Bayesian Modeling of Extreme Precipitation in Mindanao, Philippines

Abstract

The Philippines is one of the countries that is prone to typhoons and heavy rainfall. Various atmospheric measures are monitored to inform its citizens regarding climate and weather events. One of these is the r-year return level. In this study, the researchers developed precipitation return level maps with uncertainty measures for selected provinces in Mindanao, Philippines. Using advanced statistical and computational techniques, results demonstrated that a Generalized Pareto Distribution-based Bayesian hierarchical model can effectively estimate r-year precipitation return levels and their associated uncertainty. The hierarchical models efficiently handled the uncertainties in the estimation and easily integrated key covariates in the modeling. It is recommended that more parameters and other covariates be considered to extend the complexity of the model.