Simulated proxy imagery for the future Advanced Baseline Imager (ABI) has been valuable in a number of GOES-R activities, including the Algorithm Integration Team (AIT), Proving Ground, the GOES-R Analysis Facility for Instrument Impacts on Requirements (GRAFIIR), and the Algorithm Working Group (AWG) Sounding, Winds, Clouds, Aviation, Visibility/Aerosol, Imagery/Visualization, and Fire Detection teams. The quality of the imagery will depend on the realism of the NWP model simulation and the integrity of the forward radiative transfer model (RTM) and surface radiative properties. Because clouds have a major impact on the simulated imagery, it is important to know how well cloud absorption/scattering properties, which provide the crucial link between microphysics and the radiative transfer equation, are characterized.

This study focuses on evaluating the single-scattering properties of cirrus for the standard infrared window band of the ABI (band 14; 11.2 micrometers) in the Community Radiative Transfer Model (CRTM) and the Cooperative Institute for Meteorological Satellite Studies (CIMSS) RTM, two of the most widely used RTMs for generating simulated ABI imagery. Observations from MODIS band 31 (11.03 micrometers) are used as a proxy for ABI band 14, while MODIS particle effective radius products and ice water content (IWC) profiles from the Cloud Profiling Radar (CPR) on CloudSat are used as input to the RTMs. Cirrus cases were selected using CloudSat cloud classification data over ocean regions for 2007.

Early results show that when compared to MODIS observations, the CIMSS RTM has a positive offset in the simulated brightness temperatures (BTs) of about 5 K, but is within errors expected from uncertainties in the IWC and effective radius products. While the CRTM compares well to observations at warm BTs (i.e., where cirrus is optically thin), there are large biases (up to -20 K) in BTs for cirrus with optical depths of 1 to 2 that exceed errors from uncertainties in the input cloud products. Since particle extinction and single-scattering albedos in the respective RTMs are not drastically different, it is proposed that the bias may be caused by the treatment of the scattering phase function within the CRTM.