Evidence-based Modeling of Landslide Hazard in Wahig-Inabanga Watershed, Bohol, Philippines using GIS and Statistical Models

Abstract

The study combined geographic information system (GIS) and statistical models in predicting landslide hazard in Wahig-Inabanga Watershed, Bohol, Philippines. The bivariate statistical analysis (BSA) and logit regression (LR) were employed for class and factor weighting, respectively, to determine landslide prone areas using eleven significant landslide-related instability factors such as elevation, slope, aspect, lithology, soil order, soil type, fault line proximity, river proximity, road proximity, rainfall, and land cover. The very satisfactory results of model evaluation warranted the application of the LR model in evidence-based landslide hazard assessment. Out of the eleven instability factors, only soil order and soil type were determined not significant. The first three most important instability factors based on the values of regression coefficients were elevation, slope, and lithology. Landslide hazard assessment revealed around 7,063 ha or 11.33% of the total area of the watershed had high to very high landslide hazard ratings. The study showed that GIS, in tandem with useful models, provided pertinent results which could be used as scientific basis for watershed management and land use planning in relation to landslide disaster risk reduction and management.