Spatiotemporal Evolution and Drivers of Soil Erosion in the Yarlung Tsangpo River Basin: A RUSLE-Based 30-Year Analysis and Future Scenario Projections

This dataset includes gridded soil erosion data for the Yarlung Tsangpo River Basin from 1990 to 2019, derived using the Revised Universal Soil Loss Equation (RUSLE) model. It also contains projected rainfall erosivity (R-factor) rasters for four future periods (2021–2040, 2041–2060, 2061–2080, and 2081–2100) under three CMIP6 emission scenarios (SSP126, SSP245, and SSP585), based on the bias-corrected outputs of 25 GCMs using the quantile delta mapping (QDM) method. In addition, the dataset includes C-factor rasters simulated using the XGBoost machine learning algorithm, trained with historical WordClim climate variables for each future scenario and time slice. Finally, projected average soil erosion rasters for each period and scenario across the entire basin are also provided.

本数据集涵盖1990年至2019年雅鲁藏布江流域（Yarlung Tsangpo River Basin）的网格状土壤侵蚀数据，该数据集通过修正通用土壤流失方程（Revised Universal Soil Loss Equation, RUSLE）模型计算得到。此外，数据集还包含基于25个全球气候模式（General Circulation Model, GCM）经分位数增量映射（Quantile Delta Mapping, QDM）方法进行偏差校正后的输出结果，由此生成三种CMIP6排放情景（SSP126、SSP245、SSP585）下四个未来时段（2021–2040年、2041–2060年、2061–2080年及2081–2100年）的预测降雨侵蚀力（R因子）栅格数据。除此之外，数据集还包含采用XGBoost机器学习算法模拟得到的C因子栅格数据，该算法以各未来情景与时段对应的历史WorldClim气候变量进行训练。最后，数据集还提供了全流域各时段与排放情景下的预测平均土壤侵蚀栅格数据。