Landslide susceptibility assessment using hybrid geospatial, frequency ratio, and AHP models in Souk Ahras province, Northeastern of Algeria

This study looks at how likely landslides are to happen in Souk Ahras, eastern Algeria. This area is known for having unstable slopes and landslides that damage property and crops and move people out of their homes. The study uses remote sensing and Geographic Information System (GIS) methods along with two predictive models, the Frequency Ratio (FR) and the Analytical Hierarchy Process (AHP), to find and map locations that are likely to have landslides. Nine things that could have caused the problem were looked at, such as slope, rock type, height, rainfall, and how far away it was from faults, rivers, and roadways. There were 248 historical landslide incidents found, and 60% of them were utilized to train the model and 40% to check its accuracy. Using ArcGIS 10.8, the landslide susceptibility index (LSI) was divided into five classes of susceptibility. IDRISI Selva 17.0 was used to figure out the weights of the factors. The FR model said that 26.03% of the region was very likely to be affected, while the AHP model noted that 9.06% was very likely to be affected and 4.63% was highly likely to be affected. The FR and AHP models were found to be 77% and 68% accurate, respectively, according to the validation analysis. The FR model did a better job of predicting what would happen because it used data from past landslides. On the other hand, subjective weight allocations affected how well the AHP model worked. These maps of landslide susceptibility are essential for managing land and preventing disasters. They help people in Souk Ahras make wise choices by showing them where areas are most in danger.