EOS Microwave Limb Sounder Retrievals: New Approaches, Theory and Results From Simulations.

An overview is given of the retrieval and forward model algorithms that form the `Level 2 processing' for the Microwave Limb Sounder (MLS) instrument on the EOS Aura spacecraft. These are the algorithms used to convert the calibrated observations of thermal microwave emission from the Earth's limb into estimates of atmospheric temperature and composition for use by the scientific community. The algorithms take into full consideration atmospheric inhomogeneities along the line of sight, in what is essentially a `tomographic' view of the limb sounding optimal estimation retrieval problem. Results are shown from pre-launch simulations of the EOS MLS measurement system. These give the expected precision of atmospheric data products from the MLS instrument, and have been key to the MLS science team's preparation for the EOS Aura launch. Great benefit has been obtained from designing the software to be flexible enough for both routine data processing and the generation of simulated data and various other pre- and post- launch support and testing.