Spatiotemporal Dynamics of Soil Erosion and Its Interactive Mechanisms with NDVI in the Yellow River Basin, China

Land Use/Land Cover (LUCC) Data Annual land use/cover maps for 1995, 2000, 2005, 2010, 2015, and 2020 were obtained from the Resource and Environment Data Cloud Platform (https://www.resdc.cn/ ) developed by the Chinese Academy of Sciences. These maps were derived from Landsat TM/ETM+/OLI imagery with an overall accuracy exceeding 90%. Considering the relatively small interannual variation, the land use data for 1995, 2000, 2005, 2010, 2015, and 2020 were used to represent adjacent years for SE estimation. Vegetation Index (NDVI) Data The MODIS NDVI product (MOD13A3, 1 km, monthly, 2001–2022) was acquired from the NASA Earthdata portal (https://earthdata.nasa.gov/ ). Monthly data were composited to annual mean NDVI to analyze vegetation cover dynamics. The Sen’s slope estimator and Mann–Kendall test were applied to detect long-term trends and change points in vegetation recovery. Digital Elevation Model (DEM) Elevation data with a 30-m resolution were sourced from the Shuttle Radar Topography Mission (SRTM) provided by the United States Geological Survey (USGS, https://earthexplorer.usgs.gov/ ). Slope and flow direction were derived from the DEM to support soil erosion modeling in the InVEST SDR module. Meteorological Data Daily precipitation and temperature data (1995–2022) were obtained from the China Meteorological Data Service Center (CMDC, https://data.cma.cn/ ). Annual mean precipitation was interpolated using the Inverse Distance Weighting (IDW) method to match the spatial resolution of other datasets. Soil Data Soil texture and organic matter content were derived from the Harmonized World Soil Database (HWSD v1.2), complemented by data from the National Second Soil Survey of China. These parameters were used to calculate soil erodibility factors required by the InVEST SDR model.土地利用/土地覆被（Land Use/Land Cover, LUCC）数据 1995、2000、2005、2010、2015及2020年的年度土地利用/覆被图，源自中国科学院搭建的资源环境数据云平台（https://www.resdc.cn/）。该系列图件基于Landsat TM/ETM+/OLI遥感影像解译生成，总体分类精度超过90%。考虑到各年份土地利用数据的年际波动相对较小，本次研究采用上述年份的土地利用数据代表相邻年份，用于SE估计。 植被指数（Vegetation Index, NDVI）数据 本研究使用的MODIS NDVI产品（MOD13A3，空间分辨率1km，月度数据，时间跨度为2001-2022年），获取自NASA地球数据门户（https://earthdata.nasa.gov/）。为分析植被覆盖动态变化，将月度NDVI数据合成为年度平均NDVI；并采用森斜率估计器（Sen’s slope estimator）与曼-肯德尔检验（Mann–Kendall test），识别植被恢复过程中的长期趋势与突变点。 数字高程模型（Digital Elevation Model, DEM） 30m分辨率的高程数据，源自美国地质调查局（United States Geological Survey, USGS）提供的航天飞机雷达地形测绘任务（Shuttle Radar Topography Mission, SRTM）数据集。基于该DEM数据提取坡度与流向信息，用于支撑InVEST SDR模块中的土壤侵蚀建模工作。 气象数据 1995-2022年的日降水量与日气温数据，获取自中国气象数据服务中心（China Meteorological Data Service Center, CMDC，https://data.cma.cn/）。为匹配其他数据集的空间分辨率，采用反距离权重（Inverse Distance Weighting, IDW）法对年平均降水量进行空间插值处理。 土壤数据 土壤质地与有机质含量参数源自世界土壤数据库（Harmonized World Soil Database v1.2, HWSD v1.2），并补充了中国第二次全国土壤普查的相关数据。上述参数用于计算InVEST SDR模型所需的土壤可蚀性因子。