Characterizing Surfaces of Terrestrial Exoplanets during Atmospheric Retrieval of Emission Spectra

Upcoming missions such as JWST will allow for spectroscopically characterizing the atmospheres of terrestrial exoplanets with unprecedented detail. These planets pose unfamiliar challenges to the current Bayesian atmospheric retrievals. One such challenge is the presence of a planetary surface, which may impact the observed emission spectrum if the planet has an optically thin atmosphere. In Whittaker et al. (submitted), we found that the albedo spectrum of the surface can bias the inferred atmospheric abundances of existent gases and mimic the signatures of absent ones. In the work presented here, we incorporate realistic spectral surfaces into our atmospheric retrieval model and test the feasibility of identifying the surface composition and surface pressure via emission spectroscopy. We explore planets of various atmospheric and surface properties. We find that a grey surface albedo model does not correctly retrieve these properties and thus a wavelength-dependent albedo is necessary. We find that the surface pressure is a robustly estimated parameter. We find that, broadly speaking, simultaneously retrieving the atmospheric chemistry, thermal structure, and surface composition is prohibitive, but which piece of information is correctly estimated depends on the specific planet. I will talk about how to thus best implement and interpret retrievals for the thermal emission of rocky planets. We also test various parameterizations of the surface albedo spectrum and discuss degeneracy with aerosols. We also identify potential targets for surface characterization with JWST to test the connection between interior evolution and surface types, and a temperature-dependent dichotomy between solid surfaces and magma oceans.