Mapping long-term irrigated cropland dynamics in the Yellow River Basin by integrating a nonlinear LST–NDVI feature space with precipitation constraints
Remote sensing analysis uncovers stable basin-wide irrigated cropland with localized upstream expansions, indicating reliable monitoring for regional water resource management.
Key Points
To establish an accurate, long-term monitoring framework for mapping irrigated cropland dynamics across the climatically heterogeneous Yellow River Basin from 2002 to 2020.
Constructed climate-zone-specific quadratic polynomial dry and wet edge equations to capture nonlinear relationships in the LST–NDVI feature space.
Integrated MODIS-derived Temperature Vegetation Dryness Index (TVDI) with CHIRPS precipitation using a 7-day cumulative threshold to separate irrigation signals from rainfall events.
Quadratic dry–wet edge fitting significantly outperformed linear fitting (p < 0.01), showing an average R² increase of 18.6% and overall R² values exceeding 0.7.
Basin-wide irrigated cropland remained stable between 6,500 × 10³ hm² and 7,800 × 10³ hm², with upstream expansions in the Hetao (11.37 × 10³ hm²/year) and Qingtongxia (2.08 × 10³ hm²/year) irrigation districts.