Effects of changing spatial scale on debris-flow hazard assessment: A case study in the Dadu River basin, China

Debris flows are a prevalent mountain hazard that poses severe risks to human life and property. Debris-flow hazard assessments at the regional scale are vital for risk management, which establish spatial associations between debris flows and their influencing factors based on specific evaluation units. Different spatial scales of evaluation units can influence the spatial attributes and associations obtained by statistics, and further affect the accuracy of hazard assessments. However, there is limited consensus regarding the optimal spatial scale of evaluation units for debris-flow hazard assessment. To address this issue, six different scales of grid cells and forty influencing factors related to topography, material sources, hydrology, and human activities are analyzed by the geographical detector model to assess the debris-flow hazards in the Dadu River basin, China. The results reveal that over 92 % of debris-flow points fall within hazardous zones across all spatial scales, confirming the effectiveness of the assessment model. Topography, particularly local gully topography, dominates the debris- flow occurrence in the study area, while human activities also significantly contribute. As the spatial scale of evaluation units increases, the explanatory power of the influencing factors improves, with the 90 % quantile ranging from 0.23 to 0.46. This result suggests that larger spatial scales weaken the spatial characteristics of the factors. The finer and more informative the factors are, the more sensitive to spatial scale effects. The 10 km x 10 km is identified as the optimal spatial scale, which effectively preserves the local spatial characteristics while avoiding information loss or overload. These findings provide valuable insights for enhancing the accuracy of hazard assessments and improving the efficiency of risk management.