Klasifikasi Wilayah Risiko Bencana Kekeringan Berbasis Citra Satelit Landsat 8 Oli Dengan Kombinasi Metode Moran’s I dan Getis Ord G* (Studi Kasus : Kabupaten Boyolali dan Klaten)

Nadya Inarossy (Fakultas Teknologi Informasi Universitas Kristen Satya Wacana)
Sri Yulianto Joko Prasetyo (Universitas Kristen Satya Wacana)

Indonesian Journal of Computing and Modeling
Vol 2 No 2 (2019)

The purpose of this study is to classify areas that have a high risk of drought with the vegetation index and SPI index and see spatial connectivity between regions with the method of Moran's I and the formation of hotspots with the method of Getis Ord G *. Experiments with vegetation indices in 45 sub-districts in Boyolali and Klaten districts showed that the average was in class 4 including the classification of medium green. As for the SPI method on rainfall data that is interpolated with IDW techniques, all observation areas are included in the normal drought index class. The results of the analysis with Moran's I show Positive Spatial Autocorrelation, namely the drought phenomenon has spatial connectivity between regions observed. The results of Getis Ord's analysis show the formation of hotpsots and spatial connectivity between regions. The results show that 2017 drought spatial connectivity experienced a broad increase from the previous year. Based on the analysis that has been done with the vegetation index and SPI index, the areas prone to drought are Karanggede, Klego, Andong, Kemusu, Wonosegoro and Juwangi