ABSTRACT
At-site flood frequency analysis is a direct method for design flood estimation and selection of an appropriate probability distribution function is of prime importance for flood frequency analysis. In this study, data of 19 flow gauging sites of the Mahanadi River Basin, India are screened using the independence test and then seven probability distributions namely Exponential, Gumbel, Logistics, Generalised Extreme Value (GEV), Generalized Pareto (GP), Pearson Type 3, and Lognormal were derived to explore the suitability of various flood frequency models using L-moment technique. Based on the L-moment ratio diagram i.e. (theoretical relationships between L-coefficient of skewness and L-coefficient of kurtosis), General Extreme Value (GEV) distribution is identified as the best distribution for the study area. Moreover, the variability in the statistical properties including mean and Coefficient of Variation (Cv and parameters values like location (ξ), scale (α) and shape (k) of the GEV distribution were examined. The GEV was used for estimation of floods of various return periods for all 19 gauging stations of Mahanadi river basin, India.
