PEMETAAN KERENTANAN LONGSORAN MENGGUNAKAN METODE WEIGHT OF EVIDENCE DAN INFORMATION VALUE DI SUB-DAS CISANGKUY KABUPATEN BANDUNG JAWABARAT

• Main Author: Vikki Refaldo, Muhammad
• Format: Final Project
• Language: Indonesia
• Subjects: Geologi, hidrologi & meteorologi
• Online Access: https://digilib.itb.ac.id/gdl/view/61138
• Institution: Institut Teknologi Bandung

Cisangkuy Sub-watershed is located in Bandung Regency, West Java. In West Java, there were 1,578 out of 7,861 landslides that occurred in Indonesia in 2014. In 2018, there was an increase in the number of landslides that occurred in West Java to 1,824 out of 10,246 landslides that occurred throughout Indonesia. Therefore, landslide susceptibility mapping is required in Cisangkuy Sub-watershed to determine the safe zone for land use by the community and to mitigate the risk of landslides in order to minimize losses caused by landslides. This research was conducted using the weight of evidence (WoE) and information value (IV). Landslide data was conducted by using satellite imagery, field observation and PVMBG reports. The total number of landslides obtained was 406 data. The data was divided randomly into two to 70% as training dataset (284 data) and 30% as test dataset (122 data). This research used 20 parameters that cause landslides: flow direction, aspect, rainfall, distance from fault, distance from road, distance from river, fault density, road density, river density, elevation, flow accumulation, lithology, soil type, soil thickness, curvature, normalized difference vegetation index (NDVI), slope, stream power index (SPI), land cover and topographic wetness index (TWI). Based on AUC value of parameters of landslides that are more than 0.6, the parameters that could be used in the study were flow direction, aspect, rainfall, distance from fault, distance from road, fault density, road density, density river, elevation, flow accumulation, lithology, soil type, NDVI, slope, stream power index (SPI), land cover and topographic wetness index (TWI). The validation results of the WoE method show that the AUC value for the success rate is 0.947 (excellent) and the AUC value for the prediction rate is 0.930 (excellent). The validation result of the IV method show the AUC value for the success rate is 0.944 (excellent) and the AUC value for the prediction rate is 0.931 (excellent). Based on landslide susceptibility zoning map using WoE method, it is shown that the percentage of the high vulnerability zone is 7.12%, the medium vulnerability zone is 8.72%, the low vulnerability zone is 6.78% and the very low vulnerability zone is 77.38%. Based on the landslide susceptibility zoning map using the IV method, it is shown that the percentage of the high vulnerability zone is 7.36%, the medium vulnerability zone is 9.14%, the low vulnerability zone is 10.06%, and the very low vulnerability zone is 73.44%.