星载毫米波云雷达反演液态水云含水量的算法性能分析

Simulation analysis on microphysical parameters retrieval of liquid water clouds with satellite-based millimeter-wave radar

为了充分了解毫米波雷达遥感液态水云微物理参数的性能,介绍了一种基于最优估计理论从星载毫米波雷达反演液态水云微物理参数的方法,并通过模拟试验对算法性能做了初步分析。模拟分析表明,假定云中粒子半径的垂直分布不均一,而数密度和谱宽分布均一,当观测不确定度等于1.5 dBZ时,整层廓线几何平均半径、数密度、谱宽以及液态水含量LWC(liquid water content)的相对不确定度分别为:13%、59%、15%和50%。各谱参数之间存在着一定的相关性,在计算LWC不确定性时,考虑这种互相关性可以使廓线的反演不确定度显著减小。介绍了联合雷达反射率和可见光厚度信息反演云微物理参数的方案。对比试验表明,不同的参数假定方案下,联合反演方法对反演结果的影响不同。若假定云中粒子半径的垂直分布不均一,而数密度和谱宽分布均一,则联合反演方法可使LWC反演不确定度从50%减小到27%;而若假定云中3个参数的垂直分布均不均一,联合反演对于结果的改进相对较小,仅从50%减小到了48%。

To investigate the characteristics of microphysical parameters, a method based on the optimal estimation theory was used by the retrieval from spaceborne millimeter-wave radar. In order to study the performance of the method, a series of simulations were performed. The results show that the relative uncertainties of the geometric mean radius （rg）, droplet number density （N,）, width parameter （σlog） and liquid water content （LWC） are about 13%, 59%, 15% and 50% for the assumption of stratiform clouds when the radar observation error is 1.5 dBZ, and that considering the correlation among the rg, N, andσlog, the uncertainty of LWC will reduce obviously. Another retrieval scheme combining the radar reflectivity and the visible optical depth was also performed. The comparison between the two schemes show that the retrieval schemes have different impacts on the results under different assumptions, that the precision of LWC can be improved from 50% to 27% with homogeneous distribution of σlog and N, while rg are not. However, the precision of LWC can only be improved from 50% to 48% with inhomogeneous distribution of all the three parameters.