Urban Climate 16 (2016) 2542

Contents lists available at ScienceDirect

Urban Climate
journal homepage: www.elsevier.com/locate/uclim

Modeling the urban geometry influence on outdoor thermal

comfort in the case of Moroccan microclimate
Alaoui Sosse Jihad , Mohamed Tahiri
Dept. of Process Engineering, Mohammadia School of Engineers, Mohammed V University, Agdal, Rabat 10000, Morocco

a r t i c l e i n f o a b s t r a c t

Article history: Thermal indoor conditions depend not only on the building envelope but also on the ther-
Received 27 June 2015 mal outdoor conditions. It is therefore important that a great interest should be paid to out-
Revised 9 December 2015 door thermal conditions in the future Moroccan urban design. In the present work we
Accepted 17 February 2016
investigate the influence of the aspect ratio (AR), on both the heat fluxes of a canyon
and the thermal comfort of pedestrians for three Moroccan climatic zones (Agadir,
Errachidia and Fez). At first the Town Energy Balance model coupled with a turbulence clo-
Keywords:
sure model is evaluated using experimental data. The model is then used for determining
Outdoor thermal comfort
Town Energy Balance
the influence of heat fluxes generated by the canyon surfaces on the predicted mean vote
Aspect ratio (PMV) chosen as a thermal comfort indicator. The simulations results obtained using actual
High thermal capacity Moroccan envelope building characteristics show that in the three climatic zones it is best
to promote a medium urban structure to insure a comfortable street for the two cold and
hot seasons. The AR must be lower than 1.2 for Errachidia, between 2.5 and 3.4 for Agadir
and between 1.2 and 1.9 for Fez. However, canyons having higher AR can also insure a com-
fortable climate using historical thick walls with high thermal capacity.
2016 Elsevier B.V. All rights reserved.

1. Introduction

During the last two decades, Morocco is undergoing an important economic and social development and concomitantly
an increase in energy consumption estimated to a yearly rate of 7% for the coming years. Part of this energy consumption
results from a growing demand for HVAC in buildings destined for various uses in expanding urban centers. Outdoor thermal
conditions have a great influence on indoor thermal comfort and consequently on energy demand for HVAC. On the other
hand outdoor thermal comfort is becoming a growing concern in its own right, especially in urban areas because of UHI
effect. In Shanghai for example an intensity of UHI of 6.5 C was measured (IPCC, 1990), therefore during urban planning care
should be taken through designs that mitigate UHI effects to promote comfortable indoor and outdoor environments with
minimum energy requirements.
Urban geometry is henceforth of considerable importance in city planning and design for thermal comfort. It can be ascer-
tained by adjusting the fundamental morphological unit of urban geometry, namely the aspect ratio (AR), defined as the ratio
of the canyon height to canyon width AR = H/W. This ratio is an important parameter that can be used to investigate the
influence of urban geometry on outdoor environments, in terms of thermal comfort and external temperature increase. Sev-
eral investigations were conducted to study the impact of the AR of the street on the outdoor thermal comfort. In Ghardaia,
Algeria, deep streets were more appropriate to insure a good outdoor thermal comfort (Ali-Toudert and Mayer, 2006). In

Corresponding author.
E-mail addresses: sys.energie@gmail.com (A.S. Jihad), tahiri@emi.ac.ma (M. Tahiri).

http://dx.doi.org/10.1016/j.uclim.2016.02.002
2212-0955/ 2016 Elsevier B.V. All rights reserved.
26 A.S. Jihad, M. Tahiri / Urban Climate 16 (2016) 2542

Nomenclature

S net solar radiation budget for (wall, road or roof) (W/m2)

L net infrared radiation budget for (wall, road or roof) (W/m2)
Satm downward direct solar radiation on an horizontal surface (W/m2)
D downward diffuse solar radiation on an horizontal surface (W/m2)
L# downward infrared radiation on an horizontal surface (W/m2)
He sensible heat flux for urban envelopes (W/m2)
Le latent heat flux for (wall, road or roof) (W/m2)
G conduction heat flux for (wall, road or roof) (W/m2)
C heat capacity of the (wall, road or roof) (J m3 K1)
Ca heat capacity of dry air (J m3 K1)
qa the air density (kg/m3)
d thickness of the (wall, road or roof) (m)
kk thermal conductivity of the kth layer of the (roof, road or wall) (W/m K)
k the solar zenith angle (rad)
k0 the zenith angle for which the sun begins to illuminate the road (rad)
h the angle between the sun direction and the canyon axis (rad)
h0 the critical canyon orientation for which the road is no longer in the light (rad)
r the StefanBoltzmann constant (W/m2/K4)
e emissivity of the (wall, road or roof) ()
T wall, road or roof temperature (K)
Tw sunny wall temperature (K)
Tsw shaded wall temperature (K)
Tr road temperature (K)
TR roof temperature (K)
Ti the building internal temperature (K)
Tatm the air temperature at the top building level (K)
Tcan average temperature of the air canyon (K)
qcan the average humidity of the air canyon (kg/kg)
qsat the humidity of a saturated humid air (kg/kg)
Lv the latent heat of vaporization (J/kg))
Ta the air temperature (K)
h sun height (rad)
I solar constant (W/m2)
In the direct solar radiation at normal incidence (W/m2)
TL the trouble of Linke ()
X horizontal coordinate of space (m)
Y vertical coordinate of space (m)
U can average horizontal wind speeds of the street canyon (m/s)
W can average vertical wind speeds of the street canyon (m/s)
Re Reynolds number ()
u air velocity in the direction of the x coordinate (m/s)
v air velocity in the direction of the y coordinate (m/s)
Vc the canyon air velocity (m/s)
a albedo of (wall, road or roof) ()
P pressure (Pa)
pa the vapor pressure of the surrounding air (Pa)
t time (s)
H average building height (m)
W street width (m)
M the air mass ()
Me the metabolic rate (W/m2)
Li the thermal load for a person at an activity level (W/m2)
PMV predicted mean vote ()
f cl the ratio of the surface of the clothed body to the surface area of the hide body ()
hc the convective heat transfer coefficient (W/m2 K)
Icl the clothing insulation (m2 K/W)
Rcl the thermal resistance of clothing (m2 K/W)
t cl the average surface temperature of clothing (K)
tr the mean radiant temperature (K)
wp the external work done by the pedestrian in the street (W/m2)
 A.S. Jihad, M. Tahiri / Urban Climate 16 (2016) 2542 27

Abbreviations
TEB town energy balance
AR aspect ratio
HVAC heating ventilation and air conditioning
UHI urban heat island

Colombo, Sri Lanka, a higher AR ratio has a positive effect in increasing outdoor thermal comfort (Perera and Weerasekara,
2014) and in Cairo, Egypt, a medium value of the AR is better for improving thermal comfort (Fahmy and Sharples, 2008).
Several other investigations were conducted to study the impact of the AR on the daily solar gain, explaining that miti-
gation measures must focus on the canyon geometry with AR greater than 1.5 (Takebayashi and Moriyama, 2012). The dis-
persion of the pollutant is favorable for the canyon when the AR is in the wake interference regime of flow (0.5 < AR < 0.1)
(Chan et al., 2001). The ventilation efficiency of the canyon becomes largest in the wake interference regime with
(0.3 < AR < 0.65) (Barlow and Belcher, 2002). For the case of Morocco in order to insure thermal comfort very deep canyons
is preferable by including some wider streets or open spaces or both to provide solar access in winter period (Johansson,
2006). But there is no exact AR value included to insure a good evacuation of heat fluxes generated by the building envelope
on the outside of the canyon during the summer period and allow these flows to be trapped inside the canyon during the
winter period.
The purpose of the present work is at first the evaluation of an urban climate model in the three climatic zones of Mor-
occo. The model used is the AROME model (Seity et al., 2011) coupled with a misoscale urban canopy model, which is the
Town Energy Balance model based on Masson 2000 urban parameterization (Masson, 2000), including the surface energy
budget and canopy layer variables. The TEB model is the same as developed by Masson (2000) in most of its features but
differs in a few important aspects. In particular, the street canyon walls are treated separately, which leads to a different
budget of radiation within the canyon walls and provides a better estimation of the level of comfort in the street. The pro-
gram is coded by MATLAB software (MATLAB, 2010) and the model results are evaluated using the measured data of external
air temperature made in Fez, Morocco, by Johansson (2006).
The model is then used to investigate the effect of a various values of the AR on the external thermal comfort expressed in
terms of the PMV factor as developed by Fanger (1967, 1970), and on the heat fluxes generated by the components of the

Fig. 1. Climatic zoning of Morocco adapted to the needs of the thermal regulation in residential building (ADEREE).
28 A.S. Jihad, M. Tahiri / Urban Climate 16 (2016) 2542

urban street canyon (walls, roof and road). Two types of building envelope are investigated, the ordinary brick walls and the
historical thick walls like used in the old cities of the three climatic zones. Fig. 1 presents the three climatic zones of Morocco
Agadir, Errachidia and Fez, where the study is conducted. The three areas are selected because of the similarity of their cli-
mate during the extreme climatic periods of the year. The area of Fez and like the two others is characterized by a climate
marked by a large variation in temperature, summer is very hot where the maximum temperature can exceed 46 C and the
winter is very cold where the temperature can reach 5 C.
The present paper is divided in four sections: In the first section a short reminder of the studies conducted in relation to
the urban geometry and some aspect regarding the study area (Morocco). Secondly the model used is presented, evaluated
and all hypothesis and equations are given. Thirdly it will be shown how the AR recommended can insure a thermal comfort
in the outdoor conditions for two kinds of building envelope, and finally the fourth section has concluded about the impor-
tant results of this study, limitations of the work, possible uncertainties and future prospects of the study.

2. Model

The urban canopy is represented in a non-dimensional model such as the TEB model (Masson, 2000) by an array of infi-
nite street canyons and coupled with the atmospheric turbulence closure model AROME implemented by Seity et al. (2011)
and used by the Meteorological Agency of Morocco with a resolution of 0.025 (2.5 km). The turbulence closure model is a 1D
prognostic-turbulence scheme on 6 vertical levels inserted between the surface and the lowest atmospheric model level, it
provides wind speed, temperature and humidity at a reference height, about 10 m above the ground level, which represent
the height of building used in the model evaluation. The coupling of the TEB model with the turbulence closure model is
carried out via the temperature, wind and humidity at the top of the buildings in the study cases. The TEB computes the
energy fluxes in a similar manner as described by Masson (2000). However, the urban climate model described in this paper
differs from TEB in one important aspect, which is the addition of the temperature of the shaded wall Tsw (K) by treating the
street canyon walls separately, which leads to a different budget of radiation, in order to take into consideration the temper-
ature of the non-sunny wall as depicted in Fig. 2. The shaded wall affects considerably the profile of temperature inside the
canyon and therefore it will greatly affect the comfort in the street canyon. The TEB model is used to determine the heat
fluxes through the surfaces of the street canyon. These are the sensible heat flux, the latent heat flux, the heat flux by con-
duction through the layers of their envelope, the net solar radiation flux and the net infrared radiation flux.
The TEB model hypothesis are assumed according to which all road orientations and wind directions are possible with an
equal probability (Masson, 2000). This allows calculation of the average fluxes for the road, roof and walls. On the other hand,
in addition, we assume the external surface of the walls to be dry so as to neglect the latent heat flux. For the longwave

Fig. 2. The components of the street canyon used by the TEB model (Masson, 2000) and the solar radiation received in a canyon perpendicular to the sun
direction: the solar zenith angle k (rad), the zenith angle for which the sun begins to illuminate the road k0 (rad), the average building height H (m) and the
street width W (m).
 A.S. Jihad, M. Tahiri / Urban Climate 16 (2016) 2542 29

radiation flux L (W/m2) just one re-emission is taken into account (Johnson et al., 1991). Massons formula (Masson, 2000) is
used to compute the average total solar radiation flux S (W/m2) absorbed by each surface type and the sky view factors are
calculated for a canyon assumed to have an infinite length (Masson, 2000).
The heat budget for the wall, roof, and road surfaces discretized into three layers, is written as follows (Masson, 2000):
@T 1
C  S L  He  Le  G 1
@t d
where T (K) is the temperature of the considered surface (road, roof or wall), C (J m3 K1) represents the heat capacity and d
(m) the layer thickness with which the surfaces is composed, the fluxes S, L, He , Le and G (W/m2) denotes net solar radiation,
net infrared radiation, sensible heat flux, latent heat flux, and conduction heat flux between the surface layer and the under-
lying layer. The other layer temperatures evolve according to a simple heat conduction equation. For the kth layer:
@T k 1
Ck   Gk1;k  Gk;k1 2
@t d
where C k (J m3 K1) is the heat capacity of the kth layer, the version presented here uses three layers to describe the heat
conduction into the road, walls and the roof. Due to large temperature gradients that can exist, and because of the multi-
layer structure of the walls or the roofs, three layers are used to represent each surface, similar to Masson 2000 solid surface
composition:
T k  T k1
C k;k1 kk;k1  1 3
2
 dk dk1
With
dk dk1
kk;k1 4
dk =kk dk1 =kk1
where kk (W/m K) is the thermal conductivity of the kth roof, road or walls layer.
The lower boundary conditions for the roof and walls are given by the building internal temperature T i (K), assumed to be
constant by using the HVAC building systems (equal to 26 C and 19 C in summer and winter respectively) and that for the
road being represented as a zero flux lower boundary. The fluxes between the nth layer (the inner layer) and the underlying
material are:
TR  Ti
GRn;n1 kn  1 n 5
2
 dRn

Tw  Ti
Gwn;n1 kn  1 n 6
2
 dwn

Grn;n1 0 7

where n is the number of layers, the sky-view factors are computed for the TEB geometry (an infinite canyon) according to
Noilhan (1981):
"
2 #12
H H
wr 1  8
W W

 h  i12 
H 2
H
W
1 W
1
1
ww H
9
2 W

These factors represent the fraction of sky seen from the road and one wall respectively, the net longwave radiation
absorbed by the road and wall surfaces is given as (Masson, 2000):

Lr er wr L#  er rT 4r er ew 1  wr rT 4w er 1  ew 1  wr ww L# er ew 1  ew 1  wr 1  2ww rT 4w
er 1  ew 1  wr ww rer T 4r 10

Lw ew ww L# ew ww rer T 4r rer T 4w ew 1  er ww wr L# ew 1  ew ww 1  2ww L# e2w 1  ew 1  2ww 2 rT 4w

ew 1  ew ww 1  2ww rer T 4w 11

where e is the emissivity of the (wall, road or roof), r the StefanBoltzmann constant (W/m2/K4) and L# (W/m2) the infrared
radiation flux from the sky, computed using the Idso and Jackson (1969) empirical equation.
30 A.S. Jihad, M. Tahiri / Urban Climate 16 (2016) 2542

L# f1  0:261: exp7:77  104  273  Tatm2 g  rTatm4 12

where Tatm (K) is the air temperature at the top building level, the mean direct solar fluxes received by both walls and by the
road, for a street direction perpendicular to the sun are computed in the same manner as in Masson (2000):
(
 p 1 W
Satm if k > k0
Sr h 2 H
13
2 1
2
tankSatm if k < k0
(
 p 0 if k > k0
Sw h   14
2 1W
H
tankSatm if k < k0

where h (rad) is the angle between the sun direction and the canyon axis, k (rad) the solar zenith angle (from zenith), k0 (rad)
the zenith angle for which the sun begins to illuminate the road (Fig. 2) and Satm (W/m2) the direct solar radiation flux
received by a horizontal surface, calculated using the Davies et al. (1975) equation (Davies et al., 1975; Barbaro et al., 1977).
Satm I  sinh 15
2
where I (W/m ) is the solar constant and h (rad) the sun height, averaging a flux with respect to the canyon orientation is
performed with two integrations (Masson, 2000), one between h 0 and h h0 , and the other one between h h0 and h p2
and by applying corrections because the walls are treated separately. The direct solar fluxes for walls (sunny and shaded),
roads and roofs then read:


2h0 2 H
Sr Satm  tank1  cosh0 16
p pW
8 W 1 h0  1 p
< Satm H 2  p  p tank0 1  cosh0 if  2 < k < k0
>
h
Sw Satm W 1
 p0 p1 tank1  cosh0 if  k0 < k < 0 17
>
: H 2
0 if 0 < k < p2
8
> 0 if  p2 < k < 0
< W 1 h  1
Ssw Satm H 2  p0 p tank1  cosh0 if 0 < k < k0 18
>
:  W 1 h0  1
Satm H 2  p p tank0 1  cosh0 if k0 < k < p2

SR Satm 19
where h0 (rad) is the critical canyon orientation for which the road is no longer in the light, or for which the radiation is min-
imum when the sun is high enough:



W 1
h0 arcsin min ;1 20
H tank

The diffuse solar fluxes received by walls, road and by the roof are directly deduced from the sky-view factors (Masson,
2000):
Dr;w;R wr;w;R D 21
2
where D (W/m ) is the horizontal diffuse solar radiation flux for (road, roof or walls) calculated by using the Barbaro et al.
(1977) empirical equation (Barbaro et al., 1977; Idso and Jackson, 1969).
q q
D 54:8  sinh  TL  0:5  sinh 22

where TL is the trouble factor of Linke used to characterize the atmospheric turbidity (due to water vapor, mist, fumes,
dust. . .), it is calculated using the following empirical equation of Ineichen and Perez (2002):


1 In
TL ln 23
0:128  0:054  logM  M I

where In (W/m2), I (W/m2) and M are respectively the direct solar radiation at normal incidence, solar constant and the air
mass.
Finally, the heat fluxes between the canyon surfaces and the canyon air read:

C a qa T r  T can
Her 24
RES
 A.S. Jihad, M. Tahiri / Urban Climate 16 (2016) 2542 31

C a qa T w  T can
Hew 25
RES

Lv qa qsat T r  qcan
Ler 26
RES

Lew 0 27
where qa (kg/m ) is the air density at the first atmospheric level, C a (J m K ) the heat capacity of dry air, qsat (kg/kg) the
3 3 1

humidity of a saturated humid air, Lv (J/kg) the latent heat of vaporization, qcan (kg/kg) the canyon air humidity, T can (K) the
canyon air temperature and RES (s/m) the aerodynamic resistance between the canyon surfaces and the canyon air, com-
puted using the semi-empirical resistance formulation of Rowley and co-workers 1932 adopted by Masson 2000 (Rowley
et al., 1930; Rowley and Eckley, 1932).

q1
RES 11:8 4:2 U 2can W 2can 28

where U can (m/s) and W can (m/s) are the average horizontal and vertical wind speeds of the street canyon respectively.
Eq. (1) written for each surface type leads to a system of nonlinear ordinary differential equations. The solution of this
system allows computing the surface temperatures and the various heat fluxes.
A two-dimensional model is coupled with the results of the TEB model (Tw, Tsw, Tr, and TR) to have better exactitude of
the PMV indicator of comfort, to calculate the temperature in the same location in the street of the measurement materials
(3.5 m above the ground for the deep street and 6.2 m for the wide street) and to analyze the heat dispersion regimes in the
canyon, the air temperature is a function of the width (x), the height (y) and time t, Tax; y; t, based on the Navier-Stokes,
continuity and heat equations. The equations are given in the Appendix.

3. Results and discussion

3.1. Influence of AR on the external thermal comfort (PMV)

3.1.1. Model evaluation

The evaluation of the present model is performed using experimental data obtained at two sites by Johansson (2006). The
first site is a deep canyon in the old city of Fez (the Seffarine district (S)) and the second is a large canyon in the new city of
Fez (the Adarissa district (A)). The shallow and narrow streets, close to the measurement location, are shown in Fig. 3a and b
respectively. Evaluation is done using continuous data measured on a hot day (6-August, 2001), 3.5 m above the ground for
the shallow street and 6.2 m for the narrow street.
The simulations were performed on putting the model under the same climatic conditions during the day studied which
presents the hottest day in summer were the latent flux from the walls can be neglected, as assumed by the TEB model
(Masson, 2000).

Fig. 3. Shallow (a) and narrow (b) street in the city of Fez, close to the measurement location (Johansson, 2006).
32 A.S. Jihad, M. Tahiri / Urban Climate 16 (2016) 2542

Table 1
Global heat transfer coefficients of the residential building envelope for the two sites (Adarissa and Seffarine) without compliance with Moroccan thermal
regulation (ADEREE).

Sites The overall transmission coefficient of exposed roofs (W/m2 K) The overall transmission coefficient of external walls (W/m2 k)
Adarissa 2.28 1.17
Seffarine 0.55 0.65

The two chosen neighbourhoods have the same building average height (10 m) with different distance between the build-
ings, where AR = 10 for the narrow street and AR = 0.6 for the shallow street, and are well suited for a comparison of the
influence of geometry since for two different urban envelope, apart from the difference in urban geometry and envelope,
they are similar. For instance, they have no anthropogenic heat sources. The characteristics of the different envelope of
the two canyons have been respected, thick walls in the case of Seffarine (1 m of width and constructed with a raw earth
bricks) and ordinary brick walls in the case of Adarissa. Values of global heat transfer coefficients for both sites are given
in Table 1. The thermal characteristics of the building materials that are used in both ordinary and historical thick walls
are shown in Table 2.
The temperature variation of the outside air for both types of streets, with their different thermal conditions of the envel-
ope, was compared with the corresponding experimental data in the two Figs. 4 and 5.
During the measurement period, the daily maximum temperature exceeded 30 C for the site of Seffarine and 43 C for
the site of Adarissa as depicted in Figs. 4 and 5. During the afternoon, which is the hottest part of the day, the deep canyon
was cooler than the shallow one and the average difference in maximum daytime temperatures between the two canyons
was 4.4 C. The reason for the large daytime cool island of the deep canyon is related to its high AR. By day, the lower part of
the canyon is in complete shade and consequently, surfaces remain cool and the air is not warmed up. During the nocturnal
period the shallow canyon was the least warm, explained by the lower AR giving the much lower solar view factor of the
canyon and consequently, leading to lower radiative trapping, and promoting a drainage air in the street. However, it cannot
be excluded that cool air also enters the two canyons by horizontal air movements. Then we can notice the large difference
between the microclimate at the shallow and the narrow canyon. Hence the interest to study both canyon types in this
paper.
The level of uncertainty for the TEB model, represented by the root mean square error (RMSE) in C, is 0.35 (1.2% of the
average value) and 1.06 (3.2% of the average value) for the canyon air temperature of Seffarine and Adarissa respectively,
these results indicate that the model achieves the best match with the experimental data as indicated by the RMSE values
and as confirmed by the higher R2 values. Consequently, the TEB model can accurately predict the urban microclimate for the
case of Morocco with sufficient accuracy. The model will therefore be used to determine the influence of the geometry of the
streets on the pedestrian thermal comfort for the three climatic zones of Morocco during the hottest and the coldest period
of the year 2013, where their climate becomes practically similar.
In what follows, simulation results for each of the three climatic zones were obtained for the coldest period extending
from February 1 through 3 and for the hottest period extending from August 15 through 17, both for the year 2013. Data
on a three-hourly basis for the first atmospheric layer above the urban canyon were thus computed using the AROME model
used by the Meteorological Agency of Morocco. Moroccan thermal regulations set the indoor temperatures to 26 C and 19 C
for thermal comfort in summer and winter respectively. These values are used in the simulations we carried out. In order to
reach steady state conditions and to take into account the effects of heat accumulation, simulations were carried out for a
period of 3 days for the summer and winter seasons. Computations were conducted at the centroid of each of the climatic
zones. Values of global heat transfer coefficients, recommended by the Moroccan thermal regulations for residential build-
ings given in Table 3 are used, for each of the investigated climatic zones. The thermal characteristics of the building mate-
rials that are generally used in Moroccan constructions for both ordinary and historical thick walls are as shown in Table 2.

3.1.2. Air temperature

All buildings are assumed to be of a same average height so as not have any shaded surface except for the two canyon
walls. The latter has a significant influence on the temperature profile of the urban canyon. Simulations have been made

Table 2
Thermal characteristics of old and new urban constructions used in Fez climatic zone.

Envelopes Albedos () Emissivity () Heat capacity of ordinary Heat capacity of historical Thickness (m)
urban construction urban construction
(MJ m3 K1) (MJ m3 K1)
Wall 0.25 (ordinary wall) 0.85 (ordinary wall) 1 1.8 0.2 (ordinary wall)
0.15 (historical wall) 0.96 (historical wall) 1 (historical wall)
Roof 0.3 0.85 1 1 0.2
Road 0.15 0.94 2.5 2.5 Null heat conduction
 A.S. Jihad, M. Tahiri / Urban Climate 16 (2016) 2542 33

31

R=0.87
30
RMSE=0,35

Air temerature (C)

29

28

27
Modeled (S)
26
Experimental (S) [8]
25
0 12 24
Solar Time (h)

Fig. 4. TEB model evaluation: case of deep canyon, Seffarine (S) (AR = 10) during the 6-August, 2001 day (Johansson, 2006).

45
Modeled (A)
42
Experimental (A) [8]
Air temerature (C)

39

36

33
R=0.92
30 RMSE=1,06

27

24
0 12 24
Solar Time (h)

Fig. 5. TEB model evaluation: case of wide canyon, Adarissa (A) (AR = 0.6) during the 6-August, 2001 day (Johansson, 2006).

for various aspect ratios (1/6, 1/3, 2/3, 1, 3/2, 1/2, 2, 3, 6) in the two months August and February of the year 2013, results of
the external air temperature profile for the different values of AR, are presented in Fig. 6a and b for the climatic zone of Fez
and by using an ordinary envelope construction. We can notice that the AR has a noticeable influence on the profile of the
ambient air temperature as shown in Fig. 6.
Variations of the mean surface temperature for the walls, road and roof surfaces for an aspect ratio (AR = 1) are depicted
in Figs. 7 and 8 for the winter and summer periods for the climatic zone of Fez.
The results show the temperature variations to be in accordance with the variation in solar gains. The depicted data show
also that the air temperature above the buildings has a great influence on the variation of canyon surface temperature as it
appears clearly during the nighttime of February. Coherence of the results obtained can be seen through the variations of the
temperature of the two canyon walls as they alternate from sunny to shade at starting solar midday. During August the road
temperature at night is higher that the roof temperature because of the permanent entropic fluxes existing in the street. The
same observation is made at solar midday when both the road and the roof receive the same maximal solar radiation due to
the high emissivity of the road. On the other hand, during the winter period the road temperature is lower than the roof
temperature during the day, this is explained by the fact that during that period the road receives less solar radiation than
the roof, because the low position of the sun results in a lower sky factor for the road.

3.1.3. Heat fluxes from the canyon surfaces

The balance of the various heat fluxes for the canyon envelope is shown in Fig. 9a and b.
For the two seasonal periods, the following observations can be made, solar gains are the highest for the wall, because of
radiative trapping, due to longwave infrared radiation. On the other hand, the roof loses heat because of its radiative disper-
sion towards the atmosphere. The heat flux due to shortwave solar radiation is the highest for the roof because it receives
directly these radiations without a shading effect. Sensible heat fluxes are higher through the wall and road surfaces, because
of the lower ventilation effect on these surfaces in comparison with the roof surface. Radiative net fluxes by both shortwave
and longwave radiation are the largest through the surface wall as depicted in Fig. 10a and b. Therefore, the surface wall has
the greatest influence on the outside air temperature compared to both the roof and road surfaces. The surface wall should
34 A.S. Jihad, M. Tahiri / Urban Climate 16 (2016) 2542

Table 3
Recommended global heat transfer coefficients by Moroccan thermal regulation in residential building (ADEREE).

Climatic zone The overall transmission coefficient The overall transmission coefficient Coordinates of the centroid
of exposed roofs (W/m2 K) of external walls (W/m2 k)
Longitude Latitude
Agadir 0.75 1.2 8.72 31.79
Fez 0.65 0.7 5.34 33.76
Errachidia 0.55 0.6 7.16 30.03

Fig. 6. Simulation result of canyon temperature with AR varying from 1/6 to 6 for summer (a) and winter (b) period for the Fez climatic zone at midday by
using an ordinary building envelope.

Tr TR TW TSW Tatm
48
43
38
33
T [C]

28
23
18
13
8
0 12 24 36 48 60 72
Hours

Fig. 7. Temperature variations of the sunny wall (Tw), shaded wall (Tsw), roof (TR) and road (Tr) during the summer period of 2013 for the Fez climatic zone
by using an ordinary building envelope (AR = 1).
 A.S. Jihad, M. Tahiri / Urban Climate 16 (2016) 2542 35

Tr TR TW TSW Tatm
25

20

15

T [C]
10

-5
0 12 24 36 48 60 72
Hours

Fig. 8. Temperature variations of the sunny wall (Tw), shaded wall (Tsw), roof (TR) and road (Tr) during the winter period of 2013 for the Fez climatic zone
by using an ordinary building envelope (AR = 1).

be consequently the main target for mitigating the UHI effect in order to insure outdoor thermal comfort. The same results
are found for the other climatic zones and for the two seasonal periods.
Radiative net fluxes (S + L) and the sensible heat flux through canyon surfaces, for various AR for the periods investigated
and for the climatic zone of Fez, are reported in Fig. 11a and b. Wide canyons receive higher radiative net fluxes but allow
higher sensible heat fluxes through their surfaces due to a favorable aeration effect. This observation is valid for both periods
and for the other climatic zones. On the other hand narrow streets, let out lower sensible heat fluxes thus leading to heat
accumulation, which will be absorbed later by the canyon surfaces. The global radiative net fluxes (S + L) and the sensible
heat flux (He) exhibit opposite variations with the aspect ratio. The sensible heat flux (He), can therefore be used as an indi-
cator of the relative magnitude of radiative net fluxes.

3.2. Thermal comfort indicator with ordinary buildings envelope

The thermal comfort indicator PMV predicting the mean response of a larger group of people is expressed by Fanger
(1967, 1970) as:

PMV 0:303  e0:036Me 0:028  Li 29

2 2
where M e is the metabolic rate (W/m ) supposed to be equal to 108.5 W/m for an active pedestrian in the street, and Li is the
thermal load (W/m2) a function of the air temperature, wind speed and humidity. The thermal load is defined as the differ-
ence between the internal heat production and the heat loss to the actual environment for a person at comfort skin temper-
ature and evaporative heat loss by sweating at the actual activity level. The correlation of Li is given in the Appendix.
The PMV indicator and the sensible heat flux using ordinary building envelope are plotted against the aspect ratio in
Fig. 12a and b. It is to be mentioned that the PMV is computed based on the average canyon temperature. These data show
the PMV to increase with the aspect ratio for both periods and all climatic zones. The sensible heat flux on the other hand

(a) 300
250
200
150
[W/m]

100
50
0
-50 S L He
-100
-150
Wall Road Roof

(b) 200
150
100
[W/m]

50
0
S L He
-50
-100
Wall Road Roof

Fig. 9. Average over three days of the global balance of radiative net fluxes (longwave and shortwave), sensible and conduction heat flux on the wall, roof
and the road in Fez climatic zone during the summer (a) and the winter (b) period for an ordinary buildings envelope (AR = 1).
36 A.S. Jihad, M. Tahiri / Urban Climate 16 (2016) 2542

(a) 300
250

S+L [W/m]
200

150

100

50

0
Wall Road Roof

(b) 300

S+L [W/m] 250

200

150

100

50

0
Wall Road Roof

Fig. 10. Average over the summer (a) and winter (b) period of the year 2013 of the radiative net fluxes (S + L) for the three ordinary envelopes of the street
in the climatic zone of Fez (AR = 1).

(a) 350 350

300
He [W/m] S+L 300
250
250
He [W/m]

200
150 200
100
150 S+L
50
0 100
-50
50
-100
-150 0
0.2 0.3 0.5 0.7 1.0 1.5 2.0 3.0 6.0
AR

(b) 350 300

300
He [W/m] S+L 250
250
200
He [W/m]

200
S+L

150
150
100
50 100
0
50
-50
-100 0
0.2 0.3 0.5 0.7 1.0 1.5 2.0 3.0 6.0
AR

Fig. 11. Heat fluxes variations with the aspect ratio during the summer (a) and the winter (b) period for the climatic zone of Fez for an ordinary buildings
envelope.

decreases with the AR, so the radiative net fluxes are increasing with AR. The results show that during the winter period an
aspect ratio from 1.6 to 2.5 insures a favorable thermal comfort index, which is a PMV from 1 to 1 (Fanger and Toftum,
2002). An aspect ratio outside this range leads to either an excessively cold or a hot ambiance because of a lower or a higher
heat accumulation, explained by low or high solar gain access. During the summer period, the examination of the results
depicted in Fig. 12, show that an aspect ratio less than 1.5 insures comfortable thermal conditions. This is explained by
 A.S. Jihad, M. Tahiri / Urban Climate 16 (2016) 2542 37

(a) 350 14
He [W/m]
300
PMV 12
250
200 10

He [W/m]
150 8

PMV
100
50 6
0 4
-50
2
-100
-150 0
0.2 0.3 0.5 0.7 1.0 1.5 2.0 3.0 6.0
AR

6
(b) 350 He [W/m]
300
PMV 4
250
He [W/m]

200 2

PMV
150
0
100
50 -2
0
-4
-50
-100 -6
0.2 0.3 0.5 0.7 1.0 1.5 2.0 3.0 6.0
AR

Fig. 12. PMV and sensible heat flux variations with the AR for the summer (a) and winter (b) period within the climatic zone of Fez for an ordinary buildings
envelope.

the fact that such a geometry corresponding to a wide canyon leads to less heat accumulation, due to favorable ventilation in
the street.
To insure thermal comfort conditions over the year we are compelled to use an average PMV, this is simply taken as the
arithmetic mean over winter and summer data, in order to take the value that corresponds to the most unfavorable case. This
average indicator is plotted against the aspect ratio for the three climatic zones in Fig. 13. Conclusions from the data in
Fig. 13 are summarized in Table 4 which shows that a thermal comfort over the year is obtained for the Fez climatic zone
by adopting an AR in the 1.21.9 range. For the Errachidia climatic zone an AR in the 2.53.4 range leads to an acceptable
PMV. Finally, for the Agadir climatic zone, the aspect ratio should be lower than 1.2. The AR range for Errachidia allow higher
values than the two other zones for their streets AR, because of the very hot period in this zone, therefore canyon should
minimize more solar gain in this season.

11
10 Fez
9
8 Agadir
7
6 Errachidia
Average PMV

5
4
3
2
1
0
-1
-2
-3
-4
0.2 0.3 0.5 0.7 1.0 1.5 2.0 3.0 6.0
AR

Fig. 13. Average PMV as a function of the aspect ratio for the three climatic zones (Fes, Agadir and Errachidia) using an ordinary buildings envelope.
38 A.S. Jihad, M. Tahiri / Urban Climate 16 (2016) 2542

Table 4
The values of the AR of the three climate zones of Morocco in two months February and August, as
well as the average value using an ordinary buildings envelope.

Climatic zone AR August AR February AR over the year

Fez <1.5 [1.6:2.5] [1.2:1.9]
Agadir No comfort [1.2:1.5] <1.2
Errachidia <1.2 [3.3:4] [2.5:3.4]

AR
0.2 0.3 0.5 0.7 1.0 1.5 2.0 3.0 6.0
50
February
45
August
40
Average
Tcan [C]

35
30
25
20
15
10
311 233 206 191 149 106 72 16 -128
He [W/m]

Fig. 14. The temperature of the outside air depending on the sensible heat flux of the walls, roof, and the road on the two hottest and coldest months of the
year 2013 for the climatic zone of Fez using ordinary buildings envelope.

Table 5
The increase in the average external temperature caused by an error in the AR in three climatic zones of
Morocco (Fez, Errachidia and Agadir) using an ordinary buildings envelope.

The AR error (%) The average air temperature increase (C)

Fez Errachidia Agadir
10 1.2 2.2 1.4
20 2.5 4.4 2.9
30 3.8 6.7 4.4

Dependence of the PMV on the AR can also be ascertained through the variation of the average canyon temperature as
shown in Fig. 14. A variation of 10% of the aspect ratio (from a value of 2.73) leads to an increase in the average canyon
temperature of 1.2 C in the climatic zone of Fez. Similar data for the other climatic zones are presented in Table 5, therefore
a bad choice of the AR allow an increasing of external temperature and causing a trouble in thermal comfort.

3.3. Thermal comfort indicator with historical thick walls

To improve the precision required for the choice of the best suitable geometry (AR) for other walls types we focus on the
buildings with historical thick walls like used in the old city of Fez (Tables 2 and 3).
Fig. 15a and b gives the PMV variations with the AR in a street having historical thick walls, for the climatic zone of Fez.
The value of the PMV in August offer a feeling of comfort with an AR > 2, for the month of February the comfort is assured
with a value less than or equal to 5.8, while for the average PMV over the year we are in comfort with a 1.2 < AR < 5.8. The use
of thick walls in buildings and especially in the case of hot zones, creates a favorable thermal comfort even for deep street,
because the heat flux that remain trapped in the street due to very low convective conditions, which minimize the heat flux
escaped to the outside of the street, is stored in the walls for a long time because of their very high thermal capacity. Table 6
gives values of AR to insure comfort for three Moroccan climatic zones.

3.4. Heat dispersion regimes in the urban canyon

Figs. 12 and 15 of the PMV depending on the AR in a street with historical and ordinary walls, during the summer and
winter period of 2013 for the climate zone of Fez, show clearly the three air flow regimes, isolated roughness, wake
 A.S. Jihad, M. Tahiri / Urban Climate 16 (2016) 2542 39

(a) 13
12
PMV : ordinary walls
11
10
9
PMV : thick walls
8

PMV
7
6
5
4
3
2
1
0
0.2 0.3 0.5 0.7 1.0 1.5 2.0 3.0 6.0
AR

(b) 5
PMV : ordinary walls
4
3
PMV : thick walls
2
1
0
PMV

-1 0.2 0.3 0.5 0.7 1.0 1.5 2.0 3.0 6.0

-2
-3
-4
-5
-6
AR

Fig. 15. The PMV depending on the AR in a street with historical thick walls, for ordinary walls the PMV are reported for comparison and are the same than
in Fig. 12, for the climate zone of Fez during the summer (a) and winter (b) period of 2013 using historical thick walls.

interference and skimming flow determined by Chan et al. (2001), about the quality of outside air. These flow regimes that
controls the dispersion intensity of pollutants in the street appears also controls the thermal comfort inside the street
canyon. See the following Fig. 16.
The simulation results depicted in Fig. 16 give the PMV indicator within the canyon for various values of the AR during the
summer period in the climatic zone of Fez. It is observed that the AR has a noticeable influence on the profile of the ambient
air temperature as shown in Fig. 6 where a recirculation vortex appears in all cases because of the acceleration of the shear-
ing zone of the upper air layer of the canyon. For low value of the AR (shallow street), this causes heat to be trapped in the
upstream zone, this in turn leads to less temperature homogenization in the air canyon. When the AR is increasing the flow
regime shifts from the isolated roughness regime to the wake interference regime (Fig. 16), this causes more homogeneous
air circulation, which fosters heat transfer to the inner zone of the canyon (Fig. 6). Air temperature becomes consequently
more uniform, when turbulent air conditions are reached at a threshold value the regime shifts from wake interference to
skimming flow (Fig. 16).
In this case thermal comfort is obtained through the skimming flow, the latter doesnt insure good ventilation efficiency
inside the canyon (Barlow and Belcher, 2002) and not allow good pollution dispersion (Chan et al., 2001). So the choice of AR
must take into consideration the priority factor impacting the site studied, thermal comfort or the quality of the external air.

4. Conclusion

A numerical simulation based a non-dimensional model (TEB) coupled with a turbulence closure model (AROME) was
used to investigate the influence of the urban canyon geometry on the thermal conditions of the external ambient air.

Table 6
The values of the AR for the three climatic zones of Morocco (Fez, Agadir
and Errachidia) in two months February and August as well as the average
value for the case of historical thick walls.

Climatic zone AR August AR February AR over the year

Fez >2 <5.8 [1.2:5.8]
Agadir >2.3 <3.6 [1.8:3.6]
Errachidia >4.3 <5.2 [3.5:5.2]
40 A.S. Jihad, M. Tahiri / Urban Climate 16 (2016) 2542

Average PMV
0

wake interference
-0.5

-1
isolated roughness

PMV
flow
-1.5
skimming flow
-2

-2.5

-3
0.3 0.5 0.7 1.0 1.5
AR

Fig. 16. The three heat propagation regimes in the street canyon based on the average temperature of the street, in the climatic zone of Fez during the
summer period by using an ordinary building envelope.

The investigations led to the conclusion that the most relevant parameters are the envelope of buildings and the aspect ratio
of the canyon. The results obtained for three climatic zones of Morocco (Fez, Agadir and Errachidia) by using an ordinary
building envelope and based on the average temperature of the air canyon, showed that thermal comfort can be obtained
if the AR is kept within an appropriate range. Thus, for the climatic zone of Errachidia the AR should be lower than 1.2, this
parameter should be kept in the 2.53.4 range for the climatic zone of Agadir and finally the AR should be in the 1.21.9
range to insure global thermal comfort in the climatic zone of Fez.
The buildings with a deep street can be used in urban design and still maintaining an acceptable comfort of pedestrians
over the year, as measured in the old city of Fez (Johansson, 2006), by using a thick wall with high thermal capacity and by
keeping the AR within an appropriate range. Thus, for the climatic zone of Errachidia the AR should be in the 3.55.2 range,
this parameter should be kept in the 1.83.6 range for the climatic zone of Agadir and finally the AR should be in the 1.25.8
range to insure global thermal comfort in the climatic zone of Fez. These recommendations may be limited in some urban
location characterized by an absence of vegetation, which could radically change the urban climate and affect the results
found.
The present work shows that the AR of an urban canyon can be used to improve thermal conditions within the air canyon
for three climatic zones of Morocco. This parameter can henceforth be used to mitigate some of the effects of UHI and should
be given due consideration in future urban design and planning.
Another conclusion of the present work is that external thermal comfort conditions within the air canyon can be obtained
for external flow regimes of the ambient air flow, which not allow the best choice of the AR to perform other external build-
ing efficiency like good pollution dispersion, such observation is obtained for the three climatic zones, therefore future
research must focus on an alternative choice of the canyon geometry which can perform both, a good pollution dispersion
and a favorable thermal comfort.

Acknowledgements

I gratefully acknowledge Dr. Mohammed Tahir for providing numerical modeling advice and Dr. Erik Johansson from the
Department of Architecture and Built Environment, Lund University, Sweden for providing experimental data and for his
help in revising the manuscript and The team of the Meteorological Agency of Morocco, without them the present study
could not have been completed.

Appendix A

In this study, the Navier Stokes equation is solved using the SIMPLE method (Semi Implicit Method for Pressure Linked
Equations) (Ghia et al., 1982) to calculate the pressure, the air velocity and air temperature at different locations in the street,
and especially in the same location of the measurement materials (3.5 m above the ground for the deep street and 6.2 m for
the wide street). For this, the urban canopy is divided into standard grids in order to discretize the differential equations.
2D Navier-Stokes equation:
!
@u @u2 @uv @p 1 @ 2 u @ 2 v
  
@t @x @y @x Re @x2 @y2
 A.S. Jihad, M. Tahiri / Urban Climate 16 (2016) 2542 41

!
@v @u2 @uv @p 1 @2u @2v
   
@t @y @x @y Re @x2 @y2

where u and v are the velocity components of the air assumed to be non-compressible.
Continuity equation
@u @u
0
@x @y
Heat equation neglecting the radiative term:
!
@Ta @Ta @Ta 1 @ 2 Ta @ 2 Ta
u  v  
@t @x @y Re @x2 @y2

where Ta is the air temperature profile in the street, a function of the width (x), the height (y) and time t (Tax; y; t)).
For the thermal load Li (W/m2) it is calculated using the following correlations (Fanger, 1967, 1970):

Li fMe  wp  3:96  e8  f cl  t cl 4  tr 4   f cl  hc  tcl  T a  3:05  5:73  0:007  M e  wp  pa  103 

 0:42  M e  wp  58:15  0:0173  Me  5:87  pa  103  0:0014  M e 307  T a g

With

1 0:2  Icl Icl < 0:5
f cl
1:05 0:1  Icl Icl > 0:5

tcl 35:7  0:0275  Me  wp  Rcl  fM e  wp  3:05  5:73  0:007  Me  wp  pa  103   0:42  Me

 wp  58:15  0:0173  M e  5:87  pa  103  0:0014  Me  307  T a g

Rcl 0:155  Icl

1
hc 12:1  V 2c
where f cl () is the ratio of the surface of the clothed body to the surface area of the hide body, hc (W/m2 K) the convective
heat transfer coefficient, Icl (m2 K/W) the clothing insulation, M e (W/m2) the metabolic rate, pa (Pa) the vapor pressure of the
surrounding air, Rcl (m2 K/W) the thermal resistance of clothing, T a (K) the air temperature, tcl (K) the average surface tem-
perature of clothing, t r (K) the mean radiant temperature, V c (m/s) the canyon air velocity and wp (W/m2) the external work
done by the pedestrian in the street.

References

Ali-Toudert, Fazia, Mayer, Helmut, 2006. Numerical study on the effects of aspect ratio and orientation of an urban street canyon on outdoor thermal
comfort in hot and dry climate. Build. Environ. 41 (2), 94108.
Barbaro, S. et al, 1977. An atmospheric model for computing direct and diffuse solar radiation. Sol. Energy 60 (1), 3540.
Barlow, Janet F., Belcher, Stephen E., 2002. A wind tunnel model for quantifying fluxes in the urban boundary layer. Boundary-Layer Meteorol. 104 (1), 131
150.
Chan, Andy T., So, Ellen S.P., Samad, Subash C., 2001. Strategic guidelines for street canyon geometry to achieve sustainable street air quality. Atmos.
Environ. 35, 56815691.
Davies, J.A., Schertzer, W., Nunez, M., 1975. Estimating global solar radiation. Boundary-Layer Meteorol. 9 (1), 3352.
Fahmy, Mohamad, Sharples, Stephen, 2008. Passive design for urban thermal comfort: a comparison between different urban forms in Cairo, Egypt. In: PLEA
2008-25th Conference on Passive and Low Energy Architecture, Dublin, 22nd to 24th October 2008.
Fanger, P.O., 1967. Calculation of thermal comfort: introduction of a basic comfort equation. ASHRE Trans. 73 (Pt 2).
Fanger, P.O., 1970. Thermal Comfort-Analysis and Applications in Environmental Engineering. Danish Technical Press, Copenhagen.
Fanger, P.O., Toftum, J., 2002. Extension of the PMV model to non-air-conditioned buildings in warm climates. Energy Build. 34, 533536.
Ghia, U., Ghia, K.N., Shin, C.T., 1982. High re solutions for incompressible flow using the Navier-Stokes equations and a multigrid method. J Comput Phys 48,
387411.
Idso, S.B., Jackson, R.D., 1969. J. Geophys. Res. 74, 5397.
Ineichen, P., Perez, R., 2002. A new air mass independent formulation for the Linke turbidity coefficient. Sol. Energy 73 (3), 151157.
IPCC, 1990. Working Group II.: Climate Change The IPCC Impacts assessment, pp. 35.
Johansson, Erik, 2006. Influence of urban geometry on outdoor thermal comfort in a hot dry climate: a study in Fez, Morocco. Build. Environ. 41 (10), 1326
1338.
Johnson, G.T., Oke, T.R., Lyons, T.J., Steyn, D.G., Watson, I.D., Voogt, J.A., 1991. Simulation of surface urban heat islands under ideal conditions at night. Part I:
theory and tests against field data. Boundary-Layer Meteorol. 56, 275294.
Masson, V., 2000. A physically-based scheme for the urban energy budget in atmospheric models. Boundary-Layer Meteorol. 94, 357397.
MATLAB version 7.10.0 (R2010a), 2010. Natick, Massachusetts: The MathWorks Inc.
Noilhan, J., 1981. A model for the net total radiation flux at the surfaces of a building. Build. Environ. 16, 259266.
Official site of the National Agency for the Development of Renewable Energy and Energy Efficiency (ADEREE). <http://www.aderee.ma>.
Perera, N.G.R., Weerasekara, W.M.S.B., 2014. The effect of street canyon geometry on outdoor thermal comfort in Colombo. In: Proceedings of the
International Conference on Cities, People and Places ICCPP-2014 October 31th November 2nd, 2014, Colombo, Sri Lanka.
Rowley, F.B., Eckley, W.A., 1932. Surface coefficients as affected by wind direction. ASHRAE Trans. 38, 3346.
42 A.S. Jihad, M. Tahiri / Urban Climate 16 (2016) 2542

Rowley, F.B., Algren, A.B., Blackshaw, J.L., 1930. Surface conductance as affected by air velocity, temperature and character of surface. ASHRAE Trans. 36,
429446.
Seity, Y., Brousseau, P., Malardel, S., Hello, G., Benard, P., Bouttier, F., Lac, C., Masson, V., 2011. The AROME-France convective-scale operational model. Mon.
Weather Rev. 139, 976991.
Takebayashi, Hideki, Moriyama, Masakazu, 2012. Relationships between the properties of an urban street canyon and its radiant environment: introduction
of appropriate urban heat island mitigation technologies. Sol. Energy 86, 22552262.