Impacts of Climate Change and Human Activities on Global Mountain Grasslands: Insights Into the Last Two Decades and Future Climate Scenarios

Mountain grasslands are vital ecosystems providing critical services such as carbon sequestration, water regulation, and biodiversity conservation. However, these ecosystems are increasingly threatened by climate change and human activities. This study evaluates vegetation dynamics in global mountain grasslands (2000–2021) using remote sensing data, CMIP6 climate projections, and human modification indices. By means of machine learning models (Random Forest, eXtreme Gradient Boosting (XGBoost), and Long Short-Term Memory (LSTM)) we identified dewpoint temperature, total evaporation, and human modification as dominant predictors of vegetation change, while soil water content and latent heat flux exhibited region-specific impacts. Results indicate that 35.1% of grasslands remained stable, 32.1% improved, and 32.7% degraded, with degradation hotspots identified in the Tibetan Plateau, Ethiopian Highlands, Rocky Mountains, and Andes, while the Middle Eastern Mountain Ranges showed signs of improvement. Future projections using a hybrid XGBoost–LSTM model indicate limited global-scale vegetation shifts by 2050 across SSP1-2.6, SSP2-4.5 and SSP5-8.5 scenarios. However, regional differences are notable: the Tibetan Plateau shows a substantial increase in vegetation cover, while the Andes, Rocky Mountains, and parts of East Africa exhibit slight changes. Distributional shifts, especially under SSP5-8.5, suggest increasing spatial heterogeneity in grassland responses. These findings underscore the importance of regional-scale strategies to support grassland resilience under future climate and land-use pressures.