Investigating the detectability of hydrocarbons in exoplanet atmospheres with JWST

Abstract

We investigate at what abundances various hydrocarbon molecules (e.g. acetylene (C$_2$H$_2$), ethylene (C$_2$H$_4$), and methane (CH$_4$)) become detectable when observing the atmospheres of various planets using JWST. We focused on atmospheric models based on the parameters of a small sample of planets: HD 189733b, HD 209458b, HD 97658b, and Kepler-30c. We computed model transmission spectra using the Bayesian retrieval package ARCiS. We simulated observed spectra using the PandExo package. We subsequently ran retrievals on these spectra to determine whether the molecular abundances can be accurately retrieved from these simulated spectra. We find that generally we can detect and retrieve abundances of the hydrocarbon species as long as they have a VMR above approximately 1x10$^{-7}$-1x10$^{-6}$, at least for the brighter targets. There are variations based on planet type and instrument(s) used, and these limits will likely change depending on the abundance of other strong absorbers. We also find scenarios where the presence of one hydrocarbon is confused with another; this is often improved when two instruments are combined. C$_2$H$_2$, CH$_4$, and C$_2$H$_4$ will all be detectable with JWST, provided they are present in high enough abundances, and that the optimal instruments are chosen. A combination of two instruments, either NIRSpec G395M and MIRI LRS, or NIRCam F322W2 and MIRI LRS, is best for observing these species in bright exoplanet systems, with NIRSpec G395M and MIRI LRS the best option for the HD 189733b-like atmosphere with clouds included. The use of NIRSpec Prism is tentatively found to be best for fainter targets, potentially in combination with the MIRI LRS slit mode, although the target we test is too faint to draw any strong conclusions. Instrument sensitivity, noise, and wavelength range are all thought to play a role in being able to distinguish spectral features.