🔭 astrophysics

Precise Determination of the Metallicity and C/O of WASP-39~b From a Single JWST Instrument Mode with Phase-Resolved Cross-Correlation Retrievals

This study demonstrates that applying phase-resolved cross-correlation retrievals to a single JWST/NIRSpec transit observation of WASP-39b enables the robust detection of all major carbon- and oxygen-bearing molecules and the precise derivation of its metallicity and C/O ratio, overcoming the limitations of traditional retrieval methods that fail to detect CO and yield biased results from the same data.

Original authors: Arjun B. Savel, Eliza M. -R. Kempton, Erin M. May, Matthew C. Nixon, Jegug Ih, Katherine A. Bennett, Joost P. Wardenier

Published 2026-07-22

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Arjun B. Savel, Eliza M. -R. Kempton, Erin M. May, Matthew C. Nixon, Jegug Ih, Katherine A. Bennett, Joost P. Wardenier

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the universe as a giant, cosmic kitchen where stars are the ovens and planets are the dishes being baked. For a long time, astronomers have been trying to figure out the "recipe" for these distant worlds. Just like a chef needs to know if a cake is made with too much sugar or if the flour is from a specific region, scientists want to know the chemical ingredients of a planet's atmosphere. Two of the most important ingredients are "metallicity" (how many heavy elements, like carbon and oxygen, are mixed in compared to hydrogen) and the "C/O ratio" (the balance between carbon and oxygen). Why does this matter? Because these ratios act like a fingerprint of how the planet was born. They can tell us if a planet formed close to its star or far away in the cold, icy outer regions of its solar system. To read this fingerprint, we need to catch a glimpse of the planet's atmosphere as it passes in front of its star, looking for the specific chemical "shouts" of molecules like water, carbon monoxide, and carbon dioxide.

For years, the rule of thumb was that you needed a massive, multi-tool approach to get a clear recipe. You had to combine data from different instruments covering a wide range of colors (wavelengths) to catch every molecule. If you only looked at a narrow slice of the spectrum, you might miss the most important ingredients, leading to a confused or wrong recipe. This is where the story of WASP-39 b comes in. This is a "hot Jupiter," a giant planet that orbits very close to its star, making it a perfect test kitchen for our cosmic recipes.

In this paper, the authors challenge the old rule that you need a multi-instrument buffet to get a good recipe. They show that you can actually get a very precise measurement of a planet's metallicity and carbon-to-oxygen ratio using just a single instrument mode from the James Webb Space Telescope (JWST), specifically the NIRSpec/G395H mode. The secret sauce isn't a new telescope, but a new way of listening. Instead of just looking at the average light, they used a technique called "phase-resolved cross-correlation." Think of it like this: if you are at a crowded party trying to hear one specific person's voice, you might miss them if you just listen to the general noise. But if you know exactly what that person sounds like and you track how their voice shifts as they move around the room, you can pick them out of the crowd even if the music is loud.

The authors applied this "voice tracking" method to the light from WASP-39 b. They found that by keeping the data at its highest possible resolution (native pixel resolution) and watching how the planet's motion shifts the light over time, they could detect molecules that standard methods missed. Most notably, they decisively found Carbon Monoxide (CO), which is the dominant carrier of carbon in this planet's atmosphere. Standard methods, which usually average out the data, failed to see this CO at all. Because they finally found the CO, they could calculate the planet's true chemical recipe: a metallicity of 1.2 ± 0.2 (meaning it's about 15 times richer in heavy elements than our Sun) and a C/O ratio of 0.68 (+0.10, -0.14).

The paper argues that the old way of doing things—averaging the data and using standard statistical tools—was like trying to read a book by squinting at a blurry photocopy. It missed the fine print (the CO) and led to a wrong conclusion about the carbon-to-oxygen balance. By using this sharper, more dynamic approach, the authors show that we can get a complete and accurate picture of a planet's atmosphere using just one instrument mode. This means we might be able to unlock the secrets of many more exoplanets using data we already have, without needing to wait for new, expensive observations. It's a reminder that sometimes, the key to solving a mystery isn't getting a bigger flashlight, but learning how to look at the shadows differently.

Technical Summary: Precise Determination of the Metallicity and C/O of WASP-39 b From a Single JWST Instrument Mode with Phase-Resolved Cross-Correlation Retrievals

Problem Statement
Measuring atmospheric metallicities ([M/H]) and carbon-to-oxygen ratios (C/O) is a primary objective of JWST exoplanet science, as these metrics serve as signposts for planet formation relative to protoplanetary disk snowlines. Historically, achieving robust constraints on these parameters has required broad wavelength coverage (typically ∼1–5 µm), necessitating the combination of multiple instrument modes (e.g., NIRISS/SOSS and NIRSpec/G395H) to compile a complete molecular inventory. Furthermore, a specific analytical challenge has persisted regarding the detectability of carbon monoxide (CO) in hot Jupiter atmospheres using JWST data. While CO is theoretically the dominant carbon carrier, standard low-resolution, time-integrated analyses of NIRSpec/G395H data have frequently failed to detect it, leading to biased and inaccurate inferences of C/O ratios and metallicities. Traditional methods often rely on binned data and time-integrated transit spectra, which may smear out the high-resolution spectral features of molecules like CO that are distinguishable only at native pixel resolution.

Methodology
The authors present a novel analysis technique that circumvents the need for multi-instrument data by applying phase-resolved cross-correlation retrievals to a single-mode transit observation of the hot Jupiter WASP-39 b (NIRSpec/G395H). The methodology departs from standard JWST retrieval practices in four critical ways:

Data Product: Instead of fitting a time-integrated transmission spectrum (binned to $R \approx 100$ ), the analysis utilizes the full spectroscopic lightcurve at native pixel resolution ( $R \approx 2700$ ). The data, re-reduced using the Eureka! pipeline, retains 160 exposures without temporal binning.
Forward Model: The CHIMERA code generates high-resolution ( $R \approx 10^6$ ) model spectra including H $_2$ O, CO, CO $_2$ , SO $_2$ , and collision-induced absorption. Crucially, the model is Doppler-shifted at each orbital phase to match the planet's changing radial velocity ( $K_p$ ) and systemic velocity ( $V_{sys}$ ). The model also accounts for limb darkening separately for the morning and evening limbs and includes rotational broadening.
Likelihood Function: The study employs a cross-correlation likelihood function (based on Brogi & Line 2019) rather than a standard Gaussian likelihood. This approach "nulls" the continuum and fits for noise characteristics directly, preserving the spectral continuum which standard removal procedures might erroneously strip of atmospheric signal.
Sampling: The authors utilize the nautilus sampling package instead of the commonly used MultiNest, citing literature indicating MultiNest can produce underestimated posterior widths and biased Bayesian evidence.

Key Contributions

Single-Mode Robustness: The paper demonstrates that a single JWST instrument mode (NIRSpec/G395H) is sufficient to derive robust [M/H] and C/O values, provided phase-resolved, high-resolution techniques are used.
CO Detection: The study resolves the "missing CO" puzzle in WASP-39 b's G395H data. While standard retrievals on the same data yield a non-detection ( $\Delta \ln(Z) = 0.2$ ), the cross-correlation approach decisively detects CO ( $\Delta \ln(Z) = 25$ ), confirming it as the dominant carbon carrier.
Phase-Resolved Advantage: By maintaining phase resolution, the method leverages the Doppler shift of the planet's orbital velocity to distinguish molecular signals from noise and stellar contamination, a technique previously reserved for ground-based high-resolution spectroscopy.

Results
Applying this technique to archival G395H data yields the following:

Molecular Abundances: The authors obtain bounded abundances for all major carbon- and oxygen-bearing species: H $_2$ O ( $-2.5 \pm 0.3$ ), CO ( $-2.1^{+0.2}_{-0.3}$ ), CO $_2$ ( $-3.4 \pm 0.3$ ), and SO $_2$ ( $-5.6 \pm 0.2$ ). This represents the first cross-correlation detection of SO $_2$ in the literature.
Metallicity and C/O: The derived metallicity is $[(C+O)/H] = 1.2 \pm 0.2$ (or $15^{+9}_{-6} \times$ solar), and the C/O ratio is $0.68^{+0.10}_{-0.14}$ .
Comparison to Traditional Methods: Traditional retrievals on the same G395H data (binned to $R=100$ ) produced a biased C/O of $0.04^{+0.08}_{-0.01}$ and a higher metallicity ( $30^{+18}_{-18} \times$ solar). The authors attribute the low C/O in traditional analyses to the non-detection of CO, which artificially suppresses the inferred carbon reservoir.
Velocity Constraints: The systemic velocity ( $V_{sys}$ ) measurement is bimodal, with the "bluer" mode consistent with the known system velocity and correlated with a plausible minor CO isotopologue abundance. The constraint on $V_{sys}$ is driven by the comb-like structure of the CO opacity.

Significance and Claims
The paper claims that phase-resolved cross-correlation retrievals are a powerful tool for extracting maximum atmospheric information from existing and future JWST datasets. The authors assert that:

Information Dilution: Traditional approaches that bin data in wavelength and time dilute information regarding exoplanet atmospheric composition, particularly for molecules with narrow spectral features like CO.
Bias Correction: The non-detection of CO in standard G395H analyses heavily biases inferred C/O ratios and metallicities. By recovering the CO signal, the proposed method corrects these biases.
Proof of Concept: This work serves as a proof of concept that robust chemical characterizations (specifically [M/H] and C/O) can be achieved using a single JWST instrument mode, challenging the prevailing assumption that multi-mode observations are strictly required for these measurements.
Legacy Data Utility: The technique offers a pathway to maximize the scientific return of legacy datasets that may have been previously considered insufficient for full chemical inventory analysis.

The results are generally consistent with previous multi-instrument analyses of WASP-39 b, validating the accuracy of the single-mode, high-resolution approach. The authors conclude that applying these techniques to additional G395H observations of transiting exoplanets is likely to be fruitful.

More like this