Exploring the spatiotemporal distribution characteristics and driving factors of water erosion in mountain area based on RUSLE-SDR

Abstract

Quantifying the contributions of driving factors and analyzing dynamic changes of water erosion in mountain areas are crucial for water erosion control and sustainable soil resource utilization. In this study, the Revised Universal Soil Loss Equation (RUSLE) and Sediment Delivery Ratio (SDR) model were integrated, and the Geographically Weighted Regression (GWR) and path analysis models were used to explore the contributions and interactions of key influencing factors (precipitation, NDVI, slope, soil moisture) on water erosion in Longnan City. The results showed that the RUSLE-SDR model could simulate the water erosion process effectively in Longnan City from 2000 to 2020 (R2 = 0.821, NSE = 0.67). The spatial and seasonal distribution of water erosion intensity was consistent with precipitation, showing the characteristics of weak in northwest and strong in southeast, and summer is the most serious period of water erosion. The GWR and path analysis models revealed that vegetation and slope were the main influencing factors of water erosion, and they had a strong interaction. When NDVI was below 0.67, slope had a direct impact on water erosion; when NDVI was between 0.67 and 0.82, slope and vegetation jointly influenced water erosion; and when NDVI was above 0.82, vegetation became the dominant factor, while slope indirectly affected erosion by regulating vegetation cover. Precipitation was the main factor that influenced erosion when the rainfall was less than 550 mm, but when the rainfall exceeded 550 mm, it exhibited a strong inhibitory effect on erosion through vegetation. This study reveals water erosion’s driving mechanisms in mountain areas and provides soil erosion control measures’ implementation with a scientific basis and theoretical support.