Indonesia Conference Directory

In a study of landslide vulnerability, awareness to find a fit method and simple procedures of analysis became an obsession for the researchers. Moreover, the best result of the predictive landslide is a requirement when most of the landslide data validation lay on the high class in susceptibility map. Validation compares predictive values with real data set to determine the degree of confidence of the model. This research presents an overview of the utilization of inventory maps of landslides from 2004 to 2016 using google earth such mapped landslide distributions, can then be used to perform landslide zonation onto the landslide susceptibility map to the use of satellite imagery with on-screen digitization. By using GIS and combined statistical analysis, landslide data image inventory was digitized by means of historical imagery then landslide susceptibility map was created. Landslide occurrences and their location were found and interpreted based on the period of landslide occurred. The validation shows the promising result, both by using AUC of ROC curve and the ratio of landslide that falls on high to very high class of susceptibility