Abstract

With the exception of in situ atmospheric probes, the most useful way to study the atmospheres of other planets is to observe their electromagnetic spectra through remote observations, either from ground-based telescopes or from spacecraft. Atmospheric properties most consistent with these observed spectra are then derived with retrieval models. All retrieval models attempt to extract the maximum amount of atmospheric information from finite sets of data, but while the problem to be solved is fundamentally the same for any planetary atmosphere, until now all such models have been assembled ad hoc to address data from individual missions. In this paper, we describe a new general-purpose retrieval model, Non-linear Optimal Estimator for MultivariatE Spectral analySIS (NEMESIS), which was originally developed to interpret observations of Saturn and Titan from the composite infrared spectrometer on board the NASA Cassini spacecraft. NEMESIS has been constructed to be generally applicable to any planetary atmosphere and can be applied from the visible/near-infrared right out to microwave wavelengths, modelling both reflected sunlight and thermal emission in either scattering or non-scattering conditions. NEMESIS has now been successfully applied to the analysis of data from many planetary missions and also ground-based observations.