Geological hazard susceptibility assessment under land use change: a case study of Dongchuan District, Kunming, Yunnan, China

To quantitatively reveal the dynamic impact of land use change on the susceptibility to geologic hazards. Based on Landsat images from 2000 to 2020, the relationship between land use change and geological hazard susceptibility in Dongchuan District, Kunming City, Yunnan Province, China, was analyzed. Land use types were extracted using the Support Vector Machine (SVM) method. Geological hazard susceptibility was assessed using machine learning, and impacts were comprehensively analyzed. Results show: (1) Geological hazard frequency is significantly higher in areas with land use changes than in unchanged areas. (2) Land use changes from grassland to farmland and forest to grassland increased geological hazard risk in Dongchuan District. (3) Conversions to grassland and farmland showed lagged effects on hazard occurrence, while forest conversion showed immediate effects, though the Mann–Whitney U test (U = 28.5, p = 0.620) indicated no significant difference in frequency ratio distributions between contemporaneous and time-lag analyses.