Open AccessValidation of longwave atmospheric radiation models using Atmospheric Radiation Measurement data
Author(s) -
Zhou Y. P.,
Cess Robert D.
Publication year - 2000
Publication title -
journal of geophysical research: atmospheres
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.67
H-Index - 298
eISSN - 2156-2202
pISSN - 0148-0227
DOI - 10.1029/2000jd900557
Subject(s) - longwave , outgoing longwave radiation , environmental science , atmosphere (unit) , downwelling , radiant energy , atmospheric model , radiance , radiative transfer , radiant flux , meteorology , radiation , infrared window , atmospheric sciences , atmospheric models , earth's energy budget , remote sensing , physics , geology , infrared , optics , convection , oceanography , upwelling
Data taken at the Atmospheric Radiation Measurement Program's central facility in Oklahoma and processed as part of the Clouds and the Earth's Radiant Energy System‐Atmospheric Radiation Measurement‐Global Energy and Water Cycle Experiment (CAGEX) project have been used to validate the top‐of‐the‐atmosphere and surface longwave radiative fluxes for two widely used radiation models: the Column Radiation Model from the National Center for Atmospheric Research Community Climate Model (CCM), and the Moderate Resolution Transmittance (MODTRAN3) radiation code. The results show that for clear skies the models slightly overestimate outgoing longwave radiation at the top of the atmosphere (OLR) and underestimate the surface downwelling longwave flux (SDLW). The accuracy of the radiation models is quite consistent with their respective levels of complexity. For MODTRAN3, for example, the OLR overestimate is 7.1 Wm −2 while the SDLW underestimate is 4.2 Wm −2 . For cloudy skies it is emphasized that the cloud input parameters, as determined from measurements by various instruments, require careful examination and preprocessing. Spatial and temporal averaging could result in the parameters representing different volumes of the atmosphere. The discrepancy between model calculations and observations is shown to be significantly reduced through the proper choice of input parameters.