Scientific Premise

Recent work (Savvidou et al. 2026, A&A; Rustamkulov et al. 2024, AJ) has shown that simplified stellar contamination models — especially the disk-averaged Transit Light Source Effect (TLSE) correction — can introduce systematic, wavelength-dependent biases of up to ~400 ppm in transmission spectra, particularly for active M- and K-dwarf hosts at optical wavelengths. JWST observations of systems like TOI-5205b have already revealed cases where stellar contamination dominates the transmission spectrum below ~3 μm, overpowering the planetary signal by up to an order of magnitude. Yet the community has accumulated years of published HST/WFC3 and Spitzer transmission spectra for dozens of planets around active stars — many of which used now-outdated contamination corrections or none at all. A systematic, retroactive audit of these published spectra using modern self-consistent contamination models has not been performed at scale.

Target Datasets

• All published HST/WFC3 G141 grism transmission spectra (publicly archived via MAST/Barbara A. Mikulski Archive)
• Spitzer/IRAC secondary eclipse and transit photometry from the Spitzer Heritage Archive
• Stellar activity indicators: Ca II H&K indices, photometric variability from TESS and Kepler/K2 light curves, X-ray fluxes from XMM-Newton/Chandra
• JWST Early Release and Cycle 1–3 transmission spectra for cross-validation on overlapping targets

Novelty

While individual retrieval papers sometimes mention stellar contamination as a caveat, no one has performed a population-level meta-analysis asking: "For how many published atmospheric detections does the reported molecular signature become statistically insignificant once modern TLSE models are applied?" This is distinct from simply improving retrieval codes — it is a forensic audit of the existing literature. The 2026 Savvidou et al. result that disk-averaged TLSE corrections can be wrong by hundreds of ppm makes this timely and tractable.

Concrete Workflow

1. Compile a catalog of all published HST/WFC3 and Spitzer transmission spectra (roughly 70–100 planets with published spectra as of 2025).
2. Cross-match with TESS and Kepler/K2 photometry to characterize host-star variability (spot covering fractions, facular contrasts, rotation periods).
3. For each system, apply the Rackham et al. (2018) disk-averaged TLSE correction and the Savvidou et al. (2026) self-consistent geometric TLSE model.
4. Re-run atmospheric retrievals (e.g., with petitRADTRANS or POSEIDON) on the corrected spectra.
5. Flag systems where the original claimed molecular detection (H₂O, Na, K, etc.) drops below 2σ significance after correction.
6. Cross-validate against any available JWST spectra for the same targets.

Expected Signal / Observable

A ranked list of "contamination-vulnerable" detections, quantifying how many published H₂O or alkali-metal detections in the HST era may have been partially or wholly driven by unocculted stellar heterogeneity rather than planetary atmospheres. Preliminary estimates suggest 15–30% of published WFC3 water detections around M dwarfs may be significantly affected.

Possible False Positives (in the meta-analysis itself)

• Overestimating stellar activity: if the spot covering fraction is poorly constrained (e.g., from sparse TESS coverage), the correction itself may be inaccurate in the other direction.
• Retrieval degeneracies: removing stellar contamination may shift molecular abundances rather than eliminating detections entirely, making it ambiguous whether the original detection was "wrong."
• Temporal variability: stellar activity at the epoch of the HST observation may differ from the epoch of the TESS/Kepler photometry used to estimate spots.

Why This Could Matter

This project would produce a community resource — essentially a reliability index for pre-JWST atmospheric characterizations — and directly inform target prioritization for JWST Cycles 4+ and future ARIEL observations. If a substantial fraction of legacy detections are contaminated, it changes the empirical landscape of comparative exoplanetology and the inferred prevalence of water-rich atmospheres on sub-Neptunes.