Quantifying vegetation cover effects on overland flow velocity under rainfall conditions

Flow velocity is a key parameter characterizing the hydrodynamic properties and erosion processes of overland flow. However, measuring and estimating flow velocities on vegetated slopes remains challenging. To systematically investigate how variable vegetation cover affects overland flow velocity ( V ), controlled artificial rainfall experiments were conducted using six vegetation coverage levels (0-70%), five rainfall intensities (60-120 mm·h -1 ), and three slope gradients (5°-15°). The results reveal that 60% vegetation coverage ( C v ) represents a critical threshold for reducing V , decreasing it by 29.25-39.30% compared to bare slopes. The response of V to rainfall intensity ( I ) and slope gradient ( S ) follows power-law relationships, with exponents of 0.4 and 0.3, respectively. With the increase of C v from 0 to 70%, the flow velocity coefficient ( η ) decreased by 37%. The structural equation model (SEM) revealed that C v , S , and I exerted both direct effects on V and indirect effects on V by influencing η . The direct effects of C v , S , and I on η were −0.869, −0.169, and −0.044, respectively, while their total effects on V were −0.702, 0.453, and 0.381, respectively. A predictive model of V incorporating C v , S , and I was developed. It exhibited excellent performance (adjusted coefficient of determination = 0.943, Nash efficiency coefficient = 0.913, relative root mean square error = 0.062) and was further validated against datasets from distinct vegetation species. These results deliver robust support for calculating erosion rate-driving factors (i.e., shear stress or stream power) in soil erosion models, and can be directly applied to the practices of vegetation configuration optimization and soil erosion risk assessment.