A landslide susceptibility model using the Analytical Hierarchy Process method and multivariate statistics in perialpine Slovenia

摘要: Abstract Landslides cause damage to property and unfortunately pose a threat even human lives. Good landslide susceptibility, hazard, risk models could help mitigate or avoid the unwanted consequences resulted from such hillslope mass movements. For purpose of susceptibility assessment study area in central Slovenia was divided 78 365 slope units, for which 24 statistical variables were calculated. land-use vegetation data, multi-spectral high-resolution images merged using Principal Component Analysis method classified with an unsupervised classification. Using multivariate analysis (factor analysis), interactions between factors distribution tested, importance individual occurrence defined. The results show that slope, lithology, terrain roughness, cover type play important roles susceptibility. other spatial varies depending on type. Based several developed Analytical Hierarchy Process method. These gave very different results, prediction error ranging 4.3% 73%. As final result research, weights best derived AHP probability measures, potentially hazardous areas located relation population road distribution, hazard classes assessed.