Fan-shaped landforms, particularly fluvial fans, are inherently prone to multiple geohazards because steep slopes, weak lithology, sparse vegetation, and rapid runoff response often coincide within limited spatial areas. This study evaluates multi-hazard susceptibility in the Crnička Reka catchment, eastern North Macedonia, by comparing GIS-based parametric models with FAHP-based multi-criteria analysis. Excessive soil erosion, landslide susceptibility, and flash-flood susceptibility were first assessed individually using the Erosion Potential Model, Landslide Susceptibility Index, and Flash Flood Potential Index, and were then integrated through spatial overlay to identify multi-hazard zones. The EPM results indicate an average erosion coefficient of Z = 0.5, with total erosion production of 5,304 m³/year and a specific erosion rate of 780.9 m³/km²/year, mainly concentrated in the lower catchment. LSI classified 41.7% of the catchment as high to very high landslide susceptibility, while FFPI identified 57.4% as high to very high flash-flood susceptibility. The parametric multi-hazard overlay delineated 6.3% of the catchment as overlapping high-susceptibility zones, mainly in the central and lower sections. FAHP identified broader susceptibility patterns, including 19.1% high erosion susceptibility, 32.7% high to very high landslide susceptibility, 44.8% high to very high flash-flood susceptibility, and 15.4% multi-hazard overlap. ROC-AUC validation showed higher predictive performance for the parameter-based models, with 85.1% for LSI and 84.2% for FFPI, compared with 76.7% and 72.2% for the corresponding FAHP models. These results indicate that parametric GIS models are more effective for hazard-specific prediction, whereas FAHP is more sensitive to combined conditioning factors and provides complementary information for multi-hazard interpretation. The approach should be interpreted as an overlay-based susceptibility framework rather than a dynamic cascading-hazard model. This multi-method approach enhances the robustness of multi-hazard assessment and informs integrated catchment management. Future integration of LiDAR and UAV-based monitoring could further improve understanding of sediment transport, slope instability, and flash-flood dynamics.
