{"id":"ec329305-f47c-414e-a0a6-8949a8f85a11","arxiv_id":"2411.17923","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":10,"one_line_summary":"Water is detected in WASP-43b's transmission spectrum at log10(H2O) = -2.24 (+0.57/-0.48) with an upper limit C/O < 0.95.","lead":"Four high-resolution infrared transits of WASP-43b show water vapor in the planet's atmosphere and set a loose upper limit on its carbon-to-oxygen ratio. The measurement, made with the Gemini-South telescope, agrees with earlier Hubble and JWST results and shows ground-based high-resolution spectroscopy can measure water on this benchmark hot Jupiter.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Reported C/O<0.95 is internally unsupported: Section 4 says CO is essentially unconstrained, yet uses carbon-bearing upper limits to quote a 2-sigma C/O bound; the bound must be shown robust to the CO prior or removed.","rationale":"The reader's weakest assumption was model complexity, specifically isothermal and constant-abundance retrieval assumptions that could bias the H2O abundance. I agree that is a concern for the absolute water abundance, but the more decisive and directly checkable problem is the C/O upper limit, which appears not to follow from the stated retrieval results. This is not a disagreement with previous consensus; it is an internal-consistency check of the abstract's second headline number. The H2O detection itself has supporting evidence: four nights show correlated peaks, the SVD choice was tested against injected signals, and previous HST and JWST observations independently detected water. I would therefore not reject the paper outright. However, the C/O < 0.95 constraint should not be presented without demonstrating that it is insensitive to the unconstrained CO abundance and its prior. A conditional acceptance is appropriate, conditioned on either removing or strongly re-flagging the C/O upper limit, or on providing the posterior and prior robustness check described above.","tokens_in":11124,"tokens_out":6337,"duration_ms":60309,"concrete_test":"Obtain the retrieval posterior chains and explicit priors behind Figure 5. Compute C/O sample-by-sample with a standard formula such as C/O = (n_CO + n_CO2 + n_CH4) / (n_H2O + n_CO + 2*n_CO2), and record the 97.5th percentile. Then re-run the same retrieval with the CO prior widened by at least two orders of magnitude, e.g., lowering the log10(CO) prior bound from -6 to -10, and check whether the C/O upper limit remains below 0.95. If it moves above roughly 1 or is prior-dominated, the C/O claim should be removed or re-flagged; if it is stable, the concern is resolved.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"In Section 4, after listing retrieval outputs, the paper states: 'The retrieved water abundance was log10(H2O) = -2.24+0.57/-0.48, while the CO abundance was essentially unconstrained.' It then reports 2-sigma upper limits for CO2 and CH4 and says 'From the retrieved water abundance and upper limits on the carbon-bearing species, we calculate a 2-sigma upper limit of C/O < 0.95.' This is internally inconsistent. CO is both the primary carbon carrier and an oxygen carrier in a hot-Jupiter atmosphere; an 'essentially unconstrained' CO abundance leaves the carbon budget unbounded, so no data-driven 2-sigma upper limit on C/O can be derived from H2O plus CO2/CH4 upper limits alone. If the C/O limit was computed from posterior samples, then the marginal CO posterior must in fact be strongly constrained to low values, or the prior must be doing the work; either reading contradicts the text. The same section also says the retrieved CO posterior peak agrees with previous CO detections at log10(CO) ~ -3, so ignoring CO as a carbon reservoir is not justified. This is load-bearing because the abstract presents C/O < 0.95 as a main result and as evidence that the atmosphere is not carbon-rich.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript reports high-resolution cross-correlation spectroscopy (HRCCS) of four transits of WASP-43b observed with IGRINS. The authors detect H2O at an SNR of 3.51 in the combined data, do not detect CO, CO2, or CH4, and use a retrieval to derive log10(H2O) = -2.24 (+0.57/-0.48) with upper limits log10(CO2) < -2.78 and log10(CH4) < -5.89. From these they quote a 2-sigma upper limit C/O < 0.95 and argue consistency with previous HST, JWST, and CRIRES+ results.","tokens_in":11490,"tokens_out":4640,"duration_ms":40626,"significance":"If the water-abundance measurement is taken at face value, the paper provides a useful independent transmission-spectrum constraint on WASP-43b that is consistent with several prior studies and demonstrates the application of the Sc-CHIMERA/Line et al. retrieval framework to IGRINS data. The comparison to published abundances, including the discussion of transmission versus emission geometry, is informative. However, the C/O upper limit quoted in the abstract and conclusions is not supported by the retrieval results as presented, and the modest detection SNR with several analysis choices optimized on the data makes the detection claim require additional robustness checks.","major_comments":[{"comment":"The retrieval text states that the CO abundance was 'essentially unconstrained' and Figure 5 shows a flat CO posterior, yet the same paragraph quotes a 2-sigma upper limit C/O < 0.95 computed from the retrieved water abundance and upper limits on carbon-bearing species. In a hot-Jupiter atmosphere CO is the primary carbon carrier and also an oxygen carrier; an unconstrained CO abundance leaves the carbon budget unbounded. The C/O limit must either be derived from the joint posterior samples, which would require a constrained CO posterior and would contradict 'essentially unconstrained', or it must be removed or reframed as a conditional limit. As written this is internally inconsistent and load-bearing because C/O < 0.95 appears in the abstract, Section 5, and Section 6.","section":"Section 4, Figure 5"},{"comment":"The combined H2O detection has SNR 3.51, just above the adopted 3.0 threshold, and the peak is offset by Delta-Kp = -21 km/s and Delta-Vsys = +1.696 km/s from the literature values. The authors justify the Kp offset with Wardenier et al. (2023) models but do not quantify the statistical significance of the offset or the trial factor from the Kp-Vsys grid. Because the number of SVD components was chosen to maximize the detection SNR (Section 2) and night 4 was excluded after it showed no detection (Section 3), the false-alarm probability of the 3.51 SNR should be established with injection-recovery or permutation tests that include these selection steps. The current text reports an injection test for the SVD choice but not for the full detection pipeline.","section":"Section 3, Figure 3"},{"comment":"The retrieval assumes an isothermal atmosphere with constant-with-altitude abundances and a single cloud-top pressure, justified by the weak signal. The authors do not test whether the retrieved log10(H2O) or the C/O upper limit are biased when these assumptions are relaxed, for example with a two-temperature profile or a water abundance gradient. Since Section 5 interprets differences from emission-spectrum results in terms of limb/dayside geometry, a short sensitivity test with a more flexible parameterization would substantially strengthen the abundance claim.","section":"Section 4"}],"minor_comments":[{"comment":"The introduction says the conclusions and future work are given in Section 7, but the paper has no Section 7; the conclusions appear in Section 6.","section":"Section 1"},{"comment":"The table caption states that Weaver et al. (2020) used transmission spectra, but the table classifies that entry as 'Emission'; the text should be corrected to match the table.","section":"Table 2"},{"comment":"The caption refers to components removed by 'PCA', while Section 2 describes the method as singular value decomposition; the terminology should be made consistent.","section":"Figure 1 caption"},{"comment":"The reference list contains two identical entries for Line et al. (2021); the duplicate should be removed.","section":"References"},{"comment":"There is a typo: 'tenative' should be 'tentative'.","section":"Figure 3 caption"}],"recommendation":"major_revision","confidential_remarks":"The load-bearing problem is the C/O upper limit: if the authors cannot produce a posterior-derived C/O distribution, the abstract and conclusions should be revised to remove or strongly qualify the C/O claim. The detection robustness issues (SVD optimization, night 4 exclusion, Kp offset) are addressable with additional tests, so the paper remains salvageable if those are provided."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a solid incremental measurement of H2O in WASP-43b's transmission spectrum from IGRINS, with a water abundance that agrees with HST and JWST. The paper's headline C/O upper limit is not supported by the retrieval as reported; the authors say CO is essentially unconstrained and then derive C/O < 0.95 from water plus CO2/CH4 upper limits. That doesn't hold unless the CO posterior is actually doing work, which contradicts their own text. The C/O claim needs to be removed or re-derived with the CO constraint made explicit.\n\nWhat's new: first IGRINS transmission spectrum measurement for this planet, a new independent water abundance for a well-studied hot Jupiter. The paper does several things well: clear description of data reduction, standard SVD/telluric removal, explicit discussion of the Kp and Vsys offsets, and a nice comparison table of prior water abundances. The retrieval setup is standard for the field, and the consistency with Kreidberg et al. and Yang et al. gives some confidence that the water measurement is not wildly off.\n\nSoft spots beyond the C/O issue: the detection SNR is 3.51, just above the 3.0 threshold, and the SVD component count was chosen to maximize SNR, which can inflate significance. Night 4 was excluded after showing no detection; the humidity argument is plausible but post-hoc. The retrieval assumes an isothermal, constant-abundance atmosphere and includes no systematic error budget; with a 3.5-sigma detection, model degeneracies could shift the abundance. These are worth flagging but do not sink the water measurement by themselves.\n\nBottom line: the water measurement is a defensible, incremental contribution. The C/O upper limit is the load-bearing weakness and, as written, internally inconsistent. If the authors fix or drop that claim, this is a fine paper for the exoplanet atmosphere community. I'd send it to review, with a request to address the C/O derivation and the statistical choices.","headline":"A defensible but marginal H2O measurement for WASP-43b whose C/O upper limit is internally unsupported and should be removed or re-derived.","tokens_in":12037,"tokens_out":2032,"would_cite":true,"duration_ms":17993,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Four transits of the hot Jupiter WASP-43b, observed with the high-resolution IGRINS spectrograph, reveal water vapor in the planet's limb at a mixing ratio near 0.6% and set a 2-sigma upper limit of C/O < 0.95 on its carbon-to-oxygen ratio.","keywords":["exoplanet atmospheres","hot Jupiter","WASP-43b","water abundance","high-resolution spectroscopy","cross-correlation spectroscopy","C/O ratio","atmospheric retrieval"],"falsifier":"Re-analyze the IGRINS transits with a 3D retrieval that allows altitude-dependent chemistry and inhomogeneous clouds; if the water abundance shifts by more than the quoted ±0.5 dex, the constant-abundance model is falsified. A phase-resolved map of the Vsys offset across the transit would also test whether the +1.7 km/s offset is a real dynamical feature.","tokens_in":10944,"feed_emoji":"💧","tokens_out":8616,"duration_ms":65787,"temperature":0.7,"pith_summary":"Four transits of the hot Jupiter WASP-43b, observed at high spectral resolution with the IGRINS spectrograph on Gemini-S, yield a water-vapor absorption signal in the planet's limb. From the signal the authors retrieve a water abundance of log10(H2O) = −2.24 (+0.57/−0.48), a mixing ratio of roughly 6 parts per thousand, and a 2-sigma upper limit on the carbon-to-oxygen ratio of C/O < 0.95. The result matters because water and the C/O ratio are tracers of where and how a giant planet formed, and a C/O below 0.95 rules out an extremely carbon-rich atmosphere. The measurements are consistent with earlier HST and JWST observations of the same planet.","feed_headline":"WASP-43b shows water at 0.6% mixing ratio","feed_subtitle":"Four high-resolution transits yield log10(H2O) = −2.24 and a C/O upper limit below 0.95.","key_machinery":"The analysis rests on high-resolution cross-correlation spectroscopy: individual molecular absorption lines in the 1.45–2.45 µm spectra are separated from stellar and telluric features by subtracting a set of orthogonal components, and the residual spectra are cross-correlated against synthetic templates containing only one molecule at a time. A retrieval then fits the same data with a forward model that assumes an isothermal atmosphere with constant-with-height abundances of H2O, CO, CO2, and CH4 plus a single cloud-top pressure, using a nested-sampling likelihood to deliver abundance posteriors and upper limits. The water detection and the bounds on the carbon-bearing species together produce the C/O upper limit.","core_discovery":"The paper claims to detect H2O in the transmission spectrum of WASP-43b by cross-correlating four nights of IGRINS spectra with a synthetic template, reaching a detection signal-to-noise ratio of 3.51. It reports that CH4 and CO2 are not detected, CO shows only a weak signal, and a Bayesian retrieval on the combined data returns a water volume mixing ratio of log10(H2O) = −2.24 (+0.57/−0.48), an unconstrained CO abundance, upper limits of log10(CO2) < −2.78 and log10(CH4) < −5.89, and a 2-sigma upper limit of C/O < 0.95. The paper further notes that the H2O cross-correlation peak is offset by ΔKp ≈ −21 km/s and ΔVsys ≈ +1.7 km/s relative to literature values, an offset pattern partially aligned with predictions from 3D atmospheric models.","pith_inferences":["The unexplained positive Vsys offset could point to day–night winds or patchy limb clouds; phase-resolved observations across transit would test whether the offset is a real dynamical signature.","The retrieved water abundance, obtained with an isothermal constant-abundance model, is best read as a limb-averaged value; comparing it with JWST emission retrievals of the same planet could reveal vertical abundance gradients.","The same stacking-and-retrieval pipeline could be applied to other hot Jupiters with archival IGRINS transits to build a uniform sample of water abundances and C/O upper limits for planet-formation studies."],"forward_implications":["Water vapor is present in the limb of WASP-43b at a mixing ratio near $10^{-2.24}$, roughly 0.6%.","The atmosphere is not extremely carbon-rich: the carbon-to-oxygen ratio is below 0.95 at 2-sigma confidence.","The retrieved water abundance agrees within uncertainties with previous HST and JWST results, supporting the view that transmission and emission spectra probe different atmospheric layers.","The non-detections of CH4 and CO2, with only a weak CO signal, are consistent with a near-solar composition rather than a carbon-dominated one.","Additional data that confirms the tentative CO signal would bring the transmission and dayside emission measurements of WASP-43b into closer agreement."],"supporting_citations":[{"why":"Provides the cross-correlation likelihood framework used to convert CCF strengths into detection SNRs and to compute retrieval likelihoods.","marker":"Brogi & Line 2019"},{"why":"Supplies the retrieval framework, SVD data-separation approach, and model convolution method that this paper follows.","marker":"Line et al. 2021"},{"why":"Supplies the template-model generation and the method for choosing the optimal number of singular vectors to subtract.","marker":"Weiner Mansfield et al. 2024"},{"why":"Predicts Kp and Vsys offsets for water in 3D hot Jupiter models, used to interpret the observed offset pattern.","marker":"Wardenier et al. 2023"},{"why":"Previous HST/WFC3 detection of water in WASP-43b whose retrieved abundance is compared to this paper's result.","marker":"Kreidberg et al. 2014"},{"why":"Previous CRIRES+ high-resolution dayside detection of water and CO on WASP-43b, providing the C/O ratio and CO abundance for comparison.","marker":"Lesjak et al. 2023"},{"why":"JWST MIRI full-orbit emission spectrum of WASP-43b detecting water, used as a consistency check.","marker":"Bell et al. 2024"},{"why":"JWST 2D retrieval of global water and CO abundances for WASP-43b, used for comparison.","marker":"Yang et al. 2024"}],"fun_headline_variants":["WASP-43b water measured at 0.6% mixing ratio from four transits","C/O upper limit 0.95 from WASP-43b high-res water detection","Four transits pin WASP-43b water abundance to log10(H2O) = -2.24","Water detected on hot Jupiter WASP-43b, C/O < 0.95","High-resolution spectra reveal WASP-43b water, no carbon species"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The retrieval assumes the limb is isothermal, chemically uniform with height, and covered by a single cloud deck; strong vertical gradients or patchy clouds would bias the reported water abundance and C/O limit.","fun_headline_variants_meta":{"raw":{"variants":["WASP-43b water measured at 0.6% mixing ratio from four transits","C/O upper limit 0.95 from WASP-43b high-res water detection","Four transits pin WASP-43b water abundance to log10(H2O) = -2.24","Water detected on hot Jupiter WASP-43b, C/O < 0.95","High-resolution spectra reveal WASP-43b water, no carbon species"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000398,"raw_usage":{"total_tokens":2125,"prompt_tokens":1032,"completion_tokens":1093,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":648,"completion_tokens_details":{"reasoning_tokens":975}},"tokens_in":648,"tokens_out":1093,"duration_ms":8805,"temperature":1.0,"reasoning_tokens":975,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T11:42:28.405529+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-analyze the IGRINS transits with a 3D retrieval that allows altitude-dependent chemistry and inhomogeneous clouds; if the water abundance shifts by more than the quoted ±0.5 dex, the constant-abundance model is falsified. A phase-resolved map of the Vsys offset across the transit would also test whether the +1.7 km/s offset is a real dynamical feature.","supporting_citations":[],"review_version":1}