Earth Radiation Budget Experiment (ERBE) reprocessing using Clouds and the Earth's Radiant Energy System (CERES) angular distribution models

Abstract

NASA's Earth Radiation Budget Experiment (ERBE) scanning broadband radiometers flew on board the NOAA 9 (Feb 1985 to Jan 1987) and NOAA 10 (Jan 1987 to May 1989) and measured broadband shortwave (∼0.2 μm to 5 μm), longwave (5 μm to 50 μm) and total radiances. While the observations provided solid evidence of the cooling effect on the Earth system by clouds, the uncertainty of cloud radiative effects by region or by cloud type is large compared to those derived more recently from NASA's Clouds and the Earth Radiant Energy System (CERES) observations. In ERBE, top-of-atmosphere (TOA) irradiances were derived by applying 12 scene-type dependent angular distribution models (ADMs). Scene type viewed by ERBE scanners was estimated from broadband radiances using a maximum likelihood estimate method [1]. In this study, we use data taken by Advanced Very High Resolution Radiometer (AVHRR) on board the NOAA-9 satellite to derive cloud properties similar to those produced by the CERES cloud algorithm that utilizes Moderate Resolution Imaging Spectrometer (MODIS) data collocated with CERES footprints. This allows direct application of newer CERES ADMs to ERBE scanner radiances, which in turn reduces the uncertainty in the TOA irradiances. We describe the process of applying CERES ADMs and a comparison of the reprocessed data with original ERBE data. The reprocessing of 4 months of NOAA-9 measurements indicated increase in the global monthly mean shortwave TOA irradiance by ∼4%, while longwave TOA irradiance decreased by ∼0.5%, compared to irradiances derived from ERBE ADMs. These differences are largely caused by the pixel sizes of AVHRR and MODIS that yield different cloud type probability distributions.