Quantifying vegetation effects on landslide hazard using a physically based model: implications for catchment-scale prioritization and management

Vegetation plays a critical role in slope stability, yet its effects on landslide hazard remain difficult to quantify at spatial scales relevant to management. This study evaluated the influence of vegetation-induced root cohesion on shallow landslide hazard and developed a catchment-scale prioritization framework for landslide risk management in Chunyang-myeon, Bonghwa-gun, the Republic of Korea, using the physically based revised TRIGRS model. To represent uncertainty in soil depth and isolate the mechanical contribution of vegetation, 18 simulation scenarios were constructed by combining three uniform soil-depth scenarios (1, 2, and 3 m) with six root cohesion scenarios (0, 1, 2, 3, 4, and 5 kPa). Model outputs were classified into four hazard classes ( D = 1, D = 2, D = 3, and D 3) based on the minimum soil-depth condition required for instability. Under the baseline scenario (Cr = 0 kPa), the two highest hazard classes, D = 1 and D = 2, occupied 33.1 and 31.8% of the study area, respectively. Comparison with the 2008 landslide inventory showed that 94.2% of observed landslides were located within these two classes, indicating strong spatial consistency between modeled hazard patterns and historical landslide occurrence. Increasing root cohesion substantially reduced high-hazard areas: the D = 1 class decreased by approximately 89% at Cr = 2 kPa and was nearly eliminated at Cr ≥ 3 kPa. Critical root cohesion analysis further showed that 96.3% of the study area could be downgraded from the highest hazard class at Cr ≤ 2 kPa, whereas 52.8% did not reach the stable class ( D 3) even at Cr = 5 kPa. Catchment-scale prioritization across 745 forest catchments revealed a pronounced trade-off between the effectiveness of vegetation-based mitigation and the need for structural intervention ( r = −0.740, p 0.001). These findings indicate that vegetation reinforcement can effectively reduce severe landslide hazard across much of the study area, while structural measures remain necessary in selected catchments. The proposed framework provides a practical basis for integrating forest management and structural mitigation in regional landslide risk reduction planning.