Analysis between Temperature and Wind Speed in East Java using Bivariate Extreme Value Theory

Abstract

Extreme weather now is the biggest effect of climate change. It is also a major issue over the past decade. High heat and hurricanes are some of its indicators. Both natural disasters have a big impact not only on humans but also on the presence of animals and the environment. Therefore, modelling of the disaster, especially for prediction extreme events is the basic need to minimize the damage and the losses. Both of the hazards come from extreme temperatures and wind speeds, respectively. Hence, this paper will analyze the issues using extreme value theory (EVT) approach. We will investigate it separately and then analyze both two cases simultaneously. The results show that using peaks over threshold univariate or separately approach is the best model among other approaches since providing the smallest AIC value. Meanwhile, the value of AIC for the block maxima is smaller than peaks over threshold using the bivariate approach.