地-气解耦的不确定性是陆地上空气溶胶遥感的主要误差来源之一。在气溶胶偏振遥感中，地表双向偏振反射分布函数（BPDF）模型法进行地-气解耦应用广泛，但是地表BPDF模型自身的局限性会给反演带来较大的误差。针对上述问题，结合多光谱偏振观测数据特性、气溶胶散射特性和地表偏振反射规律，研究并提出地-气逐次解耦的方法，以此为基础构建气溶胶偏振反演方案，用于提升气溶胶反演精度和改进复杂地表情况下气溶胶反演等问题。在地-气解耦引入的反演误差模拟和反演实验分析的基础上，依次开展地-气逐次解耦机理和迭代过程中参量变化规律等关键科学问题的研究。设计并优化逐次解耦循环逼近策略、气溶胶参数提取以及循环流程的反馈与控制机制，实现多光谱偏振遥感气溶胶产品的高精度反演，为陆地上空气溶胶遥感的反演精度提升和卫星气溶胶产品生产提供理论与技术支持。

The uncertainty of land-atmosphere decoupling is one of the main error sources in aerosol remote sensing. Though the polarized bidirectional reflectance distribution function (BPDF) models are widely used in land-atmosphere decoupling in polarized remote sensing, the limitations of these models introduce errors in aerosol retrieval. In allusion to the problems, the method of successive land-atmosphere decoupling (SLAD) is put forward based on characters of observations, aerosol and land surface. Then, aerosol retrieval algorithm is built based on SLAD to improve the inversion accuracy and solve the problem of aerosol retrieval over complex surface. The SLAD mechanism and variation law of parameters between loops are studied by error simulation and inversion experiments. The methods of SLAD iteration, aerosol parameters inversion and feedback are designed and optimized in the research, and the high accuracy aerosol retrieval algorithm using multi-spectral and polarized measurements is built. This study can provide theoretical and technical supports to improve aerosol remote sensing inversion precision and production of aerosol from satellite observation.