多源卫星数据仿真模型及其云/气溶胶参量反演算法研究

The synergy retrieval algorithm of aerosol/cloud parameters has become the development trend of airborne aerosol/cloud parameters retrieval algorithm, but the inversion accuracy of such model is limited by the multi-sensors’ inconsistency of temporal-spatial resolution and observation target. By utilizing three spatial databases which have been constructed for parameters of aerosol, cloud and earth's surface in terms of longtime ground-based observation and product of satellite observation, and combining the modified vector radioactive transfer model and the radar scattering model considering multiple scattering, this study intends to build a suit of algorithm which could simulate the observation energy from multi-sensor simultaneously. And the problems of coherency of data from different satellite in temporal and spatial scales could be solved. On the basic of this, covariance would be calculated from errors according to the simulation values from different positions of multi-sensors and the corresponding observation results. And the synergy retrieval algorithm based on multi-sensors’ data would be constructed in the light of the minimum covariance theory. Finally, based on inversion results of data from each sensor and products from ground measurements, the correctness and consistency of algorithm can be verified, furthermore, the algorithm would be modified to improve the precision in the process of retrieving aerosol/cloud. These results not only can establish the foundation of improving the accuracy of the synergy inversion model and its application, but also provide a scientific basis for the configuration of multi-sensor and parameter settings of single sensor which detect the aerosol/cloud particle in space.

协同反演气溶胶/云参量模型已成为空基反演气溶胶/云参量算法的发展趋势，但目前多传感器的时空分辨率和观测目标不一致性限制了该模型的反演精度。本研究拟利用在长时间地基观测和卫星观测产品基础上构建的气溶胶、云和地表参量的三个空间特征数据库，结合改进的矢量辐射传输模型和考虑多次散射的雷达散射模型，构建一套能同时模拟多传感器观测值的仿真算法，从而解决时空尺度上多源卫星数据不一致性的问题；利用拟构建的仿真算法计算多传感器不同位置时的模拟值，基于模拟值与对应传感器观测值的误差组成协方差矩阵，并依据协方差最小原理构建基于仿真数据的协同反演模型；最后基于各传感器数据的反演结果和地基测量值来验证反演结果正确性和一致性，以便修正协同反演模型，从而提高气溶胶/云参量的反演精度。本研究成果不仅可以为协同反演模型精度提高及其推广应用奠定基础，还可为观测空间气溶胶/云粒子多源传感器的搭配和各传感器参数设置提供科学依据。

云和气溶胶辐射效应及对气候变化的影响已成为近年来的一个研究热点。对云——气溶胶——辐射——气候相互作用的研究，对于了解气候系统变化规律、提高数字气象预报精度具有重要的意义。基于卫星数据协同反演气溶胶/云参量是获取/云参量的最主要手段之一，但目前这种反演精度有限。针对目前基于单一传感器数据反演气溶胶/云参量精度不够，以及反演云参量模型对复杂云团无效的问题，项目利用A-Train卫星系列中的MODIS、CloudSat和Calipso三传感器对复杂云团进行垂直分布和水平分布双重"立体"观测获取的多源数据，利用微波对云团具有穿透性、对水云和大粒子敏感；激光对云团中冰云和小粒子敏感；以及红外光谱对不同地表下垫面上云顶总体辐射探测敏感的特性，依据前向多次散射仿真模型理论值和实际测量值协方差最小原理构建协同反演模型有效处理了该问题；但同时多传感器数据的应用，带来了时空分辨率和观测目标不一致性的问题，针对该问题，项目利用在长时间地基观测和卫星观测产品基础上构建的气溶胶、云和地表参量的三个空间特征数据库，结合改进的矢量辐射传输模型和考虑多次散射的雷达散射模型，构建了一套能同时模拟多传感器观测值的仿真算法，从而有效解决时空尺度上多源卫星数据不一致性的问题。项目首先利用了构建的仿真算法计算多传感器不同位置时的模拟值，然后基于模拟值与对应传感器观测值的误差组成协方差矩阵，并依据协方差最小原理构建了基于仿真数据的协同反演模型；最后基于各传感器数据的反演结果和地基测量值来验证反演结果正确性和一致性，以便修正协同反演模型，从而提高气溶胶/云参量的反演精度。本研究成果不仅可以为协同反演模型精度提高及其推广应用奠定基础，还可为观测空间气溶胶/云粒子多源传感器的搭配和各传感器参数设置提供科学依据。