Imager‐assisted cloud detection for assimilation of Infrared Atmospheric Sounding Interferometer radiances

• Published in: Quarterly Journal of the Royal Meteorological Society
• Main Author: Eresmaa, Reima
• Other Authors: EUMETSAT
• Format: Article in Journal/Newspaper
• Language: English
• Published: Wiley 2014
• Subjects: Sea ice
• Online Access: http://dx.doi.org/10.1002/qj.2304; https://api.wiley.com/onlinelibrary/tdm/v1/articles/10.1002%2Fqj.2304; https://rmets.onlinelibrary.wiley.com/doi/pdf/10.1002/qj.2304

The operational assimilation of Infrared Atmospheric Sounding Interferometer (IASI) radiances at the European Centre for Medium‐range Weather Forecasts (ECMWF) relies primarily on the use of clear data, either in completely cloud‐free locations or restricting the assimilation to channels that are insensitive to underlying cloud. Prior to the data assimilation, cloud‐contaminated channels are identified and rejected in cloud detection, i.e. in a screening process based on observation minus background departure data. Background errors have the potential to confuse the cloud detection. On the one hand, a false alarm occurs when a background error is incorrectly interpreted as a cloud. On the other hand, cloud is missed if the background error compensates for the cloud radiative effect. This article outlines a method to improve the cloud detection by making additional use of collocated imager data from the Advanced Very High Resolution Radiometer (AVHRR). An independent cloud‐detection scheme, based only on the AVHRR data, is formulated and compared with the departure‐based scheme currently in operational use at ECMWF. The intercomparison reveals a considerable number of discrepancies, with only one of the two schemes suggesting the presence of cloud. Combining the two schemes results in an imager‐assisted scheme, where the AVHRR data are used to set an additional requirement before allowing an IASI field of view to be diagnosed completely clear of clouds. In data assimilation experiments, using the imager‐assisted scheme results in systematic lower tropospheric warming in the winter hemispheres, particularly over the Arctic sea ice. The modified cloud detection is shown to have a modestly positive impact on independent observation departure statistics and forecast scores.