Mapping of Landslide Prone Areas in the Lisu River Basin Barru Regency Based on Binary Logistic Regression
IMAM GAZALI SUDARMAN and ASMITA AHMAD.

DEPARTEMEN ILMU TANAH
FAKULTAS PERTANIAN
UNIVERSITAS HASANUDDIN

Abstract

Barru Regency Government once issued an emergency response status for floods and landslides on December 28, 2018. According data from the Barru Regency Regional Disaster Management Agency, from 2013 to 2019 there were always landslides in the Districts of Tanete Riaja, Pujananting, Tanete Rilau, and Barru. This area is included in the Lisu Watershed. This will produce a map of the distribution of landslide prone areas in the Lisu watershed. This study uses Binary Logistic Regression (BLR) and NDVI (Normalized difference vegetation index) analysis. The data used are landslide points, soil types, lithology, slopes, land use, rainfall, soil texture, and distance from the river. 137 landslide points were recorded that were successfully obtained through field surveys and NDVI image analysis. The research area has a type B climate according to the Schmidth-Ferguson classification system. The variables of rainfall, land use class, lithology type, and slope class have a significant effect on regression modeling with significance values of 0.042, 0.000, 0.003, and 0.000 respectively. Variables of rainfall, distance from the river, slope class, lithology type, and land use class have a significant effect on regression modeling. Landslides occur on slopes from a rather steep to very steep slopes. A total of 60 landslides occurred in the Camba Formation. A total of 93 landslides occurred in the scrub land use class. Landslide prone areas with a slightly hazardous class covering an area of 4386 hectares, 4031 hectares of prone, and 4275 hectares of very prone areas which are generally scattered in the southern region of the Lisu River Basin.

Keywords: Landslide, Binary Logistic Regression, Lisu Watershed, Barru

Topic: Geospatial Agriculture