CERBERUS: a Bayesian retrieval scheme for efficiently applying high dimensionality model spaces to the study of exoplanet atmospheres(NPO-50770-1)

CERBERUS is a scientific software code developed for estimating properties of exoplanet atmospheres - planets that orbit nearby stars - based on spectroscopic observations from astronomical observatories. The code implements a Bayesian approach coupled with a novel branch-pruning method which allows the efficient evaluation of a complex model space that incorporates both thermo chemical equilibrium (TCE) and non-TCE models for atmospheric chemistry. The forward models used in the retrieval process include a detailed radiative transfer equations and support for both gaseous and aerosol opacity sources. Vertical dependence for both opacity sources and the atmospheric temperature profile are supported. The model selection capability of CERBERUS is a significant advancement over the current state of the art and allows CERBERUS to efficiently evaluate the statistical evidence a much broader range of models than other codes; when combined with the new high spectral resolution observations that are now becoming available, CERBERUS allows new insights in the conditions and composition of exoplanet atmospheres and allows rapid processing of large data sets.