Spatial patterns and drivers of temperature sensitivity of winter evergreen vegetation on the Tibetan Plateau

Climate warming has enhanced the winter greenness of evergreen vegetation on the Tibetan Plateau over the past 20 years. This response is primarily determined by the temperature sensitivity of vegetation, rather than the absolute extent of warming. However, how these factors shape the spatial patterns of sensitivity of vegetation to temperature (abbreviated as vegetation sensitivity) remains poorly understood. In this study, we analyzed the spatial pattern of vegetation sensitivity and its driving factors by using partial least squares regression, pixel-scale multiple linear regression, zone-level linear mixed models, and random forest model based on the winter normalized vegetation index (NDVI_winter) and multiple driving factors datasets. We found that the dominant factors influencing winter vegetation greenness are growing-season temperature and precipitation. The 72.5% of pixels exhibited positive vegetation sensitivity, with higher values in the south of the Himalayas and southeast edge. Vegetation sensitivity varied significantly across aridity zones, with the highest and lowest values in humid and arid zones, respectively. Furthermore, mixed forests showed the highest sensitivity. Random Forest analysis revealed that downward shortwave radiation (Ssrd) and the aridity index (AI) may contribute to spatial variability in NDVI_winter sensitivity to temperature, while the Ssrd and elevation were the most important environmental factors in humid and arid zones, respectively. Partial dependence plots indicated nonlinear relationships between key driving factors and NDVI_winter sensitivity, with precipitation and AI showing negative associations, while mean temperature exhibited a positive trend. The elevation and Ssrd displayed threshold-like responses, with sharp transitions in their influence on vegetation sensitivity. These findings clarify the complex regulatory roles of radiation, water availability, temperature, and topography in shaping the spatial heterogeneity of winter vegetation temperature sensitivity, and improve our understanding of alpine ecosystem responses to climate change.