An Improved Unascertained Measure-Set Pair Analysis Model Based on Fuzzy AHP and Entropy for Landslide Susceptibility Zonation Mapping

Landslides are one of the most destructive and common geological disasters in the Tonglvshan mining area, which seriously threatens the safety of surrounding residents and the Tonglvshan ancient copper mine site. Therefore, to effectively reduce the landslide risk and protect the safety of the Tonglvshan ancient copper mine site, it is necessary to carry out a systematic assessment of the landslide susceptibility in the study area. Combining the unascertained measure (UM) theory, the dynamic comprehensive weighting (DCW) method based on the fuzzy analytic hierarchy process (AHP)-entropy weight method and the set pair analysis (SPA) theory, an improved UM-SPA coupling model for landslide susceptibility assessment is proposed in this study. First, a hierarchical evaluation index system including 10 landslide conditioning factors is constructed. Then, the dynamic comprehensive weighting method based on the fuzzy AHP-entropy weight method is used to assign independent comprehensive weights to each evaluation unit. Finally, we optimize the credible degree recognition criteria of UM theory by introducing SPA theory to quantitatively determine the landslide susceptibility level. The results show that the improved UM-SPA model can produce landslide susceptibility zoning maps with high reliability. The whole study area is divided into five susceptibility levels. 5.8% and 10.16% of the Tonglvshan mining area are divided into extremely high susceptibility areas and high susceptibility areas, respectively. The low and extremely low susceptibility areas account for 30.87% and 34.14% of the total area of the study area, respectively. Comparison with the AHP and Entropy-FAHP models indicates that the improved UM-SPA model (AUC = 0.777) shows a better performance than the Entropy-FAHP models (AUC = 0.764) and the conventional AHP (AUC = 0.698). Therefore, these results can provide reference for emergency planning, disaster reduction and prevention decision-making in the Tonglvshan mining area.