同化遥感信息和作物模型的水分胁迫对玉米产量影响的预报方法研究

Agricultural drought assessment and dynamic forecast are of great significant for the national food security. There are several limitations in the field of agricultural drought assessment and dynamic forecasting, such as the lack of the description of mechanism process regarding drought impacts on crop yield and low model forecasting accuracy for future period. Northeast of China and Northeast of Inner Mongolia， which are the important productive regions of spring maize in China is studied. We will use field experiment to improve calibration of crop growth model parameters and further regionalize crop model, which will realize the crop model’s simulation from single point scale to regional scale. We will assimilate time series of remotely sensed data into crop growth model to optimize the water stress of maize during the growing season. We will explore the impacts of drought in different phenological stages on maize yields, and establish observational and model errors datasets at different phenological stages of Maize. We will explore the downscaling method for forecasting data from numerical weather prediction(NWP) and coupled mode between crop model and NWP to improve the dynamic forecast accuracy of drought impacts on maize yields in different temporal scale of future period, through combining different temporal scale data (daily, dekad and monthly) from NWP. This project aims to establish the process-based forecasting model of water stress impacts on maize yields through integrating remotely sensed data, crop growth model and NWP model. This study achievements would be useful for agricultural drought assessment and early warning.

农业干旱对产量的动态评估和预测对于国家粮食安全具有重要意义。针对水分胁迫对产量影响评估中缺少机理过程描述和缺乏未来时段高精度预报模型等突出问题，本项目选取东北三省和内蒙东北部的雨养春玉米为研究对象，通过利用田间试验，标定作物模型参数，构建区域动态模拟的模型参数库，并进行区域化，发展作物模型从单点尺度向区域尺度的推演技术；将时间序列遥感数据引入到空间化的作物模型，研究遥感信息与作物模型的水分胁迫数据同化方法，明晰玉米关键生育期干旱对产量的影响权重，构建玉米关键生育期观测和模型误差集合；研究数值天气预报的时空降尺度方法以及与作物模型的耦合机制，致力于提高未来不同时间尺度（日、旬、月）的干旱对玉米产量影响的预报精度。本项目的特色和创新是融合遥感信息、作物模型和数值天气预报数据以期望建立机理过程性的水分胁迫对玉米产量影响的预报模型，为农业干旱灾害的评估和预警服务。

遥感与作物模型的数据同化是定量遥感与精准农业的重要研究内容，也是提高区域作物生长过程模拟最具潜力的技术方法。针对水分胁迫对产量影响评估中缺少机理过程描述和缺乏未来时段高精度预报模型等突出问题，本项目选取东北三省和内蒙东北部的雨养春玉米为研究对象，通过利用田间试验，标定作物模型参数，构建区域动态模拟的模型参数库，并进行区域化，发展作物模型从单点尺度向区域尺度的推演技术；将时间序列遥感数据引入到空间化的作物模型，研究遥感信息与作物模型的水分胁迫数据同化方法，明晰玉米关键生育期干旱对产量的影响权重，构建玉米关键生育期观测和模型误差集合；研究数值天气预报的时空降尺度方法以及与作物模型的耦合机制，致力于提高未来不同时间尺度（日、旬、月）的干旱对玉米产量影响的预报精度。通过本项目的研究，对遥感与作物模型数据同化领域中的模型不确定分析、产量同化预报方法、玉米生长水分胁迫的机理过程等关键问题开展深入研究，取得突破性的进展，为今后开展遥感与作物模型数据同化研究提供了借鉴价值，促进了农业遥感学科的发展。

内容获取失败，请点击重试