for the largest costs in municipal
maintenance budgets (Sherali et al 1998:1381). Despite often scarce resources, national
and local governments are obligated to provide this resource. Since WRNs are composed
mostly of pipes, pipe-sizing decisions have become critical during the design of cost
effective WDSs capable of handling varied demand loadings and satisfying minimum
pressure head requirements. Optimisation techniques have been proven to facilitate the
best WRN designs.
The basic equations of continuity, conservation of energy and pressure head difference
are utilised in modelling WRNs. To arrive at an optimal solution, an iterative simulation-
optimisation algorithm is employed. Efficient hydraulic simulation (both static and
dynamic) in Wadessy is based on modelling the WRN using the above-mentioned
equations and determining the unknown variables using the established Newton-Raphson
iterative procedure on simultaneous equations generated using the nodal method (Cornish
1939). Pipe sizes (which are initially assumed for new WRN designs) and other pipe
parameters, consumer demands, network layout configuration, pump constants and fixed
grade node (FGN) elevations are known prior to simulation. The Choleski Decomposition
technique (Stoer & Bulirsch 1993) is employed to generate the matrix for computing
node pressures. Based on either the Darcy-Weisbach or Hazen-Williams pipe equations,
continuity is checked at each network node. If a violation exists, a correction factor is
introduced into the procedure and the entire simulation process is repeated. Output from
the simulation include pipe flows and orientation, pipe headlosses, friction factors, node
residual pressure heads, draw-off at each source node, pumping heads and valve head
losses.
Wadessy’s design optimisation procedure is adapted from Featherstone & El-Jumaily’s
(1983) model, which is based on the concept that a hypothetical linear hydraulic gradient,
So for a balanced WRN exists by which the initial network design can be iteratively
corrected to produce optimal pipe sizes and an optimal relation between each pipe. In
addition to Featherstone & El-Jumaily’s (1983) model, the effects of hydraulic surfaces
in determining optimal designs were previously undertaken by Deb & Sarker (1971), Wu
(1975) and Alperovits & Shamir (1977). Ilemobade & Stephenson (2003:3) discuss
Featherstone & El-Jumaily’s WRN optimisation concept, the work published by the
workers mentioned above and Wadessy’s optimisation model.
3 MODEL VALIDATION
each demand node is 30m. Since several workers have used the results presented by
Alperovits & Shamir (1977) to validate their models, a comparison of the results obtained
by Alperovits & Shamir (1977), Farmani et al (1999) and Wadessy are presented in
Tables 3 and 4.
The optimal solution achieved by Wadessy was obtained after running the optimisation
procedure from several starting network designs. Flow distribution is treated as a variable
until an optimal solution is reached. A 1, 5% deviation from the minimum residual
pressure (30m) was permitted during computations, hence a computed residual pressure
head of 29, 62m for Node 6 (the critical node). The global optimum network design cost
computed by Wadessy is $262000,00
new residential developments. During analysis, Wadessy was also used to hypothetically
determine the optimal WRN design for Selebi-Phikwe in comparison to that existing,
based on October 2001 consumer demands and pipe costs (Table 5). Average peak and
night flows calculated for October 2001 were 0,270m3/s and 0,135m3/s respectively.
20Ml of storage is provided in 3 concrete cylindrical tanks situated at the southeast
section of Selebi-Phikwe. Minimum and maximum node pressure heads are 15m and
90m respectively (WUC, 1995). Minimum pipe sizes to connect users to the reticulated
mains and on which fire hydrants will be located are 63mm & 75mm respectively (WUC,
1995). Three pumps are installed in the existing pumpstation; one of which acts as
standby. The polynomial equation: Hp=AQ2+BQ+C adequately represents pumping head
within the WRN. For two pumps in parallel: A=−414, 94; B=−3,50 & C=88, 93. For
pumps operating individually: A=−1660, 30; B=−12, 31 & C=90,04.
Table 1. Pipe cost data.
Pipe diameter 1 2 3 4 6 8 10 12 14 16 18 20
(in.)
Unit cost per m 2 5 8 11 16 24 32 50 60 90 130 170
($)
Detailed results of the optimisation are presented in Figures 3, 4 and 5. Figures 4 and 5
present node residual pressure head results for the existing and Wadessy optimised WRN
based on peak and night flow conditions. It can easily be seen that although both network
designs are adequate under night flows, the existing network design is deficient under
peak flows where the minimum permissible node pressure head (15m) is violated in
nodes 4, 6 and 28 (9, 64m, 13, 42m and
4 CONCLUSIONS
5 SOFTWARE PROCUREMENT
Wadessy’s WRN design software may be obtained from the author at a minimal fee.
ACKNOWLEDGEMENT
The authors are indebted to the Richard Ward Endowment Fund (fund no. TW CIVN
ILEM) for funding this research.
REFERENCES
Alperovits, E. & Shamir, U. 1997. Design Of Optimal Distribution Systems. Water Resources
Research, 13(6):885–900.
Cornish, R.J. 1939. The Analysis of Flow in Network of Pipes. Journal Institute of Civil Engineers,
13:147.
Deb, A.K. & Sarkar, A.K. 1971. Optimisation in Design of Hydraulic Network. Journal of the
Sanitary Engineering Division. Proceedings Paper 8032, ASCE, 97(SA2):141–159.
Farmani, R., Matthew, R.G.S. & Javadi, A.A. 1999. Discrete Optimisation of Water Distribution
Networks using Genetic Algorithms. In D.A.Savic & G.A.Waters (eds.) Water Industry
Systems: modelling and optimisation applications (Vol 2), Water Engineering Management
Series, Research Studies Press, England: 427–436.
Featherstone, R.E. & El-Jumaily, K.K. 1983. Optimal Diameter Selection for Pipe Networks.
Journal of the Hydraulic Division, ASCE, 109(HY2):221–234.
Ilemobade, A. & Stephenson, D. 2002. Optimally Upgrading Small Water Reticulation Networks
using Wadessy: Case Study—Selebi-Phikwe, Botswana. Proceedings, International Conference
on Water—The Lifeblood of Mankind and 5th Biennial Congress of the African Division of the
IAHR in collaboration with the IET. Arusha, Tanzania. 11–13 December:79–88.
Ilemobade, A.A. & Stephenson, D. 2003. Generic Optimisation of Small Water Reticulation
Networks using Wadessy. Journal of the South African Institution of Civil Engineers. Paper 548.
45(4):2–9.
Sarbu, I. & Borza, I. 1997. Optimal Design of Water Distribution Networks. J. of Hydraulic
Research, 35(1): 63–79.
Sherali, H.D., Totlani, R. & Loganathan, G.V. 1998. Enhanced Lower Bounds for the Global
Optimisation of Water Distribution Networks. Water Resources Research, 34(7):1831–1841.
Stephenson, D. 1998. Water Supply Management. Water Science and Technology Library. Kluwer
Academic Publishers.
Stoer, J. & Bulirsch, R. 1993. Introduction to Numerical Analysis. Texts in Applied Mathematics
12. 2nd Ed., Springer-Verlag Inc, New York: 180.
Wu, I.P. 1975. Design of Drip Irrigation Main Lines. Journal of the Irrigation Drainage Division.
Proceedings Paper 11803, ASCE, 101(IR$):265–278.
WUC—Botswana Water Utilities Corporation, 1995. Design Standards, DS 0295. Gaborone, 22
February.
The importance of constructing a correct
conceptual model for an aquifer
G.van Tonder, I.Dennis & D.Vermeulen
Institute for Groundwater Studies, University of the Free State,
Bloemfontein, South Africa
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
5. A description of the processes and interactions taking place within the study area that
will influence the movement of groundwater including aquifer parameters and
recharge calculations.
The paper focuses on a few of the problems associated with the development of an
inappropriate conceptual model when solving groundwater problems and present some
solutions to these problems.
2.1 General
Problems associated with conceptual models can be the result of:
● Incorrect analysis of aquifer tests.
● Misunderstanding the results of aquifer tests.
Table 1. Hydraulic parameters for borehole UO5.
Parameter Value
T of formation* (m2/d) 19
K of facture zone (m/d) 3600
T of fracture zone (m2/d) 576
K of matrix (m/d) 0.17
T of matrix (m2/d) 3
*
Average for fracture & matrix.
The Campus Test Site is underlain by a series of mudstones and sandstones from the
Adelaide Subgroup of the Beaufort Group of formations in the Karoo Supergroup. There
are three aquifers present on the Site:
● The first, a phreatic aquifer, which occurs within the upper mudstone layers on the Site.
A confining layer of carbonaceous shale with a thickness of 0.5 to 4m separates this
aquifer from the second aquifer.
● The second and main aquifer, occurs in a sandstone layer is between 8 and 10m thick, a
bedding plane fracture occurs in this aquifer at approximately 21m bgl.
● The third aquifer occurs in the mudstone layers (more than 100m thick) that underlie
the sandstone unit.
The known parameters of a borehole UO5 at the Campus Test Site are listed in Table 1.
By performing and analysing a slug test, a K of 17m/d (T=330m2/d) was obtained,
which does not compare with that of the fracture, the matrix or the average for the
fracture and the matrix.
Performing constant rate discharge tests at different abstraction rates result in different
yield predictions. Borehole M11 at the Meadhurst Test Site, Bloemfontein, South Africa
is used to demonstrate this effect. Borehole M11 was drilled along a dolerite dyke with
the main water strike 30m below the rest water level, which is situated at 22m below
surface. The dolerite dyke was intersected at 28m below the rest water level. At 30m
below the rest water level a water strike of 4L/s was encountered in the dolerite. Two
constant rate tests were performed on M11—one at 3L/s and the other at 7L/s. Figure 1
shows the pumping test results, as well as the information on the water strikes.
It is clear from Figure 1 that the results from the two abstraction rates produced
different drawdown curves. The estimated T-value with the 3L/s abstraction rate is
170m2/d, while the estimated T-value for the 7L/s second abstraction rate is 20m2/d. At a
rate of 7L/s, the fractures could not sustain the abstraction rate with the result that the
estimated T-value of 20m2/d is the formation T-value, while the lower rate gives a T-
value more representative of the fractures.
Water resources of arid areas 554
between the abstraction and observation boreholes (the larger the distance, the smaller the
estimated S-value).
Neuman (1994, personal communication) gave the following possible explanation:
Consider the rock to consist of nested storage ‘reservoirs’ comprising different scale
fractures. At one end of the spectrum are a few large, permeable fractures occupying a
small relative rock volume which therefore has small porosity and storativity. On the
other end are many small, low-permeable fractures occupying a relatively large rock
volume which therefore has large porosity and storativity. Close to the pumping well,
pressure in the large fractures declines rapidly relative to its rate of decline in the small
fractures. The latter therefore release a relatively large amount of water into the large
conductive fractures due to a sizeable local pressure gradient between the small and large
fracture reservoirs. Hence S is large. Far from the pumping well, the pressure gradient
between the small and large fractures is relatively small. Therefore, water release from
the small to the large fractures occurs very slowly. Most of the initial drawdown (in the
large fractures) at a great distance is associated with water release from storage in the
large fractures. Hence S is small.
With time, local pressure differentials between the reservoirs stabilise and flow
everywhere within a given radius approaches a steady radial pattern. Therefore, it could
be expected that S should approach a uniform value representing both reservoirs.
However, as the flow pattern is now essentially stabilised and close to steady state (even
though absolute pressures may continue to decline), standard pumping tests may not
reveal this fact: the flow is sensitive to S only at early times. If there were only two
reservoirs with very different S-values, log-log time-drawdown curves close to the
pumping well would exhibit a familiar dual-porosity time inflection (of the kind analysed
by Neuman for unconfined aquifers). However, if there is a continuous hierarchy of such
reservoirs with a more or less continuous local range of T- and S-values, such inflections
cannot be seen. The early log-log time-drawdown behaviour would then just look like a
regular Theis curve. Only long pumping tests would reveal deviations from this curve,
but unfortunately, storage effects during late behaviour are usually masked by large-scale
heterogeneities and boundary effects.
An example of this incorrect estimate was illustrated by Brook (1990) for the Jwaneng
Mine in Botswana. Initial S-estimates from pumping test analyses were in the order of
1×10−4 while model calibration after a number of years of monitoring showed a reliable
S-value in the order of 1×10−2.
improve estimates of the first variable (groundwater). In other words, topography can be
sampled more frequently than groundwater levels, thus this information can be used to
improve estimates of groundwater levels. Bayesian interpolation is a method that uses
this principle.
A groundwater investigation of the Hendrina Power Station, South Africa, was to be
conducted. For this study groundwater levels had to be generated. The distribution of
available data for the study area is shown in Figure 2.
Initially Kriging was used to generate water levels (see Figure 3). However, there is a
correlation of 87% between topography and groundwater levels. Therefore Bayesian
interpolation can be used to generate groundwater levels (Figure 3). Field studies have
verified that Bayesian interpolation produced a more accurate water level map.
REFERENCES
Brooke, M.C. 1990. The reliability of a groundwater model: The history of modeling the Jwaneng
Northern Wellfield in Botswana. In: Karel Kovar (ed.). ModelCARE 90: Calibration and
reliability in Groundwater Modelling. Proc. Intern. Conference, The Hague, 3–6 September
1990, IAHS Press.
Neuman, S.P. 1994. Personal Communication. University of Arizona, Tuscon, USA.
Water resources development and risk
assessment in mountain regions of Africa
Helmut Scheuerlein
University of Innsbruck, Austria
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
Regardless whether African engineers in charge of water works are dealing with water
resources exploitation or with protection against floods or droughts, their work will be
governed by the given natural, social and economical conditions of their environment.
Whatever their particular project is aiming at, the solution has to be oriented towards
sustainability. Impediments on the way towards a sustainable solution may become
effective through budgetary constraints, social requirements, insufficient expertise,
unconsciousness of ecological coherence, ignorance of environmental aspects, etc. Very
often, budgetary constraints make it necessary to apply for external financial support and
the donors dictate conditions which may be justified in economically saturated societies
in Europe or North America but may be questionable for the African country in question.
Another problem which is also related to the notoriously limited budgets in Africa is the
competition between water resources development and disaster prevention. Investments
spent for a water resources development project can be expected to pay back as soon as
the project is operational whereas money paid for mitigation measures with respect to
disaster prevention never produces direct revenues. The benefit of this type of costs lays
in risk reduction concerning potential hazards without any visible income.
In the Usambara mountains hazardous floods can not be avoided nor can the
population pressure upon the land be released. Sustainability of water resources
development can only be reached with a joint effort of various measures on varying
levels, namely
– on legal level:
prohibition of deforestation and up and down slope cultivation at slopes steeper
than an agreed value
– on educational level:
raining of local farmers with respect to soil conservation methods and appropriate
farming practices
– on technical level:
stabilization of mountain streams in steep catchments by means of check dams,
gully plugs, drop structures, sediment retention basins before the flow enters the
plains, etc.
The measures on educational level are certainly the most difficult and time consuming
part. Fortunately enough this part is already well in progress since the early eighties of
the past century. The respective conservation programme is called SECAP (Soil Erosion
Control and Agroforestry Project), an initiative of the German foreign aid company GTZ
(Scheuerlein et al, 1998). Meanwhile, the project has been handed over to the local
authorities and is carried on by them with full responsibility (Fig. 4). The overall goal of
SECAP is stabilization and ecological balance of all West Usambara watersheds. The
project was introduced to the local people from the beginning as Farmers in the Western
Usambaras apply ecologically adapted and economically sustainable farming systems.
The meritorious activities of SECAP are a successful approach towards soil
conservation and erosion control in the catchment area, however, the water courses
draining the catchment are not included in this programme. Being aware of the fact that
all sediment eroded in the catchment sooner or later ends up in one of the creeks, streams,
or rivers of the watershed, it becomes obvious that erosion control can not ignore the
necessity to rehabilitate and stabilize the water courses, too, particularly those incised in
the steep slopes. Stabilization of water courses, however, requires engineering expertise.
The problems to be solved and the mitigation measures which should be applied have
been identified by Scheuerlein et al (1998) as
− Stabilization of steep torrential streams to withstand destruction even when charged
with highly sediment-laden flow
Measures: Strengthening of existing step-pool systems or construction of new
ones, preferably with natural local material; at very steep slopes, construction of
consecutive gully plugs or drop structures.
– Stabilization of riparian slopes potentially endangered of sliding into the streams
Measures: Drainage of the slopes; construction of retaining walls; strengthening
of the stream banks.
– Protection of the plains of the plateau against devastation through debris and mud flow
Construction of sediment retention basins by means of check dams or self-
regulating barriers capable of automatic grain sorting.
As the above mentioned measures are time consuming and costly, careful
risk assessment must be carried out beforehand to allow for optimized
selection and ranking of the measures.
Water resources of arid areas 566
world, f.i. in the Sahel region south of the Sahara desert). In northern Africa, the people
finally also gave up irrigation-based agriculture and started to move further downhill to
the attractive centers downstream of the large reservoirs where they expected more
prosperous living conditions.
Meanwhile, the abandoned soils on the high plateaus had deteriorated within short
time due to sheet, rill and gully erosion and the land was turned to badlands with
extremely high erosion rates. As a consequence, the small irrigation ponds in the valleys
also experienced rapid sedimentation and—as the inexperienced dry farming trained
people were not used to operation and maintenance of diversion dams—most of these
dams became either useless due to complete sedimentation or were destroyed by
excessive floods due to inefficient flood control operation (Fig. 6, and Scheuerlein,
2002).
Deterioration of the uncultivated plateau and failing to convert the dry farming people
to irrigation technicians led to considerably faster sedimentation of the large reservoirs
than
Water-related hazardous events in African mountains can be associated with at least one
of the following phenomena:
– torrential floods
– debris flow
– landslides
In the example dealing with the High Atlas (Morocco) torrential floods must be
considered the prime risk but debris flow events and landslides may occur as well. In
Usambara (Tanzania) debris flow is the most dangerous event. It occurs together with
floods but the highest hazard potential lays in the devastating consequences of debris
flow. Recently, also landslide events contribute to debris flow, often caused by
inappropriate land-use practices. As far as the third example (Tell Atlas, Algeria) is
concerned the hazard can not be related to one single event like in the other examples, it
rather must be seen as an accumulative process caused by series of events like floods,
debris flows and landslides in the whole catchment of a reservoir.
Mitigation measures to minimize the risk of damages or failing of water resources
development works can be divided in structural and non-structural measures
(Scheuerlein, 2000b):
– Structural measures
Structural measures usually comprise erosion control, river training works (f.i.
bed and bank protection). In lower regions of the catchment sediment retention
basins may also be considered.
– Non-structural measures
Non-structural measures comprise hazard zone mapping as the main component.
Hazard zone mapping is a rather new method, developed mainly in Alpine
countries like Austria and Switzerland. In Austria hazard zone mapping is
instrumental since more than 35 years
Water resources development and risk assessment in mountain regions of Africa 569
evaluate and compare risk potential and costs for protective measures. In Austria
a hazard zone map consists of two components, one cartographic part and one
report. The cartographic part comprises a topographical map of scale 1:50000,
1:25000 or 1:20000 which covers the area to be evaluated plus the relevant
catchment. The endangered zones (hazard zones) are indicated in a separate map
scale 1:5000 (Fig. 7), with colours indicating the degree of the hazard potential
(red and yellow as the extremely dangerous zones and brown for landslides, blue
and violet for additional information concerning special restrictions). The
classical criteria for the evaluation is an event of 150 years of reoccurrence. At
present, however, there are plans to develop a more scientific concept combining
a probability approach with systems analysis and causality chains and decision
trees instead of working with one single event only.
The risk potential of disastrous events can be significantly reduced by applying
appropriate structural measures. However, as structural measures need time and money,
the feasibility of protective measures has to be investigated, too. The respective
evaluation is in Austria always done on the basis of hazard zone maps. This tool is also
helpful for decision making purposes (i.e. for selection and ranking of structural
measures).
6 CONCLUSIONS
The Summit on Sustainable Development in Johannesburg has declared water as the raw
material of the 21st century and the access to safe drinking water as a basic human right.
Furthermore, Africa was identified for special attention and support by the international
community to better focus efforts to address the development needs of Africa.
Water resources development in Africa has to be oriented towards sustainability,
particularly in mountain regions. Impediments are budgetary constraints, social
requirements, often combined with insufficient expertise, unconsciousness of ecological
coherence and ignorance of environmental aspects.
Examples from various regions of Africa (Morocco, Tanzania, Algeria) show that the
main problems to be dealt with in mountain areas are torrential floods, debris flow and
landslides. Measures to tackle the problems may be structural and/or non-structural.
Structural measures comprise erosion control, river training works and sediment
retention basins. Non-structural measures are risk assessment with hazard zone mapping
as the main component (Austria as example).
The risk potential of disastrous events can be reduced by applying appropriate
structural measures. Assessment of the risk potential can be carried out on the basis of
hazard zone mapping.
REFERENCES
Gorner, D. 1993. Ecoulement et erosion dans des petits bassis-versants à sols marneux sous climat
semi-arid mediterraneen, Deutsche Gesellschaft für Technische Zusammenarbeit (GTZ), 207pp.
Water resources development and risk assessment in mountain regions of Africa 571
Scheuerlein, H. 2000a. Determination des zones sensibles et des mesures de protection en bassins
versants de haute montagne, Invited Paper, Séminaire Etude préparatoires a la creation de
L’Agence du Bassin Hydraulique de Tensift (ABHT), Marrakech, Maroc, Rapport final, 23
pages.
Scheuerlein, H. 2000b. Risk assessment and minimization in steep catchments, Invited Paper, 8th
International Symposium on Stochastic Hydraulics, Beijing, China, Proceedings, 413–423.
Scheuerlein, H. 2001. Hazard zone mapping in Austria, XXIX IAHR Congress, Technical Workshop
6, Risk Management in Mountainous Areas, Beijing, China.
Scheuerlein, H. 2002. Sediment problems initiated through unexpected reactions of the public upon
reservoir operation strategies, UNESCO/ICCORES Workshop on Ecological, Sociological and
Economic Implications of Sediment Management, Paestum, Italy, Proceedings: 178–198.
Scheuerlein, H. & Kommes, Ch. 1998. Sustainable water resources development in the tropical
watershed of Western Usambara, Tanzania, Conference of IAHR African Division on Coping
with Water Scarcity, Hurghada, Egypt, Proceedings: 3.4–1 to 3.4–8.
Reliability, resilience and vulnerability for
reservoir sizing and operation
J.G.Ndiritu
School of Civil and Environmental Engineering, University of the
Witwatersrand, South Africa
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
Reliability, defined as the proportion of time out of the total simulation period that there
are no shortfalls to supply has been traditionally used as the criteria for reservoir sizing
and also in stochastic optimization of reservoir operation. Reliability is definitely an
important indicator of system performance but it fails to capture some vital aspects of
performance. These include the severity of shortfalls once they occur and the ability of
Reliability, resilience and vulnerability for reservoir sizing and operation 573
the system to recover from shortfalls. As Srinivasan et al. (1999) point out; continuous
shortfalls of supply generally result in greater negative impacts than intermittent
shortfalls. Continuous shortfalls may for example cause farmers to alter their irrigating
habits permanently. Long shortfalls may also stress out the alternative water source (e.g.
groundwater) meant to supplement supply from reservoirs during droughts. The term
resilience has been used for measures quantifying the ability of the system to recover
from supply shortfalls and the term vulnerability for measures that quantify the severity
of shortfalls.
Several studies on resilience and vulnerability have been carried out (Hashimoto et al.,
1982, Srinivasan et al., 1999, Moy et al., 1986, Vogel et al., 1999). Hashimoto et al.
(1982) applied a simple seasonal model in the analysis and Moy et al. (1986) analyzed a
reservoir using a short period of 18 monthly inflows that Srinivasan et al. (1999) also
applied. Vogel et al. (1999) applied a simulation of an annual lag 1 auto regressive model
and used a yearly time scale. Arid and semiarid regions are prone to prolonged droughts
and many parts of Southern Africa are currently facing severe drought conditions.
Incorporation of resilience and vulnerability into system design and operation is therefore
even more important in such regions. This analysis of the relationships provides a sense
of the inadequacy of using reliability as the sole performance criterion and the need to
incorporate resiliency and vulnerability at the sizing and operating stages of water
resource systems.
This study aimed to obtain relationships between reliability, resilience and
vulnerability for a reservoir system in located in a semi-arid area. A realistic simulation
of a system of two reservoirs incorporating reservoir operating shortage rules is used. A
monthly time interval is applied and the system is simulated 100 times using synthetic
streamflow and rainfall sequences, each 77 years long. The possibility of applying
volumetric reliability, defined as the proportion of the volume of water supplied to the
volume demanded, in place of the traditional time-based definition of reliability is
investigated. An analysis of alternative ways of defining resilience and vulnerability is
also carried out.
Resilience is the ability of a system to recover from droughts has been quantified in a
variety of ways. Hashimoto et al. (1982) defined resilience as the average probability of a
recovery from the failure set in a single time step.
(1)
where γ is the resilience; ρ is number of times the reservoir shifts between full supply and
restricted supply as a ratio of the total number of periods of analysis and 1−α is the
number of periods of restricted supply expressed as a ratio of the total number of periods
of analysis. According to equation 1, a larger value of γ means a more resilient system.
Moy et al. (1986) defined resilience as the longest number of consecutive supply
restrictions that occur with a lower value meaning a more resilient system. Srinivasan et
Water resources of arid areas 574
al. (1999) extended this definition further to incorporate the ability to recover from all the
droughts and not the worst drought alone. This definition has implicit equivalence with
that of Hashimoto et al. (1982) defined in equation 1.
The current analysis uses two definitions; the longest period of consecutive
restrictions; and the average time taken for the system to recover from restrictions or the
average number of months that a failure is expected to occur. The reciprocal of this
definition gives the definition of equation 1. This definition is considered more practical
as it relates to performance more directly than equation 1. The lower the number of
months to recovery, the more resilient the system.
Hashimoto et al. (1982) defined vulnerability as the summation of the product of the
maximum severity of a given failure and probability of its occurrence in the failure event.
(2)
where ν is the vulnerability; j is the period of the severest failure in failure event F; sj is a
measure of this severity; and ej is the probability that the failure that corresponds to sj is
the most severe outcome in failure event F. Moy et al. (1986) and Srinivasan et al. (1999)
defined vulnerability as the largest deficit during the period of simulation.
The simulation herein includes restriction rules that reduce the supply by specified
percentages depending on the storage state of the reservoirs and the month of the year.
The definition of vulnerability as the single most severe shortage is therefore unsuitable
as it is likely to be mostly confined to these percentage reductions. A definition that
considers only the worst single period may be unrealistic especially when working with
time intervals much shorter than the probable lengths of droughts. The following two
definitions of vulnerability are therefore applied herein; the worst cumulative deficit
during a drought event; and the average of the cumulative deficits during the simulation
period.
The configuration and historical data from a system of two reservoirs located in the
Elands river catchment in South Africa was selected for analysis. The upper dam, Rust de
Winter has a catchment area of 1145km2 and a mean annual runoff of 19.8mm. The
incremental area for the downstream reservoir, Mkombo is 2578km2 and the mean annual
runoff from the incremental area is 3.9mm. The mean historical mean point rainfall at
Rust de Winter and Mkombo is 605 and 243mm respectively. Seventy seven years of
monthly streamflows and point rainfalls at the reservoir site was available. Monthly
average Symon’s pan evaporation rates were also available. This system had been applied
in another study (Ndiritu, 2004a), which aimed to obtain reservoir sizes and operating
rules that maximize yield whilst dealing with multiple constraints of supply reliability.
Reliability, resilience and vulnerability for reservoir sizing and operation 575
One of the optimized solutions obtained from this study was selected as the basis of the
numerical analysis carried out here. This solution included the optimized live storages of
the two reservoirs and the monthly operating rule curves for each reservoir and for the
total system storage. The operating rules allowed for 4 levels of supply (100%, 80%, 60%
and 30%) each month depending on the volume in storage in each reservoir and the total
system storage. Allowance for lower percentages of supply (<30%) and for regulated
flow from the upper to the lower reservoir was also incorporated. Figure 1 presents a
schematic of the system.
To achieve the aims of the study, it was necessary to perform several system
simulations each providing corresponding values of reliability, resilience and
vulnerability for each reservoir. The historical streamflow and rainfall data was used to
generate 100 synthetic sequences of the same length (77 years) using a nonparametric
stochastic data generation method. This method, developed by the author, is based on the
premise that the historical data can be rearranged a large number times to generate series
that capture the important characteristics without having to ‘decompose’ the historical
data into statistics and then build it up again. This method captures the droughts in the
historical flow and produces a synthetic flow that has have the same number and lengths
of droughts but assigns these randomly over time. The generated flow series thus
maintains the important aspects of the annual serial correlation. The cross correlation
between streamflows is captured by noting the common drought years between the
historical streamflows and generating synthetic flows that have a similar number of
common droughts lengths and periods. The historical monthly distributions are used to
disaggregate the annual flows in a manner that maintains the serial correlation at the end
of one year and the beginning of the next. Tests on this approach (Ndiritu, 2004b)
indicate it is effective and efficient.
Figures 2 to 7 present the results graphically. Rust de Winter dam is denoted by ‘RdW’
and Mkombo dam by ‘Mko’ in the Figures. The following observations are made.
Reliability, resilience and vulnerability for reservoir sizing and operation 577
– There are high correlations between the two definitions of each performance indicator
(Figures 2, 3 and 4).
– A correlation exists between reliability and resilience but with a large scatter (Figure 5).
– A correlation with a large scatter also exists between reliability and vulnerability
(Figure 6).
– A high correlation exists between resilience and reliability (Figure 7).
– Rust de Winter gives a greater scatter that Mkombo in all the relationships.
The high correlation between the two definitions of reliability implies no advantage is
likely to be obtained by replacing one with the other. Relationships between volumetric
reliability and the two other performance measures were of similar quality to those
obtained with the traditional definition of reliability. The significant difference in the
relationships for Rust de Winter and for Mkombo dam is an indication that generalization
of these relationships may be difficult except for
The relationships between reliability and two other reservoir performance measures:
resilience and reliability have been studied using data and the configuration of a system
of 2 reservoirs located in a semi-arid region of South Africa. The relationships obtained
indicate that in general, a more reliable system will have a higher resilience and a lower
vulnerability. These relationships however exhibited a large scatter and generalized
relationships that provide the resilience and vulnerability for a given reliability may not
be obtainable. It is therefore recommended that measures of resilience and vulnerability
should be incorporated into reservoir system sizing and operation studies. Water
resources planning and management is shifting towards greater stakeholder participation.
For this reason, it is desirable that measures of resilience and vulnerability should be
defined in a practical way that is easily understandable by the stakeholders and not only
the systems analysts.
Reliability, resilience and vulnerability for reservoir sizing and operation 579
REFERENCES
Hashimoto, T., Stedinger, J.R. & Loucks, D.P. 1982. Reliability, resiliency and vulnerability
criteria for water resource system performance evaluation. Water Resour. Res., 18:14–20.
Moy, W.S., Cohon, J.L. & Re Velle, C.S. 1986. A programming model for analysis of the
reliability, resilience and vulnerability of a water supply reservoir. Water Resour. Res., 22:489–
498.
Ndiritu, J.G. 2004a. Optimizing Water Supply System Yield Subject to Multiple Reliability
Constraints using Genetic Algorithms. Submitted to J. Water Resour. Plann. and Manag.,
ASCE.
Ndiritu, J.G. 2004b. ‘A pragmatic nonparametric approach for monthly multisite streamflow
generation’. Paper in preparation.
Srinivasan, K., Neelakantan, T.R., Shyam Narayan, P. & Nagarajukumar, C. 1999. Mixed-integer
Programming Model for Reservoir Performance Optimization. J. Water Resour. Plann. and
Manag., ASCE, 125(5):298–301.
Vogel, R.M., Lane, M., Ravindiran, R.S. & Kirshen, P. 1999. Storage Reservoir Behaviour in the
United States. J. Water Resour. Plann. and Manag., ASCE, 125(5):245–254.
Hydrological impact of dam construction in
an arid area
D.Stephenson
University of Botswana
Z.Chengeta
Mantswe Natural Resources Consultants, Botswana
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
A gauging weir on the lower Thune was found to be well designed with gauging staff and
a recorder house. However, the bed of the river is alluvial which is likely to suspend
during floods and there is likely to be a backwater effect from the Motloutse confluence a
few kilometers downstream. By comparing with other records on a longer term basis, we
are able to check the reliability of the records and extrapolate them to the proposed dam
site. The record is however so short that it had to be extrapolated using rainfall-runoff
modeling methods.
The catchment was inspected from the point of view of estimating hydrological runoff
as well as topographical features which will be incorporated in the catchment models.
Vegetation and soil types were inspected from the point of view of erosion and
agriculture.
Wells on the banks of river were inspected and surveyors traveled down the riverbed
to investigate the use of ground water from the riverbed.
The cross sections of the river and flood plains were surveyed for use in the flood
study.
Rainfall data was obtained from the Department of Meteorological Services, i.e.
monthly and daily records from surrounding rain stations, of which Francistown is the
most comprehensive.
Flow records were obtained from the Thune river for the limited time available as well
as Motloutse, Lotsane and Shashe rivers. These are largely on a monthly basis but the
water resources study is done on a monthly basis for mass balance purposes.
Water resources of arid areas 582
4 RESERVOIR YIELD
The effect of the reservoir on the yield of the river and in particular downstream flows
was investigated. A generalized model of the catchment was used whereby the flow
records were extended using rainfall-runoff modeling, which is relatively simple using
the Rafler model. The model was calibrated by using the existing stream gauge data
available lower down on the Thune river.
The model was able to extend the monthly flows down the river to the reservoir to 50
years which enabled a better statistical comparison to be made and better risk analysis to
be done than the existing short period records.
The results were modeled to the proposed reservoir to obtain the net downstream flow
after allowing for draft and evaporation and other losses.
The results indicated a considerable change in the flow spectrum downstream. That is,
low flows, particularly when the reservoir is low, will be absent, whereas they were
previously used to replenish the water in the sandy bed of the river. Arrangements will
therefore have to be made to release water at a programmed rate in order to maintain the
regime downstream. Water is abstracted from the bed of the river downstream,
particularly for stock and domestic use.
The effect of soil erosion upstream in the Thune catchment was examined and the
variation of erosion rate with runoff and rainfall rate was investigated. It was found that
the catchment is particularly sensitive to the rate of rain and the highest erosion rate
occurred during the more intense storms. These may not necessarily be those which result
in the maximum storm runoff.
It was found that soil erosion will significantly reduce the yield of the reservoir even if
it is constructed to many times the average annual flow of the river. The yield of the
reservoir will reduce by more than 50% over 50 years after construction owing to
sedimentation.
The reservoir is also found to significantly increase evaporation losses such that the
downstream net flows are even less than indicated for just draft from the reservoir. In
fact, the dam would spill rarely unless simulated large releases are made at intervals.
More water is going to be lost by evaporation than will be consumed using a steady draft
operating system.
The construction of a dam across a river results in the capturing of most low flows and
the only uncontrolled release is the flood which over-tops the spillway on occasions when
the dam is full. The following changes to the hydrology of the river downstream therefore
occur.
Without releases from the dam, the river dries up for months a year and the time span
between discharges increases considerably. There may only be one or even no flows
down the river in any year.
The result is that communities based along the river banks become accustomed to a
dry river and can suffer when a flood, which is largely unexpected, does occur.
Communities are also dependent on low river bed flow even during dry seasons.
Water in the sandy bed of the river can be abstracted by people and animals digging
shallow wells, and wells in the banks of the river also draw water from the bed of the
Hydrological impact of dam construction in an arid area 583
river. The average release from the dam would have to be approximately 25 litres/second
to meet these requirements downstream to the Motloutse river beyond which the
Motloutse flow will contribute to the bed flow.
The reservoir behind the dam will reabsorb some of the floods, but it appears from the
calculations that the reduction in peak floods would only be of the order of 5% which is
small and this only happens on occasions. The reduced flood flow will result in slightly
lower water levels and this will enable an additional 160 hectares downstream to be less
flooded.
5 FLOOD ANALYSIS
Floods were calculated in the Thune river for various recurrence intervals. This is to
study the routing effect of the dam on the flows downstream and also to study the
changing high flow frequencies due to retention by the dam.
We used the standard methods of flood estimating, namely the rational method, the
unit hydrograph method with data we have obtained for other flood studies in the region,
and catchment models. An average was selected from the results and flood frequencies
ranging from 20 years to a 1000 years could readily be extrapolated and the hydrographs
resulting, routed through the proposed dam.
It appears there will be an attenuation of the floods even if the dam was relatively full
and therefore the downstream flows would mean the resulting floods downstream would
be routed down the river using our hydraulic water profile program. The resulting flood
plain with and without the dam were compared.
The dam has a routing effect on floods and downstream the flood inundation is less if
the dam is constructed to the amount of 160ha. The hydrological report compares the
flood spectra with and without the dam.
The entire flood spectrum downstream of the dam changes and there is not much
effect on the larger floods, but the smaller floods and low flows will not happen unless
positive releases are made from the dam. A proper operating rule is therefore required
and the report indicates the minimum flow required for instream requirements.
Water resources of arid areas 584
The construction of the dam increases evaporation losses considerably and the graph
shows the relative effect of abstraction from the dam and evaporation for the selected
dam size. The evaporation loss can be significantly greater than the release and once the
downstream instream requirements are also released, the yield of the dam will be very
small compared with the mean annual flow of the river. This means that most of the
water is lost by evaporation and this has a serious adverse effect of the dam.
6 SEDIMENTATION
The catchment is one of fairly high erodibility even compared with other Botswana
rivers. The soil erosion upstream of the dam which occurs, irrespective of whether the
dam is built or not, is approximately 700000 cubic metres a year. A reservoir capacity of
50 million cubic litres would therefore be 2/3rds full of sediment within 50 years,
reducing the reliability of the drawoff considerably.
Whereas the bulk of the sediment transported into the reservoir is deposited, flows
over the spillway during floods will be clean of sediment and have a greater potential for
erosion of the river bed and banks downstream of the dam wall.
The sediment regime of the river is considerably affected by the presence of the dam.
It is unlikely that any operating rule of the dam will change this regime much and the
following is the main impact. The reservoir behind the dam will catch practically all the
sediment as indicated in our report. This will cause the upper reaches of the dam initially
to become blocked with sediments including silt and sand and push the flood plain further
back upstream. Within 30 years the dam will be nearly half full of sediment and its
reliable yield will be much less after 20 to 50 years.
Downstream of the dam there will be less sediment released by the dam and river.
There will therefore gradually be erosion of the existing sand bed and banks of the river,
more than the present average rate, and this sediment will be transported to the Limpopo
River via South Africa, Zimbabwe and Mozambique to the Indian Ocean. This is because
the river is impounded and it loses the carried sediment in the reservoir and has to pick up
new sediment downstream to reach its equilibrium of concentration. In addition, the
energy dissipation effect of discharge over the spillway will cause erosion downstream
even if there is construction of a stilling basin. This is because cleaner water has a higher
energy and there is a considerable energy released over the dam spillway which, without
the dam, would have been dissipated in the form of channel friction along the length of
the reservoir.
Sediment yields over all the catchment indicates a potential annual average erosion
rate of 2.5×106m3 per annum. However, the erosion modeling which allowed for
reasonable ground cover and limitations on the carrying capacity of the runoff indicated
an average sediment yield of only 6,700000m3 per year. This compares well with the
yield at times for the Shashe dam and others in the area on a per km2 basis. It should be
noted that even with the reservoir capacity of 50 million cubic metres initially, the
reservoir volume would be reduced to 20 million cubic metres after 50 years.
The corresponding yield of the reservoir would therefore decrease from about 5
million cubic metres per annum with an initial capacity of 50 million cubic metres, or 4
million cubic metres per annum with an initial capacity of 40 million cubic metres, down
Hydrological impact of dam construction in an arid area 585
to less than half after 50 years. When the releases required downstream are taken into
consideration, the yield is even smaller. The yield is based on 98% reliability and would
be higher if a higher risk of failure is acceptable.
It is anticipated that a concrete causeway would be constructed to enable trucks and
other vehicles to cross the river for construction purposes. Such causeways can block
river bed flow unless adequate discharge pipes are allowed through the causeway even
below the bed surface. The causeways should be built not much more than half a metre
above the sand level to enable floods to pass sand downstream.
The river is likely to be temporarily diverted to one side of the channel while
construction of the embankment occurs on the other side and a bypass tunnel may be
constructed under the first half of the embankment so that the second half of the
embankment can be constructed subsequently. During these diversion works, the floods
will be concentrated and cause additional erosion downstream of both the bed and the
bank of the river. Erosion protection should therefore be specified as part of the
temporary diversion works.
Any outlet gates from the proposed dam would be insufficient to discharge the full
sediment load through the dam wall. Only fine sediment in the vicinity of the gate would
scour however long the gate was opened for. Releases downstream should therefore be
primarily aimed at maintaining the ecosystem along the bank. The maximum rate of
release will depend on flood rate which may be tolerable and less losses would occur in
the form of evaporation if the releases were made for example, once a week over a few
hours, instead of continuously over 24 hours a day and 365 days a year.
The downstream instream requirements should be released in gulps to minimize
evaporation loss and to ensure a flow rate never exceeding 2m3/second to avoid
endangering livestock or people. This would mean that releases should only occur for 1%
of the time, i.e. say 1 hour in 100, or approximately 2 hours a week. The rate of the
downstream release will depend on the sizes of the outlet valves and pipework.
The presence of the reservoir behind the dam wall would increase evaporation losses
from the Thune river. The area of the reservoir is however relatively small and the effect
on the climate both locally and nationally would not be noticeable.
The effects on downstream rivers would be noticeable, particularly for the low flow.
The Motloutse river into which the Thune flows and further down, the Limpopo river,
would receive less base flow from the Thune catchment. In fact, the entire base flow
would be eliminated. On the other hand, the peak floods are largely unattenuated and
therefore the dam would have negligible effect in managing floods for downstream
rivers. There is also a potential salinization problem. Salinization is a major problem in
irrigated land globally and has severe, long-term and often permanent impacts on land,
agriculture and livelihoods where rehabilitation is not undertaken.
Water resources of arid areas 586
The dam will reach the end of its economic life within a century. This will largely be due
to sedimentation. There may also be changes in water requirements and value and the
possibility of abandoning the dam in the future should be considered.
The sediment deposited in the dam would be fine silt as opposed to the coarse sandy
bed material which is present in the river at the moment. It could therefore theoretically
be used for
To optimize the releases or drawoffs from the reservoir, the reservoir level will have to be
held as high as possible. There are however a few reasons why the water level should be
fluctuated over short periods of time.
– The releases for downstream instream requirements should preferably be made over
short periods to minimize evaporation losses in the downstream channel and enable
the release to travel as far as possible down the Thune river
– Water based insect vectors can be controlled by varying the level in a tropical reservoir.
Bilharzia snails can be left on the dry banks and killed if the water level is fluctuated,
i.e. dropped, for a few weeks over the growing seasons of the snails. Mosquitoes will
also reduce breeding if the shallow water on the banks is occasionally dropped and
banks dried.
– Benthic deposits and accumulations in the reservoir can be minimized by discharging
the heavy sediment-laden waters from the bottom of the reservoir occasionally. Heavy
Water resources of arid areas 588
metals and even nitrates can be discharged in this way. It will also provide organic
matter to the downstream channel.
11 CONCLUSIONS
A dam constructed in an arid area has many more pitfalls than one constructed in a
temperate area. The river hydrology and regional water balance are affected. There is
large evaporation loss and downstream flows are reduced during drought. There are also
probable cumulative effects that may be national and regional.
The geochemistry of fresh water supplies in
Botswana
L.Molwalefhe
Department of Geology, University of Botswana, Gaborone, Botswana
S.Vriend
Department of Earth Sciences—Geochemistry, Faculty of Geosciences,
Utrecht University, The Netherlands
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
Botswana has an arid to semi arid climate, characterized by erratic rainfall and frequent
severe droughts which gives rise to relatively limited fresh water resources. Water supply
in the major cities is largely dependent on surface water. Several major dams have been
built for this purpose in the eastern part of the country where 80% of the population lives.
In the rural areas 90% of the water supply is based on groundwater. In the northern
regions several perennial rivers provide an ample source of fresh water supply. In a
country like Botswana with limited water resources it is importance to have a good
understanding of the factors that determine the water composition. The Department of
Water Affairs routinely monitors the major chemical elements in the fresh water supplies.
This database is used in this paper to study the hydrochemistry of fresh waters in
Botswana. Through statistical and hydrochemical approaches, a survey is made of the
Water resources of arid areas 590
water types that are present in the fresh waters and the processes that control the
chemical.
2 CLIMATE
The climate is continental semi-arid. Due to the absence of major topographical barriers
there is a low spatial variability in the climate. The average of the erratic yearly rainfall is
450mm and varies from 230mm in the southwest to 650 in the northeast (Figure 1). The
dry winter lasts from May to September and the rainy summer from September to March.
Temperatures are high (38°C) in summer (Bhallotra, 1984). Daily variations in
temperature of about 20°C in winter are common. Evapotranspiration is generally high
due to the high temperatures and the low humidity. Evaporation from open water is in the
order of 2m per year.
We only have waters in the database that have some economic value i.e. are potentially
useful for drinking water or cattle drinking purposes.
3.3 Histograms
Histograms of the logtransform of a selection of constituents are given in Figures 3 and 4.
4 SCATTERGRAMS
In Figure 5, scattergrams of the different elements are plotted against chlorine. Chloride
is assumed to be entirely derived from rainwater and is a conservative element (Selaolo,
1998).
Water resources of arid areas 592
concentrations do increase with the Cl concentration. This is due to the increase in ionic
strength which allows that more Ca can remain in solution in a saturated environment.
SO4 is often enriched in relation to the seawater dilution line. Only a few waters are
supersaturated in gypsum. Also a number of waters are depleted in sulphate, which is
probably caused by sulphate reduction in the aquifers.
5 PIPER DIAGRAM
Relative amounts of meq% of major cations and anions are expressed in Piper diagrams
as shown in Figure 6. In the cation triangle the waters with the higher salinity converge at
the Na+K apices. The more dilute waters contain carbonates as major anion species,
while at higher concentrations the major anion is Cl with sometimes considerable
quantities of SO4.
is strong argument for the validity of the used cluster model. For this study, following the
above, a four cluster model has been adopted. The cluster centers are given in Table 1. A
representation of the correspondence between the FCM and NLM result is given in
Figure 7. In this interpretation the fuzziness has been removed from the classification and
the memberships have been hardened to either 0 or 1. This resulted in the identification of
some 14% as intermediate samples. A characteristic label is assigned to each cluster. This
label is used in the evaluation and discussion below.
7 CONCLUSIONS
REFERENCES
1 INTRODUCTION
The area under study is located in the semi-arid zone of north-east Nigeria, which is
characterised by low rainfall and reduced river flow, Figure 1. In the last three decades
rainfall has decrease by about 30% (Hess et al. 1995) and annual discharge by the major
headwaters of Rivers Hadejia and Jama’ are has decreased by almost 60%. The reduction
in the discharge from these rivers to the
Groundwater modelling with limited data 601
data needed to carryout ‘conventional’ modelling. This limitation of data has constrained
the development of a ‘full-scale’ model where calibration, verification and prediction are
possible. It was against this background that an exploratory groundwater model was
developed using MODFLOW to get an insight into the possible interactions highlighted
in the conceptual model of Hassan (2002).
The idealized conceptual model was used to design and simulate various scenarios using
the MODFLOW model (McDonald and Harbaugh, 1988). A combined pre and post-
processor model independent graphical interface called Groundwater Vistas was used for
data input and interactive modelling with MODFLOW.
The model area was discretized three-dimensionally. In the x-direction the size of the
grid blocks were 500m, and in the y-direction it was variable with the smallest being 35m
and the largest 150m. Each grid consists of 3 columns and 50 rows. In the vertical
direction the model consists of one layer with a thickness of 16m. Row 25 contains the
river.
Input data to model consists of the three major external stresses river stage time series
with a varying river coefficient, vertical recharge and ‘leakage’. Figures 3, 4 and 5 show
the starting conditions and time series of the input data respectively. The input for one
year consists of 36 stress periods each with a length of 10 days and a single time step.
This was repeated for 2.4 years (86 stress periods) with stress period one starting from
30th October. The choice of the number and length of stress periods and time steps was
dictated by the rapid change in the river stage. Sensitivity analysis was carried out on the
wide range of aquifer parameters to arrived at acceptable values and were also assigned
as follows:
Unconfined region the horizontal hydraulic conductivity, Kh=15m/d reducing to
0.1m/d towards the southern boundary.
Confined region: Kh=15m/d
Confined storage coefficient, Sc=0.001
Unconfined storage coefficient, Su=0.05
The clay cover was modelled as a ‘leakage’ factor that allows water to seep
continuously into the aquifer. A vertical leakage of 1.5×10−5m/d, equivalent to vertical
hydraulic conductivity of clay, is used.
independent estimates by Carter et al (1994) and Edmonds et al (2002). The recharge was
applied to rows 29 to 33 and in stress periods 32 to 35 inclusive. These stress periods
corresponds to 10th–19th September to 10th–19th October respectively when recharge is
believed to occur.
The outputs from the model consisting of groundwater heads for each of the 86 stress
periods were used for calculation of the various flow processes. The river to aquifer flow
is calculated
using the equation Qriv=Criv x HDIFF (Rushton and Tomlinson, 1979). Where Criv is
the river coefficient and HDIFF is the relative head difference between water level in the
river and the groundwater head in the aquifer.
4 RESULTS
The basic and exploratory single layer model has demonstrated an ability to serve as an
interpretation of observed field data. It has also reflected the physical processes presented
in its conceptual model. The process of generating the conceptual model and fitting field
Groundwater modelling with limited data 609
data into the numerical model has provided the ‘feel’ or insight into the magnitudes of the
system’s processes and interactions.
Despite uncertainty in the estimates of some parameters such as river coefficients and
aquifer parameters, the similarity between the model results and observed data is
encouraging. For example the model has also established that in confined areas less water
enters the aquifer and there is immediate response to changes in the application of stress
when compared to the unconfined areas. The results from the model are plausible and
that it has represented to some extent the understanding incorporated in its conception.
REFERENCES
Alkali, A.G. 1995. River-Aquifer Interaction in the middle Yobe River Basin, Northeast Nigeria.
Unpublished PhD thesis. Silsoe College.
Carter, R.C. 1998. Prospects for Sustainable Water Management Policy in Sub-Saharan Africa,
with Special Reference to the Northeast Arid Zone of Nigeria. In Water Resources
Management, A Comparative Perspective. Ed. Dhirendra K.Vajpeyi.
Carter, R.C., and Alkali, A.G. 1996. Shallow Groundwater in the Northeast Arid Zone of Nigeria.
Quarterly Journal of Engineering Geology, Vol. 29, 341–355.
Diyam Consultants. 1987. Kano State Shallow Aquifer Study. Final reports Vol. 1. P.O. Box 701,
Kano, Nigeria.
Edmunds, W.M., Fellman, E., Goni, I.B., and Prudhomme, C. 2002. Spatial and temporal
distribution of groundwater recharge in northern Nigeria. Hydrogeology Journal (2000) Vol. 10.
pp 205–215.
Hassan, M. 2002. Exploratory groundwater modelling in data-scarce environments: The shallow
alluvial aquifer of River Yobe Basin, North East Nigeria. Unpublished PhD Thesis Cranfield
University UK.
Water resources of arid areas 610
Hess, T.M. 1997. BALANCE—A soil water balance model. Unpublished, Cranfield University,
UK.
Hess, T.M., Stephens, W. and Maryah, U.M. 1995. Rainfall Trends in the North East Arid Zone of
Nigeria, 1961–1990. Agricultural and Forest Meteorology, 74, 87–97.
IWACO B.V. International Water Supply Consultants. 1985. Study of the Water Resources in the
Komadougou Yobe Basin. Report No. 5. Groundwater Resources. Rotterdam, Netherlands.
Marinof-Petkoff, M.N. 1994. A Geomorphological study of the Yobe River Floodplain: Implication
for Groundwater Recharge. Unpublished MSc. Thesis. Cranfield University.
McDonald, M.G., and Harbaugh, A.W. 1988. A modular three-dimensional finite-difference
groundwater flow model. Techniques of water resources investigations 06-A1, USGS, 576p.
NEAZDP. 1990. Northeast Arid Zone Irrigation Project: Groundwater Resources Report. PMU,
Garin Alkali, P.M.B 18. Gashua. Yobe State, N.E. Nigeria.
Rushton, K.R. and Tomlinson, L.M. 1979. Possible mechanisms for leakage between aquifers and
rivers. J. Hydrology, 40:49–65.
Theme G:
Water resources management
Apple and grape vinegar application as c-
source in water denitrification
Ş.Aslan
Cumhuriyet University Department of Environmental Engineering, Sivas,
Turkey
A.Türkman
Dokuz Eylül University, Department of Environmental Engineering,
Izmir, Turkey
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
Nitrate contamination in drinking waters is a growing problem for all over the world
because of its harmful effects on human health. Nitrates cause methemoglobinemia in
infants and pregnant women, which is also known as “blue baby syndrome”. Nitrates also
cause the formation of n-nitroso compounds with amines and amides in the human body.
These n-nitroso compounds are known to be carcinogens. To protect human health,
public health agencies set nitrate limit in drinking water standards.
There are various nitrate removal methods such as ion exchange, reverse osmosis,
electrodialysis, distillation, chemical denitrification, mebrane biorector and
biodenitrification (AWWA, 1989; Delanghe, 1994; Barreiros, 1998; Flora, 1994).
Biological process has been shown to be practical, efficient and cost-effective.
The majority of microbial denitrification treatment relies on heterotrophic bacteria
which require an organic carbon source to reduce nitrogen oxides to nitrogen gas; but
drinking water has low carbon content. Therefore an external carbon has to be supplied
for microbial growth. If the organic substances in the nitrate contaminated water are
below the stoichometric requirement for denitrification, they must be added in the form
of acetic acid or ethanol resulting in increase in treatment costs. Several types of organic
compounds have been used like methanol (Wasik et al., 2001; Gomez et al., 2000; Lee et
Water resources of arid areas 614
al., 2001; Hoek and Klapwizk, 1987), and acetic acid (Dahab and Kalagari, 1996; Bandpi
et al., 1999). Though methanol assures the highest denitrification rate (Mansell and
Schroder, 1999), it can constitute certain risk if the treated water is used for drinking
purpose (Adriaan, 1992). The main disadvantage of acetic acid as compared to other
carbon sources was its high consumption ratio and high cost (Bandpi et al., 1999).
Balszczyk et al., (1981) pointed out that using acetic acid could have a significant effect
on the production of nitrite in the reactor. Consequently, the use of ethanol as alternative
is becoming more popular (Green et al., 1994; Gomez et al., 2000; Bandpi et al., 1999;
Delanghe et al., 1994; Fonseca et al., 2000; Dahab and Sirigina, 1994; Richard, 1989, and
Gomez et al., 2000).
The possibility of using alternate substances such as volatile fatty acid (Yatong, 1996),
shredded newspaper (Volokita et al., 1996a), wheat straw (Soares and Abeliovich, 1998,
Aslan and Türkman, 2003), unprocessed short fibre cotton (Volokita et al., 1996b),
atrazine (Stucki et al.,
Table 1. Composition of medium.
1 L Medium
KNO3 361mg (50mg/l as N)
KH2PO4 150mg
NaHCO3 32.5mg
FeSO4*7H2O 0.816mg
NaMoO4 0.2365mg
MnSO4*7H20 0.1565g
CoCl2*6H2O 0.526g
H2O (pure) →1L
2000), natural gas methane (Rajapakse and Scutt, 1999), elemental sulfur (Eisentraeger et
al., 2001; Soares, 2002), sugar or glucose syrup (Nurizzo and Mezzanatte, 1992) and
sugar cane (INCO-DC, 2000) have also been studied in the biological denitrification
processes.
The aim of this study is to determine the applicability of the biological denitrification
of drinking water using locally available chemicals as carbon source, apple and grape
vinegar. Denitrification efficiencies have been determined by applying different carbon
sources in up-flow packed column. Second step of the treatment is sand filtration, which
is not used commonly in Turkey, even though it is very suitable because of its simplicity
of operation, availability of sand and the advantage of removing microorganisms that is
formed in denitrification phase.
hinder the interference of colour and turbidity. COD analyses were performed with the
clear samples according to APHA (1984). NO2-N (14776) and NO3-N (14773) were
determined using analytical kits and a photometer Merck SQ 300. Dissolved oxygen
(DO) measurements were carried out using WTW oxygen meter. Turbidity was measured
by the turbidity meter in the unit of JTU (Heck Chemical Company).
3 START-UP
Inoculum was taken from a denitrification reactor used in the laboratory. For the first
three days, the reactor was fed daily in a fill and draw mode, with recirculation. From the
third day the reactor was fed continuously. It was considered that the reactor had reached
state conditions when the NO3-N removal efficiency reached to the higher than 90%.
Water resources of arid areas 616
4 EXPERIMENTAL SET UP
Experiment was carried out in an up-flow fixed film reactor operated in continuous mode.
The packed column was filled with 10mm pieces of plastic coils materials, which
supported bacterial growth. The biodenitrification reactor consisted of a cylindrical glass
reactor, 5.4cm in diameter and 50cm in height, completely submerged an effective
volume of 0.841. The sand filter column, which had 8cm diameter and 30cm height, was
filled with filter sand of an effective diameter of 0.4mm and uniformity coefficient of
0.89 (Figure 1).
Apple and grape vinegar application as c-source in water denitrification 617
6 CONCLUSIONS
ACKNOWLEDGEMENT
The authors are thankful to INCO-DC Project (Contract ERBIC 18 CT 970167) for
financial support.
Apple and grape vinegar application as c-source in water denitrification 621
REFERENCES
Adriaan, H.S. 1992. Metabolic Pathways in Paracoccus Denitrificans and Closely Related Bacteria
in Relation to the Phylogeny of Procaryotes. Antonie van Leeuwenhoek 61:1–33.
APHA Standart Methods for the Examination of Water, Sewage and Industrial Wastes. 1984. 16th
edn, American Public Health Association/American Water Works Association/Water
Environment Federation, Washington DC, USA.
Aslan, Ş. & Türkman, A. 2003. Biological Denitrification of Drinking Water by Using Various
Natural Organic Solid Substrates. Water Science and Technology. 48. 11–12:489–495.
AWWA. 1989. American Water Works Association Water Desalting and Reuse Committee,
Membrane Desalting Technologies. Journal, AWWA, 81, 30.
Barreiros, A.M., Rodrigues, C.M., Crespo, J.P.S.G. & Reis, M.A.M. 1998. Membrane Bioreactor
For Drinking Water Denitrification. Bioprocess Engineering 18.
Bandpi, M.A., Elliott D.J. & Memeny-Mazdek A. 1999. Denitrification of groundwater using acetic
acid as a carbon source. Water Science and Technology. 40(2):53–59.
Balszczyk, M, Przytocka, M, & Mycielski, R. 1981. Denitrification of High Concentrations of
Nitrites and Nitrates in Synthetic Medium with Different Sources of Organic Carbon. Acta
Microbiologica Polonica. 30:49–58.
Dahab, M.F. & Sirigina, S. 1994. Nitrate removal from water supplies using biodenitrification and
gac-sand filter system. Water Science and Technology. 30(9):133–139.
Dahab, M.F. & Kalagari, J. 1996. Nitrate removal from water using cyclically operated fixed film
biodenitrification reactors. Water Science and Technology. 34(1–2):331–338.
Delanghe, B., Nakaruma, F., Myoga, H., Magarat, Y. & Guibal, E. 1994. Drinking water
denitrification in a membrane bioreactor. Water Science and Technology. 30(6):157–160.
Eisentraeger, A., Klag, P., Vansbotter, B., Heymann, E. & Dott, W. 2001. Denitrification of
groundwater with metan as sole hydrogen donor. Water Research. 35(9):2261–2267.
Flora, J.R.V., Suidan, M.T., Islam, S., Biswas, P. & Sakakibara, Y. 1994. Numerical Modeling Of a
Biofilm-Electrode Reactor Used For Enhanced Denitrification. Water Science and Technology.
29:10–11.
Fonseca A.D., Crespo J.G., Almedia I.S. & Reis M.A. 2000. Drinking water denitrification using a
novel ionexchange membrane bioreactor. Environmental Science Technolgy. 34:1557–1562.
Gomez, M.A., Gonzales-Lopez, J. & Hantorie-Garcia, E. 2000. Influnce of carbon source on nitrate
removal of contaminated groundwater in a denitrifying submerged filter. Journal of Hazardous
Materials. B80:69–80.
Green, M., Schnizer, M., Tarre, S., Bogdan, B., Shelef, G. & Sorden, C.J. 1994. Groundwater
denitrification using an upflow sludge blanket reactor. Water Research. 28(3):631–637.
Hoek, J.P. & Klapwizk, A. 1987. Nitrate removal from groundwater. Water Research. 21(8):989–
997.
INCO-DC. Project 2000. Development of a simple technology in drinking water treatment for
nitrate and pesticide removal, 7. Meetings, Rabat, Morocco.
Lee, D.U., Lee, S., Choi, D. & Bae, J. 2001. Effects of external carbon source an empty bed contact
time on simultaneous heterotrophic and sulfur-utilizing autotrophic denitrification. Process
Biochemistry. 36: 1215–1224.
Mansell, B.O., & Schroeder, E.D. 1999. Biological Denitrification in a Continuous Flow
Membrane Reactor. Water Research. 33(8):1845–1850.
Nurizzo, C. & Mezzanatte, V. 1992. Groundwater biodenitrification on sand fixed film reactor
using sugars as organic carbon source. Water Science and Technology. 26:827–834.
Rajapakse, J.P. & Scutt, J.E. 1999. Denitrification with natural gas and various new growth media.
Water Research. 33(18):3723–3734.
Water resources of arid areas 622
UKTNP occupies vast, flat, sandy plains and dune fields (the ‘Dune Plains’), and steep
rock domes of (‘Ayers Rock’) and (The Olgas’) and lesser rock
outcrops. Stable, parallel, reticulate and irregular dunes rise an average of 10m above the
plain. The plains slopes gently SW to NE, descending about 2.5m every 1km to Lake
Amadeus, a large salt lake nearly 60km away, and a groundwater discharge zone for
UKTNP aquifers’ (and for other aquifers, north of Lake Amadeus).
Water resources of arid areas 624
2 GROUNDWATER
storage (Read, 1999; File Note). However, the position of the contact between Cz and
extremely weathered bedrock is uncertain.
Near the surface massive, karstic, calcrete with moderate water-bearing capacity
occurs. In other places, calcrete is impermeable and perched water tables may develop;
e.g. near the New Wellfield at Yulara, a layer of fresh water in the aquifer (modern
recharge) overlies more saline water. Calcrete dissolution may directly contribute to
higher TDS of groundwater; where cavities are silicified this phenomenon is mitigated.
Aquifer recharge has occurred three or four times in the past 37 years (of record-
keeping). After each recharge event there has been prolonged recessions of the water
table.
Transmissivities and SWL responses in the Cz or bedrock aquifer are similar in the
centre and west whereas, in the south and east, water chemistry, SWLs, and drilling
suggest a separate aquifer in the bedrock. A groundwater hydrograph against local
rainfall is shown in Figure 8. After recharge events in 1974, 1982, 1989 and 2000,
prolonged recessions of the water table occurred. At RN 11547, the SWL is some 3m
higher at present than it was 20 plus years ago.
TDS increases towards the SE, i.e. along the flow path of the aquifer (Wischusen,
1999) and in bedrock is variable (TDS >3,000mg/L in places). The groundwater is at or
near saturation with respect to bicarbonate (300mg/L HCO3) (ibid). The shallow aquifer
zone within the Cz has low chloride and low sulphate waters; the intermediate aquifer
zone (weathered bedrock?) has relatively low chloride and high sulphate and the deepest
aquifer zone (bedrock?) has high chloride and relatively high sulphate. Nitrate is
generally low in the main aquifer zone. A high bicarbonate to chloride ratio in the NW
part of the SAQ is likely to indicate a recharge zone (Read, 1977 and Jolly, 1979). The
north-central part is associated with floodouts emanating from the southern face of
that cross the main Ayers Rock Fault, so indirect recharge down the associated
fracture-zone may be possible.
Isotope chemistry suggests that the bulk of stored water entered the aquifer in the
geologically-recent past (last interglacial period with a wetter climatic regime) and
relatively insignificant recharge occurs under present hydrological conditions. Tritium
(3H) indicate component of modern water in the saturated zone (Jacobson, 2000),
including a bore close to (<50 years). Direct recharge of the SAQ, is indicated by
the relative enrichment of the Oxygen 18 isotope concentration in groundwater (Jacobson
et al., 1989b). This suggested that a high degree of evaporation had occurred during
recharge. Some samples were ‘ancient’ groundwaters, recharged more than 30,000 years
ago.
The potentiometric surface showed an 8m fall in SWLs from the SW end of the wellfield
to the NE end Coffey (1999). The DPA has a somewhat higher groundwater salinity than
the SAQ and is chloride-rich rather than bicarbonate-rich. When SWLs decline, due to
long periods of limited or no recharge, groundwater quality appears to deteriorate.
Isotope dating (14C, 2H and 18O) of groundwaters in the DPA (Jacobson et. al., 1989
and 2000) indicated modern groundwater ages between less than 500 years and 1,200
years with two dates greater than 5,000 years from mixing of young, local recharge
(probably less than 50 years old) and older groundwaters flowing along the palaeovalley.
Within the UKTNP water, sewerage and waste disposal services are provided by Parks
Australia to the communities of Mutitjulu, Maruku and Rangerville, the Cultural Centre
and Ranger Station, the ‘Base of the Climb’ at , and the Park Entry
Station.
Three bores situated in the SAQ west of Mutitjulu supply water to Mutitjulu/Maruku,
the Cultural Centre and Rangerville. The safe yield of the wellfield (approximately
6.5L/s) (Wischusen, 1999) will be exceeded by demand in approximately 20 years at 4%
annual demand growth (QANTEC, 2002).
The Yulara wellfield situated in the DPA consists of five PBs each producing between
5L/s and 25L/s, including one acting in a standby capacity.
The hardness, salinity (TDS about 1500mg/L) and high nitrate content of the
groundwater requires treatment for potable supply. Approximately 1.1ML/day is treated
to 200mg/L TDS by reverse osmosis (RO) desalination. Some 25% (i.e. about
0.3ML/day) of the total water put through the RO issues as reject brine and is discharged
to an inter-dunal area between the township and Yulara airport. 30% of wellfield
production (0.5ML/day; up to 2,100mg/L TDS), raw water is supplied directly to Yulara
township for irrigation, sanitation, and fire fighting use.
There has been an average increase in water consumption of approximately 10% per
annum. In 2002, production was 707ML. Historically, larger demands occurred from
September to January. This demand cycle does not directly correlate with the temperature
Water resources management 629
variation for Yulara, and is probably due to a combination of tourist numbers and higher
temperatures (Coffey, 1999). From 1998 to present there is a noticeable decrease in water
supply seasonality, with the wellfield operating at historically higher outputs throughout
the entire year, consistent with visitor numbers increasing over the summer period (AGT,
2003).
The present water demand at Mutitjulu, Maruku and Rangerville is nearly 55ML/yr
(QANTEC, 2002). However the actual water production in year, 2001 was more than
70ML. This infers that there is unaccounted for water (system leakage and wastage)
and/or irrigation is being applied (possibly to the sports oval?).
Taking the higher population growth rate, in year 2012, a demand of, say, 120ML/yr is
considered sustainable (AGT, 2003). It is less than the estimated aquifer throughflow of
200ML/yr and represents only 2% of the estimated aquifer storage of 6,000ML.
The BOD (oxygen removing capability) of raw effluent from Mutitjulu and Yulara
sewerage discharge is similar to that elsewhere (150–200mg/L) but after treatment this
has been reduced to 9mg/L, over 90% reduction at Mutitjulu (Allen & Assoc., 1999), and
likely similar level of treatment at Yulara. This concentration of BOD is quite low, and
the level of treatment consequently is considered effective.
The main problems associated with high nutrient levels are production of nuisance
algal blooms in surface waters. Apart from algal growth in the stabilization ponds (Allen
& Assoc., 1999), this is unlikely to be a problem at Mutitjulu, and as effluent is reused
for irrigation at Yulara, this is considered to be beneficial to vegetation being irrigated.
An estimated 5ML of leachate could be expected to be produced each year from a
12ha landfill area under conditions within the Yulara area (GHD, 1993). This was based
on an anticipated infiltration rate of rainfall of 42mm/yr, given local climatic conditions.
As the landfill serving Mutitjulu is unlined, leachate generated within the landfill will
percolate into the unsaturated zone. Natural processes of geo-purification will assist in
reducing concentrations of contaminants in percolating leachate. The high concentrations
of contaminants in leachate could impact groundwater if there is a hydraulic connection
with regional groundwater beneath the clay sediments within the palaeochannel. Flow
paths within the palaeochannel are likely to be complex, and the time taken for leachate
to travel from the landfill to groundwater would be long, unless channelling of leachate
occurred. Under conditions where no channelling took place, bacteria would remove
organic contaminants, ammonium would be adsorbed on aquifer minerals and transport
of this through the unsaturated zone would be much slower than that of leachate (i.e.
ammonium would be attenuated).
Groundwater quality protection guidelines recommend against locating fuel storages
within capture (protection) zones of water supply wells to avoid contamination, as clean-
up of soil once contaminated with these is extremely difficult and very expensive. The
only known potential risk is at the Old Wellfield (DPA), where diesel fuel for the PBs is
stored in elevated steel tanks which are bunded to reduce the possibility of bulk fuel
contamination of the DPA.
Water resources of arid areas 630
Reid et al. (1993), Baker & Jarman (1995) suggested that the Mulgara (a marsupial
rat) habitat and the UKTNP’s aquifers were, in some way, ecologically connected.
Jacobson (1996) reviewed the palaeodrainage and proposed that the depth to the water
table at the Mulgara habitat might be less important than its proximity to run-off from the
Sedimentaries through the adjacent extensive Mulga shrub-land. The review emphasised
the significance of this sheetwash zone near the northern boundary of the Park may have
in concentrating surface water and nutrients in the transitional zone between Mulga and
sand-plain, the core Mulgara habitat.
Several rare and endangered fauna are associated to some extent with these run-on
mulga groves, including the Mulgara Dasycercus cristicauda, Great Desert Skink Egernia
kintorei and Hairy-Footed Dunnart Sminthopsis hirtipes (Reid & Hobbs, 1996).
Earlier assessments on the distribution of Mulgara in the area suggested that
the distribution of this endangered mammal coincided with the bands of mulga (e.g. Reid
et al., 1993; Baker & Jarman, 1995). Subsequent work suggests Mulgara are linked more
closely to hummock-forming spinifex Triodia basedowii (P.Masters, pers. comm.; J.Reid,
pers. comm.; Reid & Hobbs, 1996). T.basedowii grows in or near run-on areas.
T.basedowii, however, is unlikely to have deep roots that tap into the groundwater, since
these plants show significant dieback during extended droughts (P.Masters, pers. comm.,
J.Reid, pers. comm.). This dieback would not occur if these plants were permanently
tapping groundwater.
The aquifers of UKTNP and Yulara have a huge storage of groundwater that is large in
proportion to throughflow, recharge and usage. Wellfield management may, inevitably,
require the resource to be mined due to the rare, but significant recharge events
experienced in this semi-arid environment. Because of the extremely variable nature of
arid zone rainfall amounts/intensities, prediction of rainfall-recharge return period is
problematic for aquifer replenishment. Hydrographs/rainfall relationships empirically
suggest that rainfall events would need to be in excess of 180mm in any given month to
induce significant aquifer recharge, although this would depend on a number of variables
including rainfall intensity/duration/area, antecedent soil moisture conditions, etc. SWL
monitoring data for bores has shown a number of recharge events during the past 30
years, but most of the water in the aquifer was emplaced in earlier events over a long
period of time.
7 RECOMMENDATIONS
A water conservation and re-use strategy should be enhanced and formalised for Yulara
(ARR) and ; to be re-viewed every ten years.
Water resources management 631
● core drilling, and petrophysical analysis thereof, to provide stratigraphic control of the
aquifers to enable the determination of aquifer geometry, storage capacity
(permeability and specific yield);
● review of pre-existing test pumping data to confirm hydraulic parameters;
● hydrochemical analysis and facies differentiation to determine flow paths and aquifer
recharge and discharge zones;
● samples taken for isotopic dating at each level of the aquifer to determine residence
times and the nature of recharge, i.e. diffuse (‘direct’) or flood-out (‘indirect’);
● suite of time-constant reduced SWL measurements recorded to construct a
potentiometric surface and ‘flow-net’ to calculate throughflow;
● ‘infill’ of drilling with surface geophysics (gravity, microgravity, transient electro-
magnetic to comprehensively map the DPA and the SAQ);
● aquifer numerical modelling to examine sustainability and optimise operational
pumping publish a hydrogeological map with accompanying notes;
● risk assessment of the vulnerability of the aquifer to extended periods between critical
rainfall/ recharge events;
● observe and record flood events to assist in determining recharge return periods and
volumes potentially available for recharge to aquifers; and,
● an aquifer vulnerability study (accounting for planning proposals) to mitigate against
contamination risk;
● investigate plant physiology by measuring plant water potential at various times of day
to assess the likely effects of changes in water levels and composition of the aquifer(s)
on the terrestrial flora and fauna isotopic results from possible water sources tapped by
plants should be compared with the isotopic composition of their shoots or twigs. This
would characterise and identify the water source from which they are extracting water;
● research groundwater dependent ecosystems at UKTNP (little is known about
subterranean ecosystems in the semi-arid zone).
REFERENCES
Allen, N. & Associates, 1999. Mutitjulu Community: Wastewater Management and Proposed
Sporting Oval. Report prepared for Parks Australia.
Anderson, V.J. & Hodgkinson, K.C. 1997. Grass-mediated capture of resource flows and the
maintenance of banded mulga in a semi-arid woodland. Aust. J. Bot. 45:331–342.
Australian Groundwater Technologies, 2003. Aquifer Review, 2002—
National Park. Report prepared for Parks Australia North, Environment Australia.
Ayers Rock Resort, 2000. Ayers Rock Resort Five Year Plan 21/12/00.
Baker, L. & Jarman, P. 1995. A conservation strategy for the Mulgara, Dasycercus cristicauda, at
National Park, N.T. Report #76 for the Australian Nature Conservancy, Canberra.
Coffey Geosciences, 1999. Yulara Borefield Operational Review. Report to the Northern Territory
Power and Water Authority.
DIPE (Department of Infrastructure, Planning and Environment, Northern Territory Government)
File 58.3P3, Alice Springs, NT.
Water resources of arid areas 632
Sweet, I.P. & Crick, I.H. 1992. and —A Geological History. Australian
Geological Survey Organisation.
Wischusen, J. 1999. Hydrogeology of the Southern Aquifer, : A Preliminary Assessment of
the Mutitjulu Groundwater Resource. A report to the National Park.
Young, D.N., Duncan, N., Camacho, A., Ferenczi, P.A. & Madigan, T.L.A. 2002. AYERS ROCK
SG 52–8. 1:250,000 Geological Series. Edition 2. Northern Territory Geological Survey.
Young, D.N., Duncan, N., Camacho, A., Ferenczi, P.A. & Madigan, T.L.A. 2002. AYERS ROCK
SG 52–8. 1:250,000 Geological Series. Explanatory Notes. Northern Territory Geological
Survey.
Integrated water resources management and
agriculture in southern Africa
M.McCartney & H.Sally
International Water Management Institute, Pretoria, South Africa
A.Senzanje
University of Zimbabwe, Mt Pleasant, Harare, Zimbabwe
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
For many years experts have debated the capacity of world’s agricultural systems to
produce enough food and fibre for an expanding population. It is estimated that at present
there are approximately 840 million undernourished people in the World of whom some
210 million live in sub-Saharan Africa (FAO, 2003). This situation led the 1996 World
Food summit to set a goal, reaffirmed at the 2000 Millenium Summit, of halving the
number of hungry people by 2015. Fulfilling these objectives will have significant
implications for water use in the region.
At present, irrigated agriculture, which covers approximately 275 million hectares
globally, produces 40% of all the world’s food crops. However, less than 5% (i.e., just 13
million hectares) of the total irrigated area is in Africa (Weligamage et al., 2002). To
meet the millennium goal the World Bank estimates that globally irrigated agriculture
Integrated water resources management and agriculture in southern Africa 635
will have to grow at a rate of about 4% per year. The largest percentage rise is required in
sub-Saharan Africa (FAO, 2003).
Although, depending on efficiency, there need not be a proportional rise, obviously
any increase in irrigated area will result in greater water consumption. However, water is
increasingly viewed as a scarce (i.e., limited and finite) resource and for many countries
in sub-Saharan Africa water scarcity, either physical or economic, is an increasing
constraint to economic growth. There is growing concern about the environmental
consequences and the implications for other sectors of using more water for agriculture
and a growing realisation that future land and water management must change
significantly to support human population growth in a sustainable manner.
Against this background, Integrated Water Resources Management (IWRM) is
promoted by its advocates as the modern and preferred way of managing land and water
resources (GWP, 2000). It is proffered as an approach that will simultaneously enable
development, protect the environment and ensure sustainability. This paper investigates
the implications of IWRM for agriculture in southern Africa.
Agricultural potential in Africa is huge. It is estimated that only 24% (i.e., 28.2 million
km2) of arable land is currently under cultivation. In southern Africa, approximately 80%
of the poor live in rural areas and are dependent on agriculture for their livelihoods.
Furthermore, agriculture is the most important sector of the economy for many countries,
contributing an average of 16% of total gross domestic product (Table 1) and accounting
for 67% of employment in the region. It is an important supplier of raw materials, food
and labour and is also important as the home market for much local industrial output
(Tiffen, 2003).
The population of sub-Saharan Africa is expected to increase by nearly 3% annually to
over one billion in 2025. The Forum for Agricultural Research in Africa (FARA)
estimates that, to keep up with this increase and achieve food security by 2015,
agricultural production must increase at an annual rate of 6% (FARA, 2003). However,
since the mid-1960s, per capita food production, in sub-Saharan Africa, has fallen by
about 20% (Pretty, 1999). Substantial new investments in agriculture are needed to meet
targets for poverty alleviation and food security. The United Nations Food and
Agricultural Organisation (FAO) estimates that about 75% of the growth in crop
production in sub-Saharan Africa required by 2030 will have to come from intensification
in the form of yield increases (62%) and higher cropping intensities (13%), with the
remaining 25% coming from arable land expansion (FAO, 2002).
Currently, by far the greatest part of cultivation in southern Africa is rainfed (Table 2).
Obviously, rainfed agriculture is highly dependent on the quantity and temporal
distribution of rainfall. As a general rule, an absolute minimum of 300–400mm of
precipitation is required per year to make rainfed arable farming possible. In many
southern African countries relatively low annual totals and high rainfall variability, not
only make decisions about crop choice and planting dates extremely difficult, but also
significantly reduce productivity. Irrigation, if properly designed and managed, helps
overcome many of the disadvantages inherent in rainfed agriculture. It overcomes the
Water resources of arid areas 636
need for shifting cultivation and reduces the pressure on fragile environments. The risk of
crop failure is minimized and farmers can hope for higher and more reliable agricultural
production and better levels of income (Sally, 2003).
Table 1. GDP and percentage accounted for by
agriculture in southern African countries in 1998 &
2001.
Gross Domestic Percentage of
Product (millions GDP from
US$) agriculture
Country 1998 2001 1998 2001
Angola 6,445 9,471 13 8
Botswana 4,932 5,196 3 2
Lesotho 890 797 18 16
Malawi 1,736 1,749 34 34
Mauritius 4,146 4,526 9 6
Mozambique 3,873 3,569 32 23
Namibia 3,399 3,100 11 11
South Africa 133,663 114,174 4 3
Swaziland 1,359 1,255 17 17
Tanzania 8,383 9,341 45 45
Zambia 3,238 3,639 21 22
Zimbabwe 5,732 9,057 22 18
(Source: World Bank Development Indicators,
http://devdata.worldbank.org/data-query).
The water resources of a country are usually assessed in terms of the proportion of
rainfall that enters streams and recharges groundwater and so can potentially be
abstracted for human use. In most countries in southern Africa, because a high proportion
of rainfall is evaporated, the annual renewable water resource represents only a relatively
small fraction of the total rainfall. Nevertheless, throughout the region, current human
water demand is only a small proportion of the total resource (Table 3). However, the
resource situation is a lot more critical than these figures
Table 2. Area of cultivated and irrigated land in
countries of southern Africa.
Total Total irrigated
cultivated area
area
Country Total area Total (km2) % (km2) %
(km2) cultivable cultivated
Integrated water resources management and agriculture in southern Africa 637
area (km2)
Angola 1,246,700 – 29,000 – 750 2.6
Botswana 581,730 62,000 32,420 52.2 13.8 0.1
Lesotho 30,350 – 2,093 – 27.2 1.3
Malawi 118,480 36,000 21,055 58.5 280 1.3
Mauritius 2,040 – 854 175 20.5
Mozambique 801,590 360,000 36,000 10 1,067 3.0
Namibia 824,900 250,000 2,052 1 61.4 3.0
South Africa 1,221,040 183,200 123,560 67.4 12,700 10.3
Swaziland 17,364 – 1,915 67.4 3.5
Tanzania 945,090 400,000 63,000 15.8 1,500 2.4
Zambia 752,610 163,500 10,298 6.3 464 4.5
Zimbabwe 390,760 – 27,500 – 1,166 4.2
(Source: FAO Aquastat database,
http://www.fao.org/waicent/faoinfo/agricult/agl/aglw/aquastat/main/index.stm).
indicate because these averages mask large spatial and temporal variance in freshwater
resources and patterns of requirement. With the exception of South Africa, a lack of
investment in infrastructure for water storage means that a large proportion of the total
Water resources of arid areas 638
IWRM calls for integrated planning so that water, land and other resources are utilised in
a sustainable manner. For the agricultural sector IWRM seeks to increase water
productivity within the constraints imposed by the economic, social and ecological
context of a particular region or country. A major shift in focus under IWRM is the
concept of demand management (i.e., managing water demand rather than simply looking
for ways to increase supply). This will have to be brought about in a range of ways
including:
● Improvement in crop varieties (i.e., crops that yield more mass per unit of water
consumed, perhaps as a result of genetic modification)
● Crop substitution (i.e., switching from more to less water consuming crops)
● Improved land husbandry (i.e., soil management and pest and weed control)
● Improved within-field water management (i.e., better timing and precision in the
application of water).
Properly implemented IWRM could benefit the agricultural sector in southern Africa
through more effective and equitable water use, the development of more sustainable
practices and reduced disputes over water supplies. Specifically, if implemented
appropriately it would enable:
● The implications of water use by other sectors on all forms of agriculture to be
considered in the management process and vice-versa
● A more rational decision-making process in which costs and benefits (not just
monetary) of different options for water-use are considered in a transparent manner
● More adaptable and effective utilisation of limited water resources that facilitates the
changes required to mitigate the potential negative impacts of climate change
● More effective use of water within the sector and hence increased economic returns.
All these aspects are likely to be increasingly important in the future as the “price” for
water rises and environmental controls (and financial penalties for non-compliance)
become increasingly strict. Overall the agriculture sector will benefit through more
rational use of water and clear understanding and assessment of trade-offs associated with
different ways of using water both within and outside the sector. In addition farmers will,
like everyone else, benefit from more equitable utilization of water and an improved
environment.
Although there is a growing consensus on the need to manage land and water resources in
an integrated way, there is no universally accepted method for applying the principles and
there are a large number of impediments to practical application. Barriers to successful
implementation of IWRM within the agricultural sector arise for a variety of reasons.
Water resources of arid areas 640
These are not limited to technical issues, but also occur because of complex socio-
economic pressures and limitations in human, financial and institutional capacity.
Principal constraints include:
● Perceived “low value” and “wastefulness” of agricultural use of water. This
perception obviously ignores the importance of agriculture for the livelihoods of most
people in rural Africa and the need to make water available to enable sustainable
intensified agriculture.
● Incompleteness in water management policy and legal and regulatory frameworks.
This is particularly the case in some southern African countries where water policies
are non-existent or at best rudimentary. In other cases they may be too fragmented to
effectively address the aspirations and priorities of different categories of users at
local, national and regional levels.
● Demographic pressures. Population growth linked to poverty is a significant problem
throughout southern Africa, which in turn exerts a lot of pressure on arable land and
drives inappropriate and non-sustainable agricultural practices and associated water
utilization. Communities, even when they understand the long-term consequences of
their actions, often feel that they have no alternative.
● Lack of understanding of IWRM principles and practices. In many instances, only a
few people in the hierarchy of water management know and understand IWRM and
often there is insufficient technical support to operationalise IWRM within the
agricultural sector.
● Lack of reliable data, information and knowledge. Quite often, the data required for
detailed analysis of water use trends (e.g. temporal and spatial variations in quantities
of water diverted and return flows), cause and effect linkages between land-use
patterns and hydrological regime, and the impact of changes on downstream users and
ecosystems is not available.
● Inadequate understanding of the inter-relationships between biophysical and socio-
economic aspects of a system. Successful IWRM requires the integration of
environmental, social and economic factors, but in any specific situation the
relationships between biophysical and socioeconomic systems are not well
understood. Consequently, the social implications of management decisions are often
impossible to predict.
● Lack of incentives for change. In many places water is provided to the agricultural
sector at subsidized rates, partly because of the perceived need by many governments
for nations to be self-sufficient in food production. However, the result is that there is
little economic incentive for farmers to change long established agricultural practices
that fail to improve the productivity of water.
● Entrenched agricultural practices. Very often farmers, like other groups, are unwilling
to change practices, if they believe that others will simply continue doing what they
have always done.
Clearly there is a wide diversity of constraints to implementing sustainable water
management practices. Successful IWRM requires consideration of these diverse issues
across wide range of scales.
Integrated water resources management and agriculture in southern Africa 641
In order for IWRM to succeed in the agricultural sector, it is imperative that a number of
requirements, which straddle the socio-economic, technical, and institutional domains,
are in place. Some of the key requirements include:
● Participatory management: Water must be managed at the lowest possible level with
the effective and meaningful involvement of all users of agricultural water.
● Balancing supply and demand management options: While water demand management
can help increase the economic efficiency of water use, improve water quality, and
promote sustainable water management practices, the parallel pathway of improving
water availability through the provision of improved storage and conveyance facilities
should not be neglected.
● Capacity enhancement in IWRM: To date, most IWRM training has been directed at
civil service employees, academics, researchers and water management specialists,
people who are not directly involved in the actual use of water in agriculture. In future
IWRM training must be made accessible to agricultural water users. It should also be
included in the appropriate educational curricula of universities and agricultural
extension services.
● Incentives for water saving: To encourage more efficient utilization of water resources
in the agricultural sector, incentives must be offered to those that save water and/or
maximise water productivity.
● Improved evaluation of externalities’. Future water allocation requires much greater
consideration of both immediate and long-term environmental and health impacts. At
present these issues are not consistently assessed in water planning.
● Application of appropriate decision-support tools: IWRM requires consideration of a
large number of complex and inter-related issues. Contemporary decision support
tools can help structure decision processes, promote understanding of system
dynamics, support analysis of possible choices and facilitate the communication of
information between people of different technical understanding.
7 CONCLUDING REMARKS
Improving land and water productivity, and increasing poor people’s access to water for
domestic and productive purposes are critical elements for the development of southern
Africa. Ensuring that such developments occur in a balanced and harmonious way
requires an integrated approach to land and water resources management. IWRM offers
an opportunity for holistic water management. It allows decision-making from a multi-
disciplinary perspective involving all uses and users of water taking into account the
interactions between them and the impacts of water use by a particular sub-sector, or at a
particular location, on other sub-sectors or locations. However, it is apparent that there is
no automatic or wholesale adoption and practice of IWRM in agriculture in southern
Africa. Several factors that militate against it must be overcome and a number of
requirements have to be in place to allow effective implementation of IWRM in
agriculture.
Water resources of arid areas 642
REFERENCES
Appleton, B. 2003. Climate changes the water rules. The Netherlands: Dialogue on Water and
Climate.
Forum for Agricultural Research in Africa (FARA) 2003. Building the future for Africa’s children:
building sustainable livelihoods through integrated agricultural research for development. Sub-
Saharan Africa, Challenge Program Proposal.
Food and Agriculture Organisation (FAO) 2003. Unlocking the water potential of agriculture
Rome: FAO.
Food and Agriculture Organisation (FAO) 2002. World Agriculture: towards 2015/2030 Rome:
FAO.
Global Water Partnership (GWP) 2000. Towards water security: a framework for action London:
GWP.
Houghton-Carr, H., Fry, M., McCartney, M. & Folwell, S. 2002. Drought and drought management
in southern Africa. Proceedings of the BHS Eighth National Hydrology Symposium University
of Birmingham 8–11 September 2002. London: BHS.
Hulme, M. 1996. Climate change and Southern Africa: an exploration of some potential impacts
and implications in the SADC region. Climate Research Unit, Norwich: University of East
Anglia.
Pretty, J. 1999. Can sustainable agriculture feed Africa? New evidence on progress, process and
impacts. Environment, Development and Sustainability 1 253–274.
Sally, H. 2003. Advances in integrated water resources management research in agriculture. In
McCornick, P.G., Kamara, A.B. & Tadasse, G. (eds.) Integrated Water & Land Management
Research and Capacity Building Priorities for Ethiopia. Proceedings of a
MWR/EARO/IWMI/ILRI international workshop, Addis Ababa, 2–4 December 2002. ILRI:
Addis Ababa.
Tiffen, M. 2003. Transition in Sub-Saharan Africa: Agriculture, Urbanization and Income Growth.
World Development 31(8) 1343–1366.
Weligamage, P., Barker, R., Hussain, I., Amarasinghe, U. & Samad, M. 2002. World Irrigation and
water statistics 2002 Colombo: International Water Management Institute.
Wood, S., Sebastian, K. & Scherr, S.J. 2000. Pilot analysis of global ecosystems: agroecosystems.
Washington D.C.: World Resources Institute.
Challenges for managing water resources in
semi-arid areas: a case study from two rural
communities in Zimbabwe
F.T.Mugabe
Department of Land and Water Resources Management, Midlands State
University, Gweru, Zimbabwe
A.Senzanje
Department of Soil Science and Agricultural, University of Zimbabwe,
Harare, Zimbabwe
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 BACKGROUND
About 160 million people farm and live in arid and semi-arid areas of Africa. These areas
receive below 600mm of annual rainfall and together with increasing population and lack
of infrastructure this means that many people have inadequate access to water. Water
scarcity and irregularity in rainfall are increasing due to the effects of EL Nino and
possibly to the impacts of global warming. Water is therefore the most valuable resources
in the arid and semi-arid areas hence the need for a ‘water resource audit’ (KAWAD,
2001) if sustainable water resources management is to be achieved.
The variability and unreliability of rainfall makes the sustainable development of
water resources difficult (Griffiths, 1972) hence the need for rural communities to come
up with clear management strategies given that their relationship with their resources is
key to their survival and prospects (Soussan, 1998). Good management of and secure
rights to water resources are crucial to livelihoods and particularly to people’s capacity to
cope with variability (Soussan, 1998). Water also provides a means for the diversification
of livelihoods. It is also important for addressing poverty and rural development since it
is used for food production (Chaturvedi, 2000).
climate change is likely to increase the water stress, there is now a need to investigate
more thoroughly the links between water potential resource base and how it is managed
during the season and dry years and see if there are any opportunities for reducing water
scarcity.
This paper explores how two communities have managed their water resources amid
insufficient knowledge of the quantities of their water resources base, source of water,
ownership and shifting needs during the season. The paper further outlines the challenges
these communities are facing to better manage their water resources.
Romwe and Mutangi are both micro-catchments in Chivi (Figure 1) district that is in the
semi-arid parts of Zimbabwe. Mutangi lies about 55km southwest of Masvingo while
Romwe lies some 80km due south. The two catchments have contrasting resource
endowments (Sullivan et al., 2000) and climatic pattern, with Romwe receiving more
rainfall than Mutangi in most years. The annual average rainfall for Romwe and Mutangi
are 581 and 500mm respectively and the deviation from the long-term mean for Chivi
growth point is depicted in Figure 2. They are both characterized by water shortages
during dry seasons and dry years (Moriarty and Lovell, 1999). Water resources in
Romwe are principally groundwater while Mutangi is surface water (Sullivan et al.,
submitted) but there are 18 unlined shallow wells that tap the ephemeral soil/weathered
zone aquifer and dry soon up after the rainy season (Mugabe et al., 2003). The annual per
capita level of domestic water use per day is reported to be 30.7 and 14 litres for Romwe
and Mutangi respectively (Sullivan et al., submitted). Water is also used for washing
clothes, bathing and building.
There are 26 privately owned wells in Romwe and most of them fail in those extremely
dry years like the 1992/3 season (Moriarty and Lovell, 1999) forcing the farmers to rely
on the collector well that was constructed in 1989 by CEH (former IH) and BGS
(Waughray et al., 1889). The privately owned wells are dug by the owners and are
constructed at the homesteads or in the owners’ fields and are used for garden irrigation
and domestic purposes. Water from these wells is accessible to anyone for domestic
purposes only. Access to domestic water use from the collector well is not limited to
anyone, but productive water use is limited to the families that have vegetable beds in the
1ha garden.
The group that uses water from the collector well productively has a committee that is
chaired by a chairman and this committee decides on how to use the water. Each member
contributes to a fund every month and the fund is used to buy spares whenever the pump
breaks. With the assistance from the Institute of Hydrology (IH) some of the garden
members were trained in pump maintenance and repair and they therefore do not need
outside assistance in the case of a breakdown.
Unlike the privately owned wells, the collector well is fitted with a munro chart
recorder and water levels is monitored at any one time. Most of the privately owned wells
failed in 1992/3 and the collector well was the only source of water (Lovell et al., 1998).
The reaction of the garden members was to reduce the area under cultivation by half so
that even non-members could have access to domestic water—this demonstrated a shift in
the nature of use, patterns of access to and rights over the resources, and the intensity of
use decreased.
There are 18 privately owned wells in Mutangi and three boreholes of which only one is
functional. The functional one and the non-functional boreholes were constructed by
DDF while World Vision constructed the third one. Productive water use from the
shallow wells is done during the harvesting period up to when the wells dry up around
June and then most of the gardening takes place in the community garden that is irrigated
with water from the dam. Domestic water is from the shallow wells, river and borehole
during the wet period and is obtained from the borehole during the dry season (Mugabe et
al., 2003).
There are a number of problems associated with water use from this dam and these
include:
● The dam has a lot of silt and siltation continues during the rainy season
Challenges for managing water resources in semi-arid areas 647
Unlike the bigger water users, there are no active water allocation mechanisms in the
rural areas. The new water act (1998) has left the management of rural water resources to
the users. ZINWA should come up with a clear-policy on how water resources should be
managed in the rural areas—mostly interested in the distributing water to the bigger users
who pay in given communal area-municipalities, farmers, mines etc. Other studies
(Butterworth et al., 2001) have shown that regulatory approaches to water resources
allocation will require high levels of capacity in catchment management bodies (which
looks impossible for the Zimbabwean rural communities together with the current
legislative measures in the Zimbabwean water act, 1998)
The challenges for managing water resources can be grouped into four, viz:
insufficient knowledge of the quantity of water, ownership, pessimism about the quality
of the following season, lack of projection of their annual water requirements.
The comment by Mukherjee (KAWAD, 2001) ‘Underground water reserves are like
bank accounts which can be thoughtlessly depleted by the farmer by resorting to heavy
irrigation which can be likened to issuing a series of cheques without depositing
anything’ is correct for these two rural communities. Rural communities have insufficient
knowledge of both their groundwater and surface water resource base (Mugabe and
Hodnett, 2001) and they keep on using water till ‘they are told by the bank manager that
they no longer have any more money in their accounts’ i.e. when the wells fail to give
Water resources of arid areas 648
them water or when the levels of their surface water resources are so low. Rural
communities do not know what is ‘added to their bank accounts’ i.e. groundwater
recharge and run-in the dams and they do not know the amount of water available to them
for that given season. Apart from insufficient knowledge of their water requirements they
do not project their annual water requirements.
Ownership of water resources in communal areas depends on where the resource is
found in most cases. Wells dug by individuals in their fields belongs to them while
boreholes or wells dug by organizations such as DDF, CARE, the government etc. and
river water belongs to the community. Use for water in communal areas is guided by
ownership; hence the resource is communally owned in some cases while it is not in
other cases. Access to water from individually owned wells is mainly restricted to
domestic purposes and not productive water use (Moriarty and Lovell, 1998). In most
cases communities view water sources constructed by outside organizations as belonging
to these organizations and as such they are hesitant/unwillingly to repair them in cases of
breakdown.
There are weaknesses in managing communal area water resources because of the
prevailing notions that any shortages are temporary and limited to dry seasons. People are
pessimistic that the following rainy season will bring enough rains to ‘recharge’ the
dams, wells, rivers for use during the dry season, hence water resources are rarely
managed except during periods of water scarcity (e.g. when the borehole fails or when
people see dam water diminishing fast) when ‘crisis’ management is employed. This was
shown at Romwe during the 1992 dry year when garden owners reduced the garden area
by half in order to provide domestic water for everybody even the non-garden members.
This is further complicated by lack of knowledge of the quantity of the resources in the
water bearing body.
6 CONCLUSIONS
The points highlighted in this paper that are worth emphasising are:
● Water resources management is only limited to the dry season and dry years; there is
plenty of the resource during the wet season.
● Outsiders in both cases, IH/DR&SS in Romwe and CARE in Mutangi have facilitated
water resources management.
● Water resources management is tricky in the communal areas given that the quantity of
the resources is not known (especially groundwater resources). Need to come up with
an inventory of water resources in this rural areas—audit.
● Water point management has been different in the two case studies owing to the
differences in the nature of the water resource. Romwe is groundwater based (effect of
land use practices is not apparent), hence management is mostly restricted to pump
maintenance and water allocation in times of drought. In Mutangi, a catchment
approach has been taken since it is surface water (the effect of land use practices is
apparent).
● Ownership and access to water resources distorts whatever communal management
system is in place.
Challenges for managing water resources in semi-arid areas 649
ACKNOWLEDGEMENT
REFERENCES
Butterworth, J., Mogkope, K. & Polland, S. 2001. Water resources and water supply for rural
communities in the Sand river catchment, South Africa. 2nd WARFSA/WaterNet Symposium:
Integrated Water Resources Management: Theory, Practice, Cases; Cape Town, 30–31 2001.
Chaturvedi, M.C. 2000. Water for food and rural development. Water International, 25(1), 40–53.
Cleaver, K. & Schreiber, G. 1994. Reversing the spiral: The population, Agriculture and
Environment Nexus in Sub Saharan Africa. World Bank, Washington DC.
Griffiths, J.F. 1972. Precipitation. In J.F.Griffiths (ed.) World Survey of Climatology Vol 10:
Climates of Africa: 24–28. Amsterdam: Elsevier Publishing.
KAWAD. 2001. A fine balance: Managing Karnataka’s scarce water resources. Karnataka
Watershed Development Society, No 250 1st Main Indiranagar, Bangalore 560038.
Lovell, C.J., Butterworth, J.A, Moriarty, P.B., Bromley, J., Batchelor, C.H., Mharapara, I., Mugabe,
F.T., Mtetwa, G., Dube, T. & Simmonds, L. 1998. The effects of changing rainfall and land use
on recharge to crystalline basement aquifers, and the implications for rural water supply and
small-scale irrigation. DFID report 98/3.
Moriarty, P.B. & Lovell, C.J. 1998. Water resources development in Chivi: Results of a village
mapping exercise. DFID report 98/10.
Mugabe, F.T., Hodnett, M.G. & Senzanje, A. 2003. Opportunities for increasing productive water
use from dam water: a case study from semi-arid Zimbabwe. Agricultural Water Management,
62:149–63.
Mugabe, F.T. & Hodnett, M. 2001. Micro-catchment management and common property resources
project: A report on Mutangi hydrological and water resources study in semi-arid Zimbabwe
(1999–2001). Institute of Environmental Studies, UZ, (unpubl.) 26 pp.
Soussan, J. 1998. Water/Irrigation and sustainable rural livelihoods. in Carney (ed) Sustainable
rural livelihoods: what contribution can we make? DFID-ISBN 1 86192 082 2.
Sullivan, C., Mutamba, M. & Kozanayi, W. Water use and livelihood security: A study of rural
households in Southern Zimbabwe (submitted).
Waughray, D.K., Lovell, C.J. & Mazhangara, E. 1998. ‘Developing basement aquifers to generate
economic benefits: A case study from Southeast Zimbabwe’. World Development, 26
(10):1903–1912.
An Integrated Water Resources Management
tool for Southern Africa allowing low flow
estimation at ungauged sites
M.J.Fry, S.S.Folwell & H.A.Houghton-Carr
CEH Wallingford, Wallingford, UK
Z.B.Uka
Ministry of Water Development, Lilongwe, Malawi
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
lives among the most at risk and disadvantaged population groups are not lost, but
transformed into reality.’
As water use and water stress increase, the margin for error in assessing both water
resources and the impacts of water use becomes smaller. At every level, from small
catchments to international basins, water management needs to be carried out with the
consideration of all stakeholders and with all possible awareness of what the future may
bring. In practice, this means that water allocation decisions need to be made by water
resource managers with as much information as possible on the available natural
resource, the effects of existing water use on this resource, the water needs of
stakeholders and the predicted change in these factors in the future.
country, where there is the catchment of Lake Chilwa, which has no outlet, and a smaller
area draining eastwards into Mozambique.
The topography of Malawi is varied, and the country may be divided into four broad
hydrological zones: highlands, plateau, escarpment and rift valley. The plateau is at an
altitude of between 900 and 1200m and features broad, undulating plains. The climate is
temperate, and these are the most densely cultivated regions of the country, where little of
the original woodland remains. The highlands rise abruptly from the plateau reaching
altitudes of between 2100m and 3000m. The climate is cool, and the vegetation is forest
relicts and open grasslands. These areas are now either forest reserves or game reserves,
partly covered with exotic trees. The escarpment marks the boundary between the plateau
and the rift valley. It drops down in a series of shelves and is an area of major faulting.
Considerable portions of the escarpment are protected by forest or game reserves, but
pressure to secure arable land results in some areas of steep land being cultivated outside
these protected areas. The rift valley is mainly covered by alluvial deposits of the
Quaternary age. The climate is tropical, and the original vegetation is mixed savannah
woodland. The most favourable soils in this zone have been developed into irrigated rice
and sugar schemes.
(c) In cases where insufficient natural flow records are available, naturalisation of the
observed flows (i.e., removal of the effects of artificial influences on the pattern of
river flows);
(d) Extraction of low flow measures from data records for these catchments;
The method for estimating flows at ungauged sites is practically applicable within
licensing and water resources management operations. The flow estimation method
requires information on the extent of different soil types and on mean annual rainfall for
each catchment. These data are digitised and available within GIS systems, but can be
time-consuming to derive for a particular location. A bespoke software tool can carry out
this and other functions of the estimation process, and also allow additional information
to be displayed to maximum effect. Software, incorporating these tools, was desired to
exhibit a number of features:
– Visualisation of rivers, artificial influences, gauging stations and other spatial features
– Storage of artificial influence information
– Storage of catchment characteristics—rainfall and soil types
– Derivation of catchments at any point on the river network
– Retrieval of characteristics (rainfall and soil type extents) for given catchment
– Production and visualisation of natural flow duration curve
– Visualisation of the impacts of existing artificial influences upon natural flow regimes
– Creation of ‘scenario’ flow statistics under predicted future influences
– Gauging stations
– Spot gaugings or instantaneous flow measurements
All of these features are stored within the database, each having different data stored
against them in a varying number of fields, or attributes.
(Fig. 3). Defined catchments can be saved for later use, which is useful when a catchment
is under regular scrutiny.
statistics at ungauged sites, and tools to allow these methods to be easily and consistently
applied and to integrate spatial and temporal data on catchments, water availability, water
use and the effects on flow regimes of this use.
CEH Wallingford, as part of phase II of the Southern Africa FRIEND project, has
created a prototype method for low-flow estimation in Malawi, and GIS software tools
for its application, appropriate to the requirements of that country. These tools have been
installed in the Ministry of Water Development in Malawi to allow water resource
managers to visualise water use at a catchment scale and to effectively manage water by
modelling current and predicted catchment water use and observing the effects on
streamflow statistics. Such tools were previously unavailable in the country have the
potential to provide a huge leap forward in the quantitative management of water within
most countries in the Southern Africa region.
REFERENCES
Drayton, R.S., Kidd, C.H.R., Mandeville, A.N. & Miller, J.B. 1980. A regional analysis of river
floods and low flows in Malawi. Report No. 72. Institute of Hydrology, Wallingford, UK.
ESRI. 1998. Shapefile technical description. ESRI White paper. Environmental Science Research
Institute, Redlands, California, USA.
FAO. 1996. Digital soil map of the world and derived soil properties. CD-ROM, Food and
Agriculture Organisation of the United Nations, Rome.
GWP. 2003. Poverty Reduction and IWRM. TEC Background paper no. 8. Global Water
Partnership.
Gustard, A., Bullock, A. & Dixon, J.M. 1992. Low flow estimation in the United Kingdom. Report
No. 108. Institute of Hydrology, Wallingford, UK.
UNESCO. 1997. Southern Africa FRIEND. Technical Documents in Hydrology No. 15. United
Nations Educational, Scientific and Cultural Organization, Paris.
Organization of water services in Malawi and
strengths and weaknesses in implementing
Integrated Water Resources Management
(IWRM)
Milward Selemani
Blantyre Water Board, Malawi
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
Integrated Water Resources Management (IWRM) is important for Malawi even though
20% of the country’s 118,500 square kilometre total area is covered by water. Being
mountainous some areas lie on high altitudes far from any reliable water resources. On
the other hand there are some areas, which as a result of their low-lying topography, they
have more than adequate water and some of these low-lying areas are prone to flooding
every rainy season. Lake Malawi is the third largest lake in Africa and the Shire river
which drains the lake to the Zambezi at annual average flow of 395 cubic meters per
second is regarded as the life line for Malawi because of the major activities taking place
throughout its course: There are fisheries and irrigation on its upper course, electric
power generation and water supply abstraction in the middle course and irrigation of the
Water resources of arid areas 662
sugar estates in the lower or last course of the river. There are also other smaller lakes
and a network of rivers and streams throughout the country.
At national level the Ministry of Water Development has the overall responsibility for
management of water resources throughout the country: The National Water Resources
Board, a unit within the ministry is responsible for approving and issuing of water rights
certificates whilst a Water Supply Unit, also within the Ministry of Water Development
looks after the provision of potable water to the rural communities by provision and
construction of boreholes, protected wells and piped water schemes though out the
country.
The two cities of Blantyre and Lilongwe are under Blantyre and Lilongwe Water
Boards. The two water boards are autonomous and their responsibilities are limited to
abstraction, treatment and distribution of potable water within the city boundary and
some designated outside the city boundary areas. This responsibility does not include
sanitation and sewerage system. Therefore these responsibilities lie under the jurisdiction
of City Councils.
There are a number of NGOs providing assistance to small communal water projects
in collaboration with the Government Water Supply Unit. A notable example is the
Malawi Social Action Fund (MASAF) which funds and supervises various self-help
development projects. Some of such projects are carried out with high-density peri-urban
areas, which are under the jurisdiction of the respective Water Boards and are by nature
of their soil structure and due to poor sanitation, pollution risky. Water supply ranks high
among these projects as it is regarded as basis for any meaningful and sustainable
development. MASAF is known at times to provide assistance to water projects at local
level without approval of the controlling ministry. This can be also cause for confusion.
The three regional water boards and the two water boards for Blantyre and Lilongwe
all report to the Ministry of Water Development on technical issues and Ministry of
Statutory Corporations on administrative issues. Here again reporting to two ministries
can be regarded as an organisational problem that can cause confusion.
At community level villagers sometimes elect leaders who form committees to run and
manage various development projects. Each type of development project may have its
own committee. Water supply committees may be selected for each water source or
catchment and such committees are responsible for formulation of rules and procedures
for managing the resource sustainably.
the country’s major sugar estates in the lower Shire. The control of lake levels and flow
in the Shire River is done through opening and closing gates at Liwonde barrage. The
barrage also has another function of controlling flooding in lower Shire valley. The
ministry is also responsible for gauging flows in other main rivers throughout the country
for purposes of regulating and managing flows for various purposes e.g. irrigation. The
board also approves and issues water rights and issue licences for drilling boreholes
throughout the country. It is however doubtful if the Ministry carries out regular
inspections for the purpose of monitoring these water allocations to ensure that they are
being followed.
The ministry of Water Development and the Department of Environmental Affairs’
legist ration requires both to regulate and manage water resources by ensuring proper
maintenance of the catchments and the environment in general. However this requirement
is not given much prominence and there is sometimes confusion and duplication of
functions particularly in regards to issuance of water rights and control of water pollution.
There is also insufficient liaison between the Department of Forestry and other
stakeholder ministries in issues of conservation of forests particularly in river catchments.
There are times when catchments are laid bare without regard to its long-term effect on
the environment the water resources and with no plans for re-forestation of the affected
areas in the immediate future.
At regional and district level the five water board are mainly responsible for water
supply in their respective areas of jurisdiction. For example in Blantyre the water board
has no powers over other water sources apart from its dams: The City council is the one
which has the overall responsibility over the environment and water quality throughout
the city including the collection and treatment of sewage and monitoring discharge to
rivers from sewage treatment works and industries. The City council formulates,
maintains and enforces by-laws to this effect. The City councils’ performance in this area
has of late however been unsatisfactory. There are therefore plans to transfer these
functions to Blantyre and Lilongwe water boards.
At community level, in the rural areas, there is very little IWRM worth writing about
although irrigation farmers sometimes organise themselves into small area clubs which
can formulate their own water conservation by-laws: There are sometimes cases whereby
farmers are barred from
Water resources of arid areas 664
5 CONCLUSIONS
1. There is great need for promotion of IWRM at all levels i.e. national, regional and
community level through education and awareness campaigns.
2. Government should make deliberate policy and create enabling environment for the
implementation of IWRM.
3. There is need for better coordination and collaboration among various stakeholder
ministries and water boards as well as NGOS in the strategising of the promotion of
IWRM in the country.
Towards best water resources management
practice in small town water supply system in
Tanzania
A.Mvungi
Water and Wastewater Engineer, University of Dar es Salaam, Dar es
Salaam, Tanzania
M.Makuya
Dar es Salaar, Tanzania
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
Sustainable use of water resources requires the integration of demand management with
effective and efficient institutional arrangement. Therefore, the integration of demand
management interventions and policies within the broader water resources management
policy are essential in achieving efficient use of the scarce water resource available.
Tanzania like many African countries have their small town’s inhabitants less
educated, less wealth than urban dwellers, face an uncertain future due to increased
populations, limited water resources, and ineffective water supply and sanitation systems.
They lack possibilities to make economies of scale, and to cross subsidize the poorer
members of the community. In general unable to undertake community water supply and
sanitation management.
The water policy in Tanzania is clear on strategies for service delivery to urban and
rural communities. While service delivery in urban community is purely commercial,
Towards best water resources management practice in small town water supply system 667
3 METHODOLOGY
The town is subdivided into two by Bariadi River, which flows generally westward
into Duma River. The Bariadi Dam which is the proposed source of water for the
township is constructed about 400m upstream of the bridge connecting the two towns.
The reservoir on the Bariadi River completed around 1990 forms a major water source
for the proposed water Supply system. Based on existing 1:50,000 topographical maps
the catchments area above the reservoir is about 512km2 consisting of moderate
catchments slopes.
The current capacity of the dam is limited to about 1.5 million cubic metres, unable to
meet the present population of over 60,000 people.
Local people therefore, relied heavily on shallow wells and rainwater harvesting
shallow wells due normally dug into the back yard of house units. Now with water supply
in place different option of selling water in bulk to them so that they continue to operate
as water user groups shall be looked into assessments shall also involve major sketch
holders.
5 RESULTS
6 MANAGEMENT ARRANGEMENT
6.1 Review of existing models for small towns water supply and sanitation
services management
The preparatory study made by BCEOM in Tanzania identify four basic modes for
management of small town water supply and sanitation in terms of ownership of the
utility infrastructure, the identity of the system operator, the legal status of the system
operator, and the ownership of shares of the operating company where applicable. These
modes are briefly described below.
oil supply, day to day operation, small repairs on the scheme, spare parts supply etc) to a
private operator actually it’s just an individual. In this kind of arrangement, the private
operator collects money from users, pays all running costs, and gives a determined
amount of money to the water committee.
water point, considered as the ‘lowest level of representation’. According to the size of
the system, the water supply company can appoint a few remunerated people to ensure
the day-to-day management of the scheme.
There is currently some form of a Water Board in existence in the towns researched. This
board, which was formed under the initiative of the District Council (DC) and the District
Rural development Program (DRDP), is initially responsible for organizing and raising
funds for financing the water supply project. The funds are at the moment very limited.
The board consists of the Board Chairman, Secretary other 7 members making a total of 9
members.
Following approval of financing of the water supply project, the DC and the Water
Board organized training with the following objectives:
● To formalize structure of the Board
● Establish role and responsibility of the Board
● To prepare the board for participation in the construction of facilities.
● To prepare the board for the management of the water supply project.
The workshop concluded that the current board was representatives of the water users
and that they preferred to form a water user association. This formation of management
however is more suited for rural water supply projects than for small town water supply
systems.
8 DISCUSSIONS
sustainability of service delivery to small urban centers is provided under the Urban
Water Supply and Sewerage policy section 4.9. The policy elaborates that in order to
have improved water supply and sewerage services in small urban centers:
(a) Emphasis will be on privatizing water supply and sanitation services.
(b) Local Private Sector institutions shall be promoted and strengthened. Their access to
credit facilities will be enhanced.
Both components of the policy (Rural and Urban) insist on cost recovery as the basis for
sustainable service delivery. The tariff to be developed has to pay for full cost recovery of
the water supply system.
The Small Town Water Supply and Sanitation Project Preparation Study reviews the
existing models of private sector involvement in management of water supply projects.
Within the framework of the policy, and the context of development of Mwanhuzi water
supply the following models are possible:
● Management Contract
● Leasing Contract
● Concession to private enterprise
Whatever model is adopted issues of ownership of assets, roles and responsibilities of the
different players, duration of contract, part of the investment in charge of the operator and
regulatory issues have to be made clear.
The study insists on the board being totally free from the central government and the
board members to be elected by the council through a subsidiary legislation to be made
under the local government act.
(2) A review of the existing institutional setup of some of the water supply facilities has
also been made. The Urban water Supply and sewerage Authority set up seem to be
OK. The role of the Ministry in backstopping (regulation) is fairly strong which is
good. The powers given to the boards however seem to be overshadowed by the
involvement of the Ministry (boards not dynamic enough) and the ability of the
entities to become full-fledged commercial entities, which is the ultimate objective, is
still far fetched.
Utilities, which have been operating under boards of trustees like the Hai water supply
projects in Kilimanjaro region, have proved to be performing fairly well. These utilities
are already charging tariffs, which ensure full cost recovery for the water supply system.
The board of trustees in this respect acts as a regulator, and the success could be
attributed to the effectiveness of the regulatory mechanism put in place.
Experience from operating limited liability companies and in particular KILIWATER,
are not encouraging. The company was established as a limited liability company under
the company ordinance cap 212. The company is owned by users who acquire shares
from the company, and is totally independent from central government and the local
government. A board of directors appointed by the users oversees the operation of the
water supply system. The company however has not been operating very well due to
among other things internal problems, which require external backstopping mechanisms
to resolve. Experience from this company also points out to the need of an effective and
efficient regulatory mechanism to ensure success.
Towards best water resources management practice in small town water supply system 675
The ongoing reform and restructuring of the Government and developments in private
sector involvement in the management of the major utility companies in the country in
particular TANESCO and DAWASA has led to establishment of Energy and Water
Utility Regulatory Authority (EWURA) for the purpose of regulating delivery of services
by the private sector. This regulatory board has not yet been extended to the district level.
The mechanisms of operations of the regulatory board need to be reviewed to confirm
that it will address the needs as briefly highlighted above.
9 RECOMMENDATIONS
ACKNOWLEDGEMENTS
The authors would like to thank the Managing Director of SERVICEPLAN LTD for the
financial support offered during this study.—Thanks to all the District Councils, for their
assistance during the entire filed works.
Towards best water resources management practice in small town water supply system 677
REFERENCES
MoWLD (Ministry of Water and Livestock Development) 2002. National Water Policy Tanzania
Printing Office, Government of Tanzania.
MoWLD (Ministry of Water and Livestock Development) 1997. Waste Supply and Waste Water
Disposal Design Manual 2nd Draft Vol. 1 of 1997 United Republic of Tanzania.
MoWLD (Ministry of Water and Livestock Development) 2001. Annual Report for Urban Water
Supply and Sewerage Authorities, Government Printers, United Republic of Tanzania.
MoWLD (Ministry of Water and Livestock Development) 2002. Small Towns Water Supply and
Sanitation Project Preparatory Study, Vol. 1 of 2002, Government Printers, United Republic of
Tanzania.
Van der Zaag, P 2000. Water law notes, Department of Civil Engineering, University of
Zimbabwe, Harare.
Water management in the Mauritian textile
wet processing industry
N.Kistamah & S.Roseunee
Department of Textile Technology, Faculty of Engineering, University of
Mauritius, Reduit, Mauritius
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
Over the last 25 years, with industrialisation and population growth there has been an
increasing demand for water. Enormous pressure has been exerted on water bodies for the
continuous supply of fresh water for industrial, commercial, and domestic usage. In
Mauritius, water supply is ensured by water collected during rainfall. The island receives
on average 2100mm of rainfall with the higher elevation regions receiving up to 4500mm
(Water: Resources, Uses & Pollution, 1999).
According to the Falkenmark index of water scarcity, between 1990 and 2004,
Mauritius has moved from a water-abundant to a water stressed country. Its annual per
capita of renewable fresh water has dropped by 17%, from 1750 to 1450m3 (Allybokus et
al., 1996). In the past, most of the surface water resources were exploited but with recent
industrial development, groundwater resources are being continuously developed to
supplement the increasing demand in water. The annual volume of underground water
used by the various sectors (domestic, commercial, industrial and tourism) of the country
Water management in the Mauritian textile wet processing industry 679
is 145 million m3 out of a potential usable volume of 160 million m3. The remaining
required annual volume of 800 million m3, predominantly for agricultural and
hydroelectric power production, is supplied by the surface water sources. In the industrial
sector, textile mills are notorious for its large consumption of water, in particular the
textile wet processing industry. Water is an excellent medium, relatively cheap and safe
to use for textile preparation, colouration and chemical finishing. The textile and apparel
industry uses both surface water from rivers, government pipe supply and groundwater
from boreholes as per Table 1. In dyehouses alone, water consumption has increased
from 5000m3/day to about 30,000m3 per day with an increase in the number of dyehouses
from 6 in 1983 to 32 in 2003 (Ramgulam et al., 2000).
Table 1. Consumption of water in m3 per month in
the textile industry as per water source.
Surface river Groundwater Government
water supply
80,000 490,000 340,000
The 600% increase in water demand requires attention otherwise water availability can be
a limiting factor in the development of the sector. Therefore, comprehensive programmes
of water management may be required to ease off this situation. Integrated water
management in the textile wet processing industry, in the current context, is not just an
environmental initiative. One of its most basic premises is that it improves efficiency and
productivity for the industry. These improvements are seen in lower expenditure on
resources such as energy and water, increased efficiency in production, fewer risks
associated with environmental impacts, and decreased waste-water generation that leads
to savings in water treatment costs.
In addition, the textile wet processing industry in Mauritius has for too long relied on
end-of-pipe solutions without seriously developing and implementing a strategy of
minimizing wastewater. Water scarcity and cost is forcing the industry to recognise and
develop a relationship between business, water resources and the environment.
It is interesting to note that both the fresh water charges from government supply lines
and wastewater charges have increased over the last few years. The comparative fresh
water charges are given in Table 2. The cost of wastewater charges for factories
discharging in industrial sewerage system have increased by more than 600% over the
last five years; from $0.07 to $0.52 per m3.
Water resources of arid areas 680
Therefore, water management not only bring relief on the water supply but economic
benefits to consumers through a reduction in wastewater charges, water, chemical and
energy savings.
3 DYEING MACHINERY
5 WATER REUSE
Technologies for reuse of water also hold possibilities for the industry to reduce water
consumption. Counter-current washing in continuous processing where clean water enters
the final wash box and flows counter to the movement of the fabric through the wash
boxes and the reuse of not too contaminated rinsing water may be used for processes
which do not require very high quality water. For example, in the dyeing of cotton with
reactive dyeing on jet machines, water from the last cold rinse may be used for the
preparation of the scouring and bleaching baths.
6 PROCESS MODIFICATION
Process modification as a means to save water in textile wet processing has been widely
investigated (Vigo, 1994). The focus has been on combining textile preparatory processes
which traditionally consume large volumes of water. Desizing, scouring and bleaching of
woven textiles are mandatory when dyeing pale to medium shades. Traditional methods
of processing involve carrying out each process followed by extensive rinsing to wash
out processing chemicals, degraded starch, cellulose, and lignin-based by-products.
Nowadays, the technology of one-stage process has gained more importance with water,
chemical, energy and time savings as very attractive benefits.
The use of better performing dyes and auxiliaries has also helped to save water. The
first generation reactive dyes for natural fibres suffered from the serious drawback of low
fixation efficiency with figures in the range of 60–70%. Now, high fixation reactive dyes
with fixation levels of the order of 80–90% may be attained. Besides, following dye
application in the presence of large amounts of electrolyte, the substrate has to be
subjected to 5–6 rinsing and washing cycles to remove electrolyte, alkali and unfixed
dyes. This procedure is both time and water consuming, and generates large volumes of
coloured effluents. New generation reactive dyes are mostly polyfunctional and their
fixation efficiencies are vastly superior due to the synthesis of more stable fibre-reactive
groups. They also confer the advantages of low salt requirement and the removal of the
unfixed dyes requires less water.
Water resources of arid areas 682
7 SUPERCRITICAL FLUIDS
Supercritical fluids (SCF) such as supercritical carbon dioxide (scCO2) have been
considered as solvents for chemical processing of textile materials (Ozcan, 1998). The
underlying physics and chemistry of SCF-polymer and SCF-solute interactions are now
well understood (Johnston, 1989). Densities and viscosities in supercritical fluids are less
and diffusion into the polymer is more rapid than in liquids, shortening the process time
and improving productivity. scCO2 itself as a fluid is deemed to be low-cost,
environmentally friendly, non-flammable and potentially avoids water usage. It has low
critical parameters (31°C, 73.8 bar) and can be recycled. In textiles, it has been used for a
range of processes that include scouring, dyeing, dry-cleaning [D] and impregnation of
functional finishes.
The dyeing of polyester fibres with disperse dyes in scCO2 has been widely
investigated and was found to offer a number of advantages when compared to aqueous
dyeing (Saus, 1995). For example, no reduction clear (removal of surface deposited
disperse dye) was required. This process usually consumes large volumes of water in
order to achieve commercially acceptable wet fastness and gives rise to coloured
effluents. The colouration of polyester-type fibres in SCF is, therefore, an attractive
alternative to aqueous dyeing. The colouration of natural fibres in scCO2 is still at the
pilot scale but new knowledge in this area indicates that “dry textile dyeing” may be a
possibility in the near future.
8 AWARENESS CAMPAIGN
8.3 Education
A more long-term approach to water management can be taken through formalized
employee education. Education programs are more general and less job-oriented than
training programs. There is a need for an in-depth understanding, for example, of the
chemistry & theory of dyeing and the design & components of the dyeing equipment.
This knowledge is essential for water management and long-term improvements.
In general, the training is best conducted internally because job-related issues are very
site-specific. On the other hand, general education can be conducted either internally or
externally. The University of Mauritius has conducted a number of conferences and
seminars, in-plant courses by experts in the area of water conservation to help the
industry in this respect.
9 CONCLUSION
REFERENCES
Allybokus M.E. & Ramjeawon T. 1996. Efficient uses of water in industry and agriculture. World
Day For Water Conf., Civil Engineering Department, Faculty of Engineering, University of
Mauritius.
Bradbury M.J., Collishaw P.S. & Moorhouse S. 2000. Controlled rinsing: A step change in reactive
dye application technology. Colourage Journal India, Annual 2000, 73.
Johnston K.P. & Penninger J.M.L. 1989. American Chemical Society: Washington DC, 406, 207.
Water resources of arid areas 684
Ozcan A.S., Clifford A.A., Bartle K.D., Broadbent P.J. & Lewis D.M. 1998. J. Society of Dyers
and Colourists, 114, 169.
Ramgulam R.B., Kistamah N. & Rosunee S. 2000. Study of dyehouse effluent treatment in
Mauritius. Mauritius Research Council Project Report.
Saus W. & Jasper J. 1995. Textile Technology International, 145.
Tulsi S. 2003. Potential reuse of rinsing water in dyeing. BSc. Thesis, Department of Textile
Technology, University of Mauritius, Mauritius.
Vigo T.L. 1994. Preparatory processes. Textile processing and properties, Textile science
andtechnology series 11, T.L.Vigo, 1st ed., Elsevier Science B.V., The Netherlands, 4–31.
Water: Resources, Uses & Pollution, (1999), Ministry of Environment, Mauritius,
http://www.intnet.mu/iels/%20water_mau.htm
Analysis of the microbiological situation of
the quality of domestic water sources and
identification of the microorganisms in them,
located in the semi-arid regions of the Eastern
Cape, South Africa
M.Zamxaka, G.Pironcheva and N.Y.O.Muyima
Environmental and Natural Products Biotechnology Research Group,
Department of Biochemistry and Microbiology, University of Fort Hare,
South Africa
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
ABSTRACT: The water in Alice and Gogogo, which belong to the semi-
arid areas of the Eastern Cape province of South Africa, was characterized
by using both standard microbiological methods and the standard physical
methods to investigate its present quality in the sampling sites. For
microbial analysis, indicator bacteria, namely, heterotrophic, total and
faecal coliforms and physical parameters, such as pH, turbidity and
temperature were assessed to check whether the water from dams, rivers,
wells, etc. is safe for drinking. Almost all the indicator bacteria counts
were above the South African standards. The physical parameters, such as
alkalinity of water and high turbidity, proved favourable for bacterial
growth. In another series of experiments we identified, using the API20E
Assay kit, 54 different species of microorganisms, about 77.5% of them
being human pathogens, 53.2% belonging to the family of
Enterobacteriacae and only 22.5% being non pathogenic. Our
investigations prove that the water in the domestic water sources of Alice
and Gogogo is of poor quality and needs further purification.
1 INTRODUCTION
The lack of safe drinking water and adequate sanitation measures leads to a number of
diseases, such as cholera, dysentery, salmonellosis, typhoid, etc, which claims the life of
2 million people each year in the developing countries (WHO, 1993). The primary goal
Water resources of arid areas 686
of water quality management from a health perspective is to ensure that consumers are
not exposed to doses of pathogens that are likely to cause infectious diseases (Pegram et
al., 1998). Protection of water sources and treatment of water supplies have greatly
reduced the incidence of these diseases in developed countries (Craun, 1986). One of the
difficulties in evaluating the impact of the drinking water supply on health is the lack of
local demographic statistics, particularly in rural communities.
Therefore, it is important to know the incidences, occurring in the rural semi-arid
areas, due to polluted water. This will give the opportunity to compare the incidence of
water borne diseases between communities that have drinking water and those that do not
have.
Detection of bacteria, potentially toxic substances and other contaminants usually
requires laboratory conducted tests. Detection and enumeration of indicator organisms, is
the basic technique, used in water quality monitoring. Coliform group of bacteria can be
defined as the principal indicators of purity of water for domestic, industrial or other uses.
High faecal and total coliform counts in water are usually manifested in the form of
diarrhea, fever and other secondary complications (Fatoki et al., 2001).
In the Eastern Cape Province, belonging to the semi-arid areas of South Africa, nearly
80% of the population relies on surface water as the main water source. Almost 30% of
the population is without proper water supply services, which implies that many of the
people still utilize untreated surface water for domestic purposes. The incidence and
prevalence of water-borne pathogens is subject to geographical factors. Most of the
pathogens are distributed worldwide, but outbreaks of some diseases for instance cholera,
shigellosis and typhoid tend to be regional (Grabow et al., 1994).
Meteorological events and pollution are a few of the external factors, which affect
physicochemical parameters, such as temperature, pH, and turbidity of water. They have
major influence on biochemical reactions that occur within water.
The purpose of the present study was to determine the status of domestic water
sources, used by Alice and Gogogo rural communities.
1.1.2 Chemicals
Sodium thiosulfate
API20E kit (Bio Me’rieux, Lyon, France). Filter membranes (0,45mkm, pore size,
47mm diameter) (Microsept, Cape Town, South Africa).
Analysis of the microbiological situation 687
2 RESULTS
The results of the physical, chemical and microbial analysis (pH, Temperature, Turbidity,
HPC, TC and FC) analysis of Alice and Gogogo domestic waters show that both
microbiological and physical parameters are above the South African standards and there
is a necessity of good purification systems of the water investigated.
The pH of 17 different sites of surface water has been measured in the semi-arid areas
of Alice and Gogogo as seen from Fig. 1. The optimum pH values for indicator
microorganisms, especially coliform bacteria are from pH 3 to pH 10.5. The overall
picture of the pH graph shows that Gogogo village has sites with lower pH values as
compared to Alice. Most of the sites studied in Alice have high pH values of water. This
alkalinity of Alice waters plays a very important role inhibiting the growth of pathogenic
microorganisms such Vibrios, Salmonellas, Shigellas, etc.
The temperature of the surface water was also investigated at different sites of Alice
and Gogogo. In overall, Alice Temperatures of water are higher than Gogogo
temperatures, which favour the growth of Salmonella species, Shigella species and E.coli.
in Gogogo areas, on the other hand, the water temperature favours the growth of Vibrio
species.
The number of faecal coliforms was also investigated from different water sampling
sites of Alice and Gogogo (Fig. 4). Our results show that the most contaminated sites
appear to be those, corresponding to Gogogo area. The lowest values of contamination
with faecal coliforms are observed in the water sources belonging to the Alice area.
We have checked a total of 338 isolates from both Alice and Gogogo water supplies—
183 isolates from Alice area and 155 isolates from Gogogo water sources.
Using both standard microbiological methods and the API20E commercial kit, we
were able to identify 54 different species of microorganisms, including both pathogenic
and non-pathogenic ones. Our results indicate that 77.5% of them are human pathogens
and 53.2% of them belong to the family Enterobacteriacae. Only 22.5% of them proved
to be non pathogenic to humans. Comparing the distribution of species in the drinking
waters of Alice and Gogogo, it is observed that the percentage of species belonging to
Enterobacteriacae family is 7.84% higher in the Gogogo water sources. The high
Shannon Weaver Index (H) for both areas—Alice—2.84 and Gogogo—2.24 also
indicates high biodiversity of microorganisms from sources in both areas.
3 DISCUSSION
The purpose of the present study was to determine the microbiological quality of the
domestic water sources used by the rural communities of Alice and Gogogo. Our
experiments show that both the total and faecal coliform counts are above the South
African recommended standards for drinking water in almost all the sites studied. The
high total and faecal coliform counts, TC—3162 colonies per 100ml and FC—10000
colonies per 100ml are the results of contamination especially in dams where the water is
used by domestic animals, clothe washing and is exposed to heavy rainy falls. Another
possible contamination of water sources might be the presence of pit latrines close to
them, little environmental protection and poor catchment points management (Muyima &
Ngcakani, 1998). Microbial growth and physical quality of water are considered as
priority parameters to be monitored in the rivers, dams and boreholes catchments (Fatoki
et al., 2001). The presence of faecal coliform bacteria indicates that the water is
contaminated with faecal of humans or animal waste while the total coliforms counts
Water resources of arid areas 692
indicate that the water is contaminated with both faecal waste and other bacteria from the
soil.
Our study on the identification of the microorganisms in these domestic water sources
further shows the presence of 54 different species of microorganisms, out of which
53.2% of them belonging to the family Enterobacteriacea, responsible for serious enteric
diseases. The high number of species isolated from the drinking water of the two areas, as
well as high Shannon Weaver Indexes are an indication of high water contamination.
The water sources in both areas of Alice and Gogogo show high levels of
contamination, predominantly by species of the Enterobacteriacea family. All the water
supplies, checked need extensive and efficient purification so that they can be used for
drinking water.
REFERENCES
Craun G.F. 1986. Water borne diseases in the United States. 295–302. CRC Press, Inc. Boca
Raton, Florida
Fatoki O.S., Muyima N.Y.O & Lujiza N. 2001. Situation analysis of water quality in the Umtata
River catchment. Water SA. 2(4), 467–474.
Grabow W.O.K., Favorov M.O., Khudyakovan S., Taylor M.B. & Fields H.A. 1994. Hepatitis E
seroprevalence in selected individuals in South Africa. J. Med. Virol. 44, 384–388.
Muyima N.Y.O. & Ngcakani F. 1998. Indicator bacteria and regrowth potential of the drinking
water in Alice Eastern Cape. Water SA. 24, 29–34.
Pegram G.C., Rollins N. & Espey Q. 1998. Estimating the cost of diarrhea and epidemic dysentery
in KwaZulu Natal and South Africa. Water SA. 24, 11–21.
WHO. 1993. Guideline for drinking water quality. 1. World Health Organization. Geneva,
Switzerland.
Dry season Kalahari sap flow measurements
for tree transpiration mapping—Serowe study
case, Botswana
M.W.Lubczynski1, A.Fregoso2, W.Mapanda1, C.Ziwa1, M.Keeletsang1,3,
D.C.Chavarro1 & O.Obakeng1,3
1
International Institute of Geoinformation and Earth Observations, ITC,
Enschede, The Netherlands
2
Instituto Nacional de Ecologia,, Mexico D.C., Mexico
3
Geological Survey of Botswana, Lobatse, Botswana
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
The presence of green, well adapted to dry season water stress condition vegetation on
Kalahari, the recent information about extremely deep tree rooting depths on Kalahari
(Canadel et al. 1996, La Maitre et al. 2000) and the particular importance of transpiration
in environments with extremely low recharge (Lubczynski and Obakeng 2004, in the
same issue), initiated the idea of quantifying transpiration on Kalahari (Lubczynski
2000). This idea has been materialized in the framework of the Botswana Kalahari
Water resources of arid areas 694
Monitoring project (Lubczynski and Obakeng 2004—in the same conference issue), in
which this transpiration study was realized addressing the research question on what is
the spatial variability of transpiration in semi-arid and bio-diverse Kalahari environment.
The answer to this question is expected to support groundwater modeling and
management’s problem of spatial variability of groundwater evapotranspiration
(Lubczynski 2000).
The first contribution to spatial assessment of groundwater evapotranspiration in the
study area was made by Timmermans and Meijerink (2000), who used remote sensing
solution of energy balance to derive actual evapotranspiration. The application of this
technique, particularly with regard to dry season evapotranspiration fluxes, indicated that
the method, which proved to be suitable in the irrigated areas, does not provide sufficient
accuracy to deal with very low fluxes in dry semi-arid conditions such as Kalahari,
substantially overestimating them. Considering the important role of transpiration in
groundwater evapotranspiration (Lubczynski and Gurwin 2004), the next attempt
discussed in this study, was focused on spatial transpiration assessment through remote
sensing upscaling of field sap flow measurements. The methodology applied consists of
two steps, first related to field sap flow data acquisition and the second to RS upscaling
of field measurements. This paper presents mainly the first step, discussing methodology
of the Kalahari sap flow measurements and the results of dry season plot transpiration
upscaling in the selected, representative, 10×10km study area (see Fig.1 in Lubczynski
and Obakeng, 2004, in the same conference issue). In the second assessment step, which
is still in the development stage and therefore is not discussed here, high-resolution
IKONOS image (Keelesang 2004) and aircraft based, multiband camera image are used
for transpiration upscaling (Hussin et al. 2004a, b).
Tree transpiration can be well evaluated by direct field measurements of sap flow. Sap
flow (Qs) is a product of sap velocity (ν) and sap wood (xylem) area (Ax). Unfortunately
so far there is no method being able to directly integrate Qs, therefore Qs is typically
defined by separate measurements of ν and Ax.
So far, there were in the study area three dry season sap flow measurement campaigns
(mostly in Septembers): in 2001 when tree species such as Acacia fleckii, Boscia
albitrunca and Lonchocarpus nelsii were measured (Fregoso 2002), in 2002 when
Terminalia Sericea, Burkea Africana, Acacia erioloba, Ochna pulchra, Dischrostachys
cineria and Acacia luederitzii were measured (Mapanda 2003) and in 2003 when
Ziziphus macronata and Acacia karoo were measured (Keeletsang 2004). The eleven
above mentioned species cover the tree variety in the study area.
charge regulator were used. All the measurements, were sampled every 30 seconds and
stored every 30 minutes by four synchronized DataHog2 loggers.
where Ax is the sapwood (xylem) area and ν the sap velocity. According to Granier
(1987), ν, which is typically expressed in cm/h, can be estimated from the continuously
measured temperature difference (∆T—higher in the night and lower in the day when
heat dissipating sap flow occurs) between the upper heated and the lower non-heated
TDP probes. In calculation of ν the reference is made (Equation 2) to ∆Tmax, which is the
maximum night temperature difference between the two probes when stomata’s are
closed so no sap flow occurs.
(2)
The normalized sap flow QN of individual trees, typically expressed in l/d/m2, can be
estimated as:
QN=Qs/Ac
(3)
Water resources of arid areas 698
(4)
where ΣQS is the total sap flow of all the trees present in the investigated plot
characterized by area A, in this study case 35×35m=1225m2.
3 PLOT TRANSPIRATION
In assessment of ΣQs per plot, species-specific correlations of xylem area vs stem area
and xylem area vs crown area, combined with the mean sap velocities obtained per
species were used and provided spatial, plot transpiration variability (Figure 4).
The analyze of dry season transpiration fluxes in 35×35m plots indicates large
variability of fluxes from nearly 0 to 0.15mm/d (Fig. 4). It can be noticed that the higher
and the lower transpiration flux values are clustered which is mainly resulted by variable
density of vegetation in the area of concern.
4 CONCLUSIONS
The sap flow measurement is a useful but tedious and quite expensive method for
transpiration assessment. However once in the certain area such as Kalahari the
background research defining
Dry season Kalahari sap flow measurements for tree transpiration mapping 699
small size of the plot assessment, very sensitive to the local anomalies in the tree density
and species type and also to the generally large biodiversity of the Kalahari vegetation.
The large biodiversity in the Kalahari study area makes the RS upscaling of sap flow
measurements particularly challenging, mainly due to the difficulty in RS classification
of the individual tree species.
ACKNOWLEDGMENTS
REFERENCES
Canadell, J., Jackson, R.B., Ehleringer, J.R., Mooney, H.A., Sala, O.E.T. & Schultze, ET.D. 1996.
Maximum rooting depth of vegetation types at global scale. Oecologia 108:583–595.
Fregoso, A. 2002. Dry-season transpiration of savanna vegetation. Assessment of tree
transpiration and its spatial distribution in Serowe, Botswana. MSc thesis, Library of ITC—
International Institute for Geoinformation Science and Earth Observation, Enschede, The
Netherlands.
Granier, A. 1987. Evaluation of transpiration in Douglas-fir stand by means of sap flow
measurements. Tree Physiology, 3:309–320.
Hussin, Y.A., Chavarro, D.C., Lubczynski, M.W. & Obakeng, O. 2004a. Mapping vegetation for
up-scaling evapo-transpiration using high-resolution optical satellite and aircraft images in
Serowe, Botswana. Proc. WRASRA conf. Gaborone 3–7 August 2004, Rotterdam, Balkema.
Hussin, Y.A., Lubczynski, M.W., Obakeng, O & Chavarro, D.C. 2004b. Designing and
implementing an aircraft survey mission using high-resolution digital multi-spectral camera for
vegetation mapping for up-scaling evapo-transpiration of Serowe, Botswana. Proc. WRASRA
conf. Gaborone 3–7 August 2004, Rotterdam, Balkema.
Keeletsang, M. 2004. Assessment of dry season transpiration using IKONOS images, Serowe case
study, Botswana. MSc thesis, Library of ITC—International Institute for Geoinformation
Science and Earth Observation, Enschede, The Netherlands.
Le Maitre, D.C., Scott, D.F. & Colvin, C., 2000. Information on interactions between groundwater
and vegetation relevant to South African conditions: A review. In: Past Achievements and
Future Challenges. ISBN 9058091597, Rotterdam, Balkema: 959–961.
Lubczynski, M.W. 2000. Ground water evapotranspiration—underestimated component of
groundwater balance in a semi-arid environment—Serowe case Botswana. In: Past
Achievements and Future Challenges, ISBN 9058091597, Rotterdam, Balkema: 199–204.
Lubczynski, M.W. & Gurwin, J. 2004. Integration of various data sources for transient groundwater
modeling - Sardon study case, Spain. Journal of Hydrology—in revision.
Lubczynski, M.W. & Obakeng, O.T. 2004. Monitoring and modeling of fluxes on Kalahari—setup
and strategy of the Kalahari Monitoring project. Serowe study case Botswana.
Mapanda, W. 2003. Scalling-up tree transpiration of eastern Kalahari sandveld of Botswana using
remote sensing and geographical information system.
Dry season Kalahari sap flow measurements for tree transpiration mapping 701
1 INTRODUCTION
People of Khartoum State depend in their drinking water on the fresh surface water of the
River Nile and its main tributaries (the Blue Nile and the White Nile) which meet at
Mugran area, Khartoum State). The surface water is being treated and distributed to the
three main cities comprising the State through the pipelines, although there are many
producing ground water wells inside the three main cities. With the tremendous increase
in population, and growing demand for groundwater quantity the groundwater quality is
now should be taken into account. There are many industrial areas now lacated within the
living area which expected to be a source of pollution to the groundwater, this beside the
daily municipal waste generated.
Agricultural purposes since there are vast areas around Khartoum, as in the case of all
Sudan, suitable for agricultural practices. Both governmental and private sector efforts in
the field of groundwater in Khartoum area mainly focus on searching, exploration and
exploitation of the groundwater. Information with respect to groundwater quality,
especially heavy metals, is generally lacking. This paper is an attempt to throw light on
certain heavy metals content in the groundwater of Eastern Khartoum State, which will
give an idea about heavy metals content of the Nubian Aquifer which extend even to the
north in Egypt and Libya comprising the Great Nubian Basin of North Africa.
Heavy metals and radioactivity in the groundwater of Khartoum State, Sudan 703
The study area entirely lies in the eastern part of Khartoum State, mainly occupies vast
areas of Eastern Nile Province. The area is bounded between latitudes 15.42° and
15.78°N and longitudes 32.45° and 32.93°E, it is bounded by the River Nile and the Blue
Nile from its western side, and is about 400 square kilometers in area (Fig.1).
The area lies within the arid zone with an average annual rainfall of about 167mm
during the summer season (July–September), and evaporation is about 10mm/day
(Haggaz and Khairalla,
area, mainly to the north and the east. The Basement Complex Rocks are succeeded
unconformably by the Nubian Sandstone, which consists mainly of flat-lying or gently
dipping rocks made up of continental sediments which include sandstones, grits,
mudstones, extra-formational and intra-formational conglomerates (Khairalla, 1966;
Whiteman, 1971). The Nubian formation are overlained by alluvial deposits of the Blue
Nile and the main Nile. The alluvial deposits consist of ill-sorted clays and silts with
sandy and gravelly lenses and believed to be quaternary in age (Whiteman, 1971). In
some places the alluvial deposits and the Nubian Formation are covered by wind-blown
sands and the recent Nile silts. The groundwater occurs mainly in the Nubian Sandstone
Formation and the alluvial deposits of the Niles, and the aquifers of the Nubian
Sandstone and the alluvial deposits are believed to be hydraulically interconnected
(Khairalla, 1966). Two aquifers had been recognized in the Nubian Formation, an upper
aquifer of variable thickness (10–300m) and lower one of more than 400m with higher
values of transmissibility and permeability (Bureau of Geological Research, 1979). The
depth to the
Table 1. Showing trace elements analyses in the
study area.
Well Locality Pb Zn Ni Cu Mn
no. (ppm) (ppm) (ppm) (ppm) (ppm)
1 Um Dureiwa ND 0.01 0.01 0.03 ND
2 Daroshab South ND 0.01 0.02 0.01 0.01
3 Daroshab North ND 0.01 ND 0.01 ND
4 Samrab East ND 0.01 ND 0.01 ND
5 Halfaya – – – – –
6 Um Dawan Ban ND 0.01 ND ND ND
7 Abu Groon 0.02 0.02 0.01 ND 0.63
8 Eid Sharoom 0.02 0.01 ND 0.02 0.01
9 Sheikh Mustafa El ND 0.01 ND 0.02 ND
Fadni
10 Hillat Kuku ND 0.01 0.01 ND 0.01
11 Direisab 0.02 ND 0.01 0.01 ND
12 Idd Um Dom ND 0.01 0.04 ND ND
13 Es Sutra ND 0.02 0.03 ND ND
14 El Hur ND 0.02 0.02 ND ND
15 Um Usheira ND 0.01 0.01 0.01 ND
16 Yafa 0.03 0.01 0.02 0.01 ND
17 Idd Babikir North 0.03 0.01 0.07 0.01 ND
18 Galaat Dugo 0.03 0.01 0.03 ND ND
19 Sheikh A.Rahman 0.01 ND 0.03 ND ND
20 Gabarona ND ND 0.03 0.01 0.01
21 Hattab ND ND 0.05 0.01 ND
22 EL Shafeiab ND ND 0.02 0.01 ND
23 Allogab ND ND 0.02 0.01 0.05
24 Kadaro ND ND ND ND ND
Heavy metals and radioactivity in the groundwater of Khartoum State, Sudan 705
saturated zone is variable ranging from 5m near the rivers up to 10m at a distance of
about 2.3km east of the Blue Nile.
Groundwater samples were collected from 28 producing wells located in the study area
(Fig. 1) for analyses. The water samples were collected in clean 1 liter polyethylene
plastic bottles and stored in a cooler for 24 hours. Electrical conductivity and pH were
determined in the field.
Analysis of heavy metals was carried out using a Perkin Elmer Atomic Absorption
Spectrophotometer (model 1100).
Radiochemical analysis was carried out in the Sudan Atomic Energy Commission
Laboratories, and the following procedure was applied to determine radioactivities in the
samples:
About one litre of the sample is weighted in special container made of plastic, and
then loaded on the system. The low level counting of the radiation is performed using
gamma spectroscopic system. A highly pure germanium detector (HPGe) is used. It is
connected to a pre-amplifier and an amplifier. The multichannel analyzer (MCA) as a
new version of personal computer analyzer (PCA) software (Gamma Data Reduction
‘GDR’) are used for the analyses and peak identification. The time used for collection is
between 4000 to 10000 seconds. The system is calibrated using standard source of known
energies and activities.
The results of heavy metals which are shown on Table 1, indicated that the
concentrations of these ions are very low and in most cases even not detected. This
indicated that the groundwater aquifer is free from harmful elements and free from heavy
metals pollution. Heavy metals values are far below the permissible ones set by the WHO
(1984) guidelines for drinking water. Since the groundwater of the study area belonging
to the Nubian aquifer, which is part of the Great Nubian Aquifer of north Africa, it can be
suggested that the groundwater of this aquifer is free from trace elements unless there are
certain local reasons.
Water resources of arid areas 706
5 RADIOACTIVITY
The most abundant radioactive element which could be found in water is 226R isotope
(IAEA, 1990). The reason for that is relative high solubility when compared to the other
radionuclei. Although it dissociate to give 226Ru isotope, but 222Ru is a gas and could not
be found in water specially when it is exposed to air. From results of all samples the
concentration of 226R is nil which indicate that the whole study area is free from any
source of radioactive isotope. The presence of 226Ru and K40 which are showed in each
spectrum are the normal radioactive isotopes which are present in the atmosphere and the
values are even less than the standard lower limit of the natural radioactivity. Finally the
water samples collected from the study area are free from radioactive pollution Similarly
it can be suggested that the Nubian Aquifer is free from radioactivity.
6 CONCLUSION
From the results of trace elements and radiochemical analyses it can be concluded that
the groundwater is free from heavy metals and radioactive pollution. This phenomenon
can be applied to the groundwater of the Nubian Aquifer covering fast areas in North
Sudan and other areas of the country, and if there is any pollution within this aquifer can
related to localized phenomena.
REFERENCES
Bureau of Geological Research 1979. Groundwater resources in Khartoum Province, Part II, Fed.
Inst. Geosci., Nat. Resources, Hannover.
Haggaz, Y.A.S. and Khairallah, M.K. 1988. Paleohydrology of the Nubian Aquifer North East of
the Blue Nile, near Khartoum, Sudan, Jnl. of Hydrology, 99:117–125.
Khairallah, M.K. 1966. A study of the Nubian Sandstone Formation of the Nile Valley between
latitude 14°N and 17° 42′ N with reference to groundwater geology, MSc. Thesis, University of
Khartoum.
Whiteman, A.J. 1971. The Geology of the Sudan Republic, Clarendon, Oxford.
WHO 1984. World Health Organization. Guidelines for drinking water standards, vol 1,
Recommendations, Geneva; WHO.
Impediments to the effective implementation
of a groundwater quality protection strategy in
Botswana
T.R.Chaoka,1 E.M.Shemang,1 B.F.Alemaw1 & O.Totolo2
1
Department of Geology, University of Botswana, Gaborone, Botswana
2
Faculty of Science, University of Botswana, Gaborone, Botswana
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
For a water scare nation like Botswana, protection of water resources from contamination
is one of the key prerequisites for sustainable water resources management. Botswana,
Water resources of arid areas 708
Botswana has very limited surface water and groundwater resources. This is due to its
arid climate, high rates of potential evapotranspiration (2000mm/a) and low rates of
groundwater recharge. Average annual rainfall ranges from 250mm in the southwest to
650mm in the northwest. Most of the rainfall falls in summer from October to March and
varies considerably from year to year as well as spatially. In addition to the low average
annual rainfall, Botswana is prone to frequent and long periods of drought.
Most of Botswana is covered by thick sandy soils. The thick sandy soils together with
the high evapotranspiration rates and low topography severely limit the amount of runoff.
It is estimated that the amount of surface runoff that originates in Botswana represents
Impediments to the effective implementation of a groundwater quality protection 709
about 1% of the total rainfall that reaches the land surface. Estimated recharge to aquifers
over most of the country is in the order of 1mm per year (DWA, 1991b; Gabaake, 1998).
Figure 1 shows the different types of aquifer formations found in Botswana and their
areal distribution. The different types of aquifers and their properties are summarized as
shown in Table 1.
From Figure 1, it appears as if the whole of Botswana is covered with aquifers.
However, the amount of extractable groundwater resources in most aquifers is very small
due to their generally low transmissivities, storativities, and recharge rates. Some of the
extractable groundwater resources are unsuitable for human and livestock consumption
due to their high TDS, nitrate, iron, manganese, and fluoride content. Pollution of
groundwater resources has also occurred in some parts of Botswana and is still occurring
in others. For example, in the late 1970’s a number of production boreholes in the
Mochudi Wellfield were abandoned due to nitrate and bacteriological contamination from
anthropogenic sources (DWA, 2000).
The Government of Botswana initiated three studies that were conducted in 1985 and
1991 to assess the magnitude of and identify sources of water pollution in Botswana
(DWA, 1991). The results of these studies indicated that there was a severe threat to
groundwater pollution associated with the disposal of sanitary waste. Although the threat
from mining, industrial, and agricultural activities appeared insignificant, the studies
showed that it was growing. Legislation related to water pollution was also reviewed and
was found inadequate and in need of revision.
Following these studies, the Department of Water Affairs (DWA) initiated a project on
protection zones and guidelines for major wellfields, aquifers and dams in Botswana that
was concluded in 1993 (DWA, 1993a, b). The main objectives of the study were to
identify existing
Water resources of arid areas 710
water to a production well. Implicit in this definition is the assumption that contaminants
can only reach a production well if they are released within the area contributing water to
the well.
shale
Waterberg Quartzite, 40000 F-K 2–5
Supergroup shale
Transvaal Dolomite
Supergroup
W, NW and 400 F 0.2
N Botswana
Olifantshoek Quartzite,
Sequence chist,
porphyry 33000 F 0.6–2
Damara Arkose,
Sequence quartzite,
schist, 1000 F-K
mudstone,
shale
Dolomite
Archean Ventersdorp Porphyry, 50 F 0.7
Supergroup felsite,
greywacke
Archean Granite, 61000 F 0.5–2
basement of gneiss,
SE and E migmatite,
Botswana
Amphibolite
F=fractured; K=karstified; P=porous; Successful borehole rate
>0.01m3/h.
towards a pumping well within one year (Bates and Evans, 1996). The TOT is estimated
with the aid of numerical models. In the fixed radius method protection zones are
demarcated by circles of fixed radii from the production borehole. Delineation of TOT
protection zones requires reliable estimates of hydraulic properties of aquifers and
wellfields, which are rarely known with any certainty. The shapes of TOT protection
zones depend on hydraulic properties.
The Wellhead Protection areas are divided into four zones based on groundwater
travel time and distance from the production well (Table 2): an operational courtyard,
inner (Zone I), intermediate (Zone II), and outer (Zone III). As Table 2 shows, WHP
areas are rather restricted in
Impediments to the effective implementation of a groundwater quality protection 713
The preceding situation analyses show that the Botswana groundwater quality protection
program is similar to that of many countries, yet the implementation of the program is not
properly coordinated in order for it to be effective.
In order for an effective groundwater quality protection strategy to be put in place in
Botswana, the paper recommends the following:
● Production of aquifer-wide vulnerability maps for all aquifers in Botswana:- These
maps should be used as inputs in the preparation of land use maps.
● Establishment of a unit within the National Conservation Strategy Coordinating
Agency to coordinate land use planning, groundwater development, and other related
activities that may impact negatively on groundwater resources:- Some of the
functions of this unit should include public education on water use, water
conservation, and groundwater quality protection. The unit should also have the
authority to enforce all water related legislations.
● Public participation in the conceptualization, preparation and implementation of
landuse plans:- This will ensure that sensitive areas that may impact on water quality
and groundwater resources are demarcated and protected by the community in order to
ensure sustainable water supply.
● Expansion of the monitoring network of groundwater quality:- The monitoring of
groundwater quality should be performed in all boreholes (both production and
observation wells) for adequate planning, development and protection of the
groundwater resources of Botswana.
It is the opinion of the authors that the above recommendations if implemented, would
immensely strengthen the groundwater quality and protection in Botswana.
REFERENCES
Bates, J.K. and Evans, J.E. (1996). Evaluation of Wellhead Protection Area Delineation Methods,
Applied to the Municipal Well Field at Elmore, Ottawa County, Ohio. Ohio Journal of Science,
96 (1), 13–22.
DWA (1985) Ramotswa Wellfield Pollution Study.
DWA (1991a) Magnitude and Sources of Pollution in Botswana.
DWA (1991b). Botswana National Water Master Plan Study. Final Report. Volume 5.
DWA (1991c). Botswana National Water Master Plan Study. Final Report. Volume 11. pp. 55.
DWA (2000) Groundwater Monitoring Study, volume 1.
DWA (1993a) Protection Zones and Guidelines for Major Wellfields, Aquifers and Dams in
Botswana, Volume 1.
DWA, (1993b) Protection Zones and Guidelines for Major Wellfields, Aquifers and Dams in
Botswana, Volume 2.
Gabaake, G.G. (1998). International Conference on the Role of National Geological Survey in
Sustainable Development. Abstract Volume, 15–19.
MacDonnell, L.J. and Guy, D.J. (1991). Approaches to Groundwater Quality Protection in the
Western United States. Water Resources Research, 27(3), 259–265.
Impediments to the effective implementation of a groundwater quality protection 717
Mclaren, D.A., Hazell, R.T. and Gyopari, M.C. (1996). Water Quality Protection Strategy in
Botswana. Botswana Journal of Earth Sciences, Volume 3, pp. 25–28.
Spatial assessment of groundwater pollution
vulnerability of the Kanye wellfield in SE
Botswana
B.F.Alemaw, E.M.Shemang & T.R.Chaoka
Department of Geology, University of Botswana, Gaborone, Botswana
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
showing evidence of similar pollution. These include the Molepolole wellfield in the
central part of the country and the Serowe and Palapye wellfields in the northeast. There
is therefore the urgent need to ensure that the groundwater resources are of drinking
quality, by protecting them against pollution sources such as landfills, pit latrines,
industries, agricultural pesticides and waste.
The vulnerability potential of an aquifer to pollution depends to a large extent on the
susceptibility of its recharge area, as areas with high replenishment rate are potentially
more vulnerable to pollution than others (Bekesi and McConchie, 2002). Unconfined
aquifers that do not have a cover of impermeable material are also highly susceptible to
contamination. Soils overlying the water table provide the primary protection against
ground water pollution.
The objective of the study is to investigate the potential to pollution vulnerability of
Kanye wellfield based on two methods vis-à-vis soil and geology-based vulnerability
mapping (SGV method) and the DRASTIC method.
The present study was undertaken on the Kanye wellfield (located between longitudes
25°–25.4°E and latitudes 24.8°–25.5°S Fig. 1), south-eastern Botswana with the aim of
assessing the vulnerability of this well field to pollution. The Kanye wellfield comprises
three sub-areas where groundwater is abstracted by the Water Utilities Corporation:
Northwest, Kgwakgwe and Rammonedi. The main potential source of pollution of this
well field is agricultural activities (mainly livestock farming) and domestic waste
disposal.
The aquifers in the Kanye wellfield consist essentially of carbonate rocks (dolomites)
that are highly fractured and fissured at some places. Karst phenomenon is also prevalent
within these aquifers.
3 METHODOLOGY
Two methods have been employed to assess the vulnerability to pollution of the kanye
aquifer system. The first method is geology-based vulnerability mapping (SGV method).
The DRASTIC method, which considers other factors in addition to soil and geology,
was also used to compare the various vulnerability classes.
borehole log information at 82 boreholes was used to retrieve and spatially portray the
various vulnerability classes in the well field by means of Theissen polygons.
One of the most widely used groundwater vulnerability mapping method is the
DRASTIC method. Developed by the United States Environmental Protection Agency
(EPA) as a method for assessing groundwater pollution potential (Aller et al., 1987).
The DRASTIC index, designated by Di is calculated as:
Di=DrDw+RrRw+ArAw+SrSw+TrTw+IrIw+CrCw
where r and w refer to the DRASTIC ratings and DRASTIC weightings assigned to each
of the following hydrogeological settings:
D=depth to water table
R=aquifer recharge
A=aquifer media
S=soil media
Water resources of arid areas 722
T=topography (slope)
I=vadose zone
C=hydraulic conductivity
The DRASTIC score Di is a relative value, with no specific units. A speed sheet
program was written to derive the DRASTIC indices for the various boreholes studied.
The following table shows the DRASTIC ratings and weights adopted in the study.
The results indicate that about 47% of the Kanye wellfield area is overlain by soils with
high vulnerability to pollution (class 1 soils) Fig. 2a; 44% of the area is overlain by soils
of intermediate vulnerability (class 2 soils) and 9% of the area is overlain by soils of low
vulnerability (class 3 soils). The geological vulnerability maps (Fig. 2b) show that 40%
of the well field area is underlain by rocks with high permeability and significant
fracturing thus of high vulnerability (class 1 rocks), 49% of the area is underlain by rocks
of medium vulnerability (class 2 rocks) and 11% of the area is underlain by rocks of low
vulnerability (class 3 rocks).
Considering the overall study area bounded between 25–25.4°E and 22.8–23.5°S, the
ground water vulnerability map (Fig. 2c) indicates that 22% of the well field area is very
highly vulnerable to pollution, 35% of the area is highly vulnerable to pollution, 34% is
moderately vulnerable to pollution, 4% has a low vulnerability and 5% of the wellfield
area has a very low vulnerability. The above percentage areas portraying the above levels
of vulnerability were investigated in terms of the DRASTIC coefficients. Drastic
coefficients vary in the range of 110 to 190.
The results of this study indicate that the Kanye wellfield may be highly vulnerable to
pollution if not properly managed. However, one should be cautious in the interpretation
of these results due to the sparse and uncertain nature of the basic data used for this
interpretation, and the inherent variability of the aquifer parameters.
The use of the polynomial estimator enabled consideration of simple uncertainty for
the aquifer media characteristics modelling process, uncertainty being proportional to the
size of the polygons. Estimates of the aquifer media characteristics in areas with only a
few observations or few geological logs were considered to have higher uncertainty than
areas with many observations.
In general, there is a high uncertainty of the groundwater pollution vulnerability
classes in the Kanye aquifer system. The Northwest and Rammonedi wellfields have high
uncertainty in the classes, while Kgwakgwe wellfield portrays a low level of uncertainty.
From the aquifer protection and management point of view, the question is how far
one should establish a protection zone around the wellfield so as to protect the entire
aquifer from pollution. This needs more study and field checking of the aquifer media
characteristics maps. The results presented here are preliminary findings on the use of
spatially distributed point data to assess the regional variation of aquifer media
characteristics and vulnerability to groundwater pollution.
Water resources of arid areas 724
The present study shows how vulnerable the Kanye wellfield is to pollution. It
therefore underscores the need enforcement of the recommendations by the Department
of Water Affairs (1993) to establish groundwater protection zones over the most
vulnerable aquifer.
REFERENCES
Aller, L.T., Bennett, J.H. and Hackett, G., 1987.“DRASTIC: a standardized system for Evaluating
Groundwater pollution Potential using Hydrogeoloic setting”. US Environmental Protection
Agency Report EPA/600/2–87/035, 622 pp.
Beekman, H.E., Gieske, A. and Selaolo, E.T., 1996. GRES: Groundwater Recharge Studies in
Botswana, 1987–1996. Botswana Journal of Earth Sciences 1:1–17.
Bekesi, G. and McConchie, J., 2002. The use of aquifer media characteristics to model
vulnerability to contamination, Manawatu Region, New Zealand. Hydrogeology Journal,
10:322–331.
DWA, Department of Water Affairs, 1993. Protection Zones and Guidelines for Major Wellfield,
Aquifers and Dams in Botswana. Report prepared by Water Surveys, Botswana.
Alemaw, B.F., Shemang, E.M. and Chaoka, T.R., 2004. Assessment of Groundwater Pollution
Vulnerability of the Kanye Wellfield, Southeastern Botswana—A GIS Approach, B. Journal of
Physics and Chemistry of the Earth, Elsevier Science Publishers (accepted for publication).
The effect of socio-economic activities on
watershed management: the case study of
Gaborone Dam catchment in Botswana
George S.Thabeng & Daniel B.Kemiso
Department of Water Affairs, Gaborone
Water Resources of Arid Areas—Stephenson,
Shemang & Chaoka (eds),
© 2004 Taylor & Francis Group, London, ISBN
04 1535 913 9
1 INTRODUCTION
Gaborone Dam was first built in 1963 to supply water to the urban centre of Gaborone
when demand for water by the town exceeded the supply from the older Notwane Dam
and from boreholes that had been drilled in wellfields in the vicinity of the City. As the
City of Gaborone continued to grow and, with it, the demand for water by the City, the
Gaborone Dam wall was raised by 8 metres to its present height in 1986. The main
technical characteristics of the Dam are as follows:
Catchment Area: 14,300km2
CapaCity of Dam at f.s.l: 141.4×106m3
Type of Embankment: Rock Earthfill
Crest Length: 3km
Maximum Surface Area: 1900ha (19km2)
Maximum Depth: 20 metres
Average Depth of Water at 7 metres
f.s.l:
Potential Evaporation: 2000mm per
year
The catchment area covers the southern portion of southeastern Botswana, and also falls
within the neighbouring country of South Africa (Figure 1). It falls within 3 different
districts and apart
The effect of socio-economic activities on watershed management 727
2.2 Climate
The climate is generally hot with a slight drop in the temperature during the winter
months of May to August. Mean temperatures range from 12°C (July) to 28°C (January).
Considerable local variations occur so that the winter temperatures in Lobatse are lower
than in Kanye or Mogobane. Winter frost is common in places like Lobatse, Kanye and
surrounding areas.
Annual average rainfall (1925 to 1997) is 509.6mm in Kanye, and 555mm in Lobatse.
Maximum annual totals recorded in Kanye and Lobatse are 970.6mm and 1070.4mm
compared to minimum rainfalls of 104.7 and 262.1 respectively.
2.3.1 Geology
The geology of the area is complex. The catchment overlies part of the Kaapvaal Craton,
an ancient stable segment of acidic crustal rocks. These are Archaean metamorphics,
profoundly altered by heat and pressure, as well as igneous rocks. These form the floor
above which Proterozoic and younger non-metamorphic sedimentary and volcanic rocks
occur.
Of the supercrustal rocks, the Lobatse Volcanic Group is the oldest. They now form
the arc of rugged relief that extends south of Gopane to Lobaste and northwards towards
Mogobane before striking northeast to the north of Ramotswa station. The next, younger
rocks belong to the Transvaal Supergroup, one of the major early Proterozoic
sedimentary successions, some 5000m thick. The youngest of the supercrustal rock
sequences is the Waterberg SuperGroup, the orthoquartite member (Mannyelanong
Formation, 17,000–22,000 million years) being responsible for the flat cappings on the
interfluves of the project area.
Much of the catchment, including the Gaborone Dam, is underlain by the Gaborone
Granite (1600–4000m.y).
There are also other intrusive rocks in the area, which may be unrelated to the
Gaborone granite. These are the Kgoro Complex and Mmathethe Granite (1600–
4000m.y), also dolerite sills and sheets (e.g intruding the granite at Kgale Hill) of late
Proterozoic age.
Younger Palaezoic and Mesozoic rocks are completely absent from this area. Minor
calcretes occur in the soils above the volcanics of the Transvaal Supergroup and minor
pedogenic laterites occur in flat locations over the granites.
The tectonic history of these ancient rocks is complex and many of the streams of the
catchment are strongly controlled by the structures.
Unlike in other areas of Transvaal Supergroup rocks, the economic geology of this
area has proved very disappointing, despite substantial exploration. The only metal,
which has been mined in area, is manganese at Ramotswa, Otse and Kgwakgwa at
Kanye. Occurrences are, however limited in the area and thickness. All mining has now
ceased, that at Ramotswa in 1958 and at Otse in 1996. There are no foreseeable prospects
of pollution of the Gaborone Dam water by run off from metalworking. As concerns
The effect of socio-economic activities on watershed management 729
industrial rocks and minerals, clays, river sands and gravels have been and still are
worked for brick making and aggregate.
2.4 Geomorphology
2.4.1 Weathering
The catchment comprises surfaces, plains and bottomlands, of predominantly Post-
African Surface age. The African erosion Surface was the first to form when
Gondwanaland split up into the continents of Africa, South America, India, Australia and
Antarctica. It is known to have been deeply weathered and weathering was advanced in
the sense that primary minerals were converted first to 2:1 clays and in the upper parts of
the profile where weathering was more aggressive the 2:1 clays were converted to 1:1
clays, predominantly kaolinite. This weathering involved severe leaching, so that
virtually all of the bases, essential nutrients, were removed. Very small areas of African
erosion surface survive on the interfluves as small mesas or plateaux, particularly where
Waterberg quartzites form cappings. The deep weathering profiles have been stripped off,
leaving an interestingly etched rock surface. A good example is the plateau on which
Kanye sits. The plateau is deeply penetrated by fossil gorges, terminating in huge dead
waterfalls, which clearly attest to formerly much wetter conditions. The fasts of such
plateaux, only penetrated with difficulty, provide sufficient isolation for the survival of
kudu and leopard. The sharp drop to the Post-African Surface plains, provides the
perimeters of such plateaux with many excellent, but rarely visited, view points offering
wide vistas over the inselberg-dotted lowlands.
Continental uplift and erosion resulted in the formation of the Post-African Surface,
which dominates the project area. Weathering is much less advanced, although deep in
places. Clay formation is poor. Typically, weathering has only succeeded in separating
the individual mineral components of the rock, leaving them in situ as a disaggregated
mass. Such poor weathering yields saprolite (weathered rock) with very little cohesion. It
is very susceptible to erosion. The poor development of clays results in poor, thin sandy
soils, particularly where the rocks are acidic, e.g. in granitic areas. Here the soils are
sandy, dominated by quartz and orthoclase feldspar particles, the most resistant to
weathering.
The small, isolated Hills, which dot the plains are the exposed parts of the basal
surface of weathering (the boundary between weathered and fresh rock). The relative
relief on the basal surface of weathering is high in granitic areas; the boundary plunges
up and down very abruptly, controlled by the frequency of fracturing. Small outcrops
form koppies, larger outcrops inselbergs. These shed rainwater so that weathering is
facilitated around their bases.
2.5 Soils
Soils in the catchment area are generally of the sandy loams, clays and sandy clay loam
types. Good fertile soils are found at depositional areas such flood plains and depressions.
In the Gaborone area, soils on alluvial deposits are developed in Notwane and
Water resources of arid areas 730
Segoditshane flood plains where ephemeral rivers bring sand from upstream. These soils
are suitable for trees.
2.6.1 Wetlands
There is a large wetland in the valley of the Taung river, near the village of Mogobane.
This wetland is at least 750ha in extent and it currently supports dense wetland vegetation
including fragmitis. As will be discussed later, this wetland presents an opportunity for
pollution control for Gaborone Dam. All water from the western part of the catchment,
i.e. the areas of Khanye, Ranaka, Magolhwane, Ntlhantlhe, and Lotlhakane; will pass
through this wetland before entering Gaborone Dam.
If properly maintained, the Mogobane wetland will filter and improve the chemical
quality all the water from the Taung river and its tributaries before it enters Gaborone
Dam.
2.6.2 Aquifers/wellfields
The main aquifers in southeastern Botswana are well represented in the project area in the
form of two major aquifers, both of which have been targets for exploration. These are
the dolomitic aquifers in Ramotswa-Lobatse and Kanye.
2.7 Ecology
2.7.1 Vegetation
The Gaborone Dam catchment area comprises a diverse array of vegetation types.
Extensive areas of the catchment comprise typically flat sandveld savanna which has
been heavily cultivated and is covered by a high proportion of cleared fields, actively
growing crops and abandoned fields. The dominant vegetation types in this area are
The effect of socio-economic activities on watershed management 731
acacia dominated bush savannas and tree savannas which range in density from low
density areas close to settlements to high density areas in less settled and cultivated areas.
Rocky Hills and kopjes occur throughout the catchment area and are generally covered
by medium density woodland and bush which shows a high degree of variation in species
composition depending on local soil conditions. The rocky outcrops and surrounding
Hilly areas are not cultivated and have few settlements, although the runoff from their
slopes, following intense rainstorms, contributes significantly to soil erosion on the
heavily grazed and cultivated lowlands.
Apart from the rocky Hillslopes, considerable concern also surrounds the integrity of
the riparian fringes, which experience heavy grazing by domestic stock.
Vegetation along the river channels such as Notwane in Gaborone and Peleng in
Lobatse exhibit a different shrub and tree cover, which is different from the accustomed
savanna type. In view of its uniqueness, the Department of Museum and National
Archives has planned to preserve a 2km length and 100m-wide corridor along the
Notwane River.
2.7.3 Crocodiles
The occurrence of crocodiles in the Gaborone and Notwane Dams also deserves explicit
attention. The Nile Crocodile has in general shown a drastic and well documented decline
worldwide, and while not yet endangered, has slipped to a “vulnerable status” (IUCN,
1982). Botswana’s population is concentrated in the Delta, with considerable debate
surrounding the size of the pre commercial harvesting (i.e. pre- 1957) crocodile
population in the Delta, with estimates varying from a population of 21,000 up to 66,000.
2.7.4 Fisheries
Although the fisheries potential of the majority of Dams in eastern Botswana is largely
unexploited, Gaborone Dam makes a significant contribution to both the diet and
livelihoods of people. Barbus sp. tend to be an abundant and unexploited resource in
most Dams throughout the country, probably because such time consuming activities as
cleaning and post-harvest processing are essential if the fish is not to be spoilt (Nermark
and Mmopelwa, 1994). Good-sized bream appear to be in demand from the larger towns,
although local people often prefer catfish (C. gariepinus), a species which is not only
drought resistant, but can greatly increase the fish biomass of reservoirs (Nermark and
Mmopelwa, 1994).
Water resources of arid areas 732
3 THE PROBLEM
The City of Gaborone has been expanding at a phenomenal rate throughout the past
decade and it is still continuing to grow. In a period of 30 years, the population of
Gaborone has grown by more than 10 times (Table 1).
This growth is the single most important factor underlying the phenomenal growth of
population in the catchment area as a whole. The population of the catchment area has
been growing over the years. In 1971, the catchment area’s population was 133, 211; by
1981 it had increased to 237, 247 at a rate of 5.8 per cent. The 1991 population census
indicates that the area’s
Table 1. Population statistics for Gaborone.
1971a 1981a Growth 1991b Growth 2001c Growth
rate rate rate
(1971– 1981– (1991–
1981) 1991) 2001)
17,718 59,657 12.1% 133,468 8.1% 186,007 3.3%
population increased further to 338,068 at a rate of 3.5 per cent. The recent population
census held in 2001 indicates a growth of about 3.1 percent in the catchment area. The
present population is therefore estimated at 458,370 (see Table 1). The growing urban
areas such as Gaborone, Lobatse, Ramotswa, and Kanye, have been the main nuclei for
the population pull and hence their growth. These centers have now become dormitory
towns for Gaborone, with substantial numbers of people commuting from them to work
in Gaborone daily. The highest growth rates between 1991 and 2001 are recorded for
Otse, Gaborone, and Gabane ranging between 3.3 per cent and 5.5 per cent. This is as a
result of some pull factors (migration) such as economic opportunities, social,
educational and recreational activities, which have attracted people to these centres.
The growth described above is driving a scramble for land in the catchment area. This
in turn has resulted in land use and land development problems. The problem is
exacerbated by the fact that land administration within the catchment area falls under the
jurisdiction of many authorities, with very little coordination.
As a result, the catchment area is experiencing water quality problems. A case in point
is in Ramotswa, in the South East District, where although it has abundant ground water,
it has been polluted by human waste.
More importantly, the water in Gaborone Dam on which the City depends is also
affected. The present situation regarding the chemical aspects of the water quality is that
it is excellent, falling well within BOBS standards for Class 1 water (Ideal), although
aesthetic parameters (odour and taste) have, on occasions been unacceptable. There is a
long term trend towards increase in solutes, conductivity and pH, superimposed on
seasonal patterns. The cause appears to be regional rather than local, but further research
into this is needed. Of some concern is the tentative conclusion that normal rains do not
result in very substantial washing of surface materials down into the Dam. These appear
to accumulate until above average rainfall conditions are experienced. In effect the
catchment stores surface materials, possibly for many years before they are delivered to
the Dam. This enhances the already pronounced seasonal patterns of delivery. Data on
The effect of socio-economic activities on watershed management 733
the biochemisty and biology of the Dam water are, at this point, inadequately monitored
and it is largely this aspect of water quality which could be of concern (Aqualogic, 2002)
The causes of the problem should be discernible from the foregoing. Without doubt, the
problem is the direct consequence of development in the catchment area without adequate
controls. This is the sort of problem that watershed management approach to resource
development and utilization could have helped to avoid.
In the case of the Gaborone Dam catchment area, the problem can be attributed to the
land use and land administration problems. Land ownership in the country is either
freehold or tribal (customary law). A large percentage of the study area falls under tribal
land, (ie, more than 70%) and is administered under the jurisdiction of several Land
Boards (Balete, Bakwena, Ngwaketse, Malete). Usage rights can either be granted
communally or to individuals. The majority of residential and agricultural sites are
therefore held under customary law land rights, while those for commercial, industrial
and institutional are held under common law (leasehold). Customary grants are not
registered at the Deeds registry but only at the Land Board. It thus cannot be used as
collateral security for the raising of mortgage finance unless surveyed and converted into
common law.
Freehold land on the other hand, permits management of the land to rest with the
owners in accordance with legislation such as the “Agricultural Resources Conservation
Act” (Cap 35:06). This legislation puts restrictions on the owner in terms of actual tenure,
sales and leases.
The study area reflects a range of allocation patterns in various stages of
transformation from a traditional pattern to the more recent contemporary (modern)
layouts.
The study area falls within two countries—Botswana (93%) and South Africa (7%).
The land uses in Botswana’s Section can be defined in two broad categories. The first
category relates to those predominantly rural uses situated outside the built-up areas of
the villages and towns, whilst the second are those uses of an urban nature situated within
the boundaries of the villages.
The catchment/study area is made up of the built-up areas (settlements), areas of
agriculture (both cultivation/cropping and grazing), waterbodies/Dams as well as other
tourism-related areas and key public facilities.
Agriculture is the second largest sector of Botswana’s economy and is the backbone of
the rural economy, which supports two-thirds of the national population. Spatially the
cropping and grazing (agriculture) component together, makes up the largest land use
within the catchment/study area, with grazing making up the majority of the agricultural
land, eg Rankoromane Farm near Otse.
The second largest land use is the built-up area—which is all the major villages and
towns, as well as all the smaller villages like Ranaka and Ntlhantlhe.
The waterbodies/Dams land use category, occupy temporarily or permanently some
parts of the study area. A very serious problem is caused by the occasional flooding of
Water resources of arid areas 734
the settlements within the study area. Most of the flooded areas are located in settlements
eg Ramotswa.
In addition to the above land uses, is the game/nature reserve/parks component, most
of which are privately-owned, and related to tourism (eg Mmokolodi, St Clairs Lion Park,
Manyelanong Game Reserve, etc). This land use is concentrated to the north and east of
the study area, and is one of the smallest land uses within the catchment area.
All land within and around the Dam (Forest Hill 9-KO) itself is freehold land. The
land use within this area can be categorised into the following uses: industrial, quarrying,
residential, commercial, recreational, social and public uses, agricultural and
infrastructure, and fall into an area defined as a “greenbelt” and form part of a Regional
Park EHES (2002).
This portion of the catchment area falls within the jurisdiction of Lehurutshe Council,
of the Central District Municipality of the North West Province of South Africa. The land
use in this area is a mix of rural/tribal settlements, characterised by low density, scattered
homesteads, and privately-owned game farms and/or nature reserves. (A portion of the
Madikwe Game Reserve falls within this catchment area.) There are no major public or
civic institutions in this portion of the study area, and visually looks like the central and
south of the catchment area falling within Botswana.
Added to the above is the industrial development taking place in the catchment area.
The major economic activity in the catchment area is services, which employs on the
average about 45% of the labour force. This is followed by industry, which accounts for
about 35% of the labour force with construction employing most people in the industry.
Agriculture accounts for about 9% in the catchment area, but however employs more
people in the rural areas. Commerce, which entails wholesaling and retailing employs
about 11%.
Economic activities are concentrated in the Ramotswa area. Over two-thirds of the
economic activities (68.1%) are located in and around Ramotswa, and reflects
Ramotswa’s proximity to Gaborone; Ramotswa has developed a nucleus of activities in
the sectors of grain milling, clothing and furniture industries along with “common”
activities such as brick moulding and metal works;
● The most common economic activities are brick moulding, metal works, clothing and
furniture production.
● Six tanneries are located in Ramotswa and Kanye.
● Nine scrap yards exist evenly spread over the project area.
Livestock production, especially range cattle and goat production, is the main form of
agricultural activity in the Gaborone Dam catchment.
4.1.3 Dairying
There are dairy operations in some of the commercial farms in the eastern part of the
catchment area. In the rural sector of the catchment, there are two herds, one at
Magobane and the other at Kanye. The former dairy herd had 160 cows in 2001, 77 of
them in milk and producing 1463 litres per day. The Kanye dairy comprised 103 cows in
2001, 50 of them in milk and producing 640 litres of milk per day.
From an environmental point of view, dairy operations can be a source of water
pollution when animal wastes wash into public streams and lakes. The Magobane dairy is
the largest dairy close to Gaborone Dam and it is situated on the banks of the Taung
River. Fortunately, however, the dairy is adjacent to and upstream of the Magobane
wetland. Thus, any wastes that are released from the dairy into the Taung River are likely
to be trapped by the wetland, and are therefore unlikely to have a significant direct
pollution impact on Gaborone Dam.
4.1.4 Poultry
Lately there has been an upsurge in small-scale commercial poultry production in all the
villages. Small groups of farmers in most of the villages of the catchment have
constructed chicken houses that accommodate between 500 and 6000 chickens at any one
time.
Chicken manure is rich in nitrogen and therefore has a high potential for polluting
water bodies if directly released therein. It was further observed that the manure or waste
from these chicken operations is not carefully stored or disposed of. There is a risk of
some of this manure being washed directly into the Notwane River and into the Dam.
Thus in summary, the problem of pollution in the Gaborone catchment area may be
attributed to:
● Land administration problems: different authorities administering land in the same
catchment.
● Lack of coordination among the various authorities.
● Land use under different sovereignties—RSA and Botswana.
● Industrial development within the catchment area.
5 ACTION
In order to address the problem, a number of actions have been initiated within the
paradigm of watershed management. These actions are:
(i) Revision and coordination of land use plans. This study is currently underway.
Water resources of arid areas 736
6 CONCLUSION
Clearly, environmental protection of the Gaborone Dam goes far beyond the protection of
the immediate environment. Whole catchment management is increasingly recognized as
essential to the protection of any one component within the system. To achieve this, the
functioning of the system must be clearly understood. This can be best done by adopting
the paradigm of watershed management.
REFERENCES
DWA (2002), Environmental impact assessment consultancy for the feasibility study of small to
medium Dams in eastern Botswana, Aqualogic Pty Ltd. Final report.
Department of Town and Regional Planning (1995). Gaborone Landscape Masterplan 100 pp.
EHES (PTY) LTD (2002). A strategic environmental assessment for the Gaborone Dam catchment
area.
IUCN (1982) Amphibia-Reptilia Red Data Book. Part I. Testudines, Crocodilia, Rhinchocephalia.
426 pp.
Nermark, U.P. and Mmopelwa, T.G. (1994). Utilization of small water bodies, Botswana: report of
activities towards fisheries exploitation, 1992–1993. Harare (Zimbabwe). 36 pp.
Balachandran, K.K. 67
Bauer, P. 03
Bean, J. 73
Bennie, A.T.P. 397
Böttcher, J. 379
Brunner, P. 03
Davies, J. 87
Dennis, I. 73, 79, 115, 429
Dennis, S.R. 297
Diiwu, J.Y. 347
Drury, S.A. 59
du Preez, M. 297
Duijnisveld, W.H.M. 379
Ellington, R.G. 47
Engelbrecht, P. 143
Gumiremhete, R. 315
Author index 738
Hassan, M. 467
Himmelsbach, T. 379
Hötzl, H. 379
Hodgson, F.D.I. 363
Houghton-Carr, H.A. 507
Hussin, Y.A. 239, 341
Keeletsang, M. 541
Kellner, T. 315
Kemiso, D.B. 565
Kheiralla, K.M. 247, 259
Kinzelbach, W. 03
Kistamah, N. 529
Kooke, S.O. 35
Kumar, A. 59
Kumar, H. 323
Lubczynski, M. 239
Lubczynski, M.W. 97, 271, 285, 341, 541
Machingambi, M. 181
Machiridza, R. 181
Mafa, B. 133
Magombedze, L.M. 97
Magowe, M. 29
Makobo, P. 29
Makuya, M. 519
Manzungu, E. 181
Mapanda, W. 541
McCartney, M. 493
Mishra, G.C. 105
Mkwizu, Y.B. 157
Molwalefhe, L. 459
Molwalefhe, L.N. 409
Mpala, T. 201
Msangi, J.P. 207
Mudzingwa, B. 315
Mugabe, F.T. 501
Mulwa, J.K. 303
Muyima, N.Y.O. 535
Mvungi, A. 519
Paimpillil, J.S. 67
Pironcheva, G. 535
Pretorius, J. 79
Sally, H. 493
Sbeih, M.Y. 191
Scheuerlein, H. 435
Schwiede, M. 379
Selemani, M. 515
Senzanje, A. 493, 501
Sharma, T.C. 331
Shemang, E.M. 279, 323, 551, 559
Siegfried, T. 03
Sitters, C.W.M. 331
Stadler, S. 379
Staudt, M. 371
Stephenson, D. 23, 419, 451
Tesfaslasie, F. 59
Thabeng, G.S. 565
Totolo, O. 551
Tredoux, G. 143
Trifunovic, N. 175
Türkman, A. 477
Yilmaz, L. 151
Zamxaka, M. 535
Ziwa, C. 541