SPATIAL ANALYSIS OF LANDSLIDE SUSCEPTIBILITY IN JAVA ISLAND BASED ON SNI 8291:2016Â MODIFICATION

• Main Author: Jaya Kusuma, Ricky
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/73445
• Institution: Institut Teknologi Bandung

<p align="justify">Landslides are the most frequent natural disaster in Java Island with 7205 occurrences in 1998-2022 and caused 1807 fatalities also 45770 infrastructure damages (BNPB). Landslide susceptibility mapping can be implemented as the basis of landslide risk modelling for landslide mitigation purposes. This study aims to map landslide susceptibility in Java Island and validate it based on landslide occurrence data. The method used to map landslide susceptibility is spatial analysis based on SNI 8291:2016 by adding four additional landslide causative factors; elevation, aspect, distance to rivers, and river density. Meanwhile, the method used for validation is ROC (Receiver Operating Characteristic) curve analysis using historical data of landslide occurrence from BNPB and NASA. Each landslide event data from both sources divided into two parts, 70% used to generate the landslide susceptibility map, and the rest was used for validation. The results showed that based on the susceptibility map from BNPB data, Java Island is dominated by low susceptibility, which is about 58447.07 km2. Meanwhile, based on landslide susceptibility map from NASA data, Java Island is dominated by medium susceptibility, which is about 76731.61 km2. The results of model validation based on AUC (Area Under the Curve) values are 0.99 for NASA’s data and 0.92 for BNPB’s data in assessing landslide susceptibility in Java Island.