Recent studies have suggested that the current treatment of IR radiative transfer in climate models and retrieval algorithms may not be entirely realistic. These studies show that current treatments (absorption approximation and plane-parallel homogeneous cloud assumption) result in ORL biases on the order of 5-15 W m-2 from ignoring the effects of IR multiple scattering or cloud inhomogeneity. Errors of this magnitude may compromise the integrity of large-scale dynamic features simulated by climate models and/or atmospheric properties derived from space-based measurements.

Through observation and modeling the extent of IR multiple scattering and 3-D effects on TOA biases in current radiative transfer models is examined. Moreover a treatment that account for the combined effects is considered. A comprehensive set of cloud and radiation observations obtained at the Department of Energy's Atmospheric Radiation Measurement Program (ARM) Southern Great Plains (SGP) Cloud and Radiation Testbed (CART) site is employed in this investigation.