Analysis of two independent methods for retrieving liquid water profiles in spring and summer Arctic boundary clouds

• Published in: Journal of Geophysical Research - Atmospheres Vol. 108; no. D7; pp. 4219 - n/a
• Main Authors: Löhnert, U., Feingold, G., Uttal, T., Frisch, A. S., Shupe, M. D.
• Format: Journal Article
• Language: English
• Published: American Geophysical Union 16-04-2003; Blackwell Publishing Ltd
• Subjects: active and passive ground-based remote sensing; Atmospheric Composition and Structure; Boundary layer processes; Brackish; cloud liquid water algorithms; Cloud physics and chemistry; comparison of Arctic cloud liquid water profiles; Freshwater; Instruments and techniques; LES model with explicit microphysics; Marine; Meteorology and Atmospheric Dynamics; Numerical modeling and data assimilation; optimal estimation; Remote sensing; sensor synergy; PN, Arctic Ocean; PN, Arctic
• ISSN: 0148-0227; 2156-2202
• DOI: 10.1029/2002JD002861

A large number of all‐liquid, nondrizzling stratus clouds (163 hours of measurements) were observed with a dual‐channel microwave radiometer and a colocated 35‐GHz cloud radar during the spring and summer months of the Surface Heat Budget of the Arctic Ocean (SHEBA) project. An algorithm developed by Frisch et al. [1995, 1998] to derive the liquid water content (LWC) is applied to these measurements assuming constant cloud drop number density and cloud drop size distribution breadth with height. A second algorithm developed by Löhnert et al. [2001] is specifically adapted for SHEBA clouds using a priori information from a large eddy simulation (LES) model initialized with summertime SHEBA radiosondes; about 50 soundings during nondrizzling, low‐level, all‐liquid water clouds are used. Using model‐derived drop size distributions, a relationship between simulated radar reflectivity (Z) and model LWC is derived as well as an a priori LWC profile. Once the theoretical error covariance matrix of the Z‐LWC relation is derived and the covariance matrix of the LWC profile is calculated, an optimal estimation method is applied to the SHEBA data. The Frisch et al. and Löhnert et al. methods are also applied to the LES model output, resulting in overall root‐mean‐square differences on the order of 30 to 60%. Both methods are sensitive to the assumed accuracies of the microwave‐radiometer‐derived LWP. When applied to LES model output, the Frisch et al. method shows a LWC overestimation in the lower parts of the cloud. These systematic errors are induced by the assumption of constant cloud number concentration with height.