A modified thermal time model (MTM) was developed to reproduce the leaf onset for summer-green vegetation in the Northern Hemisphere. The model adopts the basic concept of a thermal time model (TM) in that leaf onset is primarily triggered by growing degree days (GDD). Based on global phenology data derived from satellite observations, a new parameterization for the critical model parameter Tb (i.e., baseline temperature for GDD calculation) has been introduced, and the spatial distribution of Tb calculated. Simulations of leaf onset during 1982–2000 in the range 30–90°N showed a significant improvement of MTM over the standard TM model with constant Tb. The mean error and mean absolute error of the climatological simulation were 1.11 and 6.8 days, respectively, and 90% of the model error (5th and 95th percentiles) was between ?12.4 and 13.7 days.

中文题名：预测北半球夏绿植物长叶日期的新方案

研究目的：改进预测夏绿植物的长叶日期的物候方案

创新要点：基于物候观测资料分析，根据发育阈温度（Tb）与局地气候（如春季均温）的相关性，建立了Tb的空间分布场，并由此建立了改进的积温模型（MTM），用于模拟夏绿植物开始长叶日期。该模型可以体现植物自身对环境的适应性。

研究方法：利用GLASS物候资料反推得到Tb的空间分布场，通过线性回归得到了依赖于春季均温的Tb的拟合方案，使用拟合的Tb作为模型需要的发育阈温度，计算GDD，判定长叶开始日期，建立了MTM。

重要结论： MTM的气候态模拟结果平均误差（ME）为1.1天，平均绝对误差（MAE）为6.8天，90%的模型误差在-12.4至13.7天之间。模型在逐年模拟方面也具有较好的模拟结果，19年整体的ME为1.87天，MAE为10.6天，在-20至20天之间的样本占总量的87.2%。