Shifts in phenological phase of global terrestrial vegetation during the past three decades

Dynamic shifts in plant phenology significantly influence global carbon cycles, biodiversity, and ecosystem resilience. While conventional phenological methods primarily focus on discrete events such as the start or end of growing seasons, they often fail to capture the continuous and interconnected nature of plant growth. In this study, we address these challenges by employing the phase method——dynamic time warping, a novel framework inspired by the concept of phase in physics, to characterize phenological dynamics as a continuous process. Leveraging satellite-derived normalized difference vegetation index (NDVI) data and model simulated leaf area index (LAI) datasets, we extracted the global phenological phase shifts from 1982 to 2016. Our results revealed well-simulated spring phenological phase advances and subtle autumn phenological phase shifts in mid-to-high latitudes. However, models exhibited limited accuracy in capturing the delayed phases of the growing season in tropical regions and the advanced growing season phases in arid regions. Therefore, these findings provide new insights into vegetation dynamic responses to climate change, underscoring the long-term and global ecological impacts. They also highlight the necessity of integrating phenological phase responses into climate models to enhance predictive accuracy.