APPLICATION OF COPULA FOR CO-OCCURRENCE PROBABILITY ANALYSIS OF MULTI-HAZARD OF TYPHOON-WIND, TYPHOON AND NON-TYPHOON PRECIPITATION

Abstract

Natural disasters often occur simultaneously, cascadingly or cumulatively. Although co-occurrence probability of multi-hazard plays an important role in disaster risk assessment, most of studies in Viet Nam have not considered the probability of simultaneous occurrence of natural hazards. This study aims at deriving trivariate probability distribution of typhoon-wind speed, typhoon and non-typhoon precipitation that occur simultaneously or cumulatively by Copula function. The observed data of tyohoon-wind speed and maximum daily typhoon and non-typhoon precipitation from 1961-2017 of 15 meteorology stations were used in the study. Based on the Kolmogorov-Smirnov and Akaike information criterion tests, the Gumbel-Hougaard Copula is found to best fit the empirical multivariate probability distribution of multi-hazard of typhoon-wind speed, typhoon and non-typhoon precipitation in central coast of Viet Nam.