基于多维地表约束的新一代静止卫星气溶胶反演方法研究

Accurately obtaining surface reflectance is the basis of retrieve aerosol from satellite remote sensing. Due to the heterogeneity and time-varying features of the land surface, it has become a bottleneck problem for improving the accuracy of aerosol retrieval from satellite. The use of the surface reflectance estimation method based on single-constraint in aerosol retrieval can lead to the issue of accuracy and applicability. To solve this problem, this project will combine the characteristics of high observation frequency and multi-spectral of the new generation geostationary satellite. The study of the surface reflectance constraint model based on correlation between bands, time series and prior knowledge will be conducted. And then build the multi-constrained surface reflectance model to apply for the low reflectivity (dark target) and high reflectivity (bright target) surface. Finally, a multi-aerosol properties joint retrieval model for the next-generation of geostationary satellite is built. This research can give a satisfied solution for the problem in single-constraint surface, improve the accuracy aerosol retrieval and enhance the aerosol monitoring ability of the new generation of geostationary satellite.

准确的获取地表反射率是利用光学卫星遥感反演气溶胶的前提，由于陆地表面的非均一性和时变特征，地表反射率估计已经成为制约气溶胶遥感反演精度的瓶颈问题。针对卫星遥感反演中利用单一地表反射率估计方法导致的气溶胶反演精度和适用性的问题，本项目利用新一代静止卫星高时相和多波段的特点，开展基于波段间相关性、时间序列和先验知识的地表反射率约束模型研究，进而构建适用于低反射率“暗目标”和高反射率“亮目标”的多维地表反射率约束模型。在此基础之上构建面向新一代静止卫星的高时相多种气溶胶特性联合反演模型。该研究成果能够弥补单一地表约束条件的不足，提高气溶胶特性遥感反演精度，进一步提升新一代静止卫星大气气溶胶遥感监测的能力。

卫星遥感是监测大气环境的重要手段。随着静止卫星技术不断发展，新一代静止卫星空间、光谱和辐射观测能力不断提升。本项目结合新一代静止卫星高时相和多波段的特点，探索了综合多种地表反射率约束条件的气溶胶反演理论和方法。本项目主要完成的内容如下1）构建了中国地区典型的气溶胶类型，并分析了其时空分布特征；2）构建了综合波段间相关性、时间序列和先验知识的地表反射率约束模型；3）构建了面向新一代静止卫星的气溶胶光学厚度、气溶胶类型、气溶胶细模式比例等多个特性参数联合反演模型。通过本项目研究可以提升静止卫星气溶胶特性反演精度和适用性，进一步提升新一代静止卫星大气气溶胶遥感监测的能力。本项目发表SCI论文4篇，中文核心2篇，申请专利2项，获得软件著作权3项。