A methodology for simultaneous retrieval of ice and liquid water cloud properties. Part 2: Near‐global retrievals and evaluation against A‐Train products. (16th November 2016)

• Record Type: Journal Article
• Authors: Sourdeval, Odran; C.‐Labonnote, Laurent; Baran, Anthony J.; Mülmenstädt, Johannes; Brogniez, Gérard
• Is Part Of: Quarterly journal of the Royal Meteorological Society. Volume 142:Number 701(2016)
• Page Start: 3063
• Page End: 3081
• Publication Date: 2016-11-16
• Subjects: satellite retrievals -- ice cloud -- liquid cloud -- multi‐layer -- A‐Train -- EarthCare -- optimal estimation -- remote sensing; Meteorology -- Periodicals; 551.5
• DOI: 10.1002/qj.2889
• Languages: English
• ISSNs: 0035-9009

Abstract

Abstract : This article presents results of a novel methodology capable of simultaneously retrieving optical and microphysical properties of multi‐level ice and liquid clouds. The method was introduced in Part I, which theoretically demonstrated its capabilities, and its results are here analysed and evaluated against A‐Train operational products. In addition to being robust to multi‐layer conditions, another advantage of the method is that rigorous uncertainties and analysis tools are attached to its retrievals. Also, the combined use of short‐wave and thermal infrared channels provides a wide range of sensitivity from moderately thin to thick ice cloud layers. Finally, the method is also novel in that the ice water path (IWP) is directly retrieved. These new retrievals should therefore be useful in providing new data for evaluating climate model predictions of IWP. In this study, our methodology has been applied to one year of A‐Train measurements, narrowed to daytime conditions over oceanic surfaces. The retrievals and their uncertainties are statistically analysed, after a thorough discussion of the filtering process. It appears that our method is sensitive to IWPs ranging between about 0.5 and 1000 g m −2, with uncertainties better than 25% between 5 and 500 g m −2 . Retrievals of the optical depth and effective radius of liquid layers have uncertainties better than 20%. Our retrievals are then compared to five independent operational A‐Train products. Very good agreements, well within a factor of 2, are found by comparisons to products from active and passive instruments. These results overall lead to the validation of our method. Additionally, the robustness of passive operational products to multi‐layer conditions is discussed. Preliminary comparisons show a possible overestimation of retrievals obtained under the single‐layer approximation. A thorough assessment of this problem will be addressed in a following study.