The use of geospatial data from GIS in the quantitative analysis of landslides

Abstract

Abstract This study was conducted in order to compare two advanced technique used in establishing landslides susceptibility maps. The study considers a method of landslides analysis using the analytical hierarchy process (AHP) to check the occurrence of landslides in the study area through the establishment of a landslides susceptibility map based on the causative factors of landslides in the area. To further check and validate the process, it was compared with a more recent approach that is the soft computing (machine learning) technique. After the comparison, the enhanced analytical hierarchy process performed wonderfully well but not better than the machine learning method of analysis. Using the AHP methods, it was able to identify rainfall precipitation to be the major trigger mechanisms while 12 other conditioning factors were also identified. From the results obtained, it was observed that a good portion of the study area can be said to be susceptible to landslides. The analysis suggested that though the slides were fully triggered by rainfall precipitation, other factors such the geological and hydrological conditions facilitate the rapid occurrences of the phenomenal landslides in the study area. Validation was carried out by comparison of obtained results with inventories.