Highlights What are the main findings? Future NDVI is projected to increase persistently in the central and eastern Tibetan plateau but decrease along northern and southeastern margins, with variability in trend projections among different models. The multi-model ensemble indicates an overall NDVI increase in the future, with higher values under SSP-245 before the 2060s and stronger increases under SSP-585 thereafter; humid basins exhibited more pronounced increases, while arid/semi-arid basins showed limited changes. What are the implications of the main findings? This study provides a scientific basis for understanding alpine ecosystem responses to future climate change. The findings offer insights for regional ecological risk management and adaptation strategy development on the Tibetan Plateau.Highlights What are the main findings? Future NDVI is projected to increase persistently in the central and eastern Tibetan plateau but decrease along northern and southeastern margins, with variability in trend projections among different models. The multi-model ensemble indicates an overall NDVI increase in the future, with higher values under SSP-245 before the 2060s and stronger increases under SSP-585 thereafter; humid basins exhibited more pronounced increases, while arid/semi-arid basins showed limited changes. What are the implications of the main findings? This study provides a scientific basis for understanding alpine ecosystem responses to future climate change. The findings offer insights for regional ecological risk management and adaptation strategy development on the Tibetan Plateau.Abstract Under global climate change, the Tibetan Plateau, as a sensitive and ecologically vulnerable region, exhibits vegetation dynamics that significantly influence regional ecological security and hydrological cycles. This study aims to project the dynamic changes in vegetation on the Tibetan Plateau under climate change and assess the associated uncertainties in projections. Coupled Model Intercomparison Project Phase 6 (CMIP6) models were used to provide climate change outputs in the future under different greenhouse gas emission scenarios. The vegetation dynamics were described by the Global Inventory Modeling and Mapping Studies (GIMMS) Normalized Difference Vegetation Index (NDVI) data. By integrating a Random Forest model with the output climate data of CMIP6 models and training the model based on the historical observation data, NDVI changes under future emission scenarios were simulated and evaluated. The key findings of this study are as follows: (1) The multimodel ensemble (MME) performed best in simulating environmental variables, while certain individual models showed significant deviations in simulating specific variables; the Random Forest model demonstrated reliable capability in NDVI simulation and prediction. (2) The future NDVI was projected to increase persistently in the central and eastern plateau but decrease along the northern and southeastern margins, with variability in the trend projections between different models. (3) The MME model indicated an overall NDVI increase in the future, with higher values under SSP245 before the 2060s and stronger increases under SSP585 thereafter; humid basins exhibited more pronounced increases, while arid/semiarid basins showed limited changes. (4) The uncertainty in the NDVI projections showed a sustained increasing trend under both scenarios, with a stronger rise under the SSP585 scenario; spatially, the uncertainty remained low across most of the Tibetan Plateau but was relatively higher in the central-eastern region and major humid basins. These results provide a scientific basis for understanding alpine ecosystem responses to future climate change and for regional ecological risk management.