Widespread increasing negative vegetation sensitivity to phenology reshapes GPP dynamics in alpine grassland

Satellite evidence indicates a significant phenological changes in alpine vegetation. However, the effect of phenological changes on vegetation gross primary productivity (GPP) is still poorly understood, which limits understanding of climate changes. Here, we analyze the impact of different phenological factors on alpine grassland GPP based on sensitivity coefficient obtained from explainable machine learning. We found that alpine grassland GPP largely exhibits negative sensitivity to the duration of spring phenology and positive sensitivity to the duration of autumn phenology. Regions with negative sensitivity show an increasing trend in GPP sensitivity to both spring and autumn phenological durations, whereas areas with positive sensitivity indicate a decreasing sensitivity trend. This trend change pattern is related to the warm and humid climate of the Northern Tibetan Plateau. Results indicate a warning sign that phenological changes potentially limit the carbon benefits of alpine grassland derived from ongoing climate changes. This study comprehensively reveals the impact of phenological changes on alpine grassland. Research findings bear significant theoretical implications for the adoption of carbon neutrality strategies in alpine regions.