Remote Sensing Diagnosis of Vegetation Dynamics in the Three-North Shelterbelt Program Area
Abstract
Vegetation is a key component of terrestrial ecosystems, and its dynamics are jointly influenced by climatic and nonclimatic factors. Quantifying the effects of nonclimatic factors remains challenging but is essential for understanding vegetation change and supporting ecological management. Using MODIS normalized difference vegetation index (NDVI) and meteorological data from 2000 to 2024, this study analyzed vegetation changes in the Three-North Shelterbelt Program (TNSP) region. An improved NDVI residual trend analysis was applied to distinguish vegetation variation explained by selected climatic variables from residual changes associated with nonclimatic or unexplained influences. Compared with the traditional precipitation-based residual framework, the improved method incorporates both temperature and precipitation and combines NDVI trends, climate-NDVI regression significance, and residual trends to identify climate-associated, residual-dominated, and jointly driven vegetation changes. The results show that the regional mean summer maximum NDVI increased from approximately 0.30 in 2000 to 0.37 in 2024. Approximately 44.83% of the study area experienced significant vegetation recovery, including 4.01% climate-associated, 27.54% residual-dominated, and 13.28% jointly driven areas, with residual-dominated recovery representing the largest attribution category. Precipitation was more spatially extensive than temperature as the identified dominant climatic factor, particularly in arid and semi-arid areas. Spatially, residual-dominated recovery was mainly concentrated in the northeastern TNSP region, the Loess Plateau, and northwestern oasis areas. By explicitly identifying jointly driven vegetation changes, the framework classified 13.73% of the TNSP region as jointly driven, providing a more quantitative basis for vegetation attribution and sustainable ecological management in arid and semi-arid northern China.