The algorithm to retrieve atmospheric legacy profiles from the Advanced Baseline Imager (ABI) onboard the next generation Geostationary Operational Environmental Satellite (GOES-R) is developed, forecast profile from the numerical weather prediction (NWP) model is used as as the first guess. The retrieval process includes two parts: a non-linear regression followed by a physical iterative approach. Since the ABI real data is not yet available and due to the similar characteristics between ABI and the Spinning Enhanced Visible and Infrared Imager (SEVIRI) onboard the Meteosat Second Generation (MSG) satellite, the SEVIRI data is used as proxy to test the algorithm. The European Centre for Medium-Range Weather Forecasts (ECMWF) forecast (12h) profiles are used as first guess in the calculations. The ECMWF analysis data as well as the radiosonde data are used for validation.

The SEVIRI data for August 2006 is processed. The results show that both the non-linear regression and the physical retrieval lead to the better root-mean-squared error (RMSE) for humidity profile than forecast. The physical retrieval leads to better results than the regression. With information contribution from the two water vapor absorption spectral bands (6.2- and 7.2-µm), the improvement from SEVIRI is significant between 300 and 700 hPa for moisture over the forecast. Since there is only one temperature-sensitive spectral band in SEVIRI, the temperature profile does not show noticeable improvement over forecast.