Infrared Cloudy Radiative Transfer Validation Using Coincident AIRS and MODIS Observations

Modeling observed cloudy radiances of satellite-based infrared sounders is challenging because of scene heterogeneity. Comparisons of observed and calculated AIRS radiances are conditioned against cloud statistics from the MODIS cloud products to characterize the accuracy of the radiative transfer and its dependence of scene complexity. Radiances are calculated with the SARTA fast radiative transfer algorithm using atmospheric temperature, water vapor and ozone profiles from the ECMWF operational forecast, but cloud fields from the ECMWF forecasts, and AIRS L2 and MODIS L2 observational products. The error arising from differences in cloud representations and their translations to input to the radiative transfer models is discussed in the first half of the presentation, while the latter half deals with the treatment of scene heterogeneity and the error this adds to the modeled radiances. Calculated radiances smoothed over an AIRS footprint from high spatial atmospheric states are compared with radiances from smoothed states and observed radiances. The goal of this study is to quantify the improvement in cloudy radiative transfer modeling when external information about scene complexity is applied.