Investigation of the best fit probability distribution for annual maximum rainfall in Kelantan River Basin

Abstract

Abstract Flooding is one of the natural disasters that happens annually in Malaysia. Flooding is induced by the extreme rainfall event and this can cause severe impacts in terms of environment, society, health and safety. Consequently, investigating the best fit probability distribution for annual maximum rainfall can provide the fundamental ideas for government departments and relevant authorities to mitigate the flooding problems. The aim of this study is to investigate the best fit probability distribution in describing the characteristics of annual maximum rainfall for the period of 1994 to 2013 at the Kelantan River Basin. The Gamma, Gumbel, Generalized Extreme Value and Log Pearson Type-III distributions were fitted to the historical rainfall series. Three goodness of fit tests, namely the Kolmogorov-Smirnov, the Anderson-Darling and the Chi-Square tests were used to evaluate the probability distributions. The performances of each probability distribution generated by the goodness of fit tests were compared. Overall, the Generalized Extreme Value distribution seems to be the best fit probability distribution for the annual maximum rainfall at most of the rainfall stations except for the RPS Kuala Betis station that had the Log Pearson Type-III distribution as its best fit.