基于遥感数据与作物生长模型同化的作物长势监测

遥感观测和作物生长模型模拟是进行农作物长势监测的2种有效手段，并具有较好的互补性，构建二者的同化系统是目前农业遥感研究领域的热点。同化涉及多学科的交叉集成，十分有必要将同化方法中的模型、算法、数据进行集成，构建基础作物模型同化系统平台，降低科学研究和应用的难度。采用模块化结构设计，将同化系统构建所需的主要模型、算法、数据等进行有机结合和无缝集成，实现基于极快速模拟退火算法的遥感数据与CERES-Wheat作物生长模型的同化原型系统构建。此外，通过所开发的系统利用地面高光谱作为遥感数据，通过同化小麦叶面积指数对同化系统进行了检验和初步应用。同化LAI与实测结果较好的拟合度，表明所开发的同化系统基本可行，能为遥感技术与作物模型的基础研究和应用提供一个平台。

Monitoring crop growth based on assimilation of remote sensing data and crop simulation model

Remote sensing observation and crop growth model simulation are two effective ways for crop growth monitoring, and they can well complement each other. Establishing assimilation system of remote sensing data and crop growth model is an interesting research field. Assimilation involves multidisciplinary integration, so it is necessary to construct a basic assimilation system integrated with models, algorithms and datasets of many kinds to reduce the difficulty of research and applications of data assimilation. In this study, in order to establish the assimilation system prototype of crop growth model and remote sensing data, modular design was used to realize the combination of CERES-Wheat model, very fast simulated annealing algorithm and remote sensing data. Ground hyper-spectral data was chosen as remote sensing data, and the system was tested and preliminarily applied to estimate wheat LAI. The results show that the assimilated LAI can well agree with the measured LAI, and the assimilation system established is feasible. The assimilation system can provide a platform for the basic research and application of the coupling of remote sensing data and crop growth model.