Catastrophe theory and AHP modelling of landslide susceptibility using multi-source geodata in southwestern Nigeria

Abstract

Landslides are among the most destructive natural hazards in mountainous and structurally complex terrains, posing significant threats to lives, infrastructure, and sustainable development. Accurate Landslide Susceptibility Mapping (LSM) is therefore essential for hazard mitigation and land-use planning. This study develops a Catastrophe Theory (CT)-based Landslide Susceptibility Mapping Index (LSMI) that integrates surface and subsurface conditioning factors derived from geological, geomorphological, geophysical, and climatic datasets. Unlike conventional approaches that rely primarily on expert judgment or data-intensive machine-learning algorithms, the proposed framework employs Catastrophe Theory and fuzzy membership transformation to derive factor weights through catastrophe normalization functions, thereby reducing subjectivity in the weighting process. Nine landslide conditioning factors, including slope, elevation, lithology, rainfall, lineament density, drainage density, transverse resistance, coefficient of anisotropy, and bedrock topography, were integrated within a GIS environment to generate a CT-based susceptibility model and compared with an Analytical Hierarchy Process (AHP)-based model. The results revealed distinct factor-ranking patterns between the two approaches, reflecting their different weighting philosophies. While AHP emphasized surface geomorphological controls, particularly slope, CT assigned the highest weight to transverse resistance, highlighting the influence of subsurface weathering, groundwater conditions, and rock competence on slope instability. The resulting susceptibility maps were classified into five categories and validated using landslide inventory data and Receiver Operating Characteristic (ROC) analysis. The CT model achieved an Area Under the Curve (AUC) value of 0.74, outperforming the AHP model (AUC = 0.63). These findings demonstrate that integrating subsurface geophysical information with Catastrophe Theory provides a promising complementary framework for landslide susceptibility assessment in structurally complex Basement Complex terrains, although further validation in different geological settings is recommended. A combination of surface and subsurface geodata, including geoelectrical, geological, and remote sensing datasets, is compiled to develop the landslide susceptibility models. Employs Catastrophe Theory (CT) and Analytical Hierarchy Process (AHP) frameworks to generate spatial susceptibility maps from nine conditioning factors. CT–derived weighting identifies transverse resistance as the most influential parameter while AHP highlights slope as the dominant factor affecting landslide occurrence. Spatial mapping reveals zones of moderate to very high susceptibility that align with documented landslide locations in the Okemesi–Ekiti area, southwestern Nigeria. Validation using landslide inventory and AUC analysis demonstrates higher predictive accuracy for the CT model (74%) compared to the AHP model (63%). The CT–based model is recommended as a robust and less subjective approach for landslide hazard assessment, supporting targeted mitigation and land–use planning. A combination of surface and subsurface geodata, including geoelectrical, geological, and remote sensing datasets, is compiled to develop the landslide susceptibility models. Employs Catastrophe Theory (CT) and Analytical Hierarchy Process (AHP) frameworks to generate spatial susceptibility maps from nine conditioning factors. CT–derived weighting identifies transverse resistance as the most influential parameter while AHP highlights slope as the dominant factor affecting landslide occurrence. Spatial mapping reveals zones of moderate to very high susceptibility that align with documented landslide locations in the Okemesi–Ekiti area, southwestern Nigeria. Validation using landslide inventory and AUC analysis demonstrates higher predictive accuracy for the CT model (74%) compared to the AHP model (63%). The CT–based model is recommended as a robust and less subjective approach for landslide hazard assessment, supporting targeted mitigation and land–use planning.