Atmospheric Reconnaissance of TRAPPIST-1 b with JWST/NIRISS: Evidence for Strong Stellar Contamination in the Transmission Spectra

Olivia Lim, Björn Benneke, René Doyon, Ryan J. MacDonald, Caroline Piaulet, Étienne Artigau, Louis-Philippe Coulombe, Michael Radica, Alexandrine L’Heureux, Loïc Albert, Benjamin V. Rackham, Julien de Wit, Salma Salhi, Pierre-Alexis Roy, Laura Flagg, Marylou Fournier-Tondreau, Jake Taylor, Neil J. Cook, David Lafreniére, Nicolas B. Cowan, Lisa Kaltenegger, Jason F. Rowe, Néstor Espinoza, Lisa Dang, Antoine Darveau-BernierThe Astrophysical Journal Letters 2023 · DOI

TL;DR

This paper presents the first transmission spectra of the exoplanet TRAPPIST-1 b using JWST/NIRISS. The observations reveal significant contamination from the host star's surface features (starspots and faculae), which masks potential atmospheric signals. While ruling out thick, hydrogen-rich atmospheres, the data's precision is currently insufficient to detect thinner or secondary atmospheres due to uncertainties in modeling stellar activity.

What Problem Does This Paper Address?

The TRAPPIST-1 system, with its seven Earth-sized planets orbiting a small M-dwarf star, is a prime target for studying exoplanet atmospheres. Transmission spectroscopy, where starlight passing through a planet's atmosphere is analyzed, is a key technique for this. However, M-dwarf stars are known for their high levels of stellar activity, including starspots and flares, which can mimic or mask atmospheric signals. This paper addresses the challenge of disentangling these stellar effects from potential planetary atmospheric signatures in JWST transmission spectra, particularly for TRAPPIST-1 b, the innermost and most irradiated planet.

Methods

The researchers utilized the James Webb Space Telescope's Near-Infrared Imager and Slitless Spectrograph (NIRISS) in Single-Object Slitless Spectroscopy (SOSS) mode to obtain transmission spectra of TRAPPIST-1 b over two separate transits. They employed three different data reduction pipelines (supreme-SPOON, NAMELESS, and SOSSISSE) to ensure robustness. To account for stellar contamination, they used two primary fitting approaches:

  1. Sequential Fit: First, a stellar contamination model (including starspots and faculae) was fitted to the transmission spectrum of each visit independently. The stellar signal was then divided out, and the resulting planet-only spectra were combined and analyzed for atmospheric signatures.
  2. Simultaneous Fit: This approach jointly fitted a model for both stellar contamination and the planetary atmosphere to the transmission spectra of each visit.

They also tested various systematics treatments, including Gaussian processes and detrending against the star's Hα emission line, to mitigate correlated noise from stellar activity.

Datasets

The primary dataset used in this study consists of transmission spectra of TRAPPIST-1 b obtained from two transits observed by the JWST/NIRISS instrument. The data covers the wavelength range of 0.6–2.8 µm. The researchers also referenced existing HST/WFC3 and Spitzer/IRAC transit depths for comparison.

Key Insights and Ideas

  • Stellar Activity Dominates: The most striking finding is that stellar contamination from unocculted starspots and faculae significantly impacts the transmission spectra of TRAPPIST-1 b, often dominating the signal.
  • Epoch-Dependent Variability: The stellar contamination signatures differed between the two observed transits, suggesting that the star's surface features evolve over time. One visit showed evidence of starspots, while the other indicated faculae.
  • Model Fidelity is Crucial: The current theoretical models for M-dwarf stellar spectra struggle to accurately reproduce the observed out-of-transit spectra. This lack of model fidelity introduces significant uncertainties (on the order of ~1800 ppm per pixel) that are much larger than the observational precision (~89 ppm for the combined spectrum).
  • Atmospheric Rejection: Despite the stellar contamination, the data confidently reject cloud-free, hydrogen-rich atmospheres for TRAPPIST-1 b, consistent with previous studies.
  • Secondary Atmospheres Remain Elusive: The precision of the current data is insufficient to confirm or reject the presence of secondary atmospheres (e.g., those dominated by heavier molecules like CO2 or H2O) or thin atmospheres.

Key Results

  • The transmission spectra from the two visits show moderate to strong evidence of stellar contamination, with amplitudes of ~750-1000 ppm.
  • The sequential fit identified unocculted starspots for the first visit and unocculted faculae for the second visit.
  • Both sequential and simultaneous fitting approaches confirmed the rejection of clear hydrogen-rich atmospheres at 1× and 100× solar metallicity.
  • The precision of the combined spectrum is 89 ppm, but the uncertainty due to stellar model fidelity is about an order of magnitude larger.
  • No significant atmospheric features were detected in the stellar-contamination-corrected spectra.

Limitations and Future Work

The authors acknowledge several limitations:

  • Stellar Model Fidelity: The primary limitation is the inability of current theoretical stellar models to accurately represent the heterogeneous surface of M-dwarf stars like TRAPPIST-1. This leads to large uncertainties in correcting for stellar contamination.
  • Observational Precision: While JWST offers unprecedented precision, the current dataset (two transits) is not sufficient to detect thinner atmospheres or secondary atmospheres, especially given the stellar contamination.

The authors suggest the following future directions:

  • More Observations: Additional observations are needed to empirically characterize stellar heterogeneities over time and across different transit geometries.
  • Improved Theoretical Models: Further theoretical work is required to develop more accurate models for the atmospheres of cool, active stars.
  • Alternative Techniques: While transmission spectroscopy is efficient for outer planets, exploring other techniques or refining current ones will be crucial for characterizing atmospheres around TRAPPIST-1 planets.
  • Characterizing Stellar Activity: Developing better methods to model and correct for stellar activity is essential for future exoplanet atmosphere studies, particularly around M-dwarfs.

Summary updated

Abstract

Abstract TRAPPIST-1 is a nearby system of seven Earth-sized, temperate, rocky exoplanets transiting a Jupiter-sized M8.5V star, ideally suited for in-depth atmospheric studies. Each TRAPPIST-1 planet has been observed in transmission both from space and from the ground, confidently rejecting cloud-free, hydrogen-rich atmospheres. Secondary eclipse observations of TRAPPIST-1 b with JWST/MIRI are consistent with little to no atmosphere given the lack of heat redistribution. Here we present the first transmission spectra of TRAPPIST-1 b obtained with JWST/NIRISS over two visits. The two transmission spectra show moderate to strong evidence of contamination from unocculted stellar heterogeneities, which dominates the signal in both visits. The transmission spectrum of the first visit is consistent with unocculted starspots and the second visit exhibits signatures of unocculted faculae. Fitting the stellar contamination and planetary atmosphere either sequentially or simultaneously, we confirm the absence of cloud-free, hydrogen-rich atmospheres, but cannot assess the presence of secondary atmospheres. We find that the uncertainties associated with the lack of stellar model fidelity are one order of magnitude above the observation precision of 89 ppm (combining the two visits). Without affecting the conclusion regarding the atmosphere of TRAPPIST-1 b, this highlights an important caveat for future explorations, which calls for additional observations to characterize stellar heterogeneities empirically and/or theoretical works to improve model fidelity for such cool stars. This need is all the more justified as stellar contamination can affect the search for atmospheres around the outer, cooler TRAPPIST-1 planets for which transmission spectroscopy is currently the most efficient technique.
BibTeX
@article{lim2023atmospheric,
  author = {Olivia Lim and Björn Benneke and René Doyon and Ryan J. MacDonald and Caroline Piaulet and Étienne Artigau and Louis-Philippe Coulombe and Michael Radica and Alexandrine L’Heureux and Loïc Albert and Benjamin V. Rackham and Julien de Wit and Salma Salhi and Pierre-Alexis Roy and Laura Flagg and Marylou Fournier-Tondreau and Jake Taylor and Neil J. Cook and David Lafreniére and Nicolas B. Cowan and Lisa Kaltenegger and Jason F. Rowe and Néstor Espinoza and Lisa Dang and Antoine Darveau-Bernier},
  title = {{Atmospheric Reconnaissance of TRAPPIST-1 b with JWST/NIRISS: Evidence for Strong Stellar Contamination in the Transmission Spectra}},
  journal = {The Astrophysical Journal Letters},
  year = {2023},
  doi = {10.3847/2041-8213/acf7c4},
}
Figure