Radiance Tuning and First Guess Regression Updates for NOAA-21 NUCAPS Augmentation

The NUCAPS enterprise algorithm developed at the NOAA Center for Satellite Applications and Research (STAR) uses the hyper-spectral infrared and microwave observations to operationally retrieve vertical profiles of temperature, moisture, ozone, CO, CH4, and CO2 profiles separately from the JPSS (using CrIS and ATMS) and the MetOp (IASI and AMSU-A/B) series of satellites. The algorithm starts with Rosenkranz’s microwave-only retrieval using microwave sensor radiances, followed by a fast eigenvector regression of hyperspectral IR radiances from the European Center for Medium-Range Forecasts (ECMWF) to provide the initial solution for the final IR physical retrieval. The physical retrieval algorithm is built on the heritage of the Atmospheric Infrared Sounder (AIRS) science team algorithm. The NUCAPS system has been augmented to use CrIS/ATMS sensor data from the recently launched NOAA-21 satellite to generate NOAA-unique NUCAPS products. The NUCAPS products retrieved from the NOAA-21 satellite are being validated through a hierarchy of validation data sets. The augmentation of the NOAA-21 NUCAPS system required updates to the first guess regression of temperature and water vapor retrievals, and the radiance tuning to adjust the observed radiances to synthetic radiances. Two methods of radiance tuning have been adapted to study the relative merits of these methods. The cloudy and clear first guess regression update is performed using four focus days of training data spanned around four different seasons to accommodate seasonality. Results of the evaluation of the radiance tuning and the first guess regression updates applied on independent data sets, statistical metrics, and impacts on the final physical retrieval will be presented at the conference.