Design Flood Estimation
Design Flood Estimation
Design Flood Estimation
1000
L0.8
9
CN
Tc =
0.5
190 S
0.7
(1)
where, Tc is the time of concentration in minutes, L is the longest flow path of the catchment in
feet, CN is the average Curve Number value of the watershed and S is the average watershed
slope. Longest flow path is the maximum value of Flowlength Grid (FlGrid). As described in the
next section, the FlGrid represents travel distance of each cell in the catchment to the outlet. In
this study two FlGrids are calculated: for weight condition and no weight condition. Average CN
value of the catchment was determined as 78. Average watershed slope is derived from slope
grid (SGrid) of the catchment, which is obtained from DEM. Using Eq. (1), the value of Tc is
estimated as 7.5 hours.
Tt Grid =
Tc
FlGrid
Max.of cell travel lengths
(2)
In Eq. (2) maximum of the cell travel lengths is the maximum value of FlGrid and Tc is the
time of concentration value of the catchment.
3|RKR
The third parameter of the Clarks methodology, the time-area histogram of the catchment is
determined from TtGrid of the catchment. First, histograms of the TtGrid are derived for
different time intervals. It is observed that as the interval gets smaller shape of the histogram
resembles complex hydrographs, and as it gets bigger the shape roughly looks like a single
peaked hydrograph. Aim of trying several interval values is to determine a histogram shape close
to a single peaked hydrograph shape with the smallest possible interval.
IUH of the catchment will be derived if the selected time interval is infinitely small. Practically it
is impossible to obtain the histogram of TtGrid with infinitely small time interval. So the
smallest possible time interval is selected to apply the Clarks technique.
Histogram of TtGrid has time values on the abscissa and number of cells on the ordinate. The
time-area histogram of the catchment is calculated from histogram of TtGrid by converting
number of cells to area. The time area graph of catchment is shown in Figure 2.
25
Area (%)
20
15
10
0
0
60
120
180
240
300
360
420
480
540
Time (min)
After instantaneous application of unit rainfall, total volume of water that will be observed at the
catchment outlet is determined by multiplying catchment area (57 sq km) with the depth of
rainfall (1 mm). Then from the time-area histogram of the catchment, percentage of total volume
contributing to the flow at the outlet in each time interval is calculated. Volumes are then
converted to discharges for corresponding time intervals. Finally by plotting these values at the
mid values of time intervals the translation hydrograph is determined (Figure 3).
4.00
3.50
3.00
2.50
2.00
1.50
1.00
0.50
0.00
0
100
200
300
400
500
600
Time (min)
t
t
Qt =
I t + 1
Qt 1
R
+
0.5
t
R
+
0.5
(3)
In Eq. (3), It is the calculated translation hydrograph, R is the storage attenuation coefficient and
t is the selected time interval for routing. Qt which is obtained after routing is the
instantaneous unit hydrograph of the catchment. The routing process is continued till an excess
flow depth of 1 mm is obtained under the hydrograph. Eq. (3) results the instantaneous unit
5|RKR
hydrograph for the catchment. Finally the ordinate of the unit hydrograph is calculated from Eq.
(4). The derived unit hydrograph for the catchment is shown in Figure 4.
U t = 0.5 (Qt + Qt +t )
(4)
3.00
UH Ordinates (m3/s/mm)
2.50
2.00
1.50
1.00
0.50
0.00
0
120
240
360
480
600
720
840
Time (min)
( Fa / Pe ) = S /( P I a )
(5)
( Fa / S ) = Pe /( P I a )
(6)
(7)
From eq. (7), substituting Pe into eq. (6) the following results,
( Fa / S ) = ( P I a Fa ) /( P I a )
Solving eq. (8) for
Fa
(8)
Fa = S ( P I a ) /( P I a + S )
P Ia
(9)
Equation (9) gives the time distribution of abstraction. Differentiating and noting that I a and S are
constants, the relationship are stated as follows.
dFa / dt = S 2 (dP / dt ) /( P I a + S ) 2
(10)
(11)
P 0.2S
P < 0.2 S
(12)
The cumulative abstraction Fa is computed from eq. (8). The relationship given in eq. (8) can be
further expressed as follows:
7|RKR
S (P Ia )
,
Fa = P I a + S
,
0
P Ia
P < Ia
(13)
(14)
Discharge (m3/s)
20
15
10
0
0
10
15
20
Time (hour)
25
8|RKR
30
35
40
4
3.5
Discharge (m3/s)
3
2.5
2
1.5
1
0.5
0
0
10
15
20
Time (hour)
25
30
35
40
30
35
40
4
3.5
Discharge (m3/s)
3
2.5
2
1.5
1
0.5
0
0
10
15
20
Time (hour)
25
9|RKR
16
14
Discharge (m3/s)
12
10
8
6
4
2
0
0
10
15
20
Time (hour)
25
30
35
40
30
35
40
30
Discharge (m3/s)
25
20
15
10
5
0
0
10
15
20
Time (hour)
25
10 | R K R
90
80
Discharge (m3/s)
70
60
50
40
30
20
10
0
0
10
15
20
Time (hour)
25
30
35
40
30
35
40
25
Discharge (m3/s)
20
15
10
0
0
10
15
20
Time (hour)
25
10
9
Discharge (m3/s)
8
7
6
5
4
3
2
1
0
0
10
15
20
Time (hour)
25
30
35
40
Discharge (m3/s)
160
140
120
100
80
60
40
20
0
0
10
20
30
40
50
60
Time (hour)
70
80
90
100
For flood frequency analysis, two types of distribution functions viz. Pearson type 3, EV1 has
been used. The parameters of the distribution function were fitted with the Method of Moments
and L-Moments techniques. The description of the two distribution function is as follows:
Pearsons distribution
The probability density function (pdf) of PT3 is:
1 x
f ( x) =
( )
x
exp
(15)
The flood quantile can be estimated using the following set of equations (eqs 16 to 21):
+ + KT 2
xT =
(16)
where KT is the frequency factor corresponding to a return period of T years. The WilsonHilferty approximation for KT is:
KT =
2
Cs
C
s
6
Cs
u
6
+
1
1 , Cs > 0
(17)
where u depends on the return period T. The Wilson-Hilferty approximation is quite accurate
for and may be sufficiently accurate for Cs as high as 2.0. The parameter u is represented as
follows.
u = w
c0 + c1w + c2 w2
1 + d1w + d2 w2 + d3 w3
(18)
w = 2ln( P)
(19)
P = 1 F
(20)
F = 1 1/ T
(21)
The parameters of the distribution using MOM and PWM are given as follows:
MOM:
The following set of equation will be used to estimate the parameters of the PT3 using
MOM.
13 | R K R
= (2 / Cs )2
(22)
= ( m2 / ) 0.5
(23)
= m1' ( m 2 ) 0.5
(24)
m3
3/ 2
m2
(25)
Cs =
where: Cs is the coefficient of skewness, m1 is the first sample moment with zero mean, m2
and m3 are the second and third moment with respect to the mean.
PWM:
The plotting position estimates for PWMs are given as follows:
1
as = M 1,0, s =
N
(1 F )
i =1
(26)
xi
(27)
l3 = a0 6a1 + 6a2
l4 = a0 12a1 + 30a2 20a3
The L-moment ratios, which are analogous to conventional moment ratios, are defined by
Hosking (1990) as:
t = l2 / l1
(28)
tr = lr / l2 , r 3
Hosking (1986) has given the parameters for Pearson 3 distribution, and are:
For t3 1/ 3 , let tm = 1 t3 and
14 | R K R
(29)
1 + 0.2906 tm
2
3
( tm + 0.1882 tm + 0.0442 tm )
= l2
(30)
( )
1
( + )
2
(31)
= l1
(32)
f ( x) =
x
x
exp
exp
(33)
x
F ( x) = exp exp
(34)
1
xT = ln ln 1
T
(35)
The parameters of the EV1(2) can be estimated using either MOM or PWM. The required
sets of expressions are given as follows:
MOM:
Using the MOM, the parameter can be estimated as follows
= 0.7797 m2
(36)
= m1 ' 0.45005 m2
(37)
ST =
2
(38)
PWM:
The PWM of the EV1(2) distribution are the form (Greenwood et al., 1979; Hosking, 1986):
= l2 / ln(2)
(39)
= l1 0.5772157
(40)
The standard error, of quantile estimate using PWM can be calculated as follows:
ST =
2
(41)
2.1 Results
The time series of the estimated peak discharge for the years from 1974 to 2011 is given in
Figure 6. The probability plot is shown in Figure 7. The estimated parameters are given in Table
1.
600
500
400
300
200
100
0
1970
1975
1980
1985
1990
1995
Year
2000
2005
2010
2015
Figure 6 Estimated annual peak discharges for the period from 1974 to 2011
16 | R K R
500
400
300
200
100
0
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
Probability of Non-exceedance
83.8
105.23
11071.2
2.21
5.15
The estimated values of the parameters for the PT3 and EV1 using the MOM and PWM are
given in Table 2. The table also includes the value of PPCC. The graphical comparison of the
observed and estimated vales of the maximum flow is shown in Figure 7.
Table 2 Parameters of the PT3 and EV1 using MoM and L-Moment
PT3
PWM
17 | R K R
EV1
PWM
68.9638
MOM
82.0398
116.796
0.7276
43.9806
36.4335
-1.1977
PPCC
0.97408
PPCC
0.92766
0.92766
PT3 (PWM)
EV1 (PWM)
EV1 (MOM)
Observed
Flood (m3/s)
1000
100
10
1
0
0.2
0.4
0.6
0.8
Probability of Non-exceedance (F)
18 | R K R
Return
Period, T
(years)
F=1-1/T
25
50
100
250
500
1000
2000
0.96
0.98
0.99
0.996
0.998
0.999
0.9995
P=1-F
0.04
0.02
0.01
0.004
0.002
0.001
0.0005
EV1: PWM
PT3:
PWM
XT
ST
(Standard
error)
XT
ST
(Standard
error)
XT
298.84
356.55
413.83
489.25
546.20
603.10
659.99
50.78
60.77
70.80
84.10
94.19
104.31
114.45
264.56
313.07
361.22
424.62
472.49
520.33
568.15
38.88
46.03
53.21
62.75
69.99
77.25
84.52
302.38
372.65
444.08
540.30
614.43
689.70
766.08