{"id":"acbd75f0-eb7b-4882-808b-4432c60f6c7a","arxiv_id":"2501.05114","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":21,"one_line_summary":"Four nights of CRIRES+ spectra of AB Pictoris b give a solar C/O ratio (0.59±0.01), a 12C/13C ratio of 102±8, and a low rotation of 3.7 km/s, with tentative nightly variability.","lead":"Astronomers observed the young giant planet AB Pictoris b on four consecutive nights with the Very Large Telescope's high-resolution spectrograph CRIRES+. They measured its atmospheric composition, found a carbon-to-oxygen ratio close to the Sun's, and detected small night-to-night changes in some atmospheric parameters.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 'Uranus-like' obliquity (86.4±18.6°) is driven by an unpropagated 1.6σ rotation period from Zhou et al. (2019), not by the CRIRES+ v sin_i measurement; resolution systematics are secondary because v_eq≈110 km/s for P=2.1 h.","rationale":"The reader's CONDITIONAL verdict is appropriate, but I would relocate the weakest assumption. The CRIRES+ data genuinely support a low v sin_i: the line profiles are narrow, the retrieval is stable across nights, and the paper independently checks the 13CO detection. Even a 20% error in the trace-width spectral resolution changes v sin_i by well under 1 km/s in quadrature, which cannot overturn the 'slow rotator or pole-on' dichotomy. What converts that dichotomy into a quantitative obliquity is the equatorial velocity, and the only adopted value comes from a 1.6σ rotation period published in Zhou et al. (2019). Section 3.4.2 fixes P=2.1 h, uses it to get v_eq≈110 km/s, and then reports ip=1.8±1.3° and a projected obliquity of 86.4±18.6°. Since ip≈arcsin(v sin_i P/2πR), the quantitative result is nearly proportional to P. The authors do consider 8.4 h and 16.8 h alternatives, but these are still fast-rotation scenarios and do not cover the plausible range of an unconfirmed period. My concrete test would settle whether the period uncertainty actually changes the conclusion: if the projected obliquity's 68% interval stays near 86° after marginalizing over the period posterior, the concern does not land and the quoted obliquity can stand as a conditional result; if the interval reaches aligned configurations, the 'Uranus-like' claim should be removed from the abstract. The reader's spectral-resolution point should also be quantified, but it is not the decisive assumption because, under the assumed 2.1 h period, v_eq is so large that any plausible v sin_i error leaves the pole-on interpretation unchanged.","tokens_in":27187,"tokens_out":14965,"duration_ms":159598,"concrete_test":"Re-run the exoSpin calculation of §3.4.2 with P sampled from the full Zhou et al. (2019) period posterior, including a prior component for 'no period detection' (e.g., uniform in log P over 2–100 h), and propagate into ip and projected obliquity. If the resulting 68% credible interval on the projected obliquity includes aligned configurations (Ψ ≲ 30°) or moves the mode by more than 30° from 86°, the '~90° misalignment' conclusion should be downgraded to a period-dependent possibility in the abstract and conclusions. As a minimal check, plot projected obliquity versus P over 2–100 h and overlay the 1.6σ allowed range.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 3.4.2 computes the spin-axis inclination using v_eq = 2πR/P with P fixed to the 2.1 h period of Zhou et al. (2019), a 1.6σ candidate, and no period uncertainty is propagated into the quoted ip = 1.8±1.3° or the projected obliquity 86.4±18.6°. For R ≈ 1.85 RJ and P = 2.1 h, v_eq ≈ 110 km/s, so the retrieved v sin_i ≈ 3.7 km/s gives ip ≈ arcsin(0.034) ≈ 1.8° almost regardless of the trace-width resolution systematic: even a 50% error in v sin_i leaves ip < 4°. The §3.2.4 resolution caveat is therefore real but not the load-bearing link. The load-bearing input is P, since ip ≈ arcsin(v sin_i P / 2πR) is nearly linear in P. The authors test P = 8.4 and 16.8 h, but not the full period posterior and not the possibility that the 1.6σ signal is spurious; the abstract's 'pole-on with ~90° misalignment' is presented without this condition. The weaker disjunctive claim—AB Pic b is either a slow rotator or viewed pole-on—is robust, but the quantitative 'Uranus-like' obliquity rests on an unverified external period.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents four nights of VLT/CRIRES+ K-band spectroscopy of the directly-imaged companion AB Pic b, analyzed with the HyDRA retrieval framework. The authors report robust per-night detections of H2O, 12CO, and 13CO (the latter at 4.8–7.2σ per night and 12.0σ combined), with broadly consistent abundances across nights but 1–3σ differences that they interpret as atmospheric variability. A combined retrieval yields C/O = 0.59 ± 0.01 (solar-like) and 12C/13C = 102 ± 8. From a low projected rotational velocity v sin i ≈ 3.7 km/s and a newly updated orbital fit, they argue that AB Pic b is either intrinsically slow-rotating or viewed pole-on, and they quote a projected obliquity of 86.4° ± 18.6° that would imply a Uranus-like spin–orbit misalignment.","tokens_in":27680,"tokens_out":4749,"duration_ms":44950,"significance":"If the detections and C/O and isotope constraints hold, this is a valuable data point for the SupJup survey: it extends high-resolution atmospheric retrievals and isotope-ratio measurements to a young, directly-imaged super-Jupiter, and it presents one of the first systematic night-to-night variability studies of such an object with CRIRES+. The paper is transparent about its parameter choices and priors, uses publicly available data and established reduction/retrieval tools, and the per-night 13CO detections are supported by both retrievals and cross-correlation functions. The variability and obliquity interpretations are more fragile, but the former is cautiously worded in the text and the latter is a testable claim that can be strengthened or softened with the period-uncertainty propagation suggested below.","major_comments":[{"comment":"The quoted spin-axis inclination ip = 1.8° ± 1.3° and projected obliquity 86.4° ± 18.6° are computed with the rotational period fixed to P = 2.1 h from Zhou et al. (2019), a candidate detected at only 1.6σ. Because ip ≈ arcsin(v sin i P / 2πR) is nearly linear in P for small angles, the period uncertainty dominates the error budget. The paper tests P = 8.4 and 16.8 h but does not propagate the period posterior or address the possibility that the 2.1 h signal is spurious. The abstract and conclusions therefore overstate the case for a '~90° misalignment'; the robust claim is the disjunctive one (intrinsically slow rotator or viewed pole-on). Please propagate the full period uncertainty or explicitly present the obliquity as conditional on the Zhou et al. period.","section":"§3.4.2, Eq. (10) from Bryan et al. (2020)"},{"comment":"The evidence for night-to-night atmospheric variability rests on (i) a CCF of median-subtracted data against the best-fit model from the combined retrieval of the same four nights, and (ii) 1–3σ differences in retrieved parameters among four nights. The |CCF| ≲ 4 values are a scaled proxy, not a detection significance, and the template is derived from the same data being tested, so the variability significance is not fully quantified. Please provide a formal significance estimate (for example, a posterior predictive test or a cross-validation where the template is built from the other three nights), or explicitly state that the variability is suggestive rather than detected.","section":"§3.1, Fig. 6, Table 3"},{"comment":"The low v sin i ≈ 3.7 km/s is derived from line broadening after convolving models with per-night resolutions inferred from the trace width under an isotropic PSF assumption. At this value, the broadening is comparable to the resolution uncertainty, and the pole-on/slow-rotator interpretation depends on v sin i being genuinely small. Although the paper acknowledges this in §3.2.4, a sensitivity test that varies the assumed R by, say, ±10–20% would show whether the qualitative conclusion is robust; without it, the systematic error on v sin i is not quantified.","section":"§2.1/Table 1, §3.2.4"}],"minor_comments":[{"comment":"The word 'pofints' in the sentence containing Eq. (7) should be 'points'.","section":"§2.2.3, Eq. (7)"},{"comment":"The text reads 'uses the the HyDRA retrieval framework'; the duplicate 'the' should be removed.","section":"§2.2"},{"comment":"Near the end of the section, 'the orbital period could be as high as ~65 hours' should read 'rotation period', since the estimate is derived from the spin–orbit alignment assumption rather than from orbital dynamics.","section":"§3.4.2"},{"comment":"The caption contains 'We divide the the values by the standard deviation'; the duplicate 'the' should be removed.","section":"Fig. 6 caption"},{"comment":"The notation 'PEN(3) gps' in Eq. (1) is unexplained and appears to be a typesetting artifact; please clarify or fix.","section":"§2.2.1, Eq. (1)"}],"recommendation":"major_revision","confidential_remarks":"The paper's data are public and the analysis chain (excalibuhr, HyDRA, orbitize) is established, which is a strength. The main risk is the obliquity claim resting on a 1.6σ external period; the authors should be encouraged either to propagate the period uncertainty or to soften the abstract and conclusions. The variability claim is also borderline, but the authors are appropriately cautious in the text, so I see it as fixable with a formal significance test rather than a fatal flaw. The paper is well within the scope of MNRAS."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThe four-night CRIRES+ dataset is new, and the per-night retrievals detect H2O, 12CO and 13CO each night, with a combined C/O of 0.59±0.01 and 12C/13C of 102±8. These composition results look solid. So does the v sin i measurement of about 3.7 km/s, which sharply revises the earlier SINFONI value of 73 km/s and is consistent across the nights. The paper uses public data, a documented reduction pipeline, and the HyDRA retrieval framework, so the analysis is reproducible.\n\nThe spin-axis headline is the soft spot. Section 3.4.2 computes the spin-axis inclination using the 2.1 h rotation period from Zhou et al. (2019), which is a 1.6σ candidate, and the period uncertainty is not propagated into the quoted ip or obliquity. This matters more than the spectral-resolution worry. At 2.1 h the equatorial velocity is near 110 km/s, so a v sin i of 3.7 km/s makes the geometry pole-on almost regardless of resolution systematics; the real lever arm is the period. If the period is longer, the inclination moves to ~8–14° for 16.8 h, and if the period is spurious the 'Uranus-like' obliquity goes away. The authors do test periods of 8.4 and 16.8 h, but they do not sample the period posterior, and the abstract's 'pole-on with a ~90° misalignment' is stronger than the evidence. The weaker disjunctive claim—either intrinsically slow rotation or a pole-on view—does hold.\n\nThe night-to-night variability and the tentative correlation between 12C/13C and cloud-deck pressure rest on only four epochs with differences at the 1–3σ level. The authors acknowledge this and call for more epochs, which is appropriate. I wouldn't block the paper on that.\n\nThis paper is for people working on directly-imaged companions and high-resolution spectroscopy of substellar objects. It also serves as a useful case study in how an external, low-significance period can drive an interpretation. I would send it to peer review, but I would ask for a revision that propagates the period uncertainty, frames the spin-axis result as a disjunction, and softens the abstract. The composition and kinematics work deserve publication; the obliquity claim needs more discipline.","headline":"The composition and v sin i results on AB Pic b are solid, but the spin-axis headline depends on an unpropagated 1.6σ rotation period.","tokens_in":28364,"tokens_out":5537,"would_cite":true,"duration_ms":49611,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Four nights of spectra of the young super-Jupiter AB Pic b show a solar C/O ratio, carbon isotope shifts tied to clouds, and a spin axis that is either slow or pole-on.","keywords":["AB Pictoris b","directly imaged exoplanet","high-resolution spectroscopy","atmospheric retrieval","carbon isotope ratio","C/O ratio","atmospheric variability","spin-orbit misalignment"],"falsifier":"Independently calibrate the per-night spectral resolution with a fast-rotating standard star or a telluric-line fit through the same 0.4-inch slit and re-derive the rotation speed: if the true resolution is more than about 10–15 percent above the assumed value, $v \\sin i$ would rise to roughly 5–8 km/s and overturn the pole-on interpretation, while a confirmed value near 3.7 km/s would support it.","tokens_in":26919,"feed_emoji":"🪐","tokens_out":18771,"duration_ms":152868,"temperature":0.7,"pith_summary":"Four consecutive nights of high-resolution CRIRES+ spectra of the directly imaged super-Jupiter AB Pic b are used to ask whether the atmosphere of this young, wide-orbit companion changes on day-to-day timescales. The paper aims to establish that the bulk composition is broadly stable — with a carbon-to-oxygen ratio of $0.59 \\pm 0.01$, consistent with the Sun — while the measured carbon isotope ratio ($^{12}$C/$^{13}$C) shifts by up to about $3\\sigma$ between nights in a way that tracks the depth of the cloud deck. It then argues, from atmospheric retrievals (Bayesian fits of a physical model to the spectrum), that the planet's low projected rotation velocity ($v \\sin i \\approx 3.7$ km/s) means AB Pic b is either intrinsically a slow rotator because it is young, or is seen pole-on with a projected spin-orbit obliquity near 90 degrees. The reason to care is that this is one of the first attempts to use high-resolution spectroscopy to separate genuine atmospheric variability of a directly imaged companion from retrieval systematics, and to tie isotope-ratio measurements to cloud structure.","feed_headline":"Super-Jupiter AB Pic b may be tilted 90 degrees in its orbit","feed_subtitle":"Four nights of spectra reveal solar C/O, cloud-shifting carbon isotopes, and a near pole-on spin axis.","key_machinery":"The analysis rests on high-resolution ($R \\approx 93{,}000$–$113{,}000$) CRIRES+ spectra in the K-band, processed with a dedicated reduction pipeline and high-pass filtered so that all constraints come from the shapes and depths of individual molecular lines rather than the continuum. The retrieval engine is HyDRA, a Bayesian atmospheric retrieval code that fits the pressure–temperature profile with a smoothness penalty, a Gaussian-process covariance model for correlated residuals, vertically constant chemical abundances, and a power-law cloud deck with wavelength-dependent opacity. The carbon isotope ratio is read off directly from the line strengths of $^{13}$CO relative to $^{12}$CO, while the projected rotation velocity is measured by convolving the model with a rotational broadening kernel after removing the assumed instrumental resolution; converting that $\\sim 3.7$ km/s broadening into a pole-on spin-axis and obliquity then uses a published probability-density method that takes the rotation period from earlier photometry.","core_discovery":"Using the HyDRA retrieval framework on high-pass-filtered spectra, the paper detects H$_2$O, $^{12}$CO, and $^{13}$CO in each of the four nights, with $^{13}$CO detected at $12\\sigma$ in the combined dataset. The individual nights give broadly consistent abundances that vary at the ${\\sim}2\\sigma$ level, with the variation tracking the steepness of the deep temperature profile, and $^{12}$C/$^{13}$C ratios that vary between $70^{+16}_{-9}$ and $120^{+16}_{-15}$, anti-correlated with the retrieved cloud-deck pressure. Combining all nights gives C/O $= 0.59 \\pm 0.01$, matching the solar value, and $^{12}$C/$^{13}$C $= 102 \\pm 8$, slightly higher than the ISM and Solar System values. From the low $v \\sin i \\approx 3.7$ km/s, the paper argues that AB Pic b is either an intrinsically slow rotator due to its young age or is seen pole-on, with a revised orbit (inclination $98^{+12}_{-5}$ degrees) yielding a projected spin-orbit obliquity of $86.4 \\pm 18.6$ degrees; the paper flags that this rotation measurement is highly dependent on the assumed spectral resolution.","pith_inferences":["If the cloud-deck correlation with $^{12}$C/$^{13}$C is confirmed in other objects, multi-epoch high-resolution spectroscopy could turn isotope-ratio retrievals into a diagnostic of vertical mixing and cloud patchiness in giant-planet atmospheres, since $^{13}$CO forms deeper than $^{12}$CO.","The pole-on scenario predicts little rotational variability, so a survey comparing low-$v \\sin i$ companions with fast rotators of similar age could test whether low projected rotation speeds systematically suppress spectroscopic variability.","A Doppler-imaging campaign spanning several rotation periods could break the slow-rotator-versus-pole-on degeneracy: a fast rotator seen pole-on would show a symmetric polar velocity field, while an intrinsically slow rotator would show essentially no line-profile distortion.","Explicitly modeling telluric contamination night by night could turn the reported 2–3-sigma parameter shifts into a quantitative test of whether the variability is atmospheric or instrumental, since the nights with the lowest signal-to-noise show the largest deviations from the median."],"forward_implications":["If the C/O $= 0.59 \\pm 0.01$ result holds, AB Pic b formed from material with a solar carbon-to-oxygen ratio, placing its formation near the CO snowline rather than in a strongly carbon-enriched disk region.","If the night-to-night isotope shifts are real, single-epoch $^{12}$C/$^{13}$C measurements of variable companions mix formation chemistry with cloud structure, so repeat observations are needed before interpreting isotope ratios as formation tracers.","If the projected obliquity of $86.4 \\pm 18.6$ degrees is correct, AB Pic b joins a small list of directly imaged companions with Uranus-like spin-orbit misalignment, constraining dynamical histories with wide-orbit scattering or disk migration.","If AB Pic b is intrinsically slow-rotating because it is young, then it should spin up as it contracts, making its rotation period evolve measurably over time.","The revised orbit (semi-major axis about 307 AU, inclination about 98 degrees) strengthens the case that AB Pic b is on a very wide, highly inclined orbit, with the companion currently moving toward its host star in projection."],"supporting_citations":[{"why":"Provides the previous characterization of AB Pic b — P-T profile, radius, rotation velocity, and orbit — that this paper revises with higher-resolution data and a new astrometric epoch.","marker":"Palma-Bifani et al. (2023)"},{"why":"Establishes the retrieval setup for non-irradiated companions in the SupJup survey: likelihood, Gaussian-process noise model, P-T smoothness penalty, and high-pass filtering.","marker":"de Regt et al. (2024)"},{"why":"Supplies the HyDRA retrieval framework and the isotope-ratio method used to convert $^{13}$CO and $^{12}$CO abundances into $^{12}$C/$^{13}$C.","marker":"Gandhi et al. (2023b)"},{"why":"Gives the probability-density relation that converts $v \\sin i$ and the rotation period into the spin-axis inclination and projected obliquity.","marker":"Bryan et al. (2020)"},{"why":"Provides the orbitize! OFTI rejection-sampling method used for the revised orbital fit.","marker":"Blunt et al. (2017)"},{"why":"Supplies the 2.1-hour rotation period of AB Pic b used in the pole-on and obliquity calculation.","marker":"Zhou et al. (2019)"},{"why":"Justifies deriving the nightly spectral resolution from the trace width on the detector, the assumption underpinning the $v \\sin i$ measurement.","marker":"Holmberg & Madhusudhan (2022)"},{"why":"Defines the solar C/O ratio benchmark against which AB Pic b's 0.59 value is compared.","marker":"Asplund et al. (2021)"},{"why":"Provides the ISM $^{12}$C/$^{13}$C reference value used to assess AB Pic b's elevated isotope ratio.","marker":"Wilson (1999)"},{"why":"Provides the Solar System $^{12}$C/$^{13}$C comparison value used alongside the ISM reference.","marker":"Milam et al. (2005)"}],"fun_headline_variants":["AB Pic b may have a 86° spin–orbit misalignment","Four nights of spectra reveal shifting carbon isotopes on AB Pic b","Super-Jupiter AB Pic b: solar C/O but variable 13C/12C","AB Pic b's spin axis may be nearly perpendicular to its orbit"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that each night's spectral resolution, estimated from the width of the trace on the detector assuming the blur is symmetric in all directions, is correct; because the measured rotation speed is only about $3.7$ km/s, comparable to the resolution uncertainty, any error in that resolution directly changes the rotation speed and, with it, the pole-on and near-90-degree obliquity interpretation.","fun_headline_variants_meta":{"raw":{"variants":["AB Pic b may have a 86° spin–orbit misalignment","Four nights of spectra reveal shifting carbon isotopes on AB Pic b","Super-Jupiter AB Pic b: solar C/O but variable 13C/12C","AB Pic b's spin axis may be nearly perpendicular to its orbit"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000696,"raw_usage":{"total_tokens":3262,"prompt_tokens":1173,"completion_tokens":2089,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":789,"completion_tokens_details":{"reasoning_tokens":2009}},"tokens_in":789,"tokens_out":2089,"duration_ms":16786,"temperature":1.0,"reasoning_tokens":2009,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T21:18:53.403596+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Independently calibrate the per-night spectral resolution with a fast-rotating standard star or a telluric-line fit through the same 0.4-inch slit and re-derive the rotation speed: if the true resolution is more than about 10–15 percent above the assumed value, $v \\sin i$ would rise to roughly 5–8 km/s and overturn the pole-on interpretation, while a confirmed value near 3.7 km/s would support it.","supporting_citations":[{"cited_title":"The ESO SupJup Survey I: Chemical and isotopic characterisation of the late L-dwarf DENIS J0255-4700 with CRIRES$^+$","cited_arxiv_id":"2405.10841","evidence_quote":"Establishes the retrieval setup for non-irradiated companions in the SupJup survey: likelihood, Gaussian-process noise model, P-T smoothness penalty, and high-pass filtering."}],"review_version":1}