基于叶绿素荧光的青藏高原植被枯黄期对极端气候变化的响应研究

The climate change for Qinghai-Tibet Plateau (QTP) is characterized by high sensitivity, and the extreme climate events are perceived as the principle factor in driving the transformation of environment and ecosystem. Most previous studies of vegetation autumn phenology are limited in its responses to average climate factors, and pay little attention to the influence of extreme climate factors, which limited the understanding of climatic response mechanism of vegetation autumn phenology, thus largely reduced the accuracy of models predictions for autumn phenology. Therefore, it is key important to quantify the responses of vegetation autumn phenology to extreme climate change for the mechanism research of vegetation autumn phenology. The proposed study here is about the response mechanism of vegetation autumn phenology to extreme climatic factors on the QTP. (1) First, the solar-induced chlorophyll fluorescence (SIF) remote sensing data was used to analyze the spatial and temporal distribution of vegetation autumn phenology with the creation of a dataset of vegetation autumn phenology on QTP; (2) Then, based on extreme climate index, quantifying the relationship between vegetation autumn phenology and extreme temperature and extreme precipitation events, then discuss the underlying mechanisms; (3) Finally, develop the predict model for vegetation autumn phenology based on artificial neural network and random forest method, and predict the variation trend in vegetation dormancy. The implementation of this study will be of guiding significance to the study of the driving mechanism of vegetation phenology and the development of vegetation phenology models.

青藏高原气候变化高度敏感，复杂多变的极端气候是高原生态和环境变化的重要驱动。由于当前对植被枯黄期的研究局限于其与平均气候的关系，忽略了极端气候的影响，导致对植被枯黄期的气候响应机制缺乏深刻理解，直接限制了植被枯黄期的准确模拟和预测。因此，明晰植被枯黄期对极端气候变化的响应规律是研究青藏高原植被枯黄期变化驱动机制的关键。本项目将围绕植被枯黄期对极端气候变化的响应，（1）基于日光诱导叶绿荧光（SIF）遥感数据，建立青藏高原植被枯黄期数据集，分析植被枯黄期的时空格局；（2）基于多个极端温度和极端降水指标，量化植被枯黄期和极端气候指标的关系，探寻植被枯黄期变化的主要影响因素及其影响方式；（3）建立植被枯黄期预测模型，预测植被枯黄期的变化趋势，探讨气候变化背景下未来植被枯黄期对气候变化的响应规律。本项目的实施将对青藏高原植被物候的驱动机制研究和植被物候模型的发展均具有重要意义。

当前植被物候研究多集中在春季物候，而对秋季物候的气候响应机理仍然缺乏深刻理解，严重限制了植被物候的准确模拟和陆地生态系统碳收支评估。本项目围绕青藏高原等典型陆地生态系统植被秋季物候，结合日光诱导叶绿素荧光等多源遥感监测及通量监测，系统研究植被秋季物候对气候变化和极端气候的响应规律及潜在机理，并构建耦合多个极端气候因素的植被秋季物候模型。结果发现：（1）季前温度对青藏高原植被枯黄期（EGS）具有延迟效应，而降水的效应具有区域性；冷和热胁迫显著影响研究区植被EGS，季前干旱和高热胁迫导致EGS提前，而在农业和草原生态区，中热和降雨导致EGS延迟。（2）基于机器学习和统计学方法建立包含多个极端气候因素的物候模型，其预测的2081-2100年青藏高原植被秋季物候要早于传统积温模型预测的结果，在RCP4.5情景下提前1.9-3.1天，在RCP8.5情景下提前12.2-14.3天。（3）极端气候对植被秋季物候的影响在北半球多个生态系统具有差异性，约80%的区域在冷胁迫下出现秋季物候提前。与变暖相关的气候指标在高纬度地区延迟了秋季物候，随着纬度的降低，延迟效应减弱甚至逆转。（4）日光诱导叶绿素荧光遥感在捕捉亚热带常绿森林光合物候方面优于传统植被指数。中国亚热带森林秋季光合物候呈现延迟趋势，其主要受到季前最低温度的显著控制。研究成果系统地阐明了青藏高原、亚热带等典型陆地生态系统植被秋季物候对气候变化的响应机理，为植被物候模型发展和生态系统碳收支评估提供了理论支撑。在本项目资助下，以第一标注发表SCI论文9篇，包括《Global Ecology and Biogeography》、《Agricultural and Forest Meteorology》、《Forest Ecosystems》、《Nature Plants》等一区TOP期刊5篇。较好地完成项目既定研究目标，为开展接下来研究工作奠定扎实基础。