[Paper Review] A DETECTION OF WATER IN THE TRANSMISSION SPECTRUM OF THE HOT JUPITER WASP-12b AND IMPLICATIONS FOR ITS ATMOSPHERIC COMPOSITION

Laura Kreidberg, Michael R. Line+9 more|arXiv (Cornell University)|Apr 21, 2015

Stellar, planetary, and galactic studies2 references5 citations

TL;DR

This study presents a high-precision near-infrared transmission spectrum of the hot Jupiter WASP-12b using Hubble Space Telescope/WFC3, detecting water vapor absorption at 7σ confidence. Despite prior claims of a carbon-rich atmosphere (C/O > 1) from dayside emission data, the transmission spectrum retrieval constrains C/O to 0.5+0.2−0.3 at 1σ, ruling out carbon-rich composition at >3σ, highlighting the need for multi-technique, high-precision observations to resolve atmospheric composition tensions.

ABSTRACT

Detailed characterization of exoplanets has begun to yield measurements of their atmospheric properties that constrain the planets' origins and evolution. For example, past observations of the dayside emission spectrum of the hot Jupiter WASP-12b indicated that its atmosphere has a high carbon-to-oxygen ratio (C/O $>$ 1), suggesting it had a different formation pathway than is commonly assumed for giant planets. Here we report a precise near-infrared transmission spectrum for WASP-12b based on six transit observations with the Hubble Space Telescope/Wide Field Camera 3. We bin the data in 13 spectrophotometric light curves from 0.84 - 1.67 $μ$m and measure the transit depths to a median precision of 51 ppm. We retrieve the atmospheric properties using the transmission spectrum and find strong evidence for water absorption (7$σ$ confidence). This detection marks the first high-confidence, spectroscopic identification of a molecule in the atmosphere of WASP-12b. The retrieved 1$σ$ water volume mixing ratio is between $10^{-5}-10^{-2}$, which is consistent with C/O $>$ 1 to within 2$σ$. However, we also introduce a new retrieval parameterization that fits for C/O and metallicity under the assumption of chemical equilibrium. With this approach, we constrain C/O to $0.5^{+0.2}_{-0.3}$ at $1\,σ$ and rule out a carbon-rich atmosphere composition (C/O$>1$) at $>3σ$ confidence. Further observations and modeling of the planet's global thermal structure and dynamics would aid in resolving the tension between our inferred C/O and previous constraints. Our findings highlight the importance of obtaining high-precision data with multiple observing techniques in order to obtain robust constraints on the chemistry and physics of exoplanet atmospheres.

Motivation & Objective

• To obtain a high-precision transmission spectrum of WASP-12b in the 0.84–1.67 µm range using Hubble Space Telescope/WFC3.
• To test the hypothesis of a carbon-rich atmosphere (C/O > 1) in WASP-12b, previously inferred from dayside emission data.
• To assess the consistency of atmospheric composition constraints derived from transmission spectroscopy versus dayside emission spectroscopy.
• To evaluate the impact of chemical equilibrium assumptions and 1D atmospheric models on retrieval-derived C/O and water abundance.
• To advocate for multi-technique, high-precision observations to resolve discrepancies in exoplanet atmospheric composition.
• To improve constraints on atmospheric chemistry and thermal structure through advanced retrieval modeling.

Proposed method

• Acquired six transits of WASP-12b using the Hubble Space Telescope’s WFC3 instrument with G102 and G141 grisms for spectral coverage from 0.82–1.65 µm.
• Performed time-series spectroscopy in both spatial scan and staring modes, with data reduction using custom pipelines to correct for instrumental systematics.
• Binned the data into 13 spectrophotometric light curves and measured transit depths with a median precision of 51 ppm.
• Applied a Bayesian atmospheric retrieval framework to infer atmospheric properties, including water volume mixing ratio and C/O ratio.
• Used a novel retrieval parameterization that jointly fits for C/O and metallicity under chemical equilibrium assumptions.
• Constrained uncertainties using Markov Chain Monte Carlo sampling and compared results to prior dayside emission constraints.

Experimental results

Research questions

• RQ1 Does the transmission spectrum of WASP-12b provide unambiguous evidence for water vapor absorption?
• RQ2 Is the previously inferred carbon-rich atmosphere (C/O > 1) in WASP-12b consistent with high-precision transmission spectroscopy?
• RQ3 How do constraints on C/O derived from transmission spectroscopy compare to those from dayside emission spectroscopy?
• RQ4 To what extent do assumptions about chemical equilibrium and 1D atmospheric structure affect the retrieved atmospheric composition?
• RQ5 Can multi-technique observations (transmission, emission, phase curves) resolve discrepancies in exoplanet atmospheric composition?

Key findings

• The transmission spectrum reveals strong water vapor absorption with a 7σ confidence level, marking the first high-confidence spectroscopic detection of a molecule in WASP-12b's atmosphere.
• The retrieved 1σ water volume mixing ratio ranges from 10⁻⁵ to 10⁻², consistent with a C/O > 1 composition within 2σ uncertainty.
• Using a new retrieval model that jointly fits for C/O and metallicity under chemical equilibrium, the study constrains C/O to 0.5+0.2−0.3 at 1σ confidence.
• The carbon-rich atmosphere (C/O > 1) is ruled out at greater than 3σ confidence, contradicting prior dayside emission-based inferences.
• The discrepancy between transmission and emission spectroscopy results suggests potential issues with atmospheric modeling, thermal structure, or disequilibrium chemistry.
• The study underscores the necessity of combining high-precision data from multiple observing techniques—transmission, emission, and phase curves—to robustly determine exoplanet atmospheric composition.