From the Shadows: The Impact of Nightside Thermal Emission on Ultra-hot Jupiter Transmission Spectra Retrievals

TL;DR

This paper demonstrates that neglecting nightside thermal emission in transmission spectra retrievals of ultra-hot Jupiters can lead to significant biases, and proposes a method to include this effect for more accurate atmospheric characterization.

Contribution

It introduces a modified retrieval approach that accounts for nightside thermal emission, improving the accuracy of atmospheric property estimates for ultra-hot Jupiters.

Findings

Neglecting nightside emission can overestimate molecular abundances by nearly tenfold.

Omitting nightside thermal emission can underestimate dayside temperature by over 400 K.

Including nightside emission in models yields more accurate atmospheric parameters.

Abstract

Transmission spectroscopy is the most widely used technique for studying exoplanet atmospheres. Since the planetary nightside faces the observer during a transit, highly irradiated giant exoplanets with warm nightsides emit thermal radiation that can contaminate transmission spectra. Observations of ultra-hot Jupiters in the near- and mid-infrared with JWST are especially susceptible to nightside contamination. However, nightside thermal emission is generally not considered in atmospheric retrievals of exoplanet transmission spectra. Here, we quantify the potential biases from neglecting nightside thermal emission in multidimensional atmospheric retrievals of an ultra-hot Jupiter. Using simulated JWST transmission spectra of the ultra-hot Jupiter WASP-33b (0.8-12 $\mu$m), we find that transmission spectra retrievals without nightside emission can overestimate molecular abundances by almost an order-of-magnitude and underestimate the dayside temperature by $\gtrsim$ 400 K. We show that a modified retrieval prescription, including both transmitted light and nightside thermal emission, correctly recovers the atmospheric properties and is favored by Bayesian model comparisons. Nightside thermal contamination can be readily implemented in retrieval models via a first-order approximation, and we provide formulae to estimate whether this effect is likely to be significant for a given planet. We recommend that nightside emission should be included as standard practice when interpreting ultra-hot Jupiter transmission spectra with JWST.