{"id":"887474aa-ab84-4411-9e3d-d3eb2f0c066c","arxiv_id":"2507.18955","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A new analysis pipeline cross-correlates the time-varying stochastic gravitational-wave background with gamma-ray, X-ray, optical, and radio light curves to find electromagnetic counterparts of sub-threshold mergers, with mock-data false alarm rates that drop when multiple bands are combined.","lead":"The paper proposes a time-domain cross-correlation pipeline, MC^2, that searches for electromagnetic counterparts to gravitational-wave events that are too faint to detect individually but contribute to the stochastic background. The authors show in simulations that requiring coincidences in multiple EM bands sharply lowers false alarm rates, and they suggest observational follow-up strategies for current and future telescopes.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"FAR values in Tables 3-5 are single-trial p-values, not per-year rates; without a trials-factor correction the claimed multi-band false-alarm reduction is not established.","rationale":"The reader's weakest assumption correctly identifies that the null distribution omits the astrophysical SGWB and that the EM transient rates and profiles are uncertain. My stress-test pass finds a sharper and more concrete problem in the same Section 7: the quantity called \"FAR (per year)\" is computed as a single-trial complementary CDF and is never converted to a rate by multiplying by the number of EM candidates or independent trials. The paper's own \"min\" floor, delta_t / (seconds in a year), is the single-bin p-value floor; using it as a per-year FAR floor implicitly assumes exactly one trial, contradicting the Table 1 rates of 120-1000 transients per year. A trials-corrected floor for the all-band search is roughly N_trials times larger. This is an internal inconsistency in the central quantitative claim, not merely a disagreement with plausible astrophysical rates. The qualitative direction of the result (more bands reduce false coincidences) may survive, so the appropriate verdict remains conditional rather than rejection: the paper can be fixed by recomputing FARs with a proper trials factor and with the SGWB included in the null, and by reporting detection significances with error bars. My recommendation of CONDITIONAL aligns with the reader's verdict, though the specific load-bearing concern differs in emphasis, hence agreement_with_reader is partial.","tokens_in":25074,"tokens_out":13542,"duration_ms":148430,"concrete_test":"Recompute the FAR values in Tables 3-5 as p * N_trials, where p is the complementary-CDF value actually computed and N_trials is the number of independent EM candidates searched (use the Table 1 event rates; for multi-band columns, use the resulting coincidence rate). Also regenerate the null distribution with the simulated astrophysical SGWB (Section 5.1) added to A+ detector noise, keeping the same injected false counterparts. If any corrected FAR exceeds 1e-3 per year, or if any \"<= min\" entry exceeds the corrected floor N_trials * delta_t / T, then the central FAR-reduction claim is quantitatively unsupported as stated.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"In Section 7, the paper defines the false alarm rate as the complementary CDF of the TDCC distribution and reports it in units of \"per year\" (Tables 3-5). The text states: \"This area is equivalent to the value of the complementary CDF at the point of TDCC of the event, which is the FAR for that event.\" A complementary CDF is a per-trial probability, not a rate. A search that examines the EM transient population of Table 1 (total ~2045 events/yr) must multiply this p-value by the number of independent candidates or trials to obtain an expected false-alarm rate per year. The quoted \"min\" floor, delta_t / (seconds per year) = 9.5e-6 for 300 s, is the single-time-bin p-value floor; the corresponding trials-corrected floor is N_trials * delta_t / T, which is ~0.02/yr for all Table 1 events and ~0.001/yr for GRBs alone. Thus entries reading \"<= min\" do not imply false-alarm rates below 1e-5 per year. Additionally, the null distribution is generated from A+ detector noise only (Section 7), omitting the fluctuating astrophysical SGWB of Section 5.1; unrelated EM transients can correlate with random SGWB bursts, and this background is not counted. Both omissions inflate the quoted significance and directly undermine the central quantitative claim of drastic FAR reduction via multi-band coincidence.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents MC^2, a time-domain cross-correlation (TDCC) pipeline that correlates a time-dependent estimate of the astrophysical stochastic gravitational-wave background (SGWB), Omega_GW(f,t), with electromagnetic light curves (Eq. 4.4) in order to identify EM counterparts of sub-threshold compact binary coalescences and unmodeled GW bursts. The authors simulate the SGWB from CBC populations (BNS, NSBH, BBH) using the temporal fluctuation formalism of Mukherjee & Silk and Sah & Mukherjee, inject EM counterparts with GW170817-inspired profiles in gamma-ray, X-ray, UVOIR, and radio bands, and report detection spikes and false-alarm rates for single-band and multi-band combinations (Tables 3-5). They also demonstrate time-delay recovery, optimum averaging-time selection, glitch mitigation in two-detector cross-correlation, and propose an observation strategy for current and future multi-messenger facilities.","tokens_in":25398,"tokens_out":6294,"duration_ms":73981,"significance":"If the quantitative claims survive revision, the framework would provide a useful new tool for multi-messenger searches of sub-threshold GW events, particularly because the TDCC statistic avoids requiring individually detected GW triggers and can exploit nearly full-sky gamma-ray monitoring. The paper is clearly structured, the simulation recipes are described in enough detail to be reproduced, and the demonstration that two-detector cross-correlation suppresses uncorrelated glitches (Appendix B.3) is a genuinely useful addition. However, the central quantitative claim—drastic false-alarm reduction reaching the ~1e-5 per-year floor—rests on a per-trial p-value being reported as a per-year rate, and on a null distribution that omits astrophysical SGWB fluctuations. Those two issues materially weaken the current numerical support for the headline conclusion, although both are fixable in a revision.","major_comments":[{"comment":"The quoted 'FAR (per year)' values are single-trial complementary-CDF probabilities, not per-year rates. The text states: 'This area is equivalent to the value of the complementary CDF at the point of TDCC of the event, which is the FAR for that event.' A complementary CDF is a per-trial p-value; converting to an expected false-alarm rate per year requires multiplication by the number of independent trials per year, e.g., the total unassociated EM transient rate of Table 1 (~2045 events/yr) or the number of independent time offsets searched. After this correction, entries reading '≤ min' with min = delta_t/(seconds per year) = 9.5e-6 for 300 s imply a floor of roughly N_trials x 9.5e-6, about 0.02/yr for all Table 1 events and about 0.001/yr for GRBs alone. The claimed multi-band FAR reduction to the 1e-5 per-year floor is therefore not established by the presented numbers.","section":"Section 7, Tables 3-5"},{"comment":"The false-alarm null distribution is generated from A+ detector noise only, as stated at the start of Section 7 ('We generate the GW background noise of A+ design sensitivity for a duration of one year'). The TDCC in Eq. (4.4) is applied to the reconstructed Omega_GW(f,t), which in a real search contains the fluctuating astrophysical SGWB of Section 5.1. Unrelated EM transients can therefore produce nonzero TDCC through chance alignment with SGWB fluctuations, and this contribution is not counted in the quoted FAR. The null simulations should include the astrophysical SGWB, or the authors should provide a quantitative argument that its contribution to the null TDCC distribution is negligible relative to detector noise.","section":"Section 7, null distribution"},{"comment":"The detection demonstrations are injection-recovery tests with known time delays, and the TDCC peaks in Figures 3 and 4 are presented without a detection threshold, significance measure, or error bars. The statement in Section 6.1 that the correlations are 'statistically significant' is not quantified, and no criterion is given for declaring a candidate multi-messenger event in a blind search. The FAR analysis in Section 7 is applied only to the specific injected events of Table 2, not to the peaks shown in the detection figures, so the pipeline's detection capability and its false-alarm rate are not connected by an explicit decision rule.","section":"Section 6.1 and 6.2, Figures 3-4"}],"minor_comments":[{"comment":"The conclusion states that the multi-band approach 'makes possible the precise localization and redshift determination from their EM counterparts,' but the paper does not present a localization or redshift-estimation procedure, nor does it quantify the claimed precision; this statement should be softened or supported.","section":"Section 8"},{"comment":"The assumed unassociated EM event rates (Table 1) and the assumption that false counterparts have the same smooth, modeled light-curve shapes as genuine counterparts directly control the quoted FAR values. Since these rates and shapes are acknowledged to be uncertain, the paper should provide a sensitivity analysis showing how the reported FARs change under plausible alternative rates and noisier light-curve profiles.","section":"Table 1 and Section 7"},{"comment":"In the radio model, the frequency range is given as 'between 3x10^11 Hz and 9x10^3 Hz,' which appears to have the bounds reversed; this is likely a typographical error and should be corrected.","section":"Appendix A.4"},{"comment":"The notation for the estimators and averages, such as the distinction between the observed Omega_GW(f,t), the time-averaged Omega_GW(f), and the EM mean I_nu, is introduced informally; a short table of symbols or explicit definitions would improve clarity.","section":"Equations (4.1) and (4.4)"},{"comment":"The paper announces a Python-based software package MC2 but provides no code repository, version, or data availability statement; including these would aid reproducibility and is expected for a pipeline paper.","section":"General"},{"comment":"Several figures lack complete axis labels and color-scale descriptions, and the TDCC panels are normalized to unit maximum, which prevents quantitative comparison of peak heights across bands; adding scale bars or caption descriptions would help the reader assess the signal visibility.","section":"Figures 2-4"}],"recommendation":"major_revision","confidential_remarks":"The manuscript builds heavily on the authors' own previous formalism (refs. [31], [46], [54]), and the new contribution is mainly the MC^2 implementation and the FAR analysis. The trials-factor issue in Section 7 is the most serious barrier: it is a statistical interpretation error that can likely be corrected by multiplying the complementary-CDF values by the number of trials per year and rerunning the tables. The omitted astrophysical SGWB in the null distribution is also fixable but requires additional simulations. If the authors address these points quantitatively, the paper could be suitable for publication; in its present form the headline FAR claims are not supported."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The MC² framework is a sensible extension of earlier time-domain cross-correlation work, but the paper's headline quantitative claim—that multi-band analysis drastically reduces false alarm rates—rests on a definitional error. What they label FAR is a complementary CDF, a single-trial p-value, and it's not converted to a rate by multiplying by the number of trials. The 'min' entries in Tables 3–5 are single-bin p-value floors, not per-year rates. In addition, the null distribution is generated from detector noise only, so chance correlations between SGWB fluctuations and unrelated EM transients are not counted. Both issues inflate the quoted significances.\n\nWhat is actually new: the pipeline implementation (MC²), the multi-band FAR tables (even if the numbers need reworking), the mock-data demonstration for unmodeled sources, and a practical observation strategy. The EM waveform models based on GW170817 are careful and well documented, and the glitch-mitigation discussion is a genuinely useful addition. The paper is clearly written, and the formalism is sound in broad strokes.\n\nSoft spots in proportion: the FAR issue is the biggest and is fixable—multiply the p-values by the number of EM candidates per year (or by the number of independent time bins), and the claimed suppression shrinks. The omission of the astrophysical SGWB in the null is more than a detail, because that background carries its own fluctuations that can correlate with unrelated transients. The detection demonstrations show peaks at the injected delays, but there are no significance thresholds or error bars, so we can't tell how meaningful the peaks are. And the injections use the same models the method is built to find—acceptable for a first demonstration, but it doesn't prove real-world performance. The EM transient rates are assumptions, not measurements.\n\nNone of this kills the idea. The TDCC statistic itself originates in the authors' prior work, so the novelty is modest but real: the pipeline, the FAR study, and the observation strategy are new. This paper deserves a serious referee, but it needs revision—specifically, code release, a proper trials-aware FAR calculation that includes the astrophysical foreground, and explicit detection criteria. With those changes it would be a useful method paper for the multi-messenger community.\n\nRecommendation: send it to peer review; engage with the content.","headline":"A plausible method paper whose headline false-alarm rates are single-trial p-values, not per-year rates; worth refereeing after a real FAR recalculation.","tokens_in":25911,"tokens_out":4253,"would_cite":false,"duration_ms":42772,"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":"The paper claims that cross-correlating the stochastic gravitational-wave background with multi-band electromagnetic light curves recovers the counterparts of sub-threshold mergers, with multi-band coincidence reducing false alarms to…","keywords":["multi-messenger astrophysics","stochastic gravitational wave background","time-domain cross-correlation","gravitational wave counterparts","gamma-ray bursts","false alarm rate","compact binary coalescences","sub-threshold gravitational wave searches"],"falsifier":"Generate one year of simulated SGWB that includes realistic astrophysical fluctuations, time-shift EM transients so none is physically associated with a GW event, and count how many TDCC peaks survive the multi-band test; a rate far above $9.5\\times10^{-6}$ per year would show the false-alarm suppression is overestimated.","tokens_in":24877,"feed_emoji":"🔭","tokens_out":10936,"duration_ms":107156,"temperature":0.7,"pith_summary":"This paper proposes a pipeline, MC$^2$ (Multi-messenger Cross-Correlation), for finding electromagnetic counterparts to gravitational-wave sources that are too faint to show up as individual detections. Instead of matched filtering for single events, it cross-correlates the time-varying stochastic gravitational-wave background with light curves in gamma-ray, X-ray, UV/optical/IR, and radio bands. On simulated data, the method recovers the buried GW-EM association for both compact-binary coalescences and unmodeled burst waveforms. Requiring coincidence across multiple independent EM bands drives the false-alarm rate down to the discretization floor, about $9.5\\times10^{-6}$ per year for 300-second integration, a level at which a surviving candidate would be statistically meaningful.","feed_headline":"New cross-correlation finds EM counterparts in the GW background","feed_subtitle":"Multi-band coincidence in the time-domain search lowers false alarms so faint mergers become recoverable.","key_machinery":"The load-bearing object is the time-domain cross-correlation (TDCC) coefficient, implemented in the MC$^2$ pipeline, which correlates the fluctuating part of the SGWB energy density with the fluctuating part of an EM light curve at a scanned delay: $$C_\\nu(t,\\$\\Delta$ t_\\nu)=\\$\\Delta$ f\\sum_f\\left[\\hat{\\$\\Omega$}_{\\mathrm{GW}}(f,t)-\\overline{\\$\\Omega$}_{\\mathrm{GW}}(f)\\right]\\frac{1}{\\delta t}\\int_{t-\\delta t/2}^{t+\\delta t/2}\\left[\\hat{I}_\\nu(t'+\\$\\Delta$ t_\\nu)-\\bar{I}_\\nu\\right]dt'.$$ The subtraction of time averages means only temporary excesses contribute, and scanning over $\\Delta t_\\nu$ and the averaging time $\\delta t$ converts a buried GW episode plus its EM counterpart into a detectable peak. Extending the same test to several EM bands is what drives the reported reduction in false alarms.","core_discovery":"The paper's central claim is that a single time-domain statistic, the multi-messenger cross-correlation between the short-time-averaged stochastic gravitational-wave energy density $\\hat{\\Omega}_{\\mathrm{GW}}(f,t)$ and the electromagnetic flux $\\hat{I}_\\nu$, can pull a sub-threshold merger out of the stochastic background and attach it to the correct EM transient. In mock data the statistic peaks at the physical GW-EM delay for both modeled compact binaries and unmodeled sources, even when the event is invisible in the raw strain spectrogram. The quantitative edge comes from multi-band coincidence: combining gamma-ray, X-ray, UV/optical/IR, and radio channels suppresses false correlations until the false-alarm rate reaches the integration-time floor. The paper concludes that this multi-band route should make precise sky localization and redshift determination possible for otherwise undetectable GW events.","pith_inferences":["A stress test the paper does not run is to include the fluctuating astrophysical SGWB itself in the null ensemble; chance alignments between SGWB episodes and unrelated EM transients would likely raise the single-band false-alarm rates before multi-band coincidence suppresses them.","The same cross-correlation logic should transfer to other messengers: a neutrino light curve could play the role of an EM band in the TDCC, with the expected neutrino arrival delay standing in for $\\Delta t_\\nu$.","The appendix's minimum-sample-length result, roughly twice the delay time, implies that long-delay bands such as radio may need continuous monitoring rather than triggered follow-up to accumulate enough samples for a reliable correlation."],"forward_implications":["Faint compact-binary mergers that cannot be individually detected by matched-filter searches can still be recovered as time-correlated excesses in the stochastic gravitational-wave background, extending the reach beyond the loud-event horizon.","Multi-band coincidence is the main statistical lever: single-band false-alarm rates that are already small fall to about $9.5\\times10^{-6}$ per year for 300-second integration, so a candidate passing all channels is far more credible than a single-band spike.","The statistic is waveform-agnostic, so the same pipeline covers unmodeled or unknown burst sources, not only compact-binary coalescences.","When a counterpart is found, the EM signal provides precise localization and redshift, converting a sub-threshold GW detection into a cosmological measurement.","Because only archival light curves and stored GW data are needed, the search can be re-run on past observation periods to uncover missed events."],"supporting_citations":[{"why":"First proposed temporal correlations between the faint GW signal in the background and EM signals, the idea that this paper generalizes to a multi-band pipeline.","marker":"[46]"},{"why":"Established that the astrophysical stochastic GW background is time-dependent on short intervals, which makes a time-domain search meaningful.","marker":"[31]"},{"why":"Supplies the simulation method for the non-stationary SGWB used to generate the mock data in this paper.","marker":"[54]"},{"why":"Documents the GW170817 binary neutron star merger and its EM counterpart, the physical scenario the search is designed to recover.","marker":"[3]"},{"why":"Gives the A+ detector sensitivity curve used for the simulated GW noise in all analyses.","marker":"[45]"},{"why":"Provides the standard formula for the SGWB energy density from compact binary populations that the pipeline estimates in time bins.","marker":"[51–53]"},{"why":"Supplies the sine-Gaussian, ringdown, and Gaussian pulse waveform models used to test unmodeled sources.","marker":"[61]"},{"why":"Provides the gamma-ray burst pulse profile used to simulate the GRB light curves for TDCC.","marker":"[94]"},{"why":"Supply the electromagnetic transient event rates adopted for injecting false counterparts in the false-alarm calculation.","marker":"[67, 68]"},{"why":"Establishes the short prompt gamma-ray delay and the multi-messenger timing framework used for the delay windows in the search.","marker":"[47]"}],"fun_headline_variants":["Time-domain cross-correlation links GW background to EM flares","MC^2 pipeline pairs GW background with gamma-ray maps","Multi-band coincidence cuts false alarms in GW background search","Faint mergers recovered by multi-messenger cross-correlation","Sub-threshold mergers found via GW-EM cross-correlation"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The quoted false-alarm rates depend on the assumed rates of unrelated EM transients, on the smooth template shapes used for counterparts, and on treating detector noise alone as the comparison case rather than the real astrophysical gravitational-wave background.","fun_headline_variants_meta":{"raw":{"variants":["Time-domain cross-correlation links GW background to EM flares","MC^2 pipeline pairs GW background with gamma-ray maps","Multi-band coincidence cuts false alarms in GW background search","Faint mergers recovered by multi-messenger cross-correlation","Sub-threshold mergers found via GW-EM cross-correlation"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001184,"raw_usage":{"total_tokens":4864,"prompt_tokens":898,"completion_tokens":3966,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":514,"completion_tokens_details":{"reasoning_tokens":3884}},"tokens_in":514,"tokens_out":3966,"duration_ms":29841,"temperature":1.0,"reasoning_tokens":3884,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T18:05:04.906904+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Generate one year of simulated SGWB that includes realistic astrophysical fluctuations, time-shift EM transients so none is physically associated with a GW event, and count how many TDCC peaks survive the multi-band test; a rate far above $9.5\\times10^{-6}$ per year would show the false-alarm suppression is overestimated.","supporting_citations":[{"cited_title":"Mukherjee and J","cited_arxiv_id":null,"evidence_quote":"First proposed temporal correlations between the faint GW signal in the background and EM signals, the idea that this paper generalizes to a multi-band pipeline."},{"cited_title":"Sah and S","cited_arxiv_id":null,"evidence_quote":"Supplies the simulation method for the non-stationary SGWB used to generate the mock data in this paper."},{"cited_title":"Barsotti, L","cited_arxiv_id":null,"evidence_quote":"Gives the A+ detector sensitivity curve used for the simulated GW noise in all analyses."},{"cited_title":"Powell, D","cited_arxiv_id":null,"evidence_quote":"Supplies the sine-Gaussian, ringdown, and Gaussian pulse waveform models used to test unmodeled sources."},{"cited_title":"Norris, R.J","cited_arxiv_id":null,"evidence_quote":"Provides the gamma-ray burst pulse profile used to simulate the GRB light curves for TDCC."},{"cited_title":"GBM, INTEGRAL, I","cited_arxiv_id":null,"evidence_quote":"Establishes the short prompt gamma-ray delay and the multi-messenger timing framework used for the delay windows in the search."}],"review_version":1}