融合遥感与气象数据的玉米高温风险精细评价方法研究

The main body of the new management is the main force of modern agricultural production. How to scientifically prevent the meteorological risk is the rigid demand of these large-scale operators. Because of the spatial difference of phenological period, meteorological risk is significantly different between different regions, crops and even plots. However, it is difficult to guide the management of the risk assessment without the fine time scale study of "growth period - plot" of these risks. In this project, we study the data fusion of MODIS LST data and meteorological observation in the Huanghuaihai summer sowing maize area as an example. The landfill distribution and crop classification is used to study the extraction high temperature from plot to plot. And the total accumulated temperature and total radiation model of maize seedling to flowering stage were studied. Based on the spatial clustering and hotspot analysis, a new method for estimating the temperature of the maize flowering stage was established. Based on the high temperature response rule of maize flowering stage, Revealing the geometric differentiation of maize high temperature risk. The project innovation: Combinating geographic information, remote sensing technology and agricultural applications, researching point-surface temperature data fusion method, extracting accurate high temperature and maize growth period by each plot, building the fine time scale model of "growth period - plot" of high temperature for Maize.

新型经营主体是现代农业生产的主力军，如何科学防范气象风险是这些规模化经营者的刚性需求。由于物候期的空间差异，气象风险在不同区域、作物，甚至地块之间都有显著差别；但因缺少“物候期-地块”这一精细时空尺度下的风险评价研究，很难指导经营者防范风险。本项目以我国黄淮海夏播玉米区的高温风险评价为例，研究MODIS LST数据与气象观测等数据的融合方法，进行缺失数据填补及精度评价；利用耕地分布与作物分类信息，研究提取地块级的高温信息；构建玉米出苗到花期的有效积温和总辐射模型，研究玉米地块级花期提取方法；根据玉米花期高温响应规律，构建高温风险精细评价模型；基于空间聚类与热点分析，揭示玉米高温风险地域分异规律。本项目创新点：将地理信息、遥感技术与农业应用相结合，研究点-面温度数据融合方法、精确到地块的高温和玉米生育期提取方法以及地块-生育期时空尺度下的高温风险精细评价方法，具有重要理论意义和应用价值。

高温热害是影响范围最广的频发农业气象灾害之一，敏感生育期高温热害会导致玉米减产甚至绝产，给玉米新品种推广和粮食安全带来巨大挑战。本项目以黄淮海夏玉米种植区为例，基于高温热害对玉米生长发育影响机理，构建多源点-面数据的组织管理与计算框架，提取融合遥感与气象数据的高温信息，研究玉米高温热害风险指标体系与评价模型，完成了玉米花粒期高温热害胁迫综合分析。达成原计划书的各项目标。. 取得的成果主要有：（1）完成了基于地理数据网格体系的多源点-面数据管理框架，为大范围、长时序高温热害监测提供高效的数据组织、存储、检索和分析支持。（2）证明了气温-遥感反演温度在黄淮海区域存在显著的正相关性，为融合气温和遥感温度综合评估玉米高温热害风险提供了理论基础。（3）系统总结了高温热害作用于玉米授粉和结实的深层次机理，为高温热害模型的构建提供了理论依据。（4）完成了黄淮海夏玉米多年多点生育时期缺失数据的填补，有助于各亚区锁定玉米高温热害的花粒期敏感期。（5）研究和确定了高温热害风险的指标体系、量化计算的数学模型和流程，构建了玉米花粒期高温胁迫分级规则，计算了公里尺度的黄淮海夏玉米高温风险数据。（6）结合作物分布、耕地边界矢量数据，实现了地块尺度玉米花粒期高温热害精细评价，能够准确辨识地形地貌、水系等要素影响下的邻近地块玉米高温热害差异，大范围高温风险的分布规律。. 本项目将玉米高温风险的遥感监测、农学机理和空间分析相结合，充分发挥了气象连续观测和遥感大面积观测的优势，构建了可精确到地块的玉米高温热害风险精细评估方法。提高了对于黄淮海夏玉米区高温热害时空分异规律的认识，可为耐高温玉米品种选育和精准推广方案制订提供技术支持。