International Journal on Advanced Science, Engineering and Information Technology, Vol. 12 (2022) No. 4, pages: 1462-1468, DOI:10.18517/ijaseit.12.4.16221

GIS-Based Binary Logistic Regression for Landslide Susceptibility Mapping in the Central Part of Pacitan Regency, East Java, Indonesia

Muhammad Dimyati, Iqbal Putut Ash Shidiq, Dimas Bayu Ichsandya


Landslides are among the most hazardous phenomena in the Pacitan Regency, especially in the Sub-Districts of Pacitan, Kebonagung, Tulakan, and Arjosari, where the landslide mainly occurs. Strategic planning through GIS analysis can be applied to minimize potential losses and strengthen resilience to natural disasters. This study combined the binary logistic regression method and GIS to map the landslide susceptibility in the Sub-Districts of Pacitan, Kebonagung, Tulakan, and Arjosari, Pacitan Regency, East Java, Indonesia. An inventory map of 293 landslides was randomly divided into 80%-20% basis for model training and testing. Fourteen landslide conditioning factors including elevation, slope, aspect, plan curvature, profile curvature, topographic wetness index (TWI), land use, proximity to roads, proximity to rivers, proximity to faults, soil types, lithology, normalized difference vegetation index (NDVI) and rainfall was used. Analysis shows that fourteen landslide conditioning factors are contributed to 22.7%. The analysis shows that 36.59% or 17,734.95 Ha of the study area has high-very high susceptibility. The area of high-very high susceptibility is mainly located in the western part of the study area. It is related to high slope value and volcanic and sediment-volcanic rock from the formation of Arjosari and Mandalika. The validation using AUROC showed an excellent fit of 0.806. Validation of susceptibility map using testing data showed 0.711 accuracy value and 0.694 precision value, which meant that the susceptibility model was quite sensible. This information could be helpful to support the local government for hazard mitigation efforts.


Landslide susceptibility mapping; GIS; binary logistic regression.

Viewed: 93 times (since abstract online)

cite this paper     download