Gaussian feature analysis reveals limited evidence for spectral features in K2-18b's transmission spectrum.
Determining the composition of an exoplanet atmosphere relies on the presence of detectable spectral features. The strongest spectral features, including DMS, look approximately Gaussian. Here, I perform a suite of Gaussian feature analyses to find any statistically significant spectral features in the recently published MIRI/LRS spectrum of K2-18b (N. Madhusudhan et al. 2025). In N. Madhusudhan et al. 2025, they claim a 3.4-σ detection of spectral features compared to a flat line. In 5 out of 6 tests, I find the data preferred a flat line over a Gaussian model, with a χ²ν of 1.06. When centering the Gaussian where the absorptions for DMS and DMDS peak, I find ln(B) = 1.21 in favour of the Gaussian model, with a χ²ν of 0.99. With only ~2-σ in favour of Gaussian features, I conclude no strong statistical evidence for spectral features.
No takes yet. Share an insight, caveat, or question.
Jake Taylor (2025) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: