Satellite-Based monitoring of Agricultural Drought and its effects on vegetation dynamics in the Yellow River Basin

We constructed a consistent multi-source dataset for the Yellow River Basin from 2000 to 2023. Canopy-scale photosynthetic activity was characterized using the machine-learning-based Global OCO-2 SIF product (0.05°). Agricultural drought intensity and its spatial heterogeneity were quantified using a monthly Temperature–Vegetation Drought Index (TVDI, 1 km) derived from MODIS land surface temperature and NDVI, and independently validated with FLDAS surface soil moisture (0–10 cm, 0.1°). Land-use and land-cover data (1 km) were used to distinguish cropland, forest, and grassland ecosystems, while high-resolution precipitation and air temperature (1 km), MODIS evapotranspiration (500 m), and TerraClimate vapor pressure deficit (~4 km) were integrated to resolve climatic forcing and climate–water–vegetation coupling mechanisms. The dataset reveals strong spatial contrasts in drought severity, pervasive lagged and cumulative vegetation responses, declining SIF sensitivity under extreme drought, and substantially higher vulnerability of forests and grasslands relative to croplands, indicating that drought effects on vegetation productivity are mediated mainly through indirect regulation of evapotranspiration and surface moisture conditions rather than direct thermal stress alone.Method: integrated quantification using trend analysis, drought sensitivity assessment, and structural equation modeling.