The Key Laboratory of Land Surface Pattern and Simulation, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing, China.
University of the Chinese Academy of Sciences, Beijing, China.
Glob Chang Biol. 2023 Mar;29(6):1557-1573. doi: 10.1111/gcb.16569. Epub 2022 Dec 29.
The unprecedented warming that has occurred in recent decades has led to later autumn leaf senescence dates (LSD) throughout the Northern Hemisphere. Yet, great uncertainties still exist regarding the strength of these delaying trends, especially in terms of how soil moisture affects them. Here we show that changes in soil moisture in 1982-2015 had a substantial impact on autumn LSD in one-fifth of the vegetated areas in the Northern Hemisphere (>30° N), and how it contributed more to LSD variability than either temperature, precipitation or radiation. We developed a new model based on soil-moisture-constrained cooling degree days (CDD ) to characterize the effects of soil moisture on LSD and compared its performance with the CDD, Delpierre and spring-influenced autumn models. We show that the CDD model with inputs of temperature and soil moisture outperformed the three other models for LSD modelling and had an overall higher correlation coefficient (R), a lower root mean square error and lower Akaike information criterion (AIC) between observations and model predictions. These improvements were particularly evident in arid and semi-arid regions. We studied future LSD using the CDD model under two scenarios (SSP126 and SSP585) and found that predicted LSD was 4.1 ± 1.4 days and 5.8 ± 2.8 days earlier under SSP126 and SSP585, respectively, than other models for the end of this century. Our study therefore reveals the importance of soil moisture in regulating autumn LSD and, in particular, highlights how coupling this effect with LSD models can improve simulations of the response of vegetation phenology to future climate change.
近几十年来，前所未有的变暖导致整个北半球的秋季叶片衰老日期 (LSD) 推迟。然而，这些延迟趋势的强度仍然存在很大的不确定性，特别是在土壤湿度如何影响它们的方面。在这里，我们表明，1982-2015 年土壤湿度的变化对北半球五分之一有植被地区（>30°N）的秋季 LSD 产生了重大影响，以及它对 LSD 变化的贡献如何超过温度、降水或辐射。我们开发了一种基于土壤湿度约束冷却度日 (CDD) 的新模型来描述土壤湿度对 LSD 的影响，并将其性能与 CDD、Delpierre 和春季影响秋季模型进行了比较。我们表明，温度和土壤湿度输入的 CDD 模型在 LSD 建模方面优于其他三个模型，并且在观测值和模型预测之间具有更高的相关系数 (R)、更低的均方根误差和更低的 Akaike 信息准则 (AIC)。这些改进在干旱和半干旱地区尤为明显。我们使用 CDD 模型在两种情景 (SSP126 和 SSP585) 下研究了未来的 LSD，并发现与其他模型相比，在本世纪末，SSP126 和 SSP585 下预测的 LSD 分别提前了 4.1±1.4 天和 5.8±2.8 天。因此，我们的研究揭示了土壤湿度在调节秋季 LSD 中的重要性，特别是强调了如何将这种效应与 LSD 模型耦合可以提高对植被物候对未来气候变化响应的模拟。
