Abstract: ObjectiveThe "National Key Soil Erosion Control Area in the Lower Reaches of the Jinsha River" is a critical segment of China's ecological functional zoning, where severe soil erosion poses significant challenges to local agricultural production and ecological security. This study aims to reveal the spatiotemporal evolution patterns and driving mechanisms of soil erosion under the unique geomorphological conditions of the dry-hot valley. MethodsFocusing on Yuanmou County in the Jinsha River dry-hot valley, we employed the RUSLE model framework, integrating satellite remote sensing imagery with in-situ runoff plot measurements to comprehensively assess soil erosion across the entire county and its cultivated land.Results(1) Model validation showed a coefficient of determination (R²) of 0.871 between RUSLE predictions and measured values, confirming the model's applicability in the dry-hot valley region. (2) Temporal analysis revealed a 2.9% decrease in total erosion area from 2018 to 2021, but a sharp increase to 47% in 2022, indicating a significant intensification of erosion. (3) Soil erosion on cultivated land showed an overall declining trend, particularly in areas with mild and moderate erosion. Spatial autocorrelation analysis demonstrated significant clustering of high- and low-erosion zones under varying topographic conditions, highlighting the strong influence of terrain and vegetation on erosion patterns.ConclusionThese findings underscore the importance of appropriate farming practices and land conservation measures in mitigating soil erosion. The results provide a scientific basis for soil erosion assessment in Yuanmou County and similar regions, offering valuable support for the rational implementation of soil and water conservation measures and further optimization of erosion models.