BOWIE-ALIGN: Hot Jupiter Atmosphere Survey

• BOWIE-ALIGN is a JWST survey comparing hot Jupiter atmospheres to test whether migration history, indicated by orbital alignment, imprints measurable compositional signatures.
• It employs transmission spectroscopy and multiple independent retrieval methods to measure C/O ratios and metallicity while addressing retrieval degeneracies.
• The program’s controlled sample of aligned and misaligned planets around F-type stars probes differences between disc migration and post-disc high-eccentricity migration.

Searching arXiv for the BOWIE-ALIGN survey papers to ground the article in the latest relevant literature. BOWIE-ALIGN is a JWST comparative survey of hot Jupiter atmospheres designed to test whether atmospheric composition can be used to trace planet formation and migration history. Its central comparison is between planets on aligned orbits and planets on misaligned orbits around F-type stars above the Kraft break, using transmission spectroscopy to measure atmospheric carbon-to-oxygen ratio and metallicity. The programme is motivated by the premise that aligned planets are more likely to have migrated through the disc, whereas misaligned planets are more likely to have migrated after disc dispersal, and therefore may have accreted different mixtures of gas and solids. The survey was constructed specifically to determine whether those different histories produce measurable differences in atmospheric composition, while also quantifying the extent to which retrieval degeneracies complicate that inference (Kirk et al., 2024, Penzlin et al., 2024).

1. Survey concept and scientific scope

BOWIE-ALIGN was established as a controlled comparative programme rather than a sequence of isolated target studies. The survey compares 8 hot Jupiters around F stars, 4 with orbits aligned with the stellar rotation axis and 4 misaligned, with the explicit aim of testing whether migration history produces notable differences in composition between the two samples of planets (Penzlin et al., 2024). In the observational design paper, the primary objective is stated as determining whether differences in atmospheric composition can be reliably traced to differences in evolution, with particular attention to the atmospheric carbon-to-oxygen ratio, $\mathrm{C/O} = n_\mathrm{C}/n_\mathrm{O}$, and metallicity (Kirk et al., 2024).

The survey’s logic depends on using spin-orbit alignment as a proxy for migration history. In the adopted framework, aligned planets are associated with smooth disc migration, whereas misaligned planets are associated with high-eccentricity migration after disc dispersal. The host-star restriction is consequential: only F-type stars above the Kraft break are included because their radiative envelopes make tidal realignment inefficient, so the measured obliquity is more likely to preserve information about the migration pathway (Kirk et al., 2024).

This design makes BOWIE-ALIGN a population-level test of comparative exoplanetology. A single atmosphere can often be interpreted in multiple ways, but a matched sample can, in principle, discriminate between competing formation models even when individual retrievals remain partially degenerate. The survey therefore sits at the intersection of atmospheric retrieval, disc chemistry, migration dynamics, and statistical inference over small but deliberately curated samples.

2. Formation-history hypothesis and compositional predictions

The theoretical basis of BOWIE-ALIGN is laid out in a companion planet-formation modelling paper that traces the accretion of chemical components from gas and dust across a broad parameter space to estimate the range of final atmospheric compositions (Penzlin et al., 2024). The key prediction is not merely that aligned and misaligned planets differ, but that the sign and amplitude of the difference depend on the details of how solids are incorporated into the observable atmosphere.

Under the baseline interpretation, planets that migrate through the inner disc can accrete oxygen-rich silicates and other solids that reduce the atmospheric C/O ratio. The survey design paper states that the act of migrating through the inner disc should cause a measurable difference in the C/O between aligned and misaligned planets, and that the amplitude and sign of this difference depend on the amount of planetesimal accretion and whether silicates accreted from the inner disc release their oxygen (Kirk et al., 2024). The formation-model paper makes this dependence explicit: for high metallicity atmospheres, defined there as $\mathrm{O/H} > 10$ times solar, aligned and misaligned planets diverge in C/O, with aligned planets having lower C/O, specifically $<0.25$, due to accretion of oxygen-rich silicates from the inner disc (Penzlin et al., 2024).

A critical complication is the silicate rain-out alternative. If silicates do not enrich the observable atmosphere because they rain out instead of releasing their oxygen, then the predicted trend reverses: aligned planets can reach $\mathrm{C/O} > 0.6$, inverting the aligned–misaligned contrast (Penzlin et al., 2024). This is not a minor correction but a structural ambiguity in the interpretation of atmospheric abundances. A common misconception is therefore that BOWIE-ALIGN predicts a unique one-way relation between obliquity and C/O. The published model framework does not support that simplification. It predicts a composition difference, but not one with a universally fixed sign.

This suggests that the survey is best understood as a discriminator between families of formation models rather than as a simple lookup table from observed C/O to migration pathway. The value of the comparison lies precisely in testing whether the aligned and misaligned populations separate in the way expected under specific assumptions about solid accretion, volatile partitioning, and silicate processing.

3. Target selection, observing strategy, and analysis architecture

The final BOWIE-ALIGN sample was selected through explicit criteria. Planets were classified as aligned when $|\lambda| < 30^\circ$ and misaligned when $|\lambda| > 45^\circ$, with only systems having precisely measured obliquities, $\sigma(|\lambda|) < 15^\circ$, retained. The planetary cuts required hot Jupiters with mass $>0.3\,M_\mathrm{Jup}$, radius $>0.5\,R_\mathrm{Jup}$, and equilibrium temperature $1000~\mathrm{K} < T_\mathrm{eq} < 2050~\mathrm{K}$. A high Transmission Spectroscopy Metric in the $\mathrm{O/H} > 10$0-band, $\mathrm{O/H} > 10$1, was also required to ensure strong atmospheric signals (Kirk et al., 2024).

Observationally, the programme uses JWST/NIRSpec G395H transit spectroscopy over approximately $\mathrm{O/H} > 10$2–$\mathrm{O/H} > 10$3 at $\mathrm{O/H} > 10$4, with single-transit observations tailored to the target and a typical total duration of about five hours per transit (Kirk et al., 2024). The early target papers report closely related wavelength ranges such as 2.8–5.2 $\mathrm{O/H} > 10$5m, 2.87–5.16 $\mathrm{O/H} > 10$6m, 2.84–5.18 $\mathrm{O/H} > 10$7m, and 2.8–5.2 $\mathrm{O/H} > 10$8m for individual systems, reflecting target-specific reductions and binning choices rather than a change in core instrumental strategy (Kirk et al., 2024, Ahrer et al., 15 Sep 2025, Fairman et al., 18 Mar 2026).

A defining feature of the survey is methodological redundancy. Published analyses employ independent reductions and retrieval suites, including Tiberius, Eureka!, and ExoTiC-JEDI for data reduction, and petitRADTRANS, BeAR, PLATON, and POSEIDON for atmospheric retrievals. Additional modelling layers include the Met Office Unified Model for general circulation and VULCAN for photochemistry in the WASP-15b study (Kirk et al., 2024, Ahrer et al., 15 Sep 2025, Fairman et al., 18 Mar 2026). This architecture is intended to test whether the major inferences persist across alternative pipeline choices, chemistry parameterizations, cloud treatments, and thermal-profile assumptions.

The survey design paper also incorporated forward modelling and noise simulations. Synthetic JWST spectra were generated with PandExo, and Bayesian retrievals were used to estimate sensitivity to C/O and metallicity. For the selected sample of four aligned and four misaligned hot Jupiters, the survey was projected to be sensitive to the predicted differences in C/O between aligned and misaligned hot Jupiters for a wide range of model scenarios (Kirk et al., 2024).

4. Published atmospheric results

The published target studies show that BOWIE-ALIGN does not reduce to a single compositional outcome for either alignment class. Instead, it has produced a heterogeneous set of atmospheres, some strongly constrained and some dominated by muted-feature degeneracies. The currently published systems include one of the initial misaligned planets, WASP-15b, several aligned planets, and a second misaligned system, HAT-P-30b (Kirk et al., 2024, Meech et al., 31 Mar 2025, Ahrer et al., 15 Sep 2025, Claringbold et al., 19 Jan 2026, Fairman et al., 18 Mar 2026).

Target	Alignment class	Reported atmospheric inference
WASP-15b	Misaligned	Super-solar metallicity and solar-consistent C/O; evidence for SO$\mathrm{O/H} > 10$9 and possible OCS
TrES-4b	Aligned	H$<0.25$0O, CO, and CO$<0.25$1 detected; C/O $<0.25$2–$<0.25$3; metallicity $<0.25$4–$<0.25$5 solar
KELT-7b	Aligned	Weak features; high cloud deck or low-metallicity atmosphere
HAT-P-30b	Misaligned	H$<0.25$6O and CO$<0.25$7 evidence; C/O $<0.25$8–$<0.25$9; metallicity $\mathrm{C/O} > 0.6$0–$\mathrm{C/O} > 0.6$1 solar
NGTS-2b	Aligned	Muted spectrum with cloud-metallicity degeneracy and prior-sensitive H$\mathrm{C/O} > 0.6$2O posterior

WASP-15b, the first BOWIE-ALIGN result, yielded significant absorption by H$\mathrm{C/O} > 0.6$3O at $\mathrm{C/O} > 0.6$4 and CO$\mathrm{C/O} > 0.6$5 at $\mathrm{C/O} > 0.6$6. Independent reductions and retrievals inferred a super-solar atmospheric metallicity, described in the abstract as $\mathrm{C/O} > 0.6$7 solar and in the detailed summary as best estimates $\mathrm{C/O} > 0.6$8–$\mathrm{C/O} > 0.6$9 solar with an interior-structure upper bound at $|\lambda| < 30^\circ$0 solar, together with a C/O ratio consistent with solar. Those properties were interpreted as implying planetesimal accretion (Kirk et al., 2024).

TrES-4b, an aligned hot Jupiter, provided one of the clearest aligned-atmosphere constraints. Free-chemistry retrievals reported $|\lambda| < 30^\circ$1 at $|\lambda| < 30^\circ$2, $|\lambda| < 30^\circ$3 at $|\lambda| < 30^\circ$4, and $|\lambda| < 30^\circ$5 at $|\lambda| < 30^\circ$6. The observations were found to be consistent with chemical equilibrium, with $|\lambda| < 30^\circ$7–$|\lambda| < 30^\circ$8 and atmospheric metallicity in the range $|\lambda| < 30^\circ$9–$|\lambda| > 45^\circ$0 solar (Meech et al., 31 Mar 2025).

KELT-7b and NGTS-2b represent the muted-spectrum regime. KELT-7b showed only tentative evidence for H$|\lambda| > 45^\circ$1O and CO$|\lambda| > 45^\circ$2, no strong evidence for CO, and two plausible atmospheric scenarios: a high-altitude cloud deck muting all features or an extremely low metallicity atmosphere. The retrieved C/O ratios ranged from $|\lambda| > 45^\circ$3 to $|\lambda| > 45^\circ$4, while the atmospheric metallicity was suggested to be $|\lambda| > 45^\circ$5–$|\lambda| > 45^\circ$6 solar, but the posteriors were too broad for detailed conclusions about formation history (Ahrer et al., 15 Sep 2025). NGTS-2b likewise showed weak H$|\lambda| > 45^\circ$7O and CO$|\lambda| > 45^\circ$8 absorption, with free-chemistry retrievals converging on high-mean-molecular-weight solutions produced by significant H$|\lambda| > 45^\circ$9O mixing ratios; some of those solutions exceeded the $\sigma(|\lambda|) < 15^\circ$0 solar upper limit from interior structure models (Fairman et al., 18 Mar 2026).

HAT-P-30b, a misaligned system, complicates any simple aligned–misaligned dichotomy. Independent reductions and retrievals found evidence for H$\sigma(|\lambda|) < 15^\circ$1O and CO$\sigma(|\lambda|) < 15^\circ$2, with abundances consistent with equilibrium chemistry and implying a sub-solar C/O ratio of $\sigma(|\lambda|) < 15^\circ$3–$\sigma(|\lambda|) < 15^\circ$4 and a sub-solar, sub-stellar metallicity of $\sigma(|\lambda|) < 15^\circ$5–$\sigma(|\lambda|) < 15^\circ$6 solar, compared to a stellar metallicity of $\sigma(|\lambda|) < 15^\circ$7–$\sigma(|\lambda|) < 15^\circ$8 solar (Claringbold et al., 19 Jan 2026).

5. Sulphur chemistry, atmospheric structure, and limb representativeness

One of the more distinctive contributions of BOWIE-ALIGN to exoplanet atmosphere studies is its treatment of sulphur-bearing species. In WASP-15b, the published spectrum showed evidence for SO$\sigma(|\lambda|) < 15^\circ$9 absorption and an absorption feature at 4.9 $>0.3\,M_\mathrm{Jup}$0m for which the current leading candidate is OCS, although with several caveats. The paper states that, if confirmed, this would be the first detection of OCS in an exoplanet atmosphere and would point toward complex photochemistry of sulphur-bearing species in the upper atmosphere (Kirk et al., 2024).

The sulphur interpretation in WASP-15b is coupled to photochemical and metallicity arguments. The detailed summary reports that photochemical models with VULCAN show the SO$>0.3\,M_\mathrm{Jup}$1 abundance to be highly sensitive to metallicity, and that reproducing the observed feature amplitude requires high metallicity, around $>0.3\,M_\mathrm{Jup}$2 solar in best-fit models, consistent with the broader retrieval picture (Kirk et al., 2024). This places sulphur chemistry in the survey not as an isolated curiosity, but as a potential tracer of both enrichment history and upper-atmosphere photochemistry.

The NGTS-2b study illustrates the converse case: a statistically favoured model containing SO, but one that is probably not astrophysically feasible. In that analysis, the SO preference was driven by only two data points, required implausibly high SO abundance, and was associated with low atmospheric temperatures not expected for a hot Jupiter with $>0.3\,M_\mathrm{Jup}$3 K. The authors therefore treated the SO interpretation as a retrieval artifact rather than a secure compositional result (Fairman et al., 18 Mar 2026).

BOWIE-ALIGN has also addressed a separate structural question: whether terminator abundances measured in transmission are representative of the broader photosphere. For WASP-15b, general circulation model simulations with the Met Office Unified Model suggested that H$>0.3\,M_\mathrm{Jup}$4O, CO$>0.3\,M_\mathrm{Jup}$5, and CO are spatially uniform due to turbulent mixing and hot-Jupiter chemistry, and therefore that the C/O ratio measured at the limb is likely representative of the entire photosphere (Kirk et al., 2024). This is an important result because the programme’s formation inferences depend on interpreting transmission-derived elemental ratios as meaningful bulk atmospheric diagnostics rather than purely local limb properties.

6. Degeneracies, interpretive limits, and emerging significance

A recurrent theme across BOWIE-ALIGN is that muted transmission spectra generate strong degeneracies between molecular abundance, mean molecular weight, cloud opacity, and temperature structure. The KELT-7b analysis framed the problem as a choice between a high-altitude cloud deck and an extremely low metallicity atmosphere, with weak Bayes-factor support and wide abundance posteriors (Ahrer et al., 15 Sep 2025). The NGTS-2b study made the same issue more explicit by showing that the H$>0.3\,M_\mathrm{Jup}$6O abundance posterior was flat and uninformative, such that the retrievals were biased by the prior, and that some of the preferred high-metallicity solutions were ruled out only after comparison to interior structure models (Fairman et al., 18 Mar 2026).

These limitations were anticipated at the programme-design stage. The survey paper explored both clear and cloudy scenarios and estimated that, for the larger predicted population differences, the survey would be sensitive at $>0.3\,M_\mathrm{Jup}$7–$>0.3\,M_\mathrm{Jup}$8 confidence, with typical C/O uncertainties that depend on metallicity and target class (Kirk et al., 2024). The published target papers show that this sensitivity forecast was realistic but conditional: strong detections are achievable for some atmospheres, while others remain dominated by cloud-metallicity degeneracy and limited wavelength leverage.

The early sample has therefore produced a deliberately mixed evidentiary picture. WASP-15b links misalignment, super-solar metallicity, approximately solar C/O, and a planetesimal-accretion interpretation (Kirk et al., 2024). TrES-4b, although aligned, shows sub-stellar metallicity and carbon depletion, which its paper describes as challenging traditional models and potentially arising from oxygen-rich gas accretion or from a combination of low-metallicity gas and carbon-poor solid accretion (Meech et al., 31 Mar 2025). HAT-P-30b, despite being misaligned, also shows sub-solar C/O and sub-stellar metallicity, challenging models of continuous migration and accretion within a steady disc of stellar metallicity (Claringbold et al., 19 Jan 2026). KELT-7b and NGTS-2b further indicate that some aligned systems cannot yet be assigned robust formation narratives from NIRSpec/G395H transmission data alone (Ahrer et al., 15 Sep 2025, Fairman et al., 18 Mar 2026).

A plausible implication is that BOWIE-ALIGN is already demonstrating both of the propositions on which it was founded: first, atmospheric composition does preserve formation-relevant information in some systems; second, that information cannot be read off from a single abundance ratio without accounting for silicate processing, solid accretion, clouds, interior constraints, and the population context. In that sense, the programme’s significance lies less in any individual target than in the controlled comparison across aligned and misaligned hot Jupiters that the full survey was designed to deliver (Kirk et al., 2024, Penzlin et al., 2024).