Determination of bestfit distribution and rainfall events in Damansara and Kelantan, Malaysia

Abstract

This paper presents a study to determine the best-fit distribution to represent the rainfall process in Damansara and Kelantan, Malaysia. Three probability density functions, namely Wakeby distribution, Generalized Extreme Value function (GEV) and two-parameter Weibull distributions are selected and compared. The parameters of the distributions are estimated using L-moments method while the best-fit distribution is determined by using Kolmogorov-Smirnov goodness-of-fit test. In addition, weighted-average algorithm which is based on the probability values from the stations in Damansara and Kelantan is used to identify the occurrence of wet and dry events in the rainfall data. The impact of different distributions used in the determination of rainfall events is evaluated by making comparison between the actual and the reconstructed rainfall data. The results indicate that the Wakeby distribution is the best-fit distribution to explain the rainfall patterns in Damansara and Kelantan. However, Wakeby, GEV and Weibull distributions perform equally well in the estimation of wet and dry events in Damansara and Kelantan.