作物模型应用与遥感信息集成技术研究进展

作物模型和遥感技术以各自独有的优势在作物生产监测、评估及未来预测等方面发挥着关键作用。作物模型与遥感信息集成技术在大尺度、高精准的农业生产监测、评估与预测上具有明显的应用优势和广阔的发展前景。为了促进这些技术在区域尺度上的作物产量预测、农业气象灾害影响评估及农业应对气候变化决策等方面更加广泛地应用，本文采用文献综述的方法，系统归纳了欧洲、美国、澳大利亚及中国作物模型的发展与应用，总结了当前主流的数据集成方法的原理、特点和不足，概述了作物模型与遥感信息集成技术的实际应用，探讨了提升数据集成精度存在的问题，并对未来研究方向进行展望。结果表明，国内外对于作物模型及其与遥感数据集成的研究与应用广泛而深入，利用同化方法能够有效提高作物模型模拟精度，为作物模型实现区域尺度作物生长及产量评估、气候变化对产量影响、农田管理决策等提供技术支撑。作物模型模拟结果及遥感反演数据的不确定性、数据同化策略的多样性以及尺度效应是进一步提高集成系统精度与效率的限制因素。因此，遥感数据多源融合、同化过程多变量协同、作物模型多类型耦合以及数据高性能并行计算是未来作物模型与遥感数据集成研究的发展趋势。

Progresses of Crop Model Application and Its Integration with Remote Sensing Technology

Crop model remote sensing and play important roles in agricultural production monitoring, evaluation, and future prediction with their unique advantages. The integration technologies of crop model and remote sensing information have obvious application advantages and broad development prospects in monitoring, evaluation and prediction of large-scale and high-precision agricultural production. In order to promote the wider applications of these technologies in crop yield prediction, impact assessments of agrometeorological disaster, and agricultural decision-making to deal with climate change on a regional scale, the method of literature review were adopted in this paper. The development and application of crop models in Europe, United States, Australia and China were systematically summarized. The principle, characteristics and shortcomings of the current mainstream data integration methods were concluded. The practical applications of integration technologies of crop model and remote sensing information were summarized. The existing problems in improving the accuracy of data integration were discussed, and the future research direction was prospected. The results showed that the research and application of crop model and its integration with remote sensing data were extensive and intensive at home and abroad. The assimilation method could effectively improve the simulation accuracies of crop model, providing technical support for crop growth and yield evaluation on regional scales, impacts of climate change on yield, farmland management decision-making, etc. The uncertainties from crop model simulation results and remote sensing inversion data, diversities of data assimilation strategies, and scale effects were the limiting factors to further improve the accuracy and efficiency of integrated systems. Therefore, multi-source fusion of remote sensing data, multivariable cooperation in assimilation process, multi-type coupling of crop models, and high-performance parallel computing of data were the development trends of integrating crop models and remote sensing research in the future.