236

Predictions in Ungauged Basins: PUB Kick-off (Proceedings of the PUB Kick-off meeting held in Brasilia, 2022 November 2002). IAHS Publ. 309, 2007.

Design flood calculation for ungauged small basins in China

Abstract The two main methods for calculating the design flood in ungauged watersheds are introduced in this paper. The Rational Method, which is widely used in China, is described. Its design storm formula form and parameter estimation method, the derivation process for the design peak flood, Qmp, Rational Formulae, and a case study in Jiangxi province, are introduced and described in detail. A brief introduction to the second method, the Empirical Formula Method, is provided.
Keywords design flood calculation; ungauged basin; Rational Formula; Empirical Formula; case study

INTRODUCTION In China, many reservoirs have been constructed and others have yet to be constructed. The statistics of the Minister of Water Resources show that a total of about 86 000 reservoirs were constructed in the past 50 years, of which 360 are very large; the majority of the reservoirs are middle- and small-sized ones. The situation is that there is usually no data for discharge and precipitation in most of the small basins in China, especially for the years from 1960 to 1970. It is therefore necessary for us to pay more attention to the design flood calculation for ungauged small basins. A lot of studies and analyses show that there are three features to the design flood calculation for small reservoirs in China: (a) the method should be suitable for basins without flood and precipitation data; (b) the method should be simple, since the number of reservoirs is so great, and designers for design flood calculation are usually not well specialized; (c) the main focus is usually the design flood peak, not the design flood hydrograph. Four methods (Zhang Daojiang, 2002) are used to calculate the design flood in ungauged small basins in China: the Chinese Rational Formula Method; the Empirical Formula Method; use of the Nash instantaneous unit hydrograph to do some regionalization work; the so-called investigation of historical floods method.

As a contribution to the IAHS PUB meeting, the first two, the most widely used methods, are introduced.
Copyright 2007 IAHS Press

Design flood calculation for ungauged small basins in China

237

CHINESE RATIONAL FORMULA METHOD The design storm formula, the derivation of the Qmp rational formulae and a case study are described.

Design storm formula The design formula has the form:

x tp = s p t 1 n , a tp = s p t - n

(1)

where xtp is the design storm in duration t hours with an exceedence probability p (mm); atp is the design storm density in duration t hours with exceedence probability p (mm h-1); t is storm duration (usually t 24 hours); Sp is xtp with t = 1 hour; n is the storm parameter, to be determined by the regionalization method, n = 0.50.7. The parameter n is determined as follows: In a large region including the ungauged basins, the first step is to calculate the n value for basins with observed precipitation data in the region, then to determine the n value by the regionalization method for ungauged basins (by taking the mean value of n, or by division of sub-regional values of n, or use of an isogram of n). In practice, n may be obtained by reading up the relevant provincial and national hydrological handbooks which collect together the regionalization results of provincial or national ranges. The determination of n in a basin with observed data is shown in Fig. 1. In Fig. 1, the x-axis is log(t), t is storm duration; the y-axis is log(atp), and n is the negative slope of the curve. In some cases, the slopes when t 1 h and t 1 h are different, so n is usually divided into two values, i.e. n = n2, when t 1 h, n = n1 when t 1 h. For the determination of Sp, the following formula is used: Sp = X24p /241-n (2)

Fig. 1 Determination of n in a basin with observed data.

238

Chen Yuanfang et al.

where X24p is determined by the regionalization method. There are usually two ways: by using the isogram of the mean and Cv of the annual maximum 24 hour precipitation; Cs/Cv has an isogram in national and provincial ranges; by using the annual maximum daily precipitation data, x24 p = k x1dp , where k is a

coefficient that is larger than 1.

Derivation of Qmp by the Rational Method

There are four steps to deriving the Qmp Rational formulae: (a) the construction of the design storm process; (b) runoff generation time tc and amount of runoff (hr) calculation formulae; (c) empirical calculation formula of basin concentration time ; (d) Qmp formula derivation by the rational principle.
Design storm process The design storm process curve constructed is indicated in Fig. 2 where the y-axis stands for the instantaneous precipitation density ( i ), and the xaxis is accumulated time of precipitation (x). .

Fig. 2 Design storm process curve.

There are four features to the curve of design storm process: when x tends to the centre, x0,i(x0) (the instantaneous precipitation density) is infinite; it is a symmetrical curve with a centre x0; the shaded area A is equal to xtp = spt1n; there is no explicit formulation.

Design flood calculation for ungauged small basins in China

239

In fact, the same frequency method for all durations t is used to obtain the design storm process. It means that the maximal storm value for all durations t in the storm process curve is equal to the design storm value.
Runoff generation time tc and runoff hr calculation For simplicity, the loss rate of precipitation is considered as a constant . According to the analysis, the relationship between instantaneous precipitation density i and storm duration t is as follows:

Let i(t) = , in this case , t is equal to tc.

dx tp d ( s p t 1 n ) = = (1 n ) s p t n i (t ) = dt dt

(3)

(4)

The amount of runoff generating hr of design storm during tc

hr = s p t c

1n

1 1 t c = s p t cn (1 n)s p t cn t c = ns p t c

1 n

(5)

Calculation of basin concentration time, Using the empirical formula proposed by Chinese hydrologists:

= 0 . 278 L /( mJ

1/3

1 Q mp/ 4 )

(6)

where L is river length (km); J is average longitudinal gradient; m is the concentration parameter; and Qmp is the design peak discharge (m3 s-1).
Qmp formula derivation by the rational method The General Rational formulae of Qmp has two cases, tc and tc (Fig. 3). For tc , then:

h Qmp = 0.278(a ) F = 0.278 For tc , then: h Qmp = 0.278 r F

(7)

(8)

The rational formulae proposed by Chinese hydrologists use combinations of equations (1), (5), (6), (7) and (8). For tc , then: Qm p = 0.278( s p n ) F

240

Chen Yuanfang et al.

(a)

(b)

Fig. 3 Design storm process in cases of tc and tc .

1 and = 0 . 278 L /( mJ 1 / 3 Q mp/ 4 ),The former pair of equations are called equations (a). For tc , then: 1 ns p t c n Qm p = 0.278( )F

1 and = 0 . 278 L /( mJ 1 / 3 Q mp/ 4 ) . These two equations are called equations (b) In order to solve the above equations, one needs to know the following seven parameters: F, L, J, n, Sp, and m. The recommended procedure for solving equations(a)(b) is as follows: Determine the above seven parameter values. Assume tc , then use the diagram method (Fig. 4) to solve equations (a), when Qmp is solved, use equation(4) to calculate tc value. If tc , it indicates that the above assumption is correct. So the Qmp calculated is the result expected, or else turn to solve another set of equations (b)

Fig. 4 Diagram method to calculate Qmp for rational formulae.

Design flood calculation for ungauged small basins in China

241

CASE STUDY

A small reservoir is to be constructed in Jiangxi province, China. Calculation of the design peak discharge with the return period of 100 years by the rational method is required. The procedure of Qmp calculation is as follows:
Determination of basins parameters F, L and J Using to topographic map, it is easy to measure F and L. For the basin studied, F = 104 km2, L = 26 km, J = 0.0875. Determination of storm parameters n, Sp In terms of the hydrological handbook of Jiangxi province, n2 is 0.60. The parameters of annual maximal daily precipitation are:

x1d = 115mm, Cv1d = 0.42, Cs1d = 3.5Cv1d x24 p = 1.1x1d , p s p = x24, p 24 n2 1 = 84.8mm / h
Determination of and m According to hydrological handbook, the value of is 3.0 mm h-1, and m = 0.7. Use diagram method to solve rational equations Assume tc , so Qmp and can be determined by solving the following two equations: 2451.7 Qmp = 86.7 0.6

50.1 14 Qmp

Table 1 Relationship of Qmp and . Qmp = f() Qmp 8 617.4 10 529.1 12 465.3 14 416.6 = f(Qmp) Qmp 400 11.2 450 10.9 500 10.6 600 10.1

From the Fig. 3, Qmp = 510 m3 s-1, and = 10.55 h. Check tc and .
(1 n2 ) s p n2 0.4 84.8 0.6 tc = = 57 h = 3.0 since = 10.55 h < tc = 57h, the above assumption is correct. So the design peak discharge Qmp = 510 m3 s-1 (p = 1%).
1 1

242

Chen Yuanfang et al.

OTHER Qmp CALCULATION METHODS FOR SMALL UNGAUGED BASINS

The Empirical formula method is briefly described below. For other Qmp calculation methods for small ungauged basins, please refer to other references.
Single factor formula

Qmp = CpFn where Cp is the comprehensive coefficient, n is an empirical index that is related to basin area.
Multiple factors formulae

Qmp = Ch24pFn Qmp = Ch24pfFn in which f is equal to F/L2; h24p is the design annual maximal excess rain for 24 hours (mm); , and n are indexes; and C represents comprehensive coefficients.
CONCLUSION

It is very important to research the design flood for ungauged small basins. Some simple methods including the rational formula and empirical formula have been developed and are practised in China. These may also be used in basins outside of China, but it is still necessary to do some work in order to modify them because there are some assumptions which are not suitable to real cases.

REFERENCE
Zhang Daojiang, Ye Shouze et al. (2002) Engineering Hydrology. China Water Conservancy and Hydropower Publishing House, Beijing, China.