{"id":"7d9a04f4-6ed8-49ef-b8e7-76b22df084da","arxiv_id":"1908.04505","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"A homogeneous reprocessing of about 1100 transit lightcurves confirms WASP-12's inward orbit while suggesting part of the signal comes from a distant companion, and finds the WASP-4 decay claim is model-dependent.","lead":"Astronomers reprocessed 16 years of amateur and professional telescope data to measure when 17 exoplanets pass in front of their stars. They confirm that hot Jupiter WASP-12 is drifting inward, but find about 10% of that drift may be an illusion caused by an unseen outer companion, and they cast doubt on a similar claimed drift for WASP-4.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Roemer correction in Sect. 7 rests on a qualitatively preferred RV fit (c1 = -7.5 ± 2.2 from WASP-12'); the 10% shift in T_tidal is within the quoted uncertainty, so the robust TTV detection is unaffected but the exact 10% figure is model-dependent.","rationale":"The reader's weakest assumption identifies the same load-bearing concern: the reality of the radial acceleration c1 used for the Roemer correction. I find this concern is real but does not overturn the paper's central claims. The WASP-12 nonlinear TTV trend is detected at ~9 sigma in the homogeneous sample, cross-checked against Maciejewski et al. (Sect. 5.1), and is insensitive to the noise model (Sect. 5.4). The Roemer correction is a minor adjustment: switching from the preferred WASP-12' fit (c1 = -7.5) to the basic WASP-12 fit (c1 = -5.4) shifts T_tidal by only ~0.1 Myr, which is within the quoted 0.40 Myr uncertainty. Even setting c1 to zero leaves T_tidal = 3.47 ± 0.32 Myr, consistent with the quoted value. The paper appropriately hedges the RV acceleration as 'hints'. The unmodeled apsidal precession (explicitly acknowledged in the last sentence of Sect. 7) is a second, possibly more physical, alternative for the non-Roemer part of the TTV, but it does not affect the detection of the trend itself and is transparently flagged. Overall, the central claims are robust to the identified concern; the verdict remains ACCEPT.","tokens_in":36240,"tokens_out":15749,"duration_ms":137656,"concrete_test":"Recompute T_tidal from Eq. (12) using the basic split-RV fit WASP-12 in Table 5, c1 = -5.4 ± 2.0 m/s/yr, instead of the preferred WASP-12' value c1 = -7.5 ± 2.2 m/s/yr; if the resulting T_tidal is 3.70 ± 0.38 Myr (or smaller), the claimed 10% Roemer correction in Sect. 7 is dependent on a qualitative model choice and should be downgraded to a systematic uncertainty.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 7 derives the genuinely tidal decay time T_tidal = 3.80 ± 0.40 Myr using Eq. (12), which subtracts the Roemer term from the apparent T_d = 3.47 ± 0.32 Myr. The Roemer term is obtained from Eq. (11) with the radial acceleration c1 = -7.5 ± 2.2 m/s/yr, taken from the 'alternative' joint fit WASP-12' in Table 5. The basic split-RV fit WASP-12 in Table 5 yields c1 = -5.4 ± 2.0 m/s/yr. The paper prefers WASP-12' because the basic fit 'may also include a local red noise effect appearing within individual short in-transit runs' (Sect. 7), but no statistical comparison (e.g., BIC) is given; the two values differ by only ~1 sigma. Propagating either c1 value through Eq. (12) gives T_tidal = 3.70 ± 0.38 Myr (for c1 = -5.4) or 3.80 ± 0.40 Myr (for c1 = -7.5), so the 10% Roemer correction is not statistically robust. If c1 were entirely spurious, T_tidal would equal the apparent 3.47 ± 0.32 Myr, still comfortably consistent with the quoted 3.80 ± 0.40 Myr. Thus the central claim of a genuine nonlinear TTV trend in WASP-12 does not depend on this correction, but the specific '10% larger' statement is model-dependent.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a homogeneous re-analysis of roughly 1100 transit light curves for 17 exoplanet hosts, combining archival and new EXPANSION photometry with radial-velocity data for 13 targets. The authors derive transit timings with a common pipeline, empirically calibrate limb-darkening coefficients, and fit self-consistent transit-plus-RV models. Their main results are: (i) a high-significance (about 9-sigma) quadratic TTV trend in WASP-12, consistent with previous work and cross-checked against independent Maciejewski et al. timings; (ii) hints of a radial acceleration of approximately -7.5 +/- 2.2 m/s/yr for WASP-12, implying that roughly 10% of the apparent TTV is a Roemer effect and that the truly tidal decay time is T_tidal = 3.80 +/- 0.40 Myr; and (iii) a demonstration that the WASP-4 TTV trend reported by Bouma et al. is model-dependent, disappearing in the homogeneous sample and dropping to 2.8-3 sigma when the four Huitson et al. timings are removed. The paper also reports no significant periodic TTVs in the other targets and provides revised transit parameters and RV trends.","tokens_in":36643,"tokens_out":6418,"duration_ms":64268,"significance":"The paper's main strength is its careful attention to data homogeneity: the same pipeline and noise model are applied to all light curves, and the quantitative comparison with independent Maciejewski et al. timings for WASP-12 (12% RMS difference) provides a credible external validation. The empirical limb-darkening calibration and the likelihood-ratio analysis for WASP-4 are useful methodological contributions. If the WASP-12 radial acceleration is real, the Roemer correction would be an important new ingredient in interpreting this famous tidal-decay candidate, with consequences for the inferred tidal quality factor. The WASP-4 analysis is a valuable cautionary counterpoint to the earlier detection claim, and the release of the reprocessed photometry and derived timings is a community resource. The main caveats, which the authors themselves partly identify, concern the statistical support for the preferred RV fit in Section 7 and the conditional nature of the WASP-4 significance after subset removal; these points need to be addressed in revision.","major_comments":[{"comment":"The claimed 'roughly 10 per cent' Roemer correction is based on choosing the 'alternative' fit WASP-12' with c1 = -7.5 +/- 2.2 m/s/yr over the split-RV fit WASP-12 with c1 = -5.4 +/- 2.0 m/s/yr. The paper justifies this preference only qualitatively, by arguing that local red noise in short in-transit runs may contaminate the basic fit, and it gives no statistical model comparison (BIC/AIC or a formal test for local red noise). The two c1 values differ by less than one sigma. Propagating c1 = -5.4 through Eq. (12) gives T_tidal = 3.70 +/- 0.38 Myr instead of 3.80 +/- 0.40 Myr. The raw TTV detection is robust regardless of this choice, but the specific '~10 per cent larger' statement in the abstract and Section 7 is model-dependent; it should be supported by a quantitative comparison of the two fits or softened to a range that covers both values.","section":"Section 7, Eq. (12), Table 5"},{"comment":"The 2.8-3 sigma significance quoted after removing the four Huitson et al. timings is computed from a merged compilation of heterogeneous TTV datasets. This is in tension with the paper's own central methodological warning that heterogeneous TTV data should not be plainly merged. The significance therefore measures the trend under a particular ad hoc subset and merging model, not on the homogeneous sample. The paper should state explicitly that this 2.8-3 sigma value is conditional on the adopted subset and noise model, and it should present the homogeneous-sample upper limit as the primary WASP-4 result. As written, a reader may overinterpret the 2.8-3 sigma as a robust estimate of the trend's significance.","section":"Section 5.5, Fig. 4"},{"comment":"The paper demonstrates in Section 5.1 that its own TTV uncertainties are underestimated by about 22 per cent for WASP-12 (reduced chi^2 = 1.49) and applies this correction in Section 5.4. It is not stated whether the WASP-4 homogeneous-sample significance limits in Fig. 4 include an analogous correction for WASP-4's reduced chi^2 = 1.31 from Table 2. If they do not, the statement that the homogeneous WASP-4 data are consistent with a linear ephemeris 'below 1-sigma' may be overconfident. The authors should specify whether the jitter and regularization parameters used in Fig. 4 were fit from the TTV residuals and, if so, quote the significance with and without the chi^2-based scaling.","section":"Sections 5.1 and 5.5"}],"minor_comments":[{"comment":"There are typographical errors: 'homogeneus' should be 'homogeneous' in the abstract, and 'Peridograms' should be 'Periodograms' in Section 5.2.","section":"Abstract and Section 5.5"},{"comment":"The statement that the 12 per cent RMS difference between the two WASP-12 TTV series is comparable to the 'probable statistical uncertainty (1/sqrt(N) for N=73)' is imprecise; the uncertainty of a ratio of two RMS estimates is more like sqrt(2/(N-1)) ~ 17 per cent. Please rephrase to avoid giving a misleading uncertainty estimate.","section":"Section 5.1"},{"comment":"The text says the 'Clear' band limb-darkening coefficients are always fitted, but the comparison with bolometric values in Figs 3 and 4 is said to be for reference only. It would be helpful to state explicitly how the final pipeline treats the Clear band, since this band includes many amateur light curves.","section":"Section 4"},{"comment":"For Qatar-2, the assumption that the published 'BJD' times are in the TDB system is justified only by the statement that the TTV residuals look bad otherwise; the footnote could quantify how much the alternative UTC interpretation changes the derived timings, to make the assumption more transparent.","section":"Section 2, Table 1"}],"recommendation":"major_revision","confidential_remarks":"The paper is within scope and the central WASP-12 TTV detection is robust. The main weakness is the under-supported model choice in Section 7, which directly affects the headline '10 per cent' Roemer correction; this is fixable with a formal model comparison or a softened claim. The WASP-4 section is careful, but the conditional nature of the significance after subset removal should be made more prominent. No concerns about citation practices or novelty disclosure; the data release is a valuable contribution."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things you should know about this one. First, the WASP-12 TTV trend is real and robust; it is detected at about 9-sigma and matches the independent Maciejewski timings to within 12%. Second, the paper's claim that ~10% of that trend is a Roemer effect rather than tidal should be taken with a grain of salt: it rests on a radial acceleration estimate that changes from -7.5±2.2 to -5.4±2.0 m/s/yr depending on how you split the RV data, and the two fits are less than 1.5 sigma apart. The author's own caveat says the split fit may include red noise, but they don't provide a statistical comparison. The resulting tidal decay time T_tidal = 3.80±0.40 Myr is consistent with the apparent T_d = 3.47±0.32 Myr, so the central detection survives; only the decomposition into tidal vs Roemer is uncertain.\n\nWhat is genuinely new: the homogeneous reprocessing of ~3.2e5 photometric measurements for 17 targets, with careful treatment of red noise and limb-darkening biases, and the public data release. The cross-check against Maciejewski et al. is a good validation, and the authors are honest that their own TTV uncertainties are underestimated by ~22% and correct for it with a jitter term. The WASP-4 re-analysis is the most useful part: they show convincingly that the claimed quadratic trend depends on four Huitson timings and on the noise model. Without those points the trend drops to 2.8-3 sigma, and with a separated noise model it becomes non-significant. This is a strong caution against blindly merging heterogeneous TTV data.\n\nSoft spots: the RV data for WASP-12 are the weakest link. The trend is only significant at about 3-sigma in either fit, and the two fits disagree at the ~1.5-sigma level. One can also nitpick the post-hoc outlier whitelisting and the exclusion of DFOSC timings, but these are documented and don't look malicious. The absence of periodic TTVs in the other targets is a null result, which is fine.\n\nBottom line: this is a careful piece of work, worth a serious referee, and worth citing for the homogeneous catalog and for the WASP-4 warning. The WASP-12 discovery credit goes to Patra and Maciejewski, but this paper provides the cleanest homogeneous confirmation and a reasonable, if model-dependent, attempt to separate tidal and Roemer contributions.","headline":"Solid homogeneous TTV study that confirms WASP-12's TTV trend, but the ~10% Roemer correction rests on a ~1-sigma RV choice; WASP-4 skepticism is well founded.","tokens_in":37435,"tokens_out":2544,"would_cite":true,"duration_ms":29607,"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":"WASP-12 shows a genuine quadratic transit-timing trend, but about 10% of it is a light-travel (Roemer) effect; the truly tidal decay time is 3.80 ± 0.40 Myr, while WASP-4's claimed trend is model-dependent.","keywords":["transit timing variations","WASP-12b","WASP-4b","tidal orbital decay","Roemer effect","radial velocity acceleration","homogeneous transit photometry","limb-darkening calibration"],"falsifier":"Take new high-precision radial velocities of WASP-12 over the next several years: if the -7.5 ± 2.2 m/s/yr acceleration does not persist as a stable linear trend, the Roemer correction collapses and the tidal decay time returns to the apparent ~3.47 Myr. For WASP-4, a second season of homogeneous transits and simultaneous radial-velocity monitoring can decide whether any quadratic term remains once the four external transmission-spectroscopy timings are excluded.","tokens_in":36037,"feed_emoji":"🪐","tokens_out":8636,"duration_ms":78638,"temperature":0.7,"pith_summary":"This paper reprocesses about 1,100 transit light curves of 17 exoplanets through one common pipeline, so that all derived transit times are directly comparable. It confirms that WASP-12's transits arrive progressively earlier, a nonlinear trend significant at about nine sigma and consistent with the planet's orbit decaying. Adding archival radial-velocity data reveals a -7.5 ± 2.2 m/s/yr acceleration in WASP-12, attributed to an unseen distant companion; about 10 per cent of the observed trend is therefore a light-travel (Roemer) effect, and the truly tidal decay time becomes 3.80 ± 0.40 Myr rather than the apparent 3.47 ± 0.32 Myr. For WASP-4, the quadratic trend claimed from TESS data is model-dependent: it is absent from the homogeneous sample alone, and its significance drops to 2.8–3 sigma when four external transmission-spectroscopy timings are removed.","feed_headline":"WASP-12 decays in 3.8 Myr once light-travel is removed","feed_subtitle":"A homogeneous reanalysis of 1,100 transits confirms WASP-12 tidal decay and weakens WASP-4's claimed trend.","key_machinery":"The argument is carried by a quadratic transit-timing model $TTV(t)=T_0+(t-t_0)-(t-t_0)^2/(2T_d)$, where $T_d=-P/\\dot P$ is the apparent period-decay time, together with the identity $T_d^{RA} = 300/c_1$ (with $c_1$ in m/s/yr giving $T_d$ in Myr), which converts a radial acceleration into an equivalent Roemer light-travel decay time. Subtracting $1/T_d^{RA}$ from $1/T_d$ isolates the tidal part. Around this, the paper builds a homogeneous pipeline: all ~1,100 lightcurves are reprocessed with the same red-noise model, the same quadratic limb-darkening law, and empirically debiased limb-darkening coefficients, then fitted jointly with archival radial velocities, including Rossiter–McLaughlin corrections.","core_discovery":"The central discovery is that a homogeneous, uniformly reprocessed transit-timing sample, combined with Doppler measurements, can separate genuine tidal orbital decay from light-travel artifacts. In WASP-12 the quadratic TTV trend is real and robust, with apparent decay time $T_d = 3.47\\pm0.32$ Myr. Because the radial velocities show an acceleration $c_1 = -7.5\\pm2.2$ m/s/yr, part of this apparent decay is a Roemer delay induced by a distant companion; subtracting the corresponding inverse decay time yields the truly tidal value $T_d^{tidal} = 3.80\\pm0.40$ Myr. In WASP-4, the quadratic trend reported by a recent TESS-based analysis is not reproduced by the homogeneous data alone, and its significance depends heavily on four externally published timings and on how heterogeneous TTV noise is modelled, so the trend's existence and magnitude remain uncertain.","pith_inferences":["If the WASP-12 radial acceleration is confirmed by further Doppler monitoring, the unseen companion's orbit could be constrained jointly from the TTV and RV signals, turning a nuisance into a discovery.","The WASP-4 case suggests that other published TTV trend claims based on mixed-quality timings may lose significance when reanalysed with separately weighted noise models; revisiting such claims could reveal a broader class of marginal detections.","The 10 per cent Roemer correction implies that TTV-only estimates of tidal decay times for hot Jupiters may be systematically biased by roughly that amount, which is comparable to current 1-sigma uncertainties and could matter for theory comparisons.","A practical extension would be to apply the same homogeneous reprocessing to much larger archival samples, since the pipeline's power comes from removing inter-team biases rather than from any single new observation."],"forward_implications":["WASP-12's orbit is decaying by tides, and the true decay time is about 10 per cent longer than the raw TTV trend suggests; tidal quality factor estimates based on the uncorrected value are slightly too aggressive.","Unseen companions can masquerade as apparent tidal decay in TTV data; any future TTV trend detection should be checked against Doppler accelerations before interpreting it as orbital decay.","WASP-4's claimed TTV trend is not robust: the homogeneous sample is consistent with a linear ephemeris, and removing four external timings reduces the significance to about 3 sigma, so it cannot yet be used to constrain tidal physics.","The empirical limb-darkening biases measured here, especially a roughly 0.1 shift in one of the quadratic coefficients, should be applied when reprocessing archival transit lightcurves, otherwise TTV noise is underestimated.","Merging heterogeneous TTV data with adaptive weighting, rather than plain merging, can substantially change trend significances and should become standard practice."],"supporting_citations":[{"why":"Establishes the homogeneous TTV processing approach and the red-noise model that this paper extends to 17 targets.","marker":"Baluev et al. (2015)"},{"why":"Supplies the WASP-12 transit timings used as a benchmark and part of the trend confirmation.","marker":"Maciejewski et al. (2018b)"},{"why":"Earlier WASP-12 TTV decay detection whose apparent magnitude and tidal interpretation are revised here.","marker":"Patra et al. (2017)"},{"why":"Provides the TESS-based WASP-4 trend claim that this paper tests and finds model-dependent.","marker":"Bouma et al. (2019)"},{"why":"Provides four high-precision WASP-4 timings whose inclusion or removal largely controls the trend significance.","marker":"Huitson et al. (2017)"},{"why":"Constraints on WASP-12's stellar companions and tidal interpretation used to argue the radial acceleration implies unseen companions beyond the known triple.","marker":"Bailey & Goodman (2019)"},{"why":"Gives the theoretical limb-darkening tables that are empirically debiased and then applied in all lightcurve fits.","marker":"Claret & Bloemen (2011)"},{"why":"Defines the light-travel (Roemer) delay formula that connects radial acceleration to an apparent quadratic TTV.","marker":"Irwin (1952)"}],"fun_headline_variants":["TTV reanalysis cements WASP-12 decay, undercuts WASP-4","WASP-12 tidal decay real after Roemer subtraction","Homogeneous TTVs confirm WASP-12, dim WASP-4 trend","Light-travel subtracted: WASP-12 decay 3.8 Myr, WASP-4 moot"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The 10 per cent Roemer correction and the revised 3.80 Myr tidal time rest on the assumption that the measured -7.5 ± 2.2 m/s/yr radial acceleration is a real line-of-sight acceleration from an unseen companion, not red noise or an instrument artifact.","fun_headline_variants_meta":{"raw":{"variants":["TTV reanalysis cements WASP-12 decay, undercuts WASP-4","WASP-12 tidal decay real after Roemer subtraction","Homogeneous TTVs confirm WASP-12, dim WASP-4 trend","Light-travel subtracted: WASP-12 decay 3.8 Myr, WASP-4 moot"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000563,"raw_usage":{"total_tokens":2715,"prompt_tokens":1032,"completion_tokens":1683,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":648,"completion_tokens_details":{"reasoning_tokens":1592}},"tokens_in":648,"tokens_out":1683,"duration_ms":10740,"temperature":1.0,"reasoning_tokens":1592,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:41:19.917889+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take new high-precision radial velocities of WASP-12 over the next several years: if the -7.5 ± 2.2 m/s/yr acceleration does not persist as a stable linear trend, the Roemer correction collapses and the tidal decay time returns to the apparent ~3.47 Myr. For WASP-4, a second season of homogeneous transits and simultaneous radial-velocity monitoring can decide whether any quadratic term remains once the four external transmission-spectroscopy timings are excluded.","supporting_citations":[{"cited_title":"C., Winn J","cited_arxiv_id":null,"evidence_quote":"Earlier WASP-12 TTV decay detection whose apparent magnitude and tidal interpretation are revised here."},{"cited_title":"G., et al., 2019, , 157, 217","cited_arxiv_id":null,"evidence_quote":"Provides the TESS-based WASP-4 trend claim that this paper tests and finds model-dependent."},{"cited_title":"M., D\\' e sert J.-M., Bean J","cited_arxiv_id":null,"evidence_quote":"Provides four high-precision WASP-4 timings whose inclusion or removal largely controls the trend significance."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the theoretical limb-darkening tables that are empirically debiased and then applied in all lightcurve fits."},{"cited_title":"B., 1952, , 116, 211","cited_arxiv_id":null,"evidence_quote":"Defines the light-travel (Roemer) delay formula that connects radial acceleration to an apparent quadratic TTV."}],"review_version":1}