Spatiotemporal Variability of Vegetation Productivity Responses to Meteorological Factors in China's Drylands Over Two Decades

Abstract The role of meteorological factors in regulating gross primary productivity (GPP) is well‐recognized. However, the attribution of GPP trends predominantly focus on interannual timescales, potentially obscuring the differences in dynamic seasonal responses of GPP to meteorological variations. Here, a machine learning model to invert GPP in China's drylands using meteorological data including temperature, solar radiation, and precipitation from 2001 to 2020 was developed. The model was subsequently used to quantify the contributions of these meteorological factors to GPP trends and their regional variations. Results showed that declining solar radiation caused an average GPP decrease of 0.13 gC m −2 yr −1 , whereas increased precipitation contributed to an average GPP increase of 0.11 gC m −2 yr −1 , partially offsetting the radiation‐induced loss. Temperature effects were relatively minor due to seasonal compensations, with spring and autumn warming enhancing photosynthesis but summer warming suppressing vegetation growth. The combined effects of these three meteorological factors led to a slight decline of approximately 0.03 gC m −2 yr −1 in annual mean GPP across China's drylands. Spatially, GPP variations reflected the dominance of temperature in the Tibetan Plateau, precipitation in North China, and solar radiation in Northeast China. The Tibetan Plateau's low baseline temperature allowed warming to enhance GPP throughout the year. In Northeast China, the significant decline in solar radiation and the radiation sensitivity of forest canopies explained the observed GPP reduction. In North China, pronounced wetting trends established precipitation as the dominant controlling factor. These findings enhance our understanding of climate‐GPP relationships in dryland ecosystems.