Modelling Extreme Rainfall using Extended Generalized Extreme Value Distribution

This study assesses the performance of extended generalized extreme value (GEV) distribution based on Kumaraswamy generalized extreme value (KumGEV) distribution using the maximum likelihood estimates on extreme rainfall data obtained from a weather station in Phitsanulok province, a total of 408 months during January 1987 to December 2021. The findings indicate that the KumGEV distribution provides a better fit than the traditional GEV distribution, with estimated parameters µ = 41.4966 (SE = 0.6015), σ = 8.9467 (SE = 0.0797), ξ = −0.0502 (SE = 0.0308), a = 0.0310 (SE = 0.0060), and b = 0.2738 (SE = 0.0155). Additionally, the analysis of return levels derived from both GEV and KumGEV distributions shows an upward trend over return periods of 10, 20, 50, and 100 years, highlighting significant changes in rainfall patterns over time.