{"id":"85cf30c6-21ef-4e2a-8dd8-5a91cfeeefc6","arxiv_id":"2504.16872","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"FAST timing observations reveal a previously missed millisecond pulsar, M15O, whose period lies within 0.004 ms of the tenth harmonic of M15A and which is the closest known pulsar to the center of M15.","lead":"Astronomers using China's FAST telescope found a new millisecond pulsar in the globular cluster M15, hidden because its spin period almost exactly matches the tenth harmonic of a brighter pulsar in the same cluster. The discovery was confirmed by timing its pulses over four years and suggests that dense star clusters may hide more pulsars behind bright neighbors.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The phase connection depends on an unvalidated F2; no search over integer-cycle aliases is shown, so the distinct position and period of M15O are not yet uniquely established.","rationale":"The reader's weakest assumption identifies exactly the same load-bearing point: the timing solution is unique and not an integer-cycle alias. I find no sharper or more central objection. The paper has real supporting evidence—a consistent Dracula solution, manual phase-removal agreement, a folded profile, detection across many epochs, and a position offset far larger than the quoted timing uncertainties—so I would not downgrade to REJECT. But the omission of the alias search and of any goodness-of-fit comparison is a formal gap in the central claim, and it is testable. UNCHANGED means the CONDITIONAL verdict stands: the discovery is plausible but should be accepted only after the alias degeneracy is explicitly ruled out. I also note the paper's own statements in Section 4.2 acknowledge the solution appeared only after adding F2, which is the exact place where the degeneracy can hide.","tokens_in":9877,"tokens_out":5644,"duration_ms":56093,"concrete_test":"Re-run Dracula/TEMPO on the 65 TOAs (or regenerate them from Table 2) and perform an explicit alias search: for each of the 16 observing epochs, perturb the pulse count by +/-1 across every inter-epoch gap, refit F0, F1, F2, RA, Dec, and DM, and record the best chi-squared and fitted position for each cycle-count branch. If any alternative branch yields residuals within 2x the accepted RMS (108 us) or shifts the fitted position by more than 0.1 arcsec, the phase connection is degenerate and the discovery claim is not yet established.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The discovery claim reduces to the phase-connected timing solution in Table 2. Section 4.2 states that no solution could be found with F0 and F1 alone, and that only after adding the second frequency derivative F2 did a single solution emerge. The paper does not report the Dracula search space, the number of integer-cycle trials, the reduced chi-squared of the accepted solution, or the chi-squared of rejected alternative cycle counts. This matters because F2 is doing enormous work: the observed F2 = 3.3133(6)e-23 Hz/s^2 is roughly 10^6 times the value expected from magnetic-dipole spin-down, so it is equivalent to a cubic phase of order 14 turns over the 4.3-yr timing span. With 65 TOAs and irregular multi-month gaps, a different integer-cycle choice between epochs can absorb its error into F1, F2, position, and DM while leaving residuals at the 108-us level. If an alternative alias fits comparably, then the quoted spin frequency, the 0.81-arcsec separation from M15A, and the record 0.37-arcsec projected distance to the cluster center all lose their evidential force. The 'hidden in the harmonics' narrative itself does not add independent support, since the candidate was found in the power spectrum, not by timing.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript reports the discovery of a new millisecond pulsar, M15O (PSR J2129+1210O), in the globular cluster M15, based on FAST L-band observations from 2019 to 2024. The pulsar has a spin period of about 11.06686 ms and a dispersion measure of 67.44(1) pc cm^-3, with a claimed phase-connected timing solution over 65 TOAs giving f = 90.3597874705(1) Hz, fdot = 1.79191(5) x 10^-14 Hz/s, fddot = 3.3133(6) x 10^-23 Hz/s^2, and a timing position separated by 0.81 arcsec from M15A and 0.37 arcsec from the cluster center. The authors argue that M15O was previously missed because its period is extremely close to the 10th harmonic of the bright pulsar M15A, and that spectral stacking plus careful visual inspection of power spectra can reveal such hidden signals. They also report flux-density estimates and discuss the large spin-frequency derivatives of M15O in the context of the dense cluster core.","tokens_in":10088,"tokens_out":14086,"duration_ms":127360,"significance":"If the timing solution is correct, M15O is a genuine addition to the M15 pulsar census and a useful demonstration that bright-pulsar harmonics can hide physically distinct MSP signals in globular cluster searches. The paper's strengths are that the candidate was confirmed by a phase-connected timing solution, that the comparison M15A parameters come from an independent published ephemeris, and that the frequency-domain subtraction of M15A provides a direct visual confirmation of the new signal. The measured second frequency derivative and the very small projected distance from the cluster center make the source potentially interesting for dynamical studies of the M15 core. However, the significance of the positional and dynamical claims rests entirely on the uniqueness of the phase connection, which the manuscript does not yet document.","major_comments":[{"comment":"The uniqueness of the phase-connected timing solution is not established. The paper states that no solution was found with F0 and F1 alone and that adding F2 yielded a single solution, but it does not report the Dracula search space, the number of integer-cycle aliases tested, the reduced chi-squared of the accepted solution, or the fit quality of rejected aliases. Because the quoted F2 = 3.3133(6) x 10^-23 Hz/s^2 produces a cubic phase of order 14 turns over the ~4.3-year span quoted in the text (about 27 turns if the Table 2 start MJD 58404.483 is used), an alternative cycle count with slightly different F1, F2, position, and DM could plausibly fit the 65 TOAs. Please provide the search details and a comparison with rejected alias solutions; this is load-bearing for the claims that M15O is distinct from M15A and located 0.37 arcsec from the cluster center.","section":"Section 4.2 and Table 2"},{"comment":"Table 2 lists the start of timing data as MJD 58404.483, which corresponds to 2018 October 13, but Table 1 and Section 2 state that the data are 19 FAST observations from 2019 November (MJD 58796) to 2024 February. This one-year discrepancy changes the length of the phase-connected timing span and appears to be related to the statement in Section 4.2 that the observations span six years. Please correct the MJD or explain which additional data were used in the timing solution.","section":"Table 2 and Table 1"},{"comment":"The timing fit is documented only by the residual RMS of 108 microseconds; the manuscript does not give the TOA uncertainties, the number of fitted parameters, the reduced chi-squared, or the per-epoch residual rms. These quantities are needed to judge whether the quoted F2 and timing position are statistically acceptable and to compare against alternative alias solutions. Please add them to the paper.","section":"Section 3 and Table 2"}],"minor_comments":[{"comment":"The period difference between M15O and the 10th harmonic of M15A is stated as about 0.004 ms, but the quoted frequencies imply a difference of about 0.0004 ms; please correct the numerical value.","section":"Abstract and Section 3"},{"comment":"The flux densities in Table 1 are quoted to three significant figures with no uncertainties, and they are derived from the radiometer equation with an assumed pulse width W_obs = 0.15 P_spin; please state the uncertainty budget or clearly label these values as estimates.","section":"Table 1"},{"comment":"The statement that 'at least 50% of experienced persons' did not see the signal in the diagnostic plot is informal; please replace it with a quantitative description of the test or remove the anecdotal percentage.","section":"Section 4.1"},{"comment":"The comparison with Galactic-field pulsars uses a 5-arcmin diameter and 5 cm^-3 DM difference criterion, but M15 pulsars are concentrated within a single FAST beam and have a DM spread of only a few tenths of pc cm^-3; the logical connection between the field-pulsar statistics and the cluster case should be clarified.","section":"Section 4.3"},{"comment":"The stack search method is essential to the discovery narrative, but its details are deferred to Dai et al. (2025, in preparation); please include a concise description of the stacking procedure and the candidate threshold in the present paper.","section":"Section 2.2"},{"comment":"The pulsar name appears as J2129+1210O in the title and text but as J2129-1210O in the keywords; please use a consistent sign convention.","section":"Abstract and Keywords"},{"comment":"Please clarify what exactly is meant by 'chopping the M15A signal' in the frequency domain and how the removed component was identified and validated.","section":"Figure 2 caption"},{"comment":"The sentence suggesting an orbital period on the order of decades based on F1 and F2 needs a shown calculation or should be explicitly labeled as speculative; if a wide orbit is possible, a direct binary or acceleration-model fit would be the appropriate test.","section":"Section 4.2"}],"recommendation":"major_revision","confidential_remarks":"This is a standard observational discovery paper that would be suitable for RAA after the timing validation is documented. The main issue is the lack of transparency in the phase-connection search and the apparent inconsistency in the start MJD of the timing data; both are fixable. I do not see a novelty or scope concern."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe discovery of M15O is probably real, but the timing solution is less secure than the paper suggests, and the exotic F2 is doing more work than the presentation acknowledges.\n\nWhat's new: another MSP in M15, the first measured second spin-frequency derivative for that cluster, and a timing position that puts the pulsar closest to the cluster center. The harmonic mask from M15A is a genuine issue, and the suggestion to include spectra in diagnostic plots is reasonable.\n\nWhat the paper does well: the basic detection is supported by an SNR of ~11 in the power spectrum, a clean folded profile after subtracting M15A, and 65 TOAs over four years. The use of Dracula is a good choice for phase connection.\n\nThe soft spot is in Section 4.2. The paper states that no solution could be found with F0 and F1 alone, and that only after adding F2 did a single solution emerge. That F2 is about six orders of magnitude larger than the dipole spin-down expectation. With long multi-month gaps between observations, an alternative integer-cycle count could in principle absorb the phase into F1, F2, position, and DM while leaving residuals at the quoted 108 microseconds. The paper does not report the range of cycle counts tried, the reduced chi-squared of the accepted solution, or the penalty of rejected aliases. The statement that Dracula agreed with a manual check is reassuring but not quantitative. Without that information, the position, which drives the 'closest to center' claim, should be treated with caution.\n\nMinor issues: flux densities in Table 1 have no error bars; the claim that M15O could have been found earlier is not quantified; and the stack search details are deferred to a future paper.\n\nNone of this kills the discovery – the signal is there and it is distinct from M15A. But the precision of the spin parameters and the astrophysical interpretation depend on the uniqueness of the phase connection, which is not demonstrated.\n\nThis is a paper for pulsar and globular cluster people. It deserves peer review, but the referee should ask for a proper alias search and error bars on the flux densities.","headline":"New MSP in M15 is likely real, but the timing solution leans on a huge F2 and the paper doesn't demonstrate the alias search.","tokens_in":10677,"tokens_out":4053,"would_cite":false,"duration_ms":35645,"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":"M15O is a real millisecond pulsar hidden in the 10th harmonic of M15A.","keywords":["millisecond pulsar","globular cluster M15","M15O","harmonic interference","stack search","pulsar timing","spin frequency derivative","FAST"],"falsifier":"Fold a future FAST observation with the published ephemeris: if the predicted pulse phase drifts by much more than the $108\\ \\mu$s residual RMS, or if a re-fit of the existing TOAs finds another integer-cycle alias with comparable quality, then the claimed spin-down, position, and separation from M15A collapse.","tokens_in":9657,"feed_emoji":"🔭","tokens_out":12657,"duration_ms":109028,"temperature":0.7,"pith_summary":"The paper reports the discovery of M15O, a previously missed millisecond pulsar in the globular cluster M15. Its spin period, about 11.06686 ms, is so close to the tenth harmonic of the bright pulsar M15A that standard candidate screening mistook the two signals for one. The authors used a stack of 19 archival observations from the FAST telescope plus a phase-connected timing solution to show that M15O is a distinct, isolated pulsar with its own position, spin-down, and dispersion measure. If correct, the result means real pulsars can hide inside the harmonic structure of bright cluster pulsars, and that adding power spectra to candidate diagnostic plots would catch such cases.","feed_headline":"New pulsar found hiding in M15A's 10th harmonic","feed_subtitle":"M15O spins every 11.06686 ms, nearly overlapping the tenth harmonic of bright M15A, so earlier searches missed it.","key_machinery":"The load-bearing coincidence is the near-degeneracy between M15O's spin period ($\\approx 11.06686$ ms) and the tenth harmonic of M15A ($\\approx 11.06647$ ms), a difference the paper reports as roughly 0.004 ms. The mechanism that resolves the degeneracy is threefold: an incoherent stack of power spectra over many epochs raises M15O's faint peak above noise; a frequency-domain chop of M15A's harmonic removes the contaminating signal; and a phase-connected timing fit that includes $\\ddot f$ fixes the integer cycle count, giving M15O a position distinct from M15A.","core_discovery":"M15O (PSR J2129+1210O) is a genuine isolated millisecond pulsar in M15 with spin frequency $f = 90.3597874705(1)$ Hz, first derivative $\\dot f = 1.79191(5)\\times 10^{-14}$ Hz s$^{-1}$, second derivative $\\ddot f = 3.3133(6)\\times 10^{-23}$ Hz s$^{-2}$, and dispersion measure DM $= 67.44(1)$ pc cm$^{-3}$. Its projected position is $0.37''$ from the cluster center and $0.81''$ from M15A, making it the closest known pulsar to the M15 core. It was missed earlier because its period is nearly identical to the tenth harmonic of M15A, a difference the paper reports as roughly 0.004 ms. Timing of 65 arrival times with a phase-connected solution, including the second frequency derivative, is what finally separates it from that harmonic.","pith_inferences":["If harmonic proximity is not rare, some single-epoch 'harmonic' detections in other dense clusters may actually be unresolved pulsars; re-stacking archival spectra at matched dispersion measures could test this without new observations.","The large $\\dot f$ and measured $\\ddot f$ are also consistent with an unseen companion in a wide orbit; if future timing shows sinusoidal residuals on a decade timescale, M15O would become a probe of a massive central object in M15.","A practical extension would be to quantify the minimum period separation at which two same-DM pulsars can be disentangled by timing; that threshold would tell surveys how close to known harmonics they must search."],"forward_implications":["M15O must be added to the census of M15 pulsars, and its position $0.37''$ from the cluster center means it is the closest known pulsar to the core.","Earlier M15 data can be re-examined: the paper notes M15O could already have been detected by January 2021 if the spectra had been inspected.","Stacking power spectra before folding is an effective way to uncover pulsars hidden under bright harmonics, so the same approach on other globular clusters may yield more pulsars.","Because M15O is the only M15 pulsar with a measured $\\ddot f$, and its $\\dot f$ is second only to M15D, the timing solution points to strong acceleration near the core; continued monitoring may reveal whether it is gravitational acceleration or orbital motion.","Adding the power spectrum to pulsar candidate diagnostic plots is a concrete change to search practice that can prevent similar misses."],"supporting_citations":[{"why":"Provides the Fourier-domain acceleration search software and zmax criterion used to generate the spectra and fold candidates.","marker":"Ransom, Eikenberry, & Middleditch 2002"},{"why":"Introduces incoherent summation of power spectra, the basis of the stack search that detected M15O.","marker":"Anderson 1993"},{"why":"Shows stack-search detection of faint pulsars in archival data, motivating this application.","marker":"Pan et al. 2016"},{"why":"Demonstrates the stack search recovering faint globular-cluster pulsars, supporting the method's sensitivity.","marker":"Cadelano et al. 2018"},{"why":"Supplies the algorithm that resolves the integer-cycle ambiguity required for the phase-connected timing solution.","marker":"Freire & Ridolfi 2018"},{"why":"Provides the timing package used to fit the 65 TOAs and derive the ephemeris in Table 2.","marker":"Nice et al. 2015"},{"why":"Gives M15A's spin parameters and harmonic frequencies and the census of known M15 pulsars used for comparison.","marker":"Wu et al. 2024"},{"why":"Reports M15I, the analogous pulsar that needs a second frequency derivative but still lacks a phase-connected solution.","marker":"Pan et al. 2021"}],"fun_headline_variants":["M15O: pulsar hidden in M15A's harmonic for years","New millisecond pulsar M15O found in M15A's shadow","M15O: the pulsar that hid inside M15A's 10th harmonic","Pulsar M15O snuck past detection inside M15A's harmonic","M15O: closest pulsar to M15 core, hidden in harmonic"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The phase-connected timing solution is unique: the 65 arrival times have one correct set of integer pulse counts, and the fitted second frequency derivative is real rather than a stand-in for unmodeled orbital motion or phase wraps.","fun_headline_variants_meta":{"raw":{"variants":["M15O: pulsar hidden in M15A's harmonic for years","New millisecond pulsar M15O found in M15A's shadow","M15O: the pulsar that hid inside M15A's 10th harmonic","Pulsar M15O snuck past detection inside M15A's harmonic","M15O: closest pulsar to M15 core, hidden in harmonic"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000329,"raw_usage":{"total_tokens":1860,"prompt_tokens":995,"completion_tokens":865,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":611,"completion_tokens_details":{"reasoning_tokens":760}},"tokens_in":611,"tokens_out":865,"duration_ms":7512,"temperature":1.0,"reasoning_tokens":760,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T10:53:46.106992+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Fold a future FAST observation with the published ephemeris: if the predicted pulse phase drifts by much more than the $108\\ \\mu$s residual RMS, or if a re-fit of the existing TOAs finds another integer-cycle alias with comparable quality, then the claimed spin-down, position, and separation from M15A collapse.","supporting_citations":[],"review_version":1}