Pre-seasonal temperature trend break dominating the trend break in autumn grassland phenology in China

Interannual variations in the end of the growing season (EOS) play a crucial role in assessing carbon and energy cycling within grassland ecosystems. Previous studies have often fixed the trend breakpoint in autumn phenology around the year 2000 to examine the response of the vegetation EOS to long-term climate change. However, the asymmetry of climate change and the diversity of grass species may lead to spatial disparities in EOS trend breakpoints, but little research has been done to quantify their characteristics and underlying climatic driving mechanisms. Focusing on the period from 1982 to 2015, this study extracts EOS data from six different grassland subregions in China to identify EOS trend breakpoints, and then investigates the associated climatic driving mechanisms. The results highlight the presence of significant breakpoints in the EOS trend within 54.1% of China's grasslands. Prior to 1997, the grassland EOS trend exhibited a pronounced delay, with a rate of 0.29 days per year (P < 0.01), which subsequently shifted to 0.10 days per year. In addition, pre-seasonal climate factors emerged as the dominant driver, contributing a remarkable 98.8% to the timing of the grassland EOS, with preseasonal temperature and solar radiation standing out as the dominant climate variables influencing the grassland EOS. Furthermore, the main driver of the trend break in the grassland EOS was the trend break in preseasonal temperature, which contributed to 52.2% of the trend break in the grassland EOS. These results confirm the presence of breakpoints in the autumn phenological trend across Chinese grasslands, and elucidate the intrinsic climate-driven mechanism responsible for the autumn phenological trend break at the pixel scale. These findings provide valuable insights to better understand and model the complex interactions between ecosystems and the climate system.