Dedy Dwi Prastyo, Sutikno : Spatial Extreme Value Modeling Using Max-Stable Processes Approach (Case Study: Rainfall intensity in Ngawi)

Dedy Dwi Prastyo S.Si., M.Si.
Dr. Sutikno S.Si, M.Si.

Year

2016

Published in

3RD INTERNATIONAL CONFERENCE ON RESEARCH, IMPLEMENTATION AND EDUCATION OF MATHEMATICS AND SCIENCE

External link

Type

Seminar Internasional

Keywords

Spatial extreme value, GEV distribution, Max-Stable processes, extreme rainfall, return level


Abstract

Extreme events are short scale phenomena rarely happen, but almost unavoidable and give considerable impacts. Located in the tropical region close to the equator, Indonesia has high variability of rainfall intensities across regions. Prediction of rainfall, particularly the pattern and the characteristic of its extreme intensity, is expected to be able minimizing the loss caused the events. Extreme Value Theory (EVT) is a statistical method commonly used to study the behavior of extreme events in the univariate case. In this study, multivariate series of rainfall intensities from several locations were modeled by means of Max-Stable Processes and spatial extreme value approaches considering the spatial effect. The Generalized Extreme Value (GEV) distribution was employed in Max-Stable Processes with parameters estimation using Maximum Pairwise Likelihood Estimation (MPLE) method. The proposed method was applied to model the extreme rainfall in Ngawi region, East Java, Indonesia. The dependencies of rainfall intensities across location were indicated by the extremal coefficient plot. The best model selected based on Takeuchi Information Criterion (TIC) was used to predict return level of rainfall intensity. In this research, the prediction value of return level showed that the highest rainfall intensity increase in last two years for all stations