Intercomparison of Extreme Rainfall Estimates Using Normal, Gamma and Pearson, and Extreme Value Families of Probability Distributions

Estimation of extreme rainfall for a given return period is considered as one of the important parameters for planning, design, operation and management of civil and hydraulic structures. This can be achieved by performing extreme value analysis (EVA) of rainfall that consists of fitting probability distribution to the annual maximum series of the observed rainfall data. In this book chapter, the EVA of rainfall for the Dharamshala and Kangra sites in Himachal Pradesh was carried out by adopting the Log Normal, Gamma, Pearson Type-3, Log Pearson Type-3 (LP3), Extreme Value Type-1 (EV1), Extreme Value Type-2 (EV2), Generalized Extreme Value (GEV) and Generalized Pareto (PRT) belong to normal, gamma and pearson and extreme value families of probability distributions. The parameters of the distributions were determined by Method of Moments (MoM), Maximum Likelihood Method (MLM) and L-Moments (LMO), and are used for rainfall estimation. The adequacy of fitting distributions adopted in EVA of rainfall was evaluated by Goodness-of-Fit (GoF) (viz., Chi-square and Kolmogorov–Smirnov) and diagnostic (viz., D-index) tests, and the fitted curves of the estimated rainfall. The findings of the study indicated that the distributions other than LP3 (MoM and MLM) and EV2 (MoM and MLM) for Kangra whereas PRT (MoM, MLM and LMO) for Dharamshala are found to be suitable for EVA of rainfall. The outcomes of EVA results weighed with GoF and diagnostic tests indicated that the EV1 (LMO) is the most suitable distribution for estimating the rainfall in Dharamshala and GEV (LMO) for Kangra.