Deciphering the Atmospheric Composition of WASP-12b: A Comprehensive Analysis of its Dayside Emission

TL;DR

This study provides a comprehensive analysis of WASP-12b's dayside emission, confirming its carbon-rich atmosphere and resolving previous measurement discrepancies through uniform data analysis and advanced retrieval models.

Contribution

It offers the first uniform analysis of all available secondary-eclipse data and demonstrates that including C2H2 and HCN in models supports a plausible carbon-rich atmosphere.

Findings

Confirmed shallow eclipse depth at 4.5 microns.

Carbon-rich models with C2H2 and HCN fit data best.

Oxygen-rich models are inconsistent with chemical expectations.

Abstract

WASP-12b was the first planet reported to have a carbon-to-oxygen ratio (C/O) greater than one in its dayside atmosphere. However, recent work to further characterize its atmosphere and confirm its composition has led to incompatible measurements and divergent conclusions. Additionally, the recent discovery of stellar binary companions ~1" from WASP-12 further complicates the analyses and subsequent interpretations. We present a uniform analysis of all available Hubble and Spitzer Space Telescope secondary-eclipse data, including previously-unpublished Spitzer measurements at 3.6 and 4.5 microns. The primary controversy in the literature has centered on the value and interpretation of the eclipse depth at 4.5 microns. Our new measurements and analyses confirm the shallow eclipse depth in this channel, as first reported by Campo and collaborators and used by Madhusudhan and collaborators to infer a carbon-rich composition. To explain WASP-12b's observed dayside emission spectrum, we implemented several recent retrieval approaches. We find that when we exclude absorption due to C2H2 and HCN, which are not universally considered in the literature, our models require implausibly large atmospheric CO2 abundances, regardless of the C/O. By including C2H2 and HCN in our models, we find that a physically-plausible carbon-rich solution achieves the best fit to the available photometric and spectroscopic data. In comparison, the best-fit oxygen-rich models have abundances that are inconsistent with the chemical equilibrium expectations for hydrogen-dominated atmospheres and are 670 times less probable. Our best-fit solution is also 7.3*10^{6} times more probable than an isothermal blackbody model.