Abstract:
Unit hydrograph approach is one of the commonly used techniques employed for the determination of flood hydrographs. For this purpose various distribution functions have been utilized in the past. The suitability of the distribution functions are depend on the accurate estimation of their parameter. Different parameter estimation techniques are common in practice viz., method of moments, least squares, maximum likelihood, etc. However, these methods have their own limitations that induce the errors in estimation of flood hydrograph. Contrary to this, the Genetic Algorithm (GA) is used as an optimization tool which minimizes the global error in the estimation of the model parameters. In this study, the two-parameter Gamma and two-parameter Beta distribution functions were utilized to derive the unit hydrograph and the GA was used for the optimization of the parameters. The suitability of two-parameter Beta distribution function was also explored in the derivation of unit hydrograph. The methodology was found to work well in the reproduction of unit hydrograph as well as the direct runoff hydrographs.
