积雪是地球表层重要的覆盖物质，受全球气候，地表环境的影响，不同地区积雪量呈现出不同的特点，同时积雪量也影响全球气候、气象变化，局地水文、生态环境等，因此，准确估测地表积雪量变得尤为重要。而中国地区积雪量受高原气候，不同地理状况的影响，在空间和时间上存在特有的规律和特点。这样，更有必要对区域性的积雪量动态估测，为其它水文、气象等模型提供较准确的积雪量参数，为相关部门服务。由于被动微波的优势，被动微波遥感反演积雪参数，近年来成为反演积雪量的非常重要的工具。本文利用被动微波反演中国地区雪水当量，并通过开展实验，研究并验证微波辐射理论模型在中国积雪区的适用性。 首先介绍积雪的微波辐射传输理论与典型的积雪辐射传输理论模型，在此基础上，通过开展中国典型积雪区（黑河地区）积雪实验，通过实验测量数据来验证典型的积雪辐射传输模型，发现模型模拟值和实验测量值基本吻合，但是有差异，需要更进一步考虑中国积雪区积雪物理参数的特点以及辐射特性，由于积雪区垂直层内密度和温度的不均匀，下垫面的独特特点，这样积雪模型需要分层，AIEM模型不完全适用，在模型中都需要改进。 其次通过分析中国积雪区时间和空间以及地理环境的特点，基于Chang反演积雪水当量经验算法思想，增加象元积雪覆盖度和高频89GHz对反演精度的影响，考虑地表不同类型也有不同的微波响应，改进了中国地区反演积雪雪水当量算法，结果显示，新的反演算法可以提高反演中国地区雪水当量的精度。但是，由于微波空间分辨率低、地面象元内复杂的地物类型以及积雪状态随时间和空间的变化等，都给积雪雪水当量的反演带来问题，这还需要我们更进一步研究。 最后，将改进后的新的反演算法形成中国地区雪水当量产品，基于软件开发工程原理和产品用户的需要，用MATLAB语言开发，形成中国地区雪水当量反演产品原型系统。 总结本文对被动微波遥感辐射理论模型和积雪反演算法改进的研究情况，讨论进一步研究方向，为了改进中国地区积雪雪水当量的反演精度，需要综合分析积雪状态和地理环境状况的变化，联合多种传感器的遥感数据，发展基于物理理论的积雪雪水当量反演模型，这是未来的主要方向。

Snow is the important material of the earth’s surface, imposed by global climate and environment, which is different snow features in different area, and snowfall influence the global climate and weather change, hydrological and ecological environment and so on, so accurate estimation of snowfall is very important. In china, snowfall is affected by platenu climate and diversity geographic condations, so there are specific rules and characteristics in time and space. Thus, we need to estimate the snowfall to provide the accurate parameters to hydrological and climate model, in order to serve the related departments. With the predominance of passive microwave, estimating the snow parameters by passive microwave is very important methods. The aim of the article is to inverse the snow parameters by passive microwave, verify the microwave radiative model by the experiment data in china snow area. First, the introduction of snow microwave radiative transfer theory and model, then we do the snow experiment in typical area of china (heihe area), and verify the model by the snow experiment data; find that there is the accordance in simulation values and measured values, but partly difference, for the model can’t include the influence of the vertical snow layer’s non-uniform distribution of temperature and density in snow area of china. Hence, we should study further for the improving model. Sencod, analysing the snow distribution characteristics of china area in time and space and in diversity of geographical condition, then based on Chang inversion algorithm idea of snow water equivalent (SWE), added the pixel snow cover area, high-frequency (89GHz) and surface classification, I improved the SWE inversion algorithm in china area. The results showed that new SWE inversion algorithm can enhance the inversion precision. But, the lower microwave spatial resolution, complex ground coverage type in a pixel and changes of snow features with time and space, all of which can cause the difficulty in inversion, this is need to further study. Finally, we can get the SWE product of china area from the new inversion algorithm, which is based on principles of software development and needs of product users in MATLAB language. And we summed up the content and give the direction: to get more accurate SWE of china area, we need to comprehensively analyse the changes of snow features and geographical environment, and to combine the remote sensing data in a variety of sensors, and to improve SWE inversion model based on snow physical theory.