Deciphering transmission spectra by exploring the solar paradigm

TL;DR

Stellar activity introduces wavelength-dependent changes in a star's apparent radius that bias transmission spectroscopy. The authors leverage the Sun as a ground truth, combining SDO/HMI spot and faculae maps with the SATIRE model to quantify the chromatic radius change across 0.6–6 μm and compare a disc-averaged approach with a per-pixel, center-to-limb variation aware formulation. They find that neglecting CLV underestimates the radius change, with faculae dominating the signal and producing up to about 40 ppm for Jupiter-like transits (vs JWST's ~10 ppm floor) and around 0.4 ppm for Earth-like transits, indicating when stellar contamination must be modeled. The results provide scaling relations for forward models, emphasize the need for physics-based faculae and accurate CLV treatment, and propose leveraging solar data and multi-technique constraints to mitigate stellar contamination for current and future missions such as Ariel.

Abstract

Transmission spectroscopy probes exoplanet atmospheres via the wavelength dependence of transit depths, but stellar contamination from magnetic activity can significantly bias these measurements. Activity-induced changes in the chromatic apparent stellar radius represent a major challenge for atmospheric characterisation. As surface distributions of magnetic features are generally unknown for stars other than the Sun, we adopt the Sun as a benchmark to study how the chromatic effect depends on the distribution of spots and faculae. Using spot and facular masks derived from SDO/HMI magnetograms and intensitygrams, combined with the SATIRE model, we compute the chromatic dependence of the Sun's apparent radius. We test different methods of convolving surface coverage with spectra the identify physical drivers of the effect. We find that simplified approaches, which neglect the CLV, underestimate the apparent radius, particularly for faculae, whose surface coverage dominates at near-solar activity levels. Proper treatment of facular CLV is therefore essential. The activity-induced variation between solar minimum and maximum reaches around 40 ppm for a Jupiter-like transit, exceeding JWST's expected 10 ppm noise floor, while remaining at around 0.4 ppm for an Earth-like transit.

Figures (9)

• Figure 1: Dependence of the apparent solar radius as a function of solar activity for the different approaches. Top figure shows the spot area coverage as a function of the time for solar cycle 24. The colours in the bottom row plots correspond to the spot disc area coverage. Approach 1 corresponds to the disc integrated spectra and Approach 2 takes the spectra of each pixel of the HMI maps into account.
• Figure 2: Activity induced apparent radius (as indicated by the colours) as a function of spot area coverage for selected wavelengths. The grey lines indicate the position of the maximum in the apparent radius for each of the wavelengths shown. We note, that the average facular area is 1% and the mean spot area is 0.04%.
• Figure 3: A closer look at the different approaches presented in Fig. \ref{['fig:apparent_radius_change']} by taking six days presented in Tab. \ref{['tab:areas']}, with their spot distribution (first column) and the faculae distribution (second column), the apparent radius for Approach 1 (black solid lines) and Approach 2 (coloured solid lines) (column three), and scatter-plots of the apparent radius as a function of wavelength (column four). For details of why those dates were chosen we refer to the text.
• Figure 4: Flux ratio between the faculae and the quiet Sun for the different approaches. In black line Approach 1 (black line) is shown and the coloured lines indicate the flux rations for different $\mu$-positions .
• Figure 5: Similar to Fig. \ref{['fig:contrasts_faculae_mu']}, but for the spots.