A 30-meter resolution LS-factor dataset for the China Region

<p indent="0mm">Soil erosion is a significant environmental issue affecting global land degradation and sustainable development. Soil erosion models can estimate the spatial and temporal patterns of soil loss, guiding the formulation of watershed-level policies for soil and water conservation, as well as the implementation of effective strategies. The Revised Universal Soil Loss Equation (RUSLE) model, known for its simplicity and accuracy, has become one of the most widely used soil erosion models globally. Among all the input parameters of the RUSLE model, the topographic factor (LS-factor) has the most significant impact on soil erosion potential. The input data quality and calculation methods used to compute the LS-factor directly influences the overall quality of RUSLE model calculations. To address this, we developed a 30-meter resolution LS-factor dataset for the China region (LS_China). The dataset was computed using open-source tools, SAGA (System for Automated Geoscientific Analyses) and GDAL (Geospatial Data ion Library), based on publicly available SRTM 30-meter elevation data (SRTMGL1), covering the entire region of China. The data is organized into individual raster tiles, each covering a 1°×1° area, and stored in Geotiff format. In the computation process at large scales, neighborhood-dependent calculation methods were employed on the basis of spatial data decomposition, alongside a multiple flow direction algorithm, so as to ensure the accuracy of LS-factor calculations. In terms of data quality control, we validated the dataset using the coefficient of variation (CV) and compared it with other datasets for the same region. The results showed that the LS_China dataset has a CV of 1.24, indicating lower internal heterogeneity and demonstrating its high quality and reliability. This dataset serves as input for soil erosion assessments at various scales, ranging from local to regional and national levels.