Clouds
1. Arlindo Arriaga |
|
The method is based on the CO2 slicing technique for IASI, in agreement with the one previously developed and tested by Smith and Frey (1990) and recommended to EUMETSAT a long time ago by the IASI Sounding Science Working Group. (Smith, W.L., and Frey, R., 1990: On cloud altitude determination from high resolution interferometer sounder (HIS) observations, J. Appl. Meteor, 29, 658-662). The main points of that algorithm were presented at the ITSC-16, 2008, with a drafted paper (should be submitted for publication).
That algorithm does not include dedicated tests for cloud detection within IASI IFOVS. These ones are performed on the IASI L2 processing chain BEFORE calling the CO2 slicing algorithm. Therefore, the algorithm (which works with IASI radiances and co-located NWP forecasts) is supposed to run for IASI IFOVS previously identified as cloudy. The implementation of those tests for cloud detection (using IASI) was performed by Xavier Calbet. A new colleague is now reviewing the tests on cloud detection for improvement.
For specific information on validation of those clouds tests, as well as on the validation of the CO2 slicing algorithm you may be willing to take a look at the EUMETSAT document:
"Peter Schlüssel, Xavier Calbet, Tim Hultberg, Arlindo Arriaga, Thomas August, Olusoji Oduleye, and Hidehiko Murata, 2008: EPS Product Validation Report: IASI L2 PPF. EUMETSAT document EUM/MET/REP/07/0224, v2D Draft, from August 2008." (available as PDF file on
http://www.eumetsat.int/Home/Main/Documentation/Technical_and_Scientific_Documentation/ProductValidationReports/?l=en)
The fundamental description of such cloud tests and the basis of the adopted technique for implementation of the CO2 slicing method can be find on "EPS Ground Segment, IASI Level 2 Product Generation Specification, 2004: EUMETSAT document EPS.SYS.SPE.990013, Issue 5 Rev 5, WBS number 260000, 31 March 2004" (available as PDF file on http://www.eumetsat.int/Home/Main/Documentation/Technical_and_Scientific_Documentation/Technical_Notes/SP_1126189367518?l=en)
The output from the CO2 slicing method (cloud top pressure and equivalent cloud amount) might also help on the identification of the IFOV as cloudy or clear. Indeed, if the equivalent cloud amount is very low, then the IFOV might be considered to be clear. There are also cases with significant cloud amount but unfortunately noisy cloud signal - what hampers the application of the method, and thus the retrieval of the cloud top pressure and equivalent cloud amount. This happens very especially with low level clouds, namely over ice (and desert). This means however that the application of the CO2 slicing method may also help to detect (or just confirm) a cloudy IFOV but the retrieval might not be feasible with noisy cloud signal.
We also have interest on using IASI and AVHRR for retrieving the cloud top pressure of multi-cloud formations within a single IFOV. The basic idea is to work with the AVHRR information on cloudy and clear clusters identified within IASI IFOVS. I believe to have by end 2010 a validated algorithm running.