Wetness is one of the important physical parameters of snowpack. Its spatial and temporal changes play a key role in snowmelt runoff forecast, regional climate change, and agricultural irrigation. In this study, we proposed a new method to retrieve snow wetness from full-polarimetric synthetic aperture radar (SAR) data. First, the dominant scattering components, i.e. surface and volume scattering, in wet-snow conditions were obtained by polarimetric decomposition. The random rough surface scattering models and the dense media radiative transfer (DMRT) model were then used to establish the surface and volume scattering models, through which the surface and volume scattering lookup tables were created, respectively. Based on the lookup tables and the polarimetric decomposition results, the snow surface and volume wetness were retrieved, respectively. Finally, the effective snow wetness was derived from the weighted summation of surface and volume wetness. The advantage of this method mainly comes from the full consideration the snow surface roughness, the local incidence angle on complex mountain terrain, and the polarimetric information. In experiments, the GaoFen-3 data obtained on January 17, 2018 in the Kelan River Basin and the Radarsat-2 data obtained on March 19, 2014 in the Manasi River Basin were selected to verify the applicability of the proposed method at different conditions. From the analysis of experiment results, the correlation coefficient between the estimated and the ground measured snow wetness in the Kelan River Basin is 0.72. The mean absolute error (MAE) and the root mean square error (RMSE) are 3.35% and 3.89%, respectively. The correlation coefficient between the snow wetness estimated by Radarsat-2 and the measured values in the Manasi River Basin is 0.62. MAE and RMSE are 1.32% and 1.62%, respectively. These results proved that the proposed method can effectively retrieve snow wetness under different SAR data, different areas, and different snow periods.

湿度是积雪的重要物理参数之一。其时空变化在融雪径流预报，区域气候变化和农业灌溉中起着关键作用。在这项研究中，我们提出了一种从全极化合成孔径雷达（SAR）数据中检索雪湿度的新方法。首先，通过极化分解获得湿雪条件下的主要散射成分，即表面和体积散射。然后使用随机粗糙表面散射模型和密集介质辐射传递（DMRT）模型建立表面和体积散射模型，分别通过该模型创建表面和体积散射查找表。根据查找表和极化分解结果，分别获取雪面和体积湿度。最后，有效积雪湿度是由表面和体积湿度的加权总和得出的。该方法的优势主要来自于充分考虑雪面粗糙度，复杂山地上的局部入射角以及极化信息。在实验中，选择了2018年1月17日在克兰河流域获得的GaoFen-3数据和2014年3月19日在玛纳斯河流域获得的Radarsat-2数据，以验证该方法在不同条件下的适用性。通过对实验结果的分析，可兰河流域的估计雪湿度与地面实测雪湿度之间的相关系数为0.72。平均绝对误差（MAE）和均方根误差（RMSE）分别为3.35％和3.89％。Radarsat-2估算的积雪湿度与玛纳斯河流域实测值的相关系数为0.62。MAE和RMSE分别为1.32％和1.62％。这些结果证明，该方法可以有效地提取不同SAR数据，不同区域，不同雪期下的积雪。