Water and Energy International

Frequency analysis of rainfall is one of the important tools to estimate the design rainfall, which is also considered as an input to predict the design flood. This can be used for planning and management of civil and hydraulic structures viz., water use and flood control structures. For this purpose, six probability distributions such as 2-parameters Log Normal, Log Pearson Type-3, Extreme Value Type-1 (EV1), Extreme Value Type-2, Generalized Extreme Value (GEV) and 3-parameters Pareto are applied. The parameters of the distributions are determined by Method of Moments, Maximum Likelihood Method (MLM) and Method of LMoments (LMO); and are used for estimation of rainfall. Goodness-of-Fit (viz., Chi-Square and Kolmogorov-Smirnov) and diagnostic (viz., Correlation Coefficient and D-index) tests are applied for evaluating the adequacy of fitting probability distributions to the rainfall data series (viz., D30 with 30 years data, D40 with 40 years data and D50 with 50 years data) that is created from the annual 1-day maximum rainfall of Dahanu. The outcomes of the results indicate the estimated rainfall obtained from six distributions using the D30, D40 and D50 series adopted in the study are in increasing order when data length increases. The paper presents that the EV1 (LMO) is better suited distribution for rainfall estimation while D30 series is used in estimating the rainfall whereas GEV (MLM) for D40 and GEV (LMO) for D50; and these values could be considered as a design rainfall for designing civil and hydraulic structures at Dahanu.