Empirically Constraining the Spectra of Stellar Surface Features Using Time-resolved Spectroscopy

Abstract

Transmission spectroscopy is currently the technique best suited to study a wide range of planetary atmospheres, leveraging the filtering of a star’s light by a planet’s atmosphere rather than its own emission. However, as both a planet and its star contribute to the information encoded in a transmission spectrum, an accurate accounting of the stellar contribution is pivotal to enabling robust atmospheric studies. As current stellar models lack the required fidelity for such accounting, we investigate here the capability of time-resolved spectroscopy to yield high-fidelity, empirical constraints on the emission spectra of stellar surface heterogeneities (i.e., spots and faculae). Using TRAPPIST-1 as a test case, we simulate time-resolved JWST/NIRISS spectra and demonstrate that with a blind approach incorporating no physical priors, it is possible to constrain the photospheric spectrum to ≤0.5% and the spectra of stellar heterogeneities to within ≲10%, a precision that enables photon-limited (rather than model-limited) science. Now confident that time-resolved spectroscopy can propel the field in an era of robust high-precision transmission spectroscopy, we introduce a list of areas for future exploration to harness its full potential, including wavelength dependency of limb darkening and hybrid priors from stellar models as a means to further break the degeneracy between the position, size, and spectra of heterogeneities.

Toward Robust Corrections for Stellar Contamination in JWST Exoplanet Transmission Spectra

“…To assess this possibility, we recommend that these techniques be applied to real JWST data, starting with an inactive star and advancing to more active stars to understand the limits of model-based inferences with real data -keeping in mind that a "good fit" does not automatically imply "model fidelity." We also recommend the further exploration of empirical approaches for deriving the unique spectral components of a photosphere and their filling factors (e.g., Berardo et al 2024), enabling corrections that are independent of spectral models. focus on the results from the first of the five noise instances tested.…”

“…From this we conclude that when aiming to correct for the TLS effect using a model-based approach, an out-of-transit baseline that is at least double the transit duration is sufficient to shift the limiting factor in the final precision from the data to the models. However, empirical approaches to TLS corrections (e.g., TRAPPIST-1 JWST Community Initiative et al 2023;Berardo et al 2024) and other considerations, such as stellar variability, will warrant longer out-of-transit baselines in some cases.…”