Prediction model for debris flow runout distance based on dynamic parameters

Traditional methods for predicting debris flow runout distances mainly employ static analysis to establish empirical formulas, which cannot explain why debris flows with identical volumes and initial elevations exhibit different runout distances due to their neglect of dynamic parameters. This study conducted flume experiments under controlled initial elevation and channel slope conditions. Using the control variable method, we systematically investigated the effects of volumetric solid fraction, particle size, and flow volume on debris flow velocity and depth, and further analyzed their influence on runout distance. The results show that: (1) flow velocity exhibits correlation coefficients of -0.68 with solid volume fraction and -0.23 with particle size; (2) flow depth demonstrates correlation coefficients of 0.61 with solid volume fraction and 0.31 with particle size; (3) a machine learning-based prediction model for runout distance was established using dimensionless numbers to incorporate velocity and depth parameters. The findings quantitatively clarify the influence of dynamic parameters on runout distance and provide a scientific basis for more accurate runout distance prediction and optimal mitigation planning on debris flow fans.