Carbon dioxide in an exoplanet atmosphere is inferred from how the planet changes light at particular wavelengths. A useful reading of a discovery follows three steps: identify the observation, see how the molecular feature is distinguished from other effects, and separate the detection from conclusions about abundance, climate or life. The spectrum and its uncertainty are the starting point.
First identify which light was measured
During a transit, a planet crosses the face of its star. Some starlight passes through the atmosphere around the planet's edge. Wavelengths absorbed more strongly make the planet appear to block a slightly larger fraction of the star. A transmission spectrum plots that wavelength-dependent blocking.
During a secondary eclipse, the planet goes behind its star. Comparing the combined light just outside eclipse with the star alone during eclipse isolates a contribution from the planet. At infrared wavelengths this can reveal its dayside thermal emission. Temperature structure matters as well as composition.
NASA's 2008 HD 189733b report described this second approach using Hubble's NICMOS instrument. The historical claim concerned a hot giant planet's emitted light. Keep that observing geometry attached to the result when comparing it with a later transit spectrum.
Read the WASP-39 b carbon-dioxide feature
A particularly clear example is the Webb NIRSpec spectrum of WASP-39 b, observed July 10, 2022 and released August 25. Open the full graph and read the horizontal wavelength axis and the vertical amount-of-light-blocked axis. Around 4.3 micrometers the points rise into a broad feature associated with CO₂ absorption.
The rise makes sense because the vertical axis measures blocked light: stronger absorption produces a higher value. In a plot of transmitted light, the same absorption could appear as a dip. Always read the axis definition before interpreting the direction of a feature.
Separate the measured points and their error bars from the smooth model curve. The curve is calculated using atmospheric assumptions; it helps explain the observations. It is not a second set of measurements. The original WASP-39 b study reports the CO₂ feature in a 3.0–5.5 micrometer spectrum and evaluates its statistical strength through model comparison.
Ask how robust the identification is
| Check | What to look for |
|---|---|
| Instrument behavior | How detector trends, pointing changes and wavelength calibration were modeled or corrected. |
| Feature shape | Agreement across several wavelength bins and, where available, more than one molecular band. |
| Alternatives | Models with other gases, clouds, temperature structures and stellar effects. |
| Repeatability | Independent reductions, additional visits or observations with another instrument. |
| Uncertainty | Which assumptions enter a quoted significance or abundance interval. |
The history of HD 189733b illustrates why these checks matter. A subsequent analysis of its dayside spectrum examined how temperature profiles, overlapping molecular effects and the available datasets constrained composition. Read such studies as tests of a particular observation and model, rather than assigning every early molecular claim the same evidentiary status.
A good fit alone leaves a further question: can a substantially different atmosphere also explain the data? Researchers call the process of inferring atmospheric properties a retrieval. Its output is a range of model parameters conditional on the data and assumptions, which should accompany any single best-fit value.
A small signal can carry specific information
To infer abundance, the model must connect molecular opacity to atmospheric scale height, the planet's reference radius, clouds and temperature. Two planets with similar CO₂ abundance can therefore show different feature amplitudes. Comparing only the height of two published curves can erase those physical differences.
What a detection lets you conclude
CO₂ is a useful tracer of atmospheric chemistry and planetary formation. A detection establishes chemical evidence in the region sampled by the observation. Surface pressure, habitability and biological activity require further evidence and context. WASP-39 b is a hot gas giant; its molecular spectrum is a demonstration of atmospheric measurement.
When saving a discovery, record the planet, instrument, transit or eclipse geometry, observed band, paper date and remaining model uncertainty. For the wider question of oceans and habitable environments, continue with water on other planets. For image and spectrum labels, use the Webb interpretation guide.