Decreasing soil erosion in South China with uncertainties driven by NDVI estimates

• Soil erosion in South China is decreasing. • Soil erosion in South China is dominated by slight and mild erosion. • NDVI leads to substantial differences in soil erosion modulus. • AVHRR NDVI dataset indicates more severe soil erosion than the GIMMS NDVI dataset. Investigating soil erosion in South China, one of China’s most populous coastal regions, is crucial for understanding regional soil and water conservation, preventing soil degradation, and safeguarding food security. However, significant discrepancies persist among existing estimates of soil erosion, calling for further evaluation of long-term trends in its magnitude and spatial–temporal dynamics. This study utilized the Revised Universal Soil Loss Equation (RUSLE) model to assess soil erosion dynamics over 35 years, and further evaluated the influence of the Normalized Difference Vegetation Index (NDVI). Our findings revealed a general decline in soil erosion across South China, dominated by slight and mild erosion. However, the spatiotemporal patterns exhibited marked variations depending on NDVI datasets, particularly in interannual fluctuations and spatial discrepancies. The soil erosion modulus estimated from AVHRR NDVI demonstrated higher values and greater variability than those based on GIMMS NDVI. Spatially, three out of five datasets indicated a consistent reduction in erosion intensity, while two AVHRR datasets showed an initial decline followed by a resurgence over the past decade. Variations in NDVI data can lead to order-of-magnitude differences in soil erosion estimates, highlighting the need for careful dataset selection for soil erosion analysis. A comprehensive analysis and understanding of these differences are needed to provide valuable insights into the applicability of various NDVI datasets in future soil erosion modeling and risk assessment.