Exoplanet Atmosphere Report · JWST NIRSpec/G395H

LHS 475b

An Earth-sized rocky planet interior to its M-dwarf's habitable zone, and one of JWST's first terrestrial-exoplanet targets. This report statistically tests the real, published transmission spectrum against real candidate atmosphere models — showing exactly which are ruled out and which remain possible.

Transit-discovered, 2023 (TESS + JWST validation) Real NASA Exoplanet Archive parameters Real JWST NIRSpec spectrum (Lustig-Yaeger & Fu et al. 2023)

AI-generated artist's concept of LHS 475b — not a real photograph. All data and figures in this report come from actual JWST NIRSpec observations (see below).

The planet, in numbers

Queried live from the NASA Exoplanet Archive TAP service (pscomppars).

Radius	0.991 Earth radii — essentially Earth-sized
Mass	0.941 Earth masses
Orbital period	2.029 days
Semi-major axis	0.0204 AU
Equilibrium temperature	586 K — warm, interior to the habitable zone
Host star	LHS 475, M3 dwarf, T_eff = 3300 K, 0.279 R_sun, 0.262 M_sun
Distance	12.48 parsecs (~40.7 light-years)
Discovery	2023, TESS transit photometry; independently validated with JWST

Why a "featureless" spectrum is still a real, useful result

A rocky planet's transmission spectrum shows wavelength-dependent absorption dips only if starlight is passing through a real, extended, molecule-bearing atmosphere during transit. A flat, feature-free spectrum is not a non-result — it actively rules out entire classes of atmosphere, the same way a flat line on a metal detector rules out large buried objects. JWST's precision (constraining features below 50 parts per million here) makes this a genuinely stringent test, something no previous facility could do for a planet this small.

Lustig-Yaeger & Fu et al. (2023) used two real JWST NIRSpec/G395H transit observations to independently validate LHS 475b's existence and obtain its first transmission spectrum, finding no significant molecular features and using that flatness to rule out specific real candidate atmospheres.

Close-up 3D-rendered concept of LHS 475 b's rocky surface

AI-generated 3D-render-style concept — one of several atmosphere scenarios still statistically consistent with the flat spectrum above, not a confirmed depiction.

Real statistical test, quantified

The evidence audit now uses every atmosphere curve in the public release. Panel A places representative curves over the 56-bin FIREFLy spectrum, panel B compares the wavelength-dependent shapes of all 13 scenarios after fitting one nuisance vertical offset, and panel C tests sensitivity to the Eureka!, FIREFLy, and Tiberius reductions of the same two transits.

Three-panel LHS 475 b JWST evidence audit with atmosphere models and reduction-pipeline sensitivity

A: spectrum and selected curves. B: all 13 shape-only comparisons. C: mean-subtracted reductions rebinned to a common grid. Generated by scripts/analyze_spectrum.py.

Test	χ² / dof	reduced χ²	p-value
Flat line	50.70 / 55	0.92	0.640
1× solar H2-rich, shape-only	9590.10 / 55	174.37	effectively zero
Pure CH4, shape-only	108.90 / 55	1.98	2.1×10⁻⁵
Cloudy Venus-like, shape-only	56.05 / 55	1.02	0.435
Mars-like, shape-only	55.94 / 55	1.02	0.439

The flat fit is not rejected (χ² = 50.70 for 55 degrees of freedom; p = 0.640). This rules out broad, low-mean-molecular-weight spectral structure but does not demonstrate that LHS 475 b is airless. Compact, cloudy, and high-mean-molecular-weight scenarios remain degenerate at this precision. A large p-value means only that a curve has not been rejected by this test; it is not evidence that the atmosphere exists.

Two estimands are published in the machine-readable table. The supplied-curve statistic preserves the archive's upstream depth offset and fits nothing locally (dof = 56). The shape-only statistic fits one vertical offset (dof = 55), making the uncertain reference radius explicit. These goodness-of-fit checks are not atmospheric retrievals, molecule detections, or Bayes factors.

Data and method notes

System context comes from the NASA Exoplanet Archive. The spectra and all 13 atmosphere models come from Zenodo record 7925111 (Lustig-Yaeger, Fu et al. 2023). The three pipelines analyze the same two visits: their differences measure reduction sensitivity, not independent repeatability. FIREFLy HR→LR is chiefly a self-consistency check. Elevated rebinned scatter in Eureka! or Tiberius can reflect covariance or formal-error calibration and must not be relabelled as an atmospheric feature.

Every copied file is covered by a normalized SHA-256 manifest. Reproduce the chain with python scripts/validate_sources.py followed by python scripts/analyze_spectrum.py. Download the complete outputs as model comparison CSV, pipeline comparison CSV, or analysis JSON.

AI-generated illustration of the James Webb Space Telescope, whose NIRSpec instrument took the real spectrum used in this report. Not an official mission photograph — see NASA/JWST for real imagery.

References

Lustig-Yaeger, J. & Fu, G. et al., 2023. A JWST transmission spectrum of the nearby Earth-sized exoplanet LHS 475 b. Nature Astronomy, 7, pp.1317-1328.
Ment, K. et al., 2023. LHS 475 b: A Venus-sized Planet Orbiting a Nearby M Dwarf. The Astronomical Journal (submitted), arXiv:2304.01920.
NASA Exoplanet Archive, exoplanetarchive.ipac.caltech.edu — system parameters, queried live via TAP.
Zenodo record 7925111, zenodo.org/records/7925111 — real reduced JWST spectrum and atmosphere models.