Vegetation phenology is a key indicator of ecosystem responses to climate change. This study investigates the spatial-temporal dynamics of vegetation phenology in the Changbai Mountain Nature Reserve from 2001 to 2025 and projects future changes under CMIP6 scenarios using a particle swarm optimization–least squares support vector machine (PSO-LSSVM) model. The results show that SOS exhibits an advancing trend, while EOS is delayed, leading to an overall extension of LOS. Spatially, phenological patterns are strongly controlled by elevation, with higher elevations characterized by later SOS and shorter LOS. Correlation analysis indicates that SOS is primarily driven by spring temperature, whereas EOS is influenced by both temperature and precipitation, showing more complex responses. Notably, a negative relationship between autumn temperature and EOS suggests that factors other than temperature may play an important role. Future projections reveal that phenological changes intensify with increasing emission scenarios. By the end of the 21st century, SOS is projected to advance by 0.8–3.6 days, EOS to be delayed by 0.8–7.4 days, and LOS to extend by 1.6–11.8 days. Vegetation-type-based analysis further demonstrates significant heterogeneity in phenological responses. These findings improve the understanding of vegetation phenology in mountain ecosystems and provide a useful reference for assessing ecosystem responses under future climate change.
He et al. (Fri,) studied this question.