Investigating CloudSat Retrievals Sensitivity to Forward Iterative Algorithm Parameters in the Mixed Cloud Layers

Abstract

When millimeter-wave cloud radar data are used for the forward iterative retrieval of the liquid water content (LWC) and effective radius of cloud droplets (Re) in a cloud layer, the prior values and tolerance ranges of the cloud droplet number density (Nt), scale parameter (Rg) and spectral width parameter (Wg) in the iterative algorithm are the main factors that affect the retrieval accuracy. In this study, we used data from stratus and convective clouds that were simultaneously observed by CloudSat and aircraft to conduct a sensitivity analysis of Nt, Rg, and Wg for the retrieval accuracies of LWC and Re in both stratus and convective clouds. Nt is the least sensitive parameter for accurately retrieving stratus LWC and Re in both stratus and convective clouds, except for retrieving the convective cloud LWC. Opposite to Nt, Rg is the most sensitive parameter for both LWC and Re retrievals. As to the effects of parameter tolerance ranges on the retrievals of LWC and Re, the least important parameter is the Nt tolerance range; the most important one is the Wg tolerance range for retrieving convective cloud LWC and Re, the Rg is the important parameter for retrieving stratus LWC and Re. To obtain accurate retrieved values for clouds in a specific region, it is important to use typical values of the sensitive parameters, which could be calculated from in situ observations of cloud droplet size distributions. In addition, the sensitivities of the LWC and Re to the three parameters are stronger in convective clouds than in stratus clouds. This may be related to the melting and merging of solid cloud droplets during the convective mixing process in the convective clouds.