A PEMODELAN JUMLAH KEJADIAN BANJIR DI KABUPATEN DAN KOTA PROVINSI JAWA TIMUR DENGAN METODE GEOGRAPHICALLY WEIGHTED REGRESSION (GWR)

Abstract

East Java Province is a province that experiences many flood disasters. Floods are natural disaster events that are generally affected by the inability of an area to accommodate high rainfall, where rainfall is different in each region. This study aims to determine models and factors that can significantly cause floods in East Java Province with predictable variables including population density, number of rainy days, rainfall, humidity, population growth rate and development land use. The regression method that is able to model cases with these conditions is Geographically Weighted Regression (GWR). Source of research data were obtained from the Central Statistic Agency, POWER Data Access Viewer and Ministry of Environment and Forestry. The best model can be shown by the coefficient of determination, where the GWR obtains a greater coefficient of determination, namely 65.37% compared to the coefficient of determination in linear regression, which is equal to 31.19%, and the coefficient of determination of SAR is 36.26%.