Abstract

The retrieval of information from measured limb radiance is usually done by iteration of the unknown independent parameters in a forward model until measurement matches calculation. The extremely non-linear nature of the atmospheric infrared radiative transfer equation requires a very accurate model, particularly for emission. Ideally, a rigorous line-by-line calculation should be done for each atmospheric state addressed. Past limitations in algorithms, hardware, and dollars (or a combination of the same) confined such calculations to small data sets or research simulations. However, improvements in algorithms and computer hardware have made the ideal possible, for many applications, on relatively inexpensive systems. This paper will give an overview of algorithms, methodology, and hardware now being prepared for use in the reduction of future data sets from HALOE (The Halogen Occultation Experiment) to fly on UARS and SAFIRE (Spectroscopy of the Atmosphere using Far Infra-Red Emission) scheduled to fly on the second EOS polar platform. Several high resolution multipath examples with timing benchmarks on different machines are presented.

© 1990 Optical Society of America