TR(BR)-149 : Development of regional flood frequency relationship and flood formulae for various Subzone of Zone 3 of India

Abstract

A comparative regional flood frequency study has been carried out using annual maximum peak flood series data of the small to medium size catchments defined by the bridge sites for the seven hydro meteorological subzones of Zone 3. Extreme Value type-1 (EV1) distribution, following modified U.S.G.S. method as well as probability weighted moments (PWM) fitting method; General Extreme Value (GEV) and 5 parameter Wakeby (WAKE) distributions fitted by PWM, utilizing: (i) at site data, (ii) at site and regional data combined, and (iii) regional data alone, have been considered. The homogeneity of the various subzones has been tested by the U.S.G.S. homogeneity test. A regional relationship between mean annual peak flood and catchment area has also been developed for each of the subzones for computation of mean annual peak floods for the ungauged catchments.

Descriptive ability of the various methods has been tested based on the goodness of fit criteria viz. (i) average of relative deviations between computed and observed values of annual maximum peak floods (ADF), (ii) efficiency (EFF) and (iii) standard error (SE). Performance of various methods has also been evaluated based on the predictive ability criteria viz. (i) bias (EIAS), (ii) root mean square error (RMSE), and (iii) coefficient of variation (CV) computed from samples of different sizes drawn from generated EV1, GEV and WAKE populations by considering 1000 replications of the computation procedure for each sample size. The data of two catchments in each subzone viz, second smallest and second largest in size have not been used for estimation of parameters of the flood frequency distributions as these catchments have been treated as ungauged for computing the at site mean annual peak flood values and testing the predictive ability of the various methods. The regional flood frequency curves have also been developed for the combined Zone 3 considering the mean annual peak flood data of the 101 bridge sites out of total 115 bridge sites of the 7 subzones in combined form. Effect of regional heterogeneity is studied by comparing the growth factors of various sub-zones and combined zone 3.

The relationship between mean annual peak flood and catchment area for each subzone has been coupled with the respective regional flood frequency curves of the GEV distribution, and regional flood formulae have been derived for the various subzones and combined zone 3. The developed regional flood formulae have also been expressed in the form of Dicken's formula and revised Dicken's coefficients have also been computed. For estimation of floods of different return periods for gauged catchments, the regional flood frequency curves developed for each subzone together with at site mean annual peak flood may be used; while for ungauged catchments, the regional flood formulae developed for the respective subzones may be adopted.