Abstract Variations in autumn phenology (i.e., the end of growing season, EOS) in recent decades have been suggested to be the dynamic response of terrestrial ecosystems to global environmental change. However, predicting EOS change with climate variation is still challenging, given that there are several environmental factors that can affect EOS apart from warming. In this study, EOS over the Northern Hemisphere (>30°) was calculated by three different extracting methods using the Normalized difference vegetation index (NDVI) records from the Global Inventory Modeling and Mapping Studies third generation (GIMMS3g) during 1982 to 2015. We used partial correlation and partial least-squares regression (PLSR) analyses to investigate the EOS variations in response to five global environmental change factors, including temperature, precipitation, insolation, atmospheric CO2 and nitrogen (N) deposition. We found that about 58.5% (13.6% were significant) of the study area showed earlier EOS trends and 41.5% (10.3% were significant) with later EOS. Warming played a critical role in regulating EOS. The cumulative precipitation during the preseason had the greatest positive impact on EOS in arid/semiarid regions. Sufficient insolation could contribute to the delayed EOS over high latitudes. Compared with climatic factors, EOS also showed high sensitivity to changes in the atmospheric factors, and the dominant control of atmospheric CO2 and N deposition on EOS was found at 25–30% of the areas. Our results confirmed the dominant role of climatic factors in regulating EOS, but other global environmental cues, for instance, elevated atmospheric CO2 concentration and increased atmospheric nitrogen (N) deposition should also be included in the autumn phenological model to further understand the carbon and N cycles of terrestrial ecosystems with climate change.

中文翻译：

---

确定气候和大气变化对北半球中高纬度秋季物候的贡献

摘要 近几十年来秋季物候（即生长季节结束，EOS）的变化被认为是陆地生态系统对全球环境变化的动态响应。然而，考虑到除了变暖之外，还有几个环境因素可以影响 EOS，预测随气候变化的 EOS 变化仍然具有挑战性。在本研究中，北半球 (>30°) 的 EOS 是通过三种不同的提取方法计算的，使用来自 1982 年至 2015 年全球清单建模和制图研究第三代 (GIMMS3g) 的归一化差异植被指数 (NDVI) 记录。我们使用偏相关和偏最小二乘回归 (PLSR) 分析来研究 EOS 变化对五种全球环境变化因素的响应，包括温度、降水、日晒、大气 CO2 和氮 (N) 沉积。我们发现大约 58.5%（13.6% 显着）的研究区域显示出较早的 EOS 趋势，41.5%（10.3% 显着）显示出较晚的 EOS。变暖在调节 EOS 方面发挥了关键作用。季前累积降水对干旱/半干旱地区的EOS影响最大。充足的日照可能会导致高纬度地区的 EOS 延迟。与气候因素相比，EOS对大气因素的变化也表现出较高的敏感性，在25%～30%的区域发现EOS对大气CO2和N沉积的主导控制。我们的结果证实了气候因素在调节 EOS 中的主导作用，但其他全球环境线索，例如，