Use of Meteosat Second Generation optimal cloud analysis fields for understanding physical attributes of growing cumulus clouds

This study develops an understanding on how retrieved cloud parameter fields from the Optimal Cloud Analysis (OCA) algorithm, operating on Meteosat Second Generation (MSG), Spinning Enhanced Visible and Infrared Imager (SEVIRI) data, behave at 5-min time resolutions for growing cumulus clouds. Fields retrieved by the OCA algorithm include cloud optical thickness (τ), cloud-top particle effective radius (re), cloud-top pressure (pc), and cloud-top phase. OCA is based on a one-dimensional optimal estimation methodology, and a measure of radiance fit, the cost function (Jm), is a quantity developed as part of the retrieval process and is shown to be useful in delineating mixed phase clouds; it too is evaluated (at 5-min intervals) for the information it provides. Data for 94 growing cumulus cloud events are processed. An “event” is defined as a cumulus cloud that is monitored at 5-min intervals with OCA, as it grows from the “fair weather” or “towering cumulus” stage to near the cumulonimbus stage when precipitation begins. The hypothesis is that OCA products are of high-enough quality to provide unique information about microphysical processes occurring at and near cloud top. The goal through analysis of the 94 events is to identify consistent, repeating patterns in OCA fields during cloud growth that can be in turn used to infer physical processes. Data from the Convective and Orographicallyinduced Precipitation Study (June and July 2007) and in four regions of Europe on 25 May 2009 are used. The validity of the OCA data is presented with a comparison to CloudSat Precipitation Radar and MODerate resolution Imaging Spectroradiometer retrieved cloud properties, showing good statistical agreements. Subsequently, results from the analysis of OCA fields for all events show that as cumuli deepen, re values tend to increase, and then decrease in size as cloud tops glaciate and particle settling begins. The τ magnitudes generally increase as clouds deepen, while pc values and cloud-top temperatures fall as expected. The Jm values exhibit the pattern of spiking in magnitude (over a 5–10-min period), which indicates the increase “misfit” within OCA during the mixed phase, at about the time τ values increase substantially as clouds deepen.