Detecting year-to-year variability of the timing of the start of the growing season (SGS) and the end of the growing season (EGS) is an important task in accurately evaluating ecosystem functions and services under climate change in vulnerable ecosystems in Siberia. We constructed a degree-day model for estimating the SGS and EGS dates at a deciduous coniferous forest site in Siberia, based on the relationship between daily phenology images and daily mean air temperature between 2013 and 2017. We tested the model by applying it to another similar site in Siberia. The model successfully estimated the SGS and EGS dates from the cumulative effective temperature, derived from daily mean air temperatures exceeding a best-fit threshold value of 2 °C (root mean square error: RMSE = 1.00) and falling below a best-fit threshold value of 1 °C, 0 °C or −1 °C (RMSE = 2.29), respectively. The modelled SGS and EGS dates closely matched the observed dates of leaf flush and leaf fall, respectively, in the larch overstory.

检测生长季节开始（SGS）和生长季节结束（EGS）时间的逐年变化是准确评估西伯利亚脆弱生态系统在气候变化下的生态系统功能和服务的一项重要任务。我们基于2013年至2017年之间的每日物候图像与每日平均气温之间的关系，构建了一个度日模型来估算西伯利亚落叶针叶林站点的SGS和EGS日期。我们通过将其应用于另一个西伯利亚的类似站点。该模型成功地通过累积有效温度估算了SGS和EGS日期，该累积温度是从每日平均气温超过2°C的最佳拟合阈值（均方根误差：RMSE = 1.00）而降至最佳拟合阈值以下得出的值为1°C，分别为0°C或-1°C（RMSE = 2.29）。建模的SGS和EGS日期分别与落叶松上层中观测到的叶子潮红和掉落日期紧密匹配。