Radiative Flux Estimation from a Broadband Radiometer Using Synthetic Angular Models in the EarthCARE Mission Framework. Part II: Evaluation

The instantaneous top-of-atmosphere (TOA) radiance-to-flux conversion for the broadband radiometer (BBR) on board the Earth Clouds, Aerosols, and Radiation Explorer (EarthCARE) was assessed in Part I of this paper, by developing theoretical angular distribution models (ADMs) specifically designed for the instrument viewing configuration. This paper validates the BBR ADMs by comparing derived flux estimates with flux retrievals obtained from the Clouds and the Earth’s Radiant Energy System (CERES) Terra models. A CERES BBR-like database is employed in the assessment, which is an optimum dataset to validate the BBR algorithms and to determine the benefits of the multiangular conversion procedures in the BBR instrument. The validation of theoretical results with empirical data is essential to prepare the conversion algorithms prior to the launch of EarthCARE. This paper demonstrates that the application of a linear combination method is not recommended when outgoing radiances do not follow the response modeled in the radiative transfer calculations. An effective radiance averaged model outperforms all other developed models, in terms of the coefficient of variation of the root-mean-square error, in the validation study of the shortwave (SW) regime (clear sky 1.9%; cloudy 7.1%) while an effective radiance along-track model obtains the best comparisons for the longwave (LW) regime (clear sky 1.4%; cloudy 1.5%). The evaluation of the multiangular models with scenes with high anisotropy shows that multiview flux conversion algorithms can statistically improve CERES ADM results when CERES flux discrepancies of a target are higher than 4 W m−2 in the LW domain and SW clear-sky scenes and higher than 20 W m−2 in scenes with cloudy conditions.