{"id":"341d9e5f-26b3-42ff-9ad6-70b4bd197327","arxiv_id":"2411.18215","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":10,"one_line_summary":"The M31 nova AT 2023tkw shows four light-curve peaks with growing time gaps, interpreted as internal shocks, and likely hosts a 0.65-solar-mass white dwarf with a K III companion.","lead":"Astronomers watched a slow nova in the Andromeda galaxy brighten, dim, and rebrighten several times over five months. They argue the repeated flashes come from shock waves inside the ejected gas and estimate the binary system's masses from the light and spectrum.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claimed power-law relation between successive peak intervals (Eq. 1, §6.3) is not robustly established: with only four maxima, an ambiguous definition of the first point, and no objective peak identification, the reported b = 0.89 ± 0.12 may be an artifact of peak selection.","rationale":"The reader's weakest assumption focuses on the HST source identification, which affects only the derived binary parameters and is not central to the shock interpretation. The paper's strongest claim, however, is explicitly about internal shocks producing the multiple peaks and the increasing interval power law. That claim rests on two pillars: (1) the spectral-model residual at the primary peak, which the paper itself notes could be free-free emission, and (2) the power-law correlation of peak intervals, which is presented as a new quantitative result. I judge (2) to be the more load-bearing because it is the only quantitative, falsifiable prediction, and it is based on only four maxima. The technical issue with Eq. (1) at i = 1, the lack of a stated normalization constant, and the subjective peak identification together mean the reported b and its uncertainty cannot be trusted as they stand. A re-analysis with an objective peak-finding algorithm and a correctly specified fitting equation would settle whether the correlation is real. This concern does not invalidate the careful photometric and spectroscopic characterization, nor does it require changing the reader's CONDITIONAL verdict; it strengthens the reasons for conditionality. I therefore keep the verdict unchanged while disagreeing with the reader's choice of the weakest assumption.","tokens_in":21142,"tokens_out":7929,"duration_ms":74328,"concrete_test":"Obtain the published GIT+ZTF light curve, smooth it with a Gaussian of width 2-3 days, and identify local maxima using a robust peak finder with a threshold of, say, 0.3 mag above neighboring minima. Independently re-fit the relation log(Δt) = b log(t - t_0) + c on the resulting peak times using bootstrap resampling of the photometric errors to propagate peak-time uncertainties. Report b, c, and the number of peaks. If b is not consistent with 0.89, or if the correlation is not significant (e.g., r² or p-value), the claimed 'first such correlation' fails; if c is significantly different from 0, Eq. (1) is misspecified.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative claim that successive peaks of AT 2023tkw follow log(t_i - t_{i-1}) = b log(t_i - t_0) with b = 0.89 ± 0.12 (Eq. 1, §6.3) is not robustly supported. First, the equation as written is singular at i = 1: for i = 1, LHS and RHS both equal log(t_1 - t_0), so equality forces b = 1 unless t_1 - t_0 = 1 in the chosen units. Since the reported b differs from 1, either the relation is intended only for i ≥ 2, or the fit included an unstated normalization constant; the printed equation is not the relation actually fitted, so the reader cannot verify the fit. Second, the relation is fitted to only four maxima, i.e., three intervals. If the i = 1 point is excluded, only two independent points constrain b; if included, it pulls b toward 1. Thus the stated uncertainty 0.12 must underestimate the true error. Third, the peak times are not determined by any objective algorithm; the light curve is noisy and sparsely sampled (daily GIT cadence, three-day ZTF cadence), so the identification of 'four peaks' is subjective. The last feature at day 155-160 is described as a rise before coverage ended, so its peak time is particularly uncertain. A shift of even a few days in one peak time can change b significantly. Since this correlation is the first reported for an M31 nova and is used to support the internal-shock interpretation via Steinberg & Metzger (2020), it is load-bearing for the paper's central claim.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a detailed photometric and spectroscopic study of the M31 classical nova AT 2023tkw, discovered by the GROWTH-India Telescope. It reports a multi-peaked light curve with four maxima, and argues that the increasing intervals between successive peaks follow the relation log(t_i - t_{i-1}) = b log(t_i - t_0) with b = 0.89 ± 0.12, which the authors interpret as evidence for internal shocks following the models of Steinberg & Metzger (2020). The paper also uses PHOENIX and Cloudy spectral modeling to derive photospheric temperatures, an ejecta mass of about 10^-4 Msun, and, by combining HST astrometry with archival and eruption properties, a binary consisting of a 0.65 Msun white dwarf and a K III giant secondary.","tokens_in":21518,"tokens_out":4693,"duration_ms":44096,"significance":"If the peak-spacing correlation is robust, this would be the first such measurement for an M31 nova and would add an extragalactic data point to the discussion of multi-peaked nova light curves and their possible origin in internal shocks. The dataset itself is valuable: dense GIT photometry, ZTF forced photometry over a long pre-discovery baseline, multi-epoch optical and NIR spectroscopy, and HST imaging of the quiescent field. The data reduction follows standard, reproducible procedures, and the authors are generally explicit about their caveats, including the uncertainty in the quiescent host identification and the speculative nature of the shock interpretation. However, the central quantitative claim on peak spacing currently rests on a fragile fit to very few points, and the printed equation does not match the relation that was actually fitted.","major_comments":[{"comment":"As printed, Eq. (1) is singular at i = 1: it reduces to log(t1 - t0) = b log(t1 - t0), which forces b = 1 unless t1 - t0 = 1 in the chosen units. The reported b = 0.89 therefore cannot come from the displayed equation if i = 1 is included. If the fit instead excludes i = 1, then with only four maxima the slope b is constrained by just two independent intervals, and the stated uncertainty of 0.12 is not credible. Please state the actual fitted form (including any additive normalization constant), the numerical peak times used, and explicitly which maxima entered the fit.","section":"§6.3, Eq. (1)"},{"comment":"The four peaks used for the power-law fit are picked by eye from a noisy light curve with daily GIT and roughly three-day ZTF cadence, and the last feature is described as a rise before coverage ended, making its peak time especially uncertain. No objective peak-finding algorithm, peak-time uncertainties, or robustness tests are provided. A shift of a few days in any single peak time can change b significantly. The authors should demonstrate that the reported b and its error are stable under alternative peak selections and should propagate peak-time uncertainties into the fit.","section":"§6.3, Figure 5"},{"comment":"The identification of the single HST/ACS source within the GIT 1σ error circle as the quiescent host has a 11.89% probability of being a random field star, which is not 'rather low' as stated in §5.2. Because the derived binary parameters (0.65 Msun WD, K III secondary, accretion rate 5e-13 Msun/yr) depend entirely on that association, the binary solution should be presented as tentative, or the analysis should be repeated using a more stringent localization than the 1σ circle (for example, the final image-subtraction position of the transient).","section":"§5.2, §6.1"}],"minor_comments":[{"comment":"The word 'ejcta' in the text describing the Cloudy model is a typo for 'ejecta'.","section":"§4.2.2, Table 2"},{"comment":"The text refers to 'V4250 Oph' in the discussion of Ak et al. (2005); this appears to be a typo for 'V2540 Oph', which is the object discussed earlier in the paper and in the cited reference.","section":"§6.3"},{"comment":"The caption of Figure 1 states that the temperature evolution is shown in the last panel, but the panel does not show any error bars or uncertainties; please add them or explicitly note that they are omitted for clarity.","section":"§3, Figure 1"},{"comment":"The units of t_i and t_0 should be stated explicitly (days since first detection), and the definition of t0 as the first maximum rather than the eruption date should be made unambiguous in the text around Eq. (1).","section":"§6.3, Eq. (1)"},{"comment":"The density parameter N is defined as ρ ∝ r^-N for PHOENIX in §4.2.1, but the Cloudy model in Table 2 lists N = -3; please clarify whether the same sign convention is used and explicitly state the radial density profile adopted in the Cloudy model.","section":"§4.2.1, §4.2.2"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: a useful, honest single-object study. The genuinely new item is the first measurement of the growing peak-interval correlation in an M31 nova (b = 0.89 ± 0.12), and the dense GIT + ZTF light curve plus multi-epoch optical/IR spectra make AT 2023tkw a well-characterized slow M31 nova. The internal-shock interpretation is plausible but not proven, and the paper mostly says so.\n\nWhat it does well: the photometry is standard, host-subtracted, and well documented; the PHOENIX fits reproduce continuum, emission, and P Cyg features except at the primary peak; the Cloudy NIR modeling gives a reasonable ejecta mass ~1e-4 Msun. The authors are unusually upfront about the HST host identification, giving the 11.89% chance of a random field star and noting the possibility that the true quiescent system is fainter. They also consider dust and wind-transition alternatives before settling on internal shocks.\n\nSoft spots: the peak-interval claim is weaker than the title/abstract suggest. Equation 1 as printed cannot hold for i = 1 unless b = 1 or the units are special, so it must be intended for i ≥ 2, but the paper doesn't say that. With four maxima there are only three intervals; excluding the first leaves two points constraining b. The 0.12 uncertainty is optimistic, and the peak times are picked from a noisy, daily-to-three-day cadence light curve. A few-day shift in one peak changes b noticeably. The same correlation already exists for Galactic novae, so the M31 detection is an extension rather than a new physical effect, but it is a legitimate one. The shock attribution rests heavily on the single unmodeled peak spectrum and the Steinberg & Metzger picture; the paper explicitly says that without gamma-ray data it can only speculate, so the title slightly overstates the evidence. The Cloudy ejecta mass and binary parameters carry no propagated uncertainties, and the binary parameters depend on that one HST source; again, the paper flags the host-source risk.\n\nVerdict: this deserves a serious referee, not a desk reject. Nova and M31 transient people will want it. A referee should ask for an objective peak-selection procedure, a corrected equation with the fitted form stated explicitly, and a more honest uncertainty on b. The shock language in the title could be softened. With those changes it's a solid contribution.","headline":"Useful single-object M31 nova study with a first peak-interval correlation that is suggestive but not robust; worth refereeing.","tokens_in":22204,"tokens_out":3422,"would_cite":true,"duration_ms":31721,"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":"AT 2023tkw's repeated peaks are caused by internal shocks that inflate and cool the photosphere, with spacing that grows as a power law.","keywords":["classical novae","M31","internal shocks","light-curve rebrightenings","white dwarf binaries","photospheric expansion","spectral modeling","shock heating"],"falsifier":"Take spectra at high cadence across all four maxima and the following minima: the internal-shock picture requires the photosphere to be cooler and more extended at each optical maximum, so a spectrum showing the hottest photospheric temperature at a maximum would contradict it.","tokens_in":20872,"feed_emoji":"🌟","tokens_out":10801,"duration_ms":87946,"temperature":0.7,"pith_summary":"This paper follows a single slow classical nova in the Andromeda galaxy that brightened, dipped, and rebrightened four times over roughly five months. The authors argue that the repeated peaks are produced by internal shocks: each shock heats the ejected envelope, inflates and cools the photosphere, and the photosphere then contracts until the next shock arrives. They find that the time between successive maxima grows with time from the first maximum according to $\\log(t_i - t_{i-1}) = b \\log(t_i - t_0)$ with $b = 0.89 \\pm 0.12$, the first time this pattern has been reported for an M31 nova. If the interpretation is right, optical rebrightenings in extragalactic novae can be read as direct tracers of shock activity, and the system is a low-mass white dwarf slowly accreting from a cool giant companion.","feed_headline":"M31 nova's four peaks point to internal shocks","feed_subtitle":"Peaks of AT 2023tkw space out on a power law, the first such correlation seen in an M31 nova.","key_machinery":"The argument is carried by the internal-shock mechanism together with a timing relation. Alternating slow and fast outflows from the eruption collide near or below the photosphere; each collision heats the envelope, inflates the photosphere, and shows up as an optical peak, after which the photosphere cools and contracts. The paper demonstrates the expansion-and-contraction cycle with full spectral fits: PHOENIX, an expanding-photosphere radiation-transfer code, reproduces the continuum, emission lines, and P Cygni profiles at all epochs except the primary peak, and Cloudy, a photoionization code, fits the optically thin infrared spectrum. The timing relation $\\log(t_i - t_{i-1}) = b \\log(t_i - t_0)$ with $b = 0.89 \\pm 0.12$ connects the growing gaps between successive maxima to the time since the first maximum, tying the peak spacing to the shock sequence.","core_discovery":"The central claim is that AT 2023tkw is a slow classical nova whose four light-curve maxima are caused by a series of internal shocks generated near or within the photosphere, rather than by dust formation, orbital modulation, or a one-time transition between expansion states. Spectral modeling with the stellar-atmosphere code PHOENIX shows that the photosphere is cooler and more extended at each optical maximum and hotter during the declines, matching the expansion-and-contraction cycle expected when shock fronts reach and stretch the photosphere. The spectrum taken at the primary peak deviates from pure photospheric models, which the paper reads as evidence of an additional shock-heated or free-free component contributing at maximum. The intervals between successive maxima satisfy $\\log(t_i - t_{i-1}) = b \\log(t_i - t_0)$ with $b = 0.89 \\pm 0.12$, consistent with earlier Galactic results and reported here for the first time for an M31 nova. From archival space-telescope photometry and comparison with nova evolution models, the authors identify the underlying binary as a $0.65\\,M_\\odot$ white dwarf accreting at roughly $5\\times 10^{-13}\\,M_\\odot\\,\\mathrm{yr}^{-1}$ from a K III giant, with an ejected mass near $10^{-4}\\,M_\\odot$.","pith_inferences":["If the source detected in archival space-telescope images is eventually shown to be unrelated to the nova, the binary parameters would need revision, but the photometric and spectroscopic case for internal shocks would stand on its own.","The same $b \\approx 0.89$ timing law found in Galactic novae and now in M31 suggests a common mechanism; a systematic search of existing M31 nova light curves could show whether the correlation is universal or a chance occurrence.","A purely photometric test is possible: if the color and temperature oscillations predicted by photospheric expansion and contraction repeat at every peak, high-cadence multi-band monitoring of future M31 novae could confirm the shock scenario without spectra.","If confirmed, the result implies that M31, with its known distance and high nova rate, can serve as a laboratory for studying nova shock physics in the optical alone, complementing the gamma-ray-emitting Galactic sample."],"forward_implications":["If the shocks are real, the optical light curve of an extragalactic nova can serve as a timing record of energy-injection events near the photosphere, without requiring gamma-ray or X-ray detections.","The power-law spacing with $b \\approx 0.89$ implies that rebrightenings should keep arriving at ever longer intervals, so continued monitoring of AT 2023tkw's decline would test whether a further peak appears near the extrapolated time.","The binary parameters imply an extremely slow accretion rate and a recurrence timescale of hundreds of millions of years, meaning this is a genuine one-shot classical nova rather than a recurrent system.","The measured ejected mass of roughly $10^{-4}\\,M_\\odot$ is typical for slow novae, so the multi-peaked shock behavior is not tied to unusually heavy ejecta.","Because the peak spectrum cannot be fitted by a single photospheric model, any model that treats the optical peak of a nova as purely photospheric is incomplete."],"supporting_citations":[{"why":"1D hydrodynamic models showing alternating slow and fast outflows produce internal shocks and multi-peaked nova light curves.","marker":"Steinberg & Metzger (2020)"},{"why":"Established the log-interval versus time-since-first-maximum correlation that this paper fits.","marker":"Pejcha (2009)"},{"why":"Derived a near-linear version of the same peak-spacing relation for Galactic novae.","marker":"Tanaka et al. (2011a)"},{"why":"V906 Car's simultaneous gamma-ray and optical peaks show shock emission coupled to optical brightening.","marker":"Aydi et al. (2020)"},{"why":"Review of nova multi-wavelength observations that establishes internal shocks as a source of high-energy emission.","marker":"Chomiuk et al. (2021)"},{"why":"Nova evolution grid used to match the derived white-dwarf mass, accretion rate, and slow recurrence timescale.","marker":"Yaron et al. (2005)"}],"fun_headline_variants":["Four peaks in M31 nova reveal shock heating","Internal shocks explain M31 nova's multi-peak light curve","AT 2023tkw: Shock series behind nova's repeated brightening","First M31 nova shows power-law spacing of peaks","M31 nova's multiple peaks linked to shock heating"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The single faint source found in archival space-telescope images inside the telescope's error circle is assumed to be the nova's quiescent binary system, even though the paper itself estimates an 11.89% probability that it is an unrelated field star.","fun_headline_variants_meta":{"raw":{"variants":["Four peaks in M31 nova reveal shock heating","Internal shocks explain M31 nova's multi-peak light curve","AT 2023tkw: Shock series behind nova's repeated brightening","First M31 nova shows power-law spacing of peaks","M31 nova's multiple peaks linked to shock heating"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001141,"raw_usage":{"total_tokens":4763,"prompt_tokens":1003,"completion_tokens":3760,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":619,"completion_tokens_details":{"reasoning_tokens":3678}},"tokens_in":619,"tokens_out":3760,"duration_ms":23829,"temperature":1.0,"reasoning_tokens":3678,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T11:24:29.662998+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take spectra at high cadence across all four maxima and the following minima: the internal-shock picture requires the photosphere to be cooler and more extended at each optical maximum, so a spectrum showing the hottest photospheric temperature at a maximum would contradict it.","supporting_citations":[],"review_version":1}