GEMS JWST: A sub-Solar metallicity atmosphere for giant planet TOI-5293Ab orbiting a rapidly changing M-dwarf

Abstract

The growing sample of Giant Exoplanets around M-dwarf Stars (GEMS) helps probe the extremes of giant planet formation. Comparing the properties of this sample with their FGK counterparts can help us understand how planet formation and migration depend on stellar mass. We initiated a large Cycle 2 JWST transmission spectroscopy survey of seven GEMS. Here we present the atmospheric characterization using two JWST transits of TOI-5293Ab, a 0.5 $M_J$ planet orbiting an early M-dwarf with a period of $\sim$ 3 days. The two NIRSpec/PRISM transits indicate the planet is eclipsing a rapidly changing (heterogeneous) stellar photosphere. We see that Visit 1 had heterogeneity crossings across the entire transit chord, rendering inferences from it to be unreliable. The Visit 1 spectrum exhibits a downward slope ${<1}$ $μ$m suggestive of stellar contamination from faculae. In contrast, for Visit 2 we are able to model the heterogeneity crossings and obtain a transmission spectrum free from stellar contamination. We therefore limit our conclusions to a detailed analysis of Visit 2, and using Bayesian free chemistry retrievals, we find a low atmospheric metallicity ($\log [\mathrm{M/H}] = -1.03^{+0.53}_{-0.44}$ $\times$ Solar) and slightly super-solar C/O ratio ($1.23^{+2.94}_{-0.75}$). The retrievals yield Bayes factors that indicate strong evidence for \ce{CH4} as well as low significance detections of \ce{CO2}, \ce{H2O}, \ce{NH3}. Finally, using thermal evolution models we find that the radius of TOI-5293Ab is inflated above theoretical expectations ($\sim$ 1.07 $R_J$), despite it having an temperature of $\sim$ 700 K, and hence we were unable to constrain its bulk composition.

Figures (16)

• Figure 1: The phase-folded white light curves from the ExoTiC-JEDI reduction, after detrending with a quadratic polynomial (see \ref{['tab:5293wlcpriors']} for details) and binning to a cadence of 10 s for clarity. The labels $T_{1}-T_4$ mark the four separate points of contact between the planetary and stellar disks (e.g., start of transit, end of ingress, start of egress, end of transit). Takeaway: Visit 1 displays a different transit shape and depth (\ref{['apptab:NoSpotTransitfit']}), suggesting that complex surface heterogeneities during Visit 1 impart signals on the entirety of the transit, including subtle differences at the limbs.
• Figure 2: JWST NIRSpec/PRISM white light curves produced using ExoTiC-JEDI. Data are binned to a cadence of 5 s. Top row: The data and best-fitting model (solid line) along with the residuals to the fit. In-transit data are blue while out of transit data are orange. These models do not account for surface heterogeneities. Bottom row: The RMS for each Visit for in-transit (blue) and out-of-transit (orange) data. The prediction for Gaussian white noise is shown as a red solid line. The black point at (0, 0.001) represents the median error of the data. The analogous models including surface heterogeneities are displayed in \ref{['fig:transitcrossing']}. Takeaway: Note the stark differences in residuals to a simple transit fit between the visits.
• Figure 3: Similar to \ref{['fig:NoSpotTransitFit']} but using spotrod to model surface heterogeneities. Top row: The data and best-fitting model (solid line) along with the residuals to the fit. Middle row: The stellar surface and the adopted heterogeneity configuration. The transparency of the heterogeneities do not reflect the flux ratio (we do not differentiate between spots or faculae). The solid line marks the center of the planet and the dashed lines mark the transit chord ($\pm~R_p$). Bottom row: The RMS for each visit for in-transit (blue) and out-of-transit (orange) data along with the median error of the data (black point).
• Figure 4: Top: The pixel-level ExoTiC-JEDI transmission spectra with Visit 1 as red triangles and Visit 2 as blue squares. Bottom: The differences between both visits, scaled by the errors of the first visit (median $\sigma_B\sim630$ ppm). The 1, 2, and 3$\sigma$ regions are shaded for reference. Takeaway: As mentioned earlier, we see substantial differences in the transit properties and transmission spectra between visits, especially at wwavelengths $\lesssim3.8$ µm.
• Figure 5: (a). The pixel-level ExoTiC-JEDI transmission spectrum (blue circles) compared to the Eureka! pixel-level transmission spectrum (orange squares) for Visit 1. The difference is shown in the bottom row and scaled by the errors of the ExoTiC-JEDI data. (b). The same but for Visit 2.