CHAPTER
SURFACE
WATER
Surface water is wate stored or flowing on the earth's surface. The surface water
system continually interacts withthe atmospheric and subsurface water systems
described in previous chapters. Tis chapter describes the physical laws governing
surface water flow and shows how hydrologic data are analyzed to provide input
information for models of surface flow.
5.1 SOURCES OF STREAMFLOW
“The watershed, or catchment othe arca of land draining ito stream ata given
location. To describe how the various surface water processes vary through ime
Rn oan 63.)
Where Ry isthe observed rainfall (in) in time interval m.
‘Example 5.3.1. Determine the dieet runoff hydrogeph, the grindex, and the
Xcess rainfall hyetograph for tbe observed sina and steamiow dats gives
in Table $3.1. The watersed area is 7.03 mi,
‘Solution. ‘The basin-averge rfl data given in cola 2 of Table 5.3.1 were
‘obianed by taking Thiessen-veighted averages ofthe rainfall data fom wo rif
‘zges in the watershed Idealy, dita from several moe gages would be used.) The
le data representation is wed for rainfall with time interval of At = 1/2 yoo
ach value shown in column 2 isthe incremental precipitation tat occurred dung
the halhour up to the time shown. The streamflow data shown were reorded
sample dat: the value shown in column 3 is the sesmfow recorded at thet ata
(of time. The observed rainfall nd sueamflow data are plated in Figs 5.3.1, fom
‘hich iis apparent that rfl prior to 9:30 PM. produced «smal ow in the
stream (approximately 400 ef) and thatthe diet inf occured following intone
rainfall between 9:30 and 1130 PM.
‘The computation of the effective rainfall hyetograh and the dieetronoft
Ihydrograph uses the follwing procedure:
‘Step 1. Estimate the basclow. A constant baseflow rte of 40 ef i eeced
tp 2. Calelate the direct renofthydrograph (DRED. The DRH, in column
6 of Table 3.3.1, i found bythe straight line method, by subtracting the $00 cf
basefiow from the observed sreamflow column 3) Eleven half-hour tine iervals
in column 4 are labeled from the fist period of non-zero direct runoff beganing
5930 Pa
‘Step 3. Compute the volame Vz and depth r of direct runoff, suwrace warm 137
‘ramus 5
Rainfall and streamflow data dapted from the storm of May 24-25, 1961, on
‘Shoal Creek at Northwest Park, Austin, Texas
Le ee
‘Observed aces rial Dirck rant
‘Time “Rafal” Stenmslow Time hycogrph (ERH) hydroprpt (DRE)
(Cn eC) )
Colmes 1 75 os 6
ame ls ge
2Me 530 PM. Po
sm ose
sso
io 133m on
ws 230 as 1923
9 2068 Sear 3 m7
90 00 os Stat
2M T2Q0AM. 009 1hoas 1062s
1220 mH 6 74
190 af so
20 me eas
200 lm 8 02
20 10 to 00
30 73 ot 33.
30 34
00 34 Tt 4.0 ess0
0 x
‘ees ainfill= observed rfl sbntons (027 npr Bacay
Dict ruff = observed steanow ~ bsefow (0 cf)
v= So
$43,550.08 12h
43,550
7.839 107
ae
Warhed aea
2139 108
TO i 53609
0.4008
4.90 in = ¢= 027 ain
Rata
7asmnaiass
Pa AM.
igure $3
Rein and steanfow for he sor of May 24-25, 1861, on Shoal Creek at Norwest Pak,
Asin, Te
hep 4, Estimate the rate of rainfall abstractions by infiltration and surface
Sorgen the watershed, Any rainfall prior tothe beginning of diet runoff is
1aken a8 initial absiracion (Le, that rainfall pie 19 9:30 PM. in Table 5.3.1)
‘The sbsaction rate, and M, the numberof nonzero pulse of exces anf, are
{oan by til an eer,
1. HEM = 1, the largest nfl plse, Re = 2.20 in is sleted, substited into
Eq, (5.3.1) using rg= 480 in and Arm 0.5, and solved fora tial vale of|
é
r= > Ra 680)
480-2.20- $x 0.5)
6-58.20 in
whch snot physically posible
2. IfM = 2, the one-hour period having te highest runfl is selected (beeen
11200 P.M. and 11-00 Pt.) and substituted into (5.3.1) to solve fra new tril eed
caer
n= Sa sa9
490 = m0 +208 9x2% 05)
= -ostwn
sain inp
3. LEM = 3, the 1} our period having pulses 1.38, 2.20, nd 2.08 ini eleted,
andthe daa is Substinted into
re YR $80
4.80 = (1.33 +2.20 + 208- x3 x 0.5)
6 = 054 inh
‘This value of is stsactory because it gives At = 0.27 in, which it greater Bh
than all of the afl! pulses in column 2 outside ofthe tree assed 10
contibut to direct rnof. i
‘Step 5. Calealate the excess rinfall byetograph. The ordinates (csluma 5)
are found by subeacting i = 0.27 In from the ordinates of the observed rainfall i
hyetograph(colamn 2, nepecting all intervals in which the observed rainfall depth j |
it han gas Tc drton of exces ral 61.3 hn seule O30
Tio rat," ep of excels checked eae tat ea et 4
sf column 5 = 400i) Te exempt of he here anal pegs ii
Sovtnctesina 33.1
at Cnet '
Ac ued ry ete fe, a |
pepo ae reer pe erty et en
FE cee ec nyt nt ee bn Ty
SSSA Reno ee cease
etc miei aon a
Seca ce aoe teeny nae cat
SPA AES OS at met Maan a ee
ata, IFS fo Re is the total rainfall and ry the corresponding depth of runoff,
So eiitics tas
me 53.2) i
dm ‘Ruumpl 5.3.2. Determine the runoff coefficient forthe sorm in Example 5.3.1,
Solution. Considering only te rainfall that occured eer the bepaning of direct
‘noft (30
Dk = 133+ 2.20 + 2008 + 0.20 + 0.09
= 590i
eam
54 USING INFILTRATION
EQUATIONS.
‘Abstractions include interception of precipitation on vegetation above the ground,
depression storage on te ground surface as water socumulates in hollows over
the surface, and injitraion of water into the sol. Interception and depression
‘Morage abstractions are estimated based an the nature ofthe vegetation end ground
surface or are assumed to be negligible in a large storm,
In the previous section, the rte of abstractions from rinfell wes determined
by using a known streamflow hydrogreph. In most hydrologic problems, the
streamflow hydrogreph isnot evslable andthe abstractions must be determined
by calculating infiltration and accounting separately for other forms of absrac
tion, such as interception, and detention ot depression storage. In this socton, it
‘assumed that al abstractions arise from infiltration, and a method for
the ponding time and inflation under a variable intensity rainfall is developed
based on the Green-Ampt infiltration equation. Equivalent relationships for use
with the Horton and Philip equations are presented in Table 5.4.1. The problem,
considered is: given a rainfall hyetograph defined using the pulse date repre-
sentation, and the perametes of an inflation equation, determine the ponding
time, the infltetion after ponding occurs, and the excess rinfll hyetograph
‘The basic principles used for determining ponding time under constant rain-
{all intensity in Sec. 4.4 are also employed here: in the absence of ponding,
cumulative inflation is calculated from cumulative rainfall; the potetal infil,
tration rate ata given time is calculated frm the cumulative infiltration a thas
time; and ponding has occured when the potential infiltration rat is less than oF
equal tothe rainfall intensity. 142 sam wrsec.ocr
Consider a time interval from 1 to 1 + Ar. Tae rainfall intensity during
this interval is denoted i, and is constant throughout the interval. The potential
inflation rate and cumulative infiltration at the beginning of the interval are f-
and F,, respectively, and the coresponding values athe end ofthe interval are
‘rosy tnd Fay. is assumed that F, is known froma given inital conditions or
‘previous computation
‘A flow chart for determining ponding time is presented in Fig 5.4.1. There
are three cases to be considered: (1) ponding occurs throughout the interval: 2)
there is no ponding throughout the interval; and (3) ponding begins part-way
hough the interval. The infiltration rate is always either decreasing or constant
with time, so once ponding is established under a given rainfall intensity, it will,
‘continue. Hence, ponding cannot cease in the middle of an interval, but only st
its end point, when the value of th rainfall intensity changes.
Following the flow char, the frst step isto calculate the current potential
infiltration rate f, from the known value of cumulatve infiltration F, For the
Green-Ampt method, one usee
west
in Weat) ea
fee me iil ly om
cs ia Sa lS
ee Ce ad bo tre
seat Pat ll
reas + WE}
Fi yao
‘This equation is derived in a manner similar to that shown in Sec. 44 for Ea,
445).
Both cases (2) and (3) have f, > i, and no ponding atthe beginning of the
{ntrval. Assume that this remains so throughout the interval then, the infiltration
‘ate js i, and a tentative value for cumulative infitaton at the end of the time
interval is
Flew = Fit iar (54.3)
Next, a corresponding infiltration rate fis caleulaed fom Fay IF fsap
greater than i, ease (2) oceurs and there is no ponding throughout the interval,
‘Thus Fras, = Fir and the problem is solved for this interval
Jf Sivas is less than or equal toi, ponding occurs during the interval (case
(@). The cumulative infitation Fat ponding time is found by setting f, 1,
and F, = F, in (5.4.1) and solving for Fy to give, for he Green Ampt equation,
Kyag
uok
‘The ponding time is then: + AY’, where
Re 6.4.4) sunce waren 143
Nr Fa0.
Coes fom
ponding occurs Negenegtne |
beta merece i
‘atom
by nro gen,
FIGURE 54.1 Bi
‘low chr fr Gtrining flonan ponding ime unde varie rial enti
ay = oF 645)
tnd the cumulative infiltration Fy found by subiaing F, = F and =
AP—Av in (54.2). The exes nal lus ae cell by Subacting 144 serum nrorccooy
cumulative infiltration from cumulative rainfall, then aking successive differences
‘of the resulting values,
Example $4.1 A rainfall hytopraph is given in columns 1 and 2 of Table 5.42.
1f this rin falls on a sandy loom sil of inal effective saturation 40 percent
letermine the excess rainfall bytograph,
Setution. From Table 43.1, for a sandy loam soil, K = 1.09 ems, = 11.01 em
snd @ = 0.412. From Ea. (4.3.10),
ABR = 5
=u =0.4,0.812)
=0.287
890.247 x 11.01
Rem
‘The time interval in Tble 5.4.2 is Ar = 10 min = 0.167 h. Column 3 of the able
sows the cumulative rainfall depths found by summing the incremental vlue
columa 2. The rainfall yetograph andthe eumlative rani jetograph ae shown
| Fig. 54.2. The rainfall intensity in column 4s found fom column 2 by dividing
by ar For example, during the fst me Interval, 0.18 cm of rainfall curso
4, 0.180.167 = 1.08 em as shown, Initially. F = 0, s0 f= = from (Sly an
Ponding doesnot occur atime 0. Hence F at tine 10 min i alulated by 3-4
thus: Fieg = Fi + Ar 0+ 0.18 = 0.18 cm. The conesponding valve of fi.
is, from (8.4.1),
1757 own
5 shown in column 5 of the table This value s greater than i; therefore, no
Ponding occurs during this inerval and cumulative infiltration equals suulaiee
‘infil as shown in column 6.1 is found that ponding does not ozcur 10 60
ine fi ta Din
ae
s(t
an
Tmt)
enh
hic is ess than = 3.88 cm forthe interval from 60 t0 70 minutes, 50 ponding
begins at 60 min sce Fig. 54.2) Hl-cansaslese
=) = = =) am = =
ewosoet anemia) Set OINTa)—eeT
roe ome Tost
* ‘ ° 5 , « z a)
‘srdurexG) wopeaba wopesnyay Wiury-ase15 ow Bun ydeBOIAC ppUIEH SaIxD Jo HONEA
Pes av During the pnded period, (5.4.2) i used to eset infiltration, The value of
Frsar st Oris is given by
eee
ru r-sen [$8 gy
Fey + 2.72
T7427
ee teens
Fey +2.
Few at
95427210
ator ane
Iattration rate
54
Zrencien
sei
ae em)
0 @ © % 0 1m 10 Wo 1m
Clave depth)
om © © © wom Ho 10 1m
Tine ie)
ours 5.42
aftsion an exces rainfall under vars rif tensity (Example 5.4), semncewicen 147
“which is solv by the method of sucesive approximation to 2ive Fu 2.21
‘Gm as shown in column 6 of Table 5.4.2. The cumulative exces runfll column
‘Tie found by subtracting cumulative infliraion (column 6) from cumulative
‘sinfll (Colucn 3). And the excess rin values in column 8 are found by taking
0.5 in, the continuing sbaraction ys computed free,
659):
su
L+s
= 250 0.5)
P=054250
= 250? ~ 05)
P40
For example, after two hours, the cumulative minal is P = 0,90 in, so
= 2.5000.9 - 0.5)
09 +20
=0.34 in
45 shown in columa 4 ofthe ble. The excess unflisthat remaining after inal
‘nd continuing abstatons From (5.5.2)
rR
TABLE 55.3
Computation of abstractions and excess rainfall hyetograph by the
‘SCS method (Example 5.5.4)
a ee ee
Cohuma:
‘Comsttve ala
‘Time fala sb8ractes Gn)
oto FE
28 yy
oo Came
120 on - 9
0.06
2 0” 050 034 005
on
31a 00 059 018
ost
sn 050 10s 076
5465 00 156 259
0.08
7536 00 165 3.21 sancewarm 155
Pp=P-le—Fe
0.90 0.50 -0.34
=0.06 in
ss shown in cok 5; The excess rain hytogagh i deemed by
thing te tfean of ucusive ales of P(iman 6.
5.6 FLOW DEPTH AND VELOCITY =~
“The flow of water over & watershed surface is a complicated process varying in
all thee space dimensions and time, It bepins when water becones ponded on,
the surface at sufficient depth to overcome surface retention fores and begins
to flow. Two basic flow types may be distinguished: overland flow and channel
flow. Overland flow has a thin layer of water flowing over a wide surface.
‘Channel flow has a much narrower steam of water flowing in a confined path,
Chapter 2 gave the physical laws applicable to these two types of flow. On a
natural watershed, overland flow is the first mechanism of surface flow but it
‘may persist for only a short distance (say up to 100 ft) before ronuniformitis,
in the watershed surface concentrate the flow into tortuous channels. Gradually,
the outflows from these small channels combine to produce recognizable stream
channel flows which accumulate going downstream to form streamflow at the
‘watershed oul
Surface water flow is govemed by the principles of continuity and
‘momentum. The application of these principles to three-dimensional unsteady
flow on a watershed surface is possible ony in very simplified situations, so one-
‘or two-dimensional flow is usually assumed.
Overland Flow -
(Overland flow is a very thin sheet flow which occurs at the upper end of slopes
before the low concentrates into recognizable channels. Figure 5.6.1 shows flow
‘down a uniform plane on which rain is falling at intensity i and infiltration
‘occurring at rate f, Sufficient time has passed since rainfall began that all flows.
are steady. The plane is of unit width and length Lo, and is inclined at angle @
to the horizontal with slope Sp = tan 8.
pee eee es ee ae
en an
‘The inflow to the contol volume from rainfall isi cos @, and the outflow is
{fla cos 8 from infiltration plus Vy from overland flow. The depth y is measured
etpendicular to the bed andthe velocity V parallel tothe bed. Thus the continuity
‘equation is writen te Vye U-Pb
FIGURE 5.6.1
‘Steady flow on nfo plane under nfl.
(ifexereconese
‘The discharge per unit with, go, is given by
d= W= i -flo cos 0 (662)
Momentum. For uniform laminar low on an inclined pline, it can be shown
Roberson and Crowe, 1985), thatthe average velocity Vis piven by
ye ess
where gis acceleration due to gravity and v's the kinematic viscosity of the uid.
For uniform flow, Sp = Sj = hy/L, and (5.6.3) can be rearanged to yield
664)
Which i inthe form ofthe Day Weisbach equation 2.5.1) fr ow resistance
Le
he Sine 665
with the friction factor f = 96/Re in which the Reynolds sumber Re = 4VRIv,
tnd the hydraulic radius R = y, For a unit widdh sheet flow, R= areal(wetted
Perimeter) =y x 1/1 = y, as required, The flow remains laninar provided Re =
2000.
For laminar sheet flow under rainfall, the friction factor increases with the
rainfall intensity. If it is assumed that f has the form C,/Re, where Cis a sunracewarex 157
resistance coefficient, experimentation carried out at the University of Ilinois
(Chow and Yen, 1976) gave
96 + 10814 56.6)
‘here isthe rainfall intensity in inches per hour.
Solving for y from (5.6.5) and using the fact that hf. = So for uniform
ie
He
=
which specifies the depth of sheet flow on a uniform plane.
Example$.61. A nfl of measity 1 nh fls on uniform, smooth, pervious
plane 100 fect long at S percent slope. Caleulate the discharge per unit with te
epi, and the veloc at the lower end ofthe plane, Take v= 1.2% 10-1,
Solution. The disctarge per unit widths given by (5.6.2) with! = 1 ih = 2.32
10 Ws, and f= 0. The angle 0 = tan-"Sq) = tan-X0.05) = 2.86%, so
cos 0 = 0.599,
=~ cos 8
= 2.32 10-0) x 100x 0.999
=2.31 107 is
‘The Reynolds number
ae
a
4231107
12x
70
tnd the flow i lamina. The resistance coefficient Cis given by (5.6.6)
cy =96 + 108%
=96 + 108()°*
=208
209/770 = 0.265, and the dept is ealeulted
‘The fiction factor is f = CUR
from Eq, (5.6.8), = 0.0048 f (0.06 in)
‘The velocity Vis given by
%
y
31 x 100.0048
0.48
Field studies of overland flow (Emmett, 1978) indicate tht the flow is
laminar but thatthe flow resistance is about ten times larger than for laboratory
studies on uniform planes. The increase inflow resistance results primarily from:
the unevenness in the topography and surface vegetation, Equation (5.6.8) can,
be rewsitten in the more general form
For laminar flow m = 2/3 and a = (f/8@Sq)"°, Emmet’s studies indicate that
the Darcy-Weisbsch friction factor fis inthe range 20-200 for overlnd flow at
eld sites.
When the flow becomes turbulent, the fiction factor becomes independent
‘of the Reynolds rumber and dependent only on the roughness ofthe surface. Ta
this case, Manning's equation (2.5.7) is applicable to describe the flow:
ve LB page 5.6.10)
‘= So for uniform flow, and go = Vy. This can be solved for y to
6.6.11)
which isin the general form of (5.6.9) with = (w/1.49S¥)% and m= 3/5
For SI units, a = n0#/s@
Example 5.62. Calculate the deph and velocity of a discharge of 231 x10?
fst (with) on turf having Dary-Weisbach f= 7S and a slope of 8 percent.
Take = 1.2810" fs,
= 790
From (5.6.9)
Solution. ‘The Reynolds number is Re = 4ggy= 42.31% 10")
(aminar flow), and a = (785) = (75(8 32.20.08)"
with m= 313 sunraceworen 15%
yoo i
= 1,902.31 « 10-3 a
031 f (0.4 ia) H
Velocity V= guy = 2.31 x 10°20.031 = 0.075 fs, I canbe see that hie flow i
smock deper and slower flowing than low on te emooth plane of Example 5.6.1.
Example §.6.3. Caleulate the discharge pec unit with, depth, and velocity athe
end of «200-4 sip of asphalt, of slope 0.02, subject wo rainfall of 10 in, with
“Maning'sn = 0.015 and Kinematic viscosity » = 1.2 x 10" £3, i
Soluion. The discharge per unt width s given by Eq, (5.6.2) with {= 10 inh =
2.32% 10+ fas, f= 0, and @ = tan™M0.02) = 1.15, for which cs @ = 1.00: i
=~ filo cos 0
2.32 x 10°F 200 1.00
0464 fe 3
“The Reynolds umber is Re = dgy/v= 4% 0,046)(1.2% 10-4 = 1533, s0 the ow
is tubulent. The depth of flow is given by Eq (5.6.9) with « = (wl. 495]9)°5— 1
[O.018(49 0.02") = 0.205 and m = 0.6: x
3
ya ogg o
= 0.205 x (0.0468)
032 (0.4 in)
Also, {i
v= i
y
0.0964
0.082
21.8 15,
Channel Flow
‘The passage of overland flow into channel canbe viewed as a lateral flow inthe
same way that the previous examples have considered rainfall asa lateral flow
‘onto the watershed surface.
Consider a channel of length L, that is fed by overland flow from a plane
as shown in Fig. 5.6.2. The overland flow has discharge qo pec unit width, 30
the discharge in the channel is Q = guLe. To Bnd the depth and velocity at
various points along the channel, an iterative solution of Maaning's equation is
‘necessary. Manning's equation is
°
LB oyraRe (56.12) FIGURE 5.62
‘Ovriand fw fom x pane ito «chan
Solution of Manning's Equation by Newton's
anning’s Equatio
‘There is no general analytical solution fo Manning's equation for determining the
flow depth given the flow rate because the area A and hydraulic radius R may be
complicated functions of the depth. Newton's method an be applied ieratively to
sive a numerical solution, Suppose that at iteration j the depthy, is selected and
the flow rate Q; is computed from (5.6.12), using the area and hydraulic radius
omesponding 10 y,. This Q; is compared withthe actual low Q2: the object is to
select y so that the error
$0) =~ 0 6.613)
4s accepubly small. The gradient off with respect toy is
de s.
oer (6.6.18)
because @ is a constant. Hence, assuming Manning's m is constant,
-t2e0)
Aap ME. pot)
aS
‘ wane 6615)
124 p23 2 1
Waragn( 2a 1a)
2ae | 14a
~ol Ret 4%)
‘where the subscript j outside the parentheses indicates that the contents are
evaluated for y= 3, ao
fy)
iaisn,
oo)
riurr 563.
Newon’s method exraplaes tt
feet of te enor Reto st the erent
ep cu th deph yn forthe
‘This expression for the gradient is useful for Newton's method, where,
siven a choice of y. 9/1 is chosen to satisfy
[af\ , = fo,
“This 1 the value of, na plot off vy, where the tangent the curve a
1 = itset the hoot ais, a used in ig 5.6.3,
Solving (5.5.19 fx yen
Rae oe 6.17
1 aid; oon
hich the fundamental equson ofthe Newton's method Retin ae con-
nue unl here io sigalcan change n'y: his wil happen wien the enor
79) 8) costo 2.
‘Stbstttng ito (6.17) fom Eos. (5.6.13) and (5.615) gives the New
tonsmetedeguton for solving Maing’ equation
- toe, (5.6.18)
(ee +44),
For a rectangular channel A = By and R = Byy/(By + 29) where By is the
‘channel width; after some manipulation, (5.6.18) becomes
10
aan =)
Values forthe channel shape function {(2/3R) (dBRldy) + (V/A) (Ald) for other
cross sections are given in Table 5.6.1
ye
yen
i
i “er a) 14°12 LEI “YOK MON AON el A ns
ns
‘Suauiap> PuUEYD 40} SUORSURY sam
tesa sare waren 163
Example 5.6.4, Calculate the flow depth in a two-footwide rectangular channel
having = 0.015, Sp = 0.025, and Q = 9.26 cf
Solution.
2 oa_Ban)™
B+
= 29 0.0252 Gw
0.0150" +
aly
aes (5.6.19)
a
Also,
1dA_ 58. +6y 10+ 6y
Ady IB. +2) ~ HAHA)
1867 + y
wl 9)
rom Bq, (5.6.18)
(= 92619) +)
rae (6.6.20)
From an abitaily chosen suring guess of y= 1.00 &, the solution to three
‘gnifican figures is achieved after thee iterations by successively solving (56.19)
and (5.6.20) for Q, and y 1. The result is y = 0.58 f.
WAH
Merton? 123
ym 100 Ow Os” OST
Gites) 19.79 982 9.25 9.26
Example 5.6.8. Compute the velocity and dep of flow at 200-f00t increments
along # 1000-fotlong rectangular channel having with 2 roughness n= 0.015,
and slope $= 0.025, supplied bya ata flow of 0.00926 cfu
Solution. The method of Example 5.6.4 is applied repesvely to compute y fr
Q = 0.008261. The velocity is V= O/B, = 0/2.
Distance along channel, Lt) 0 200 «00 6o0 BO 1000
ow ae) © 185 370 556 Tal 926
Depth 9c © 020 031 O41 O49 038
els Vn) © 463 597 686 756 nor 3
2
FIGURE 5.64
t 7 7 Dinetsoneshyogaph of ovee fow.
02 04 06 G8 10 Ty andy ow oe ated tine of
te polio (Ae lene, 146)
‘The examples inthis section have assumed steady flow on the watershed
In reality, under a constant intensity rainfall, the steady flow at equilibrium is
‘spprosched asymptotically in the manner illustrated by Fig. 5.6.4. Thus, the flow
is varying both in space and time on the watershed surface and in te steam
channel,
5.7 TRAVEL TIME
‘The travel time of flow from one point ona watershed to another canbe deduced
from the flow distance and velocity. If two points on a steam are a distance L
‘part and the velocity along the path connecting them is v0), where I is distance
slong the path, then the travel time # i given by
at=par
a
fa- [4 62.)
a
fs 672)
1 the velocity can be assumed constant at vin an increment of length Al,
Petz vsash then
ry (5.7.3)
‘Velocities for use in Eq. (5.7.3) may be computed using the methods described
in Sec. 5.6 or by reference to Table 5.7.1 sunrace waren 165
Tapes.
Approximate average velocities in Vs of runoff lw for calculating
time of concentration
eseriton of water coarse ‘Slope im percent
os « on a
ictoreee gs Eg ge
‘Unconsenid
Woodnde os sas asa aa
Pastures O25 2833 a4 als
cote 030 304s as ss ss
Prvenets oss asiss asi ne
Concert
‘et canse—deterine velocity by Maing’ formule
tell deface o2 2 4 =
Ps ence oes per ene of ward pw bowing Bow mroning
‘rE rs vy wih te te sie ater sono, Wi posite, roan eels
‘tbe ade pa snes yh ng ee eae
(Sewer Dap Mu Tess Higny Deparment, ale VI IF, 1370)
‘Because ofthe travel time tothe watershed outlet, only prt ofthe watershed
may be contributing to surface water flow at any timer afer precipitation begins,
‘The growth of the contributing area may be visualized in Fg. 5.7.1. If rental
of constant intensity besins and continues indefinitely, then the area bounded by
the dashed line labeled twill contribute to streamflow at the watershed eat.
afer time 1; likewise, the area bounded by the line labeled ry wll contribute to
FIGURE 5.7.1
Sodhons mt fin th sea contibng fw a th cut fr rfl of dari,
sd Time of soncentaton tthe tine of Bow fromthe fares pin inthe wes a
‘te ont 166 prs evonaxcor
TABLE 87.2
‘Travel time in a channel (Example 5.7.1)
Distance slong cmae, 100) 0 20089 oom
al 200 m0 meme ame
Catelued velocity VOR) O43
Avenge vei Vv) ae sm gana
Dewliim arm AV) 862752
@ ar= 208.55)
ass sy
streamflow after time fa, The boundaries ofthese contributing areas are lines of
qual tie of flow tothe outlet and are called isochrones. The time at which sll
(of the watershed begins to contribute is the time of concentration Ts this i the
time of fow from the farthest point on the watershed tothe outlet
Example 5.7.1. Calculate the time of concentration ofa waterbed ja which the
longest flow path coves 100 feet of pasture at 5 percent slope, thea enen ¢
1000-foot-long rectangular channel having width 2, roughness = O01, ead
‘ops 2.5 percent, and receiving a lateral low of 0.00926 ef
Solution. From Tible 5.7.1, pasture at $ percent slope has & velocity of flow in
the ange 2.53.5 fu: use a velocity of 3.0 us. The travel time over he 100 fect
Of pane is At = Ally = 10/3.0~ 3s, For the rectangular channel the veloc
1 20-oot intervals was calculated in Example 5.6.5. The travel tine over eh
{inter is found from the average velocity in tht interval. For example, fos oe
fis 200 ft, At = Ally = 200/28 = 86.2, This yields a oul uae ee Oe
5.8 STREAM NETWORKS
‘In Muid mechanics, the study ofthe similarity of uid flow in systems of differ-
ent sizes is an important tool in relating the results of small-scale model tuice
to large-scale prototype applications. In hydrology. the geomorphology of the
Jratershed, or quantitative study of the surface landform, is used to arive at mee,
sures of geometic similarity among watersheds, especially among their recy
networks.
‘The smallest recognizable channels are designated order 1; these channels
normally flow only during wet weather sunrace waren 167
‘Where two channels of order 1 join, a channel of order 2 results downstream;
in general, where two channels of order! join, a channel of order + 1 results,
‘Where a channel of lower order joins a channel of higher order, the ctannel
downstream retains the higher ofthe two orders,
‘The order of the drainage basin is designated as the order of the seam
draining its outlet, the highest steam order in the basin, 1.
‘An example of this classification system for a small watershed in Texas is own
in Fig. 58.1.
Key
(nse eam
FIGURE 51
Ware of Miler Creek, Bisco County, Teas, showing the delineation of seam eres, Horton (1945) found empirically that the bifurcation ratios, or ratio of
the number Ni, of channels of onder i tothe number Naf channels of ender
from one order to another. This is Horton's Law of
=k 1,2,
58.1)
As an example, in Fig. 5.8.1, Mj = 28, Me™ 5, and Ny = 1:80 NUN, = 5.6
2 NelNs = 5.0, The theoecl mina vale ofthe bce cs ic
and values typically iin he range 35 (Stabler 1968)
By measuring the lenin of each steam, tc average length of sweams of
cach ore, Lan be fund. Horn proposed Lon of ean ero
th average Ings of seams of suerte ene ae wed ty ice
Re
ha
t
By asim reasoning, Schum (1956) proposed a Law of ream dreaso rete
the average seas A rind by seam ef eeeene od
Aust
Re (5.8.2)
683)
FIGURE 5.82
Seonarholgeal ramet fr tb Manon ais. (Source: Vales, ilo, nd Reig ee,
'- 1128, 179. Copp bythe Ameren Geophys! Union sunrace waren 169
‘These ratios ae computed by plotting the values for Nj, Li. and A; on a loge
‘ithmic scale egainst steam order on a linear scale, as shown for two Vene nu. i
clan watersheds in Fig. 5.8.2. The ratios Rp, Ru, and Ra are computed from the i
slopes ofthe lines on these graphs. The Mamon'S watershed is a subbasin of the i
Mamon watershed (Fig. 5.8.3), The consistency of Rp... and Ry between the i
‘wo watersheds demonstrates their geometric similarity. Studies have been made i
to relate the characteristics of flood hydrographs to stream network parameters i
(Rodsiguez-Inube and Valdes, 1979; Gupta, Waymite, and Wang, 1980; Gupta,
Rodriguez-Iture and Wood, 1986). i
‘ther parameters useful for hydrologic analysis are the drainage density and i
‘the length of overland flow (Smart, 1972). The drainage density D is the rato of i
‘the total length of stream channels in a watershed to its area Hi
$a
684 i
Manon stn
‘aren sztm)
wore)
rnoune sa3 i
Dimes fi fe Mane warned Y(t: Vl, Fal nd Rigs
hen 1, 1-year Ge Tne 4 170 seruconvorovoor
where Ly is the length of the jth stream of order i. If the streams are fed by
Hortonian overland flow from sll of their contributing area, then the average
length of overland flow, L,, is given approximately by
1
b= 68.5)
‘Shreve (1966) showed that Horton's stream laws result from the mast likely
Ccombinstions of channele into a network if random selection ie made among all
possible combinations.
REFERENCES:
Beton, RP, What is waesbed met? J. Geophys. Re Yl 69: 9.8 154-1852, 1964
Chow, V.T.,Ope-chane Hara, Metal New York, 195,
(Chow, V.7. and BC, Yen, Uban sarmvater oof deemiaaon of volumes on wrats,
eyo EPA-0072-76-16, Maia Evinmenl Reach Laborer, Office of Rca
tnd Development, U.S. Environmenl Peecton Ageacy, Chsinat, Oo, May T97e
Dune. T.R. Moore, and. H. Tylor, Rcognon an pein of not pedis tees
in amie reins, Hdl. Si, Bull, vol 20, 0.3, pp 305-32, 1973
Emmet. W.W..Overan fm, in lillope Hydrology ed by M. J. Kikby, Wily, New ork,
Pp. 45-176, 1978,
‘rcte R.A, Rol of tsarfoe fw ingesting srfice mnt 2. Upstream youce wees, Woe
‘eso: Re. vo. 8.205, pp. 1272-1283, 1972
Free, R.A. Shtanflow gereniton, Re. Geopys. Space Ph. vl 12, m0. 4, pp, 627-647,
wie
(up, V.K... Waymie,C.7. Wang, A representation ofa tants uit yoga from
‘scomorpology, Wier Reso. Res. sl 16, 0.5, pp. 85-862, 980
(up V. Kf Roiuerute, and EF, Wood eds), Sele Probleme Hpdvoloy, D, Riel,
Dest, Holland. 1986,
Har, RD. Water fk in sland subsoil ona seep feed slope, J. Hydrol, vl. 3, 37-58,
om
‘Hewlet ID, and AR Hibben, Facts fectng the sponse of smal watered to ptipiaton
ii reas, fe Samp. on Fret Holy by W. E Soper and H.W. Lal,
aganon Pes, Oxford, pp. 275-290, 1967
eit, ID. Pipes of Foren Harolgs, Univ. of Georgia Press, thers C.. 1982.
‘oro RE. The re of infin inte bolo eye, Trane. Am. Geol Un, a
14, p. 246-40, 1938,
oon, R'E, Erion development of sueens sd his ring ati
to quainive mrpoloy Bal. Geo. Sx. Am, v.56 pp 275-970, 108
lez, CF. Hyrasof rant fom developed sures, Proceedings, 265 Aral Mtg of
‘the Highvay Research Bou, va 26, pp. 129-14, Deseo 146,
Manin, G. Ns Charatan af sinple exponential baton, dal. (New Zealand).
ol 1, mo Te pp. 37-52, 1973,
Mor-Seytoun,'H. i and J.P. Verdi, Eension ofthe Slt Conservation Service rainfall
‘oof methodology or ngaed watered, repo no, FAWAIRD-61/060, Feral phe)
‘Acminsaton, Washingtoe, D.C. evaube fom Nato Teh Ifaaton Sent,
‘Spengel, Va 22616, 198
‘Moseley: MiP. Stenmfow generation ina forse watered, New Zeland, Water Resour, Res
“et 8, no. pp 795-405, 1978
ence, A.J, nd A
