In order to study the multidimensional dynamic evolution of soil erosion under the influence of multi-factors and spatial variability of the subsurface, to realize high-precision simulation of water and sand erosion process in the watershed, to comprehensively analyze the spatial and temporal distribution characteristics of water and sand migration and to predict the occurrence of erosion risk. In this paper, we propose a modified general soil loss model equation RUSLE, water erosion prediction model WEPP and watershed soil and water evaluation model SWAT coupled watershed soil and water erosion remote sensing monitoring method, Dazu section of the Laixi River watershed as the study area, the coupled model from the surface, line and point scales to comprehensively invert the watershed water and sand from the subsurface into the slope of the fine gully, with the watershed to the spatial-temporal dynamics of the process of sedimentation in the river channel, the study shows that Based on the measured hydrological data, the multi-model coupled risk assessment of soil erosion proposed in this paper is more comprehensive than a single model, the prediction results of RUSLE model have strong consistency with the actual erosion law, the Nash efficiency coefficient NSE and the deterministic coefficient R2 of the parameterization of SWAT model have reached more than 0.6, and the evaluation factors of uncertainty analysis P-factor and R-factor are also more satisfactory, and the simulation prediction results of RUSLE model are better than that of the actual model. The evaluation factors of uncertainty analysis P-factor and R-factor are also better, and the simulation prediction results have better robustness. Secondly, there are strong correlations between any two of the three models, and the WEPP and SWAT models have reached a highly significant correlation level, with an R2 of 0.96; the R2 of the RUSLE and WEPP models is 0.77; and the R2 of the RUSLE and SWAT models is 0.58.Furthermore, in the erosion models at different scales, there are certain influences of factors on the simulation accuracy, and in the case of the Setsui River, there are some effects of factors on the simulation accuracy, such as the NSE and R-factor. certain influence, in the Laixi River watershed rainfall and elevation has a direct effect on soil erosion, with the increase in elevation caused by the increase in the risk of surface soil erosion, but elevation on the watershed slopes and channels of the water and sand erosion has less influence, however, vegetation has a positive effect on slowing down the water and sand erosion of the slopes and channels, but the better the vegetation cover, the more susceptible the soil erosion is to be affected by the smaller.