This paper describes the implementation of an efficient and accurate multiple scattering parameterization within the LOWTRAN and FASCODE transmittance/radiance models. The parameterization is based on a stream approximation in which the local radiance field needed to evaluate the multiple scattering source function is estimated from the local radiation fluxes. The latter are calculated based on a parameterized two-flux for individual layers and an adding method for combining layers. Because of the line-by-line nature of FASCODE, it is straightforward to implement the multiple scattering treatment. For LOWTRAN, an interface scheme was developed using the k-distribution method to match the multiple scattering approach to the band model calculation of gas absorption. The interface scheme represents the LOWTRAN band model by a sum of pseudomonochromatic calculations. The approach is valid for any band model for which k-distribution parameters can be evaluated. The accuracy of the multiple scattering parameterization has been demonstrated by comparing it with more detailed calculations for a variety of atmospheric conditions. RMS errors in radiance considering all possible viewing angles are <20%. In addition, to insure consistency between models, overlapping LOWTRAN and FASCODE spectral regions are compared. Finally, it is demonstrated that the implemented multiple scattering parameterization corrects LOWTRAN's previous underestimation of path radiance for long horizon paths where multiple scattering is significant.

This report covers the project progress of grant DE-FGO2-86ER60422 for the period March i, 1990 .. October I, 1990. The research program includes three tasks' GCM intercomparison and improvement, cl.imate data-model statistics, and China project ;cience coordination. This work has been performed in collaboration with our subcontractor, Dr. Wei-Chyung Wang, SUNY/Albany. Their progress is summarized below. Task i: GCM Intercomparison and Improvement This task includes three subtasks" spectral-to-grid GCM comparison, regional climate study and energy/hydrologic cycle. S__ctral-to-Grid GCM Comparison Research on this task has been completed and described in a DOE technical