基于叶绿素荧光参数的生态系统模型光合参数遥感反演方案及其区域应用

The ecosystem model which formed by combing ecological relations with the data gathered from field data is a useful tool for ecosystem management and agricultural sustainable development studies. The retrieval of the model parameter from remote sensing is the foundation of the spatial application of ecosystem model. In this study, the winter wheat and summer maize in the North China Plain is taken as an example, with the aid of chlorophyll fluorescence techniques, the empirical relationship between MODIS data and the photosynthetic parameters (quantum yield and maximum catalytic capacity of Rubisco) of the crop will be established at field scale. Based on this empirical relationship, we develop a new scheme to retrieve the photosynthetic parameters from remoter sensing and ecosystem model (Vegetation Interface Processes model). The new scheme, integrate the interval estimation theory and parameter-optimization algorithm, will narrow the search interval of photosynthetic parameters in optimum searching process. The photosynthetic parameter patterns of winter wheat and summer maize in the NCP will be obtained with the new scheme and used in the VIP model to simulate the spatiotemporal pattern of crop productivity, evapotransipiration and water use efficiency in the NCP. The field data, census data of crop yield at county scale and the multi-source remote sensing data are used in the model validation. With the photosynthetic parameter patterns, the ecosystem is hoped to provide a more accuracy map of crop productivity, evapotransipiration and water use efficiency in the NCP. This will help us to explore the spatiotemporal variation feature of energy-water and carbon exchange of agroecosystem in the NCP, and provide scientific basis for agroecosystem management, crop yield estimation and agriculture policy establishing.

基于生物物理和生物化学过程的生态系统模型是区域生态系统管理、农业可持续发展战略研究的重要工具，融合遥感信息和生态系统模型反演模型关键参数是模型区域应用的基础。本项目以华北平原为研究对象，分别以冬小麦和夏玉米为例，基于地面遥感试验，借助叶绿素荧光技术，建立表征作物光合作用能力的关键参数与MODIS特征波段、植被指数间的经验关系；以此为基础，基于生态系统模型(VIP模型)，融合遥感信息，改进模型参数区域化方案，反演模型光合作用模块关键参数组的空间格局，实现模型与中尺度分辨率多元遥感信息的融合，提高模型区域尺度农业生产力、蒸散和用水效率的模拟预测精度；结合地面实测数据和多元遥感数据，对改进的模型进行单点和区域尺度的验证；探索华北平原区域尺度作物生长和水热平衡的时空变化规律，揭示作物需水和水分生产力的区域分异规律和时空演变特征，为华北平原农业管理、产量估算和政策制定提供科学依据。

融合遥感信息和生态系统模型，获取区域尺度生态系统模型关键参数的空间分布，是提高生态系统机理模型区域模拟能力的有效途径。本项目基于VIP生态水文模型，提出了两种遥感信息与机理模型同化反演作物光合参数的新方法（经验公式法和同化法），发展了机理模型尺度扩展的理论方法。经验公式法以VIP模型为桥梁，通过作物统计产量构建MODIS-NDVI数据和光合能力参数(最大羧化率Vcmax和光量子效率)的经验关系；同化法以一维黄金分割法为基础，基于区间估计原理，引入光合参数空间先验分布的估算，融合遥感信息（植被指数和fPAR）和机理模型，获取光合参数空间分布格局和标准偏差。结合地面实测数据和多元遥感数据的模型多尺度验证表明，光合参数时空格局的反演有助于提高生态系统模型对作物生产力、水热平衡过程的模拟精度。相比于利用遥感信息获取的状态变量驱动生态系统模型，利用反演的光合参数驱动机理模型能有效避免区域产量和蒸散量模拟过程中遥感信息饱和导致的叶面积指数低估现象，提高机理模型作物生产力和需水的区域模拟精度。以反演的光合参数时空格局为基础，对作物产量及蒸散量时空变异的模拟分析表明，华北平原产量的空间格局受土壤盐碱化程度、灌溉条件和氮素施用量的综合影响，其中氮肥施用量是影响灌溉地产量的主要原因。