摘要

This study demonstrates a simple procedure for sharpening the infrared (IR) field data for 3 km sampling distance Meteosat Second Generation (MSG) pixels that are "partly cloudy," and hence are not completely cloud filled. Without a sharpened IR pixel, signals of clouds are blurred, whereas with sharpening, both partly cloudy (pixels with 1/9th to 8/9th cloud cover) and completely cloudy pixels possess similar IR channel and trend values, making for more consistent use of these data within algorithms that diagnose and predict aspects of clouds (i.e. growth, vertical development, microphysical characteristics).
For this method, I km sampling distance MSG High Resolution Visible (HRV) and IR data are first used to determine a cloud type, and then a fraction of pixels within an IR pixel is determined. Each IR pixel contains 3 x 3 HRV pixels. The focus here is on more dynamically growing convective clouds, and data for 187 cumulus cloud scenes are processed over regions of Europe on 25 May 2009 and 15 August 2010. Results show that the IR fields are normalized from partly cloudy to completely cloudy pixels. Sharpened IR channel data are subsequently used within channel differences and time trends to show the applicability of this exercise.

  • 出版日期2013-12-1

全文