The evaluation of the HIRS/4 instrument on Metop 2 by using the ECMWF data assimilation system to provide a reference to compare with other sensors.

TitleThe evaluation of the HIRS/4 instrument on Metop 2 by using the ECMWF data assimilation system to provide a reference to compare with other sensors.
Publication TypeMiscellaneous
Year of Publication2007
AuthorKelly, GA
Secondary TitleTechnical Memorandum
Number523
Abstract

The HIRS instrument was first flown onboard NIMBUS-6 in (June 1975 to May 1976). Since Oct 1978 there have been one or more of these instruments providing operational good quality inferred measurements on the NOAA operational satellites and now onboard Metop A. Although this instrument is inferior to the new multi-spectral instruments, such as AIRS and IASI, it is of sufficient accuracy to provide a 30 year global climate record of temperature and moisture. The current operational version of the ECMWF data assimilation was considered unsuitable for this study because of the use of Variational satellite bias correction that slowly evolves the satellite radiance biases with the time varying forecast error. A version of the operational system was run using non-evolving satellite bias coefficients. An additional study was undertaken to estimate the bias corrections for HIRS channel 8 (atmospheric window) on Metop A and NOAA17. The HIRS window check and the cloud detection scheme are required to produce the HIRS clear radiances. Using the data assimilation as the reference, a comparison of the HIRS clear radiances was made between HIRS/4 onboard Metop A and HIRS/3 on board NOAA17. These two instruments have very similar departure statistics (with the exception of channel 1 on HIRS NOAA17). These departure statistics are also similar for equivalent microwave channels from AMSUA and MHS on Metop A and NOAA18. There is also evidence that the smaller field of view for HIRS on Metop A produces about 20 % more clear spots. Finally a comparison is made with the current operational HIRS cloud detection and there is a need to retune the operational HIRS cloud detection.