Understanding China's greening trends and drivers: Nonlinear responses and spatial heterogeneity revealed by interpretable machine learning

This dataset supports the study “Understanding China’s greening trends and drivers: Nonlinear responses and spatial heterogeneity revealed by interpretable machine learning.” It does not contain raw data, but rather processed and aggregated variables derived from multiple publicly available datasets (fully cited in the associated manuscript).

The dataset provides county-level statistics across China for the year 2022, including vegetation conditions and their potential driving factors. Vegetation dynamics are represented by the Kernel Normalized Difference Vegetation Index (KNDVI), calculated based on the MODIS/Terra Vegetation Indices product (MOD13A1). Environmental and socio-economic variables include surface solar radiation downward (SSRD), precipitation (Pre), temperature (Tem), digital elevation model (DEM), slope, slope direction (Dir), soil organic matter (SOM), Soil depth (SD), soil porosity (POR), soil pH, population (POP), land urbanization (LU), and nighttime light (NTL).

All variables have been spatially aggregated to the county scale to ensure consistency for subsequent statistical and machine learning analyses. This processed dataset is intended to facilitate reproducibility of the study and to support further research on human–climate interactions, nonlinear relationships, and spatial heterogeneity in vegetation dynamics across China.