{"id":"648e909b-2677-48f2-9ee7-9b3995332113","arxiv_id":"2412.10143","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"A multi-instrument statistical analysis finds an excess of consecutive same-hemisphere solar flares separated by roughly 30 degrees of longitude and occurring within 1.5 hours, supporting the reality of sympathetic flares at a rate of about 5%.","lead":"This paper finds that solar flares that occur within 1.5 hours of each other and on the same hemisphere tend to be separated by about 30 degrees of longitude, suggesting they can trigger each other. If confirmed, the result would settle a long-standing debate about whether sympathetic flares are a real statistical phenomenon and could improve space weather forecasting.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 30° sympathetic-flare excess lacks a calibrated null test; the fit-ratio significance and peak-derived selection criteria leave the central claim contingent on an unverified background model.","rationale":"I read the paper in good faith and find substantial strengths: three independent instruments, a large sample, an online catalog of candidate events, and an explicit Appendix B attempt to control for the underlying active-region distribution. The PFSS footpoint-separation analysis is a plausible and interesting physical context for a 30° scale. However, the load-bearing step is the statistical claim that the Δφ≈30° excess at w≤1.5 h is real. Everything downstream—the 5% occurrence rate, the hemispheric and solar-cycle variations, the propagation-velocity distribution, and even the comparison with PFSS footpoint separations—inherits the validity of this step. The current evidence is a ratio of two fits to one histogram, which is not a calibrated significance test. The Appendix B control addresses a different object (nearest active regions) and therefore cannot rule out a null in which independent flares occur in clustered active regions with varying flare productivity. The same-active-region exclusion threshold is not specified, and the candidate sample is selected using the same fitted peak, making the occurrence-rate estimate circular. The transequatorial deficit, while interesting, rests on a small sample without significance quantification. A permutation or Poisson null test is feasible with the released catalog and would directly settle whether the excess is a flaring interaction or an artifact of the spatial distribution of flare locations. I therefore keep the reader's CONDITIONAL verdict unchanged, with the condition being the addition of a proper null-hypothesis test and explicit reporting of sample sizes and selection thresholds.","tokens_in":18753,"tokens_out":4667,"duration_ms":57255,"concrete_test":"Generate 1,000 null catalogs by randomly permuting the observed flare start times among the observed flare coordinates within each hemisphere (or, better, by drawing independent flare times from a time-varying Poisson rate fitted to the catalog), preserving locations, hemispheres, and detection selection. For each surrogate, compute the Δφ distribution for pairs with w≤1.5 h, fit Eq. 3, and record the Gaussian amplitude/background ratio near 30°. If the observed ratio is exceeded in more than 5% of surrogates, the claimed sympathetic excess is consistent with the spatial distribution of flare locations and does not require an interaction. Also rerun with a stated same-active-region exclusion threshold and bootstrap the fit ratio to report a confidence interval.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim—a real excess of hemispheric flare pairs at Δφ≈30° and w≤1.5 h—is supported only by comparing two fits (Eq. 3, with and without a Gaussian component) to the same histogram. This ratio is not a significance test: it has no null distribution, no propagated uncertainty, and no correction for the fact that the peak position, width, and the 1.5 h threshold are all read off the same data. The Appendix B control does not close this gap: it compares the distribution of nearest active-region separations, not the distribution of consecutive flare pairs expected under a null of independent flares, so the excess could in principle be produced by the spatial clustering of flare-productive active regions rather than by physical sympathy. The same-active-region exclusion threshold is never stated, and the 581-pair/7% 'occurrence rate' is counted from a sample defined by the very peak being tested, making that rate circular. The transequatorial 'unsympathetic' deficit (Fig. 8) is additionally based on a small sample with no significance quantification, so it cannot independently validate the 30° scale.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper analyzes consecutive solar flares from the SDO/AIA, RHESSI, and Solar Orbiter/STIX catalogs, computing waiting times and angular separations between successive events. It reports an excess of same-hemisphere flare pairs separated by approximately 30 degrees in Carrington longitude and triggered within less than 1.5 hours, which it interprets as a statistical signature of sympathetic flares, with an occurrence rate of roughly 5% (7% for AIA, 3.5% for RHESSI, 3.9% for STIX). It also identifies a deficit of transequatorial flare pairs at latitude separations of 25-30 degrees when the longitude separation is less than 5 degrees, termed 'unsympathetic flares.' The paper further examines variations with solar-cycle phase and hemisphere, propagation velocities, and flare magnitude correlations, and proposes a potential field source surface (PFSS) interpretation in which the longitudinal footpoint separation of closed coronal field lines peaks near 21 degrees. Appendix B uses the distribution of nearest active-region separations to argue that the flare excess is not caused by the underlying active-region distribution.","tokens_in":18967,"tokens_out":5061,"duration_ms":51492,"significance":"If the central detection is real, the paper would provide a long-sought statistical confirmation of sympathetic solar flares and identify a characteristic angular scale for flare-flare interactions, with implications for coronal connectivity and flare trigger mechanisms. The multi-instrument comparison across solar cycles and the public release of candidate event catalogs are valuable assets. However, the statistical foundation of the detection is currently insufficient to support the strength of the claims. The significance assessment rests on an ad hoc fit-ratio rather than a null-hypothesis test, the candidate selection criteria are derived from the same data used to establish the peak, and the active-region control in Appendix B is not the relevant null for consecutive flare pairs. The transequatorial 'unsympathetic' deficit is based on a very small sample with no significance quantification. These issues are load-bearing because the paper's conclusions, including the proposed 30-degree scale and the 5% occurrence rate, depend directly on them. The paper's significance is therefore conditional on a substantial reanalysis with proper statistical controls.","major_comments":[{"comment":"The claimed statistical significance of the ~30-degree peak is based on the ratio of two fits to the same histogram: one including the Gaussian term in Eq. (3) and one excluding it. This ratio has no null distribution, no propagated uncertainty, and no correction for the fact that the peak position, width, and the 1.5-hour threshold are all read off the same data. The statement 'we find an average value of around 3 times more events within the peak' is not a significance measure. The authors need to replace this with a formal hypothesis test against a null model that preserves the active-region and flare-rate distributions while randomizing the inter-AR coupling (for example, by shuffling flare longitudes/latitudes or times within the observed AR schedule), and to report a p-value or confidence interval, along with a multiple-testing correction across the waiting-time bins, cycle phases, hemispheres, and instruments.","section":"Section 2.3, Eq. (3)"},{"comment":"The candidate sympathetic flares are selected using criteria (1) Delta-phi in [Delta-phi_0 +/- 2 sigma] and (2) w <= 1.5 hours, where Delta-phi_0 and sigma are fitted to the same histogram that is then used to compute the occurrence rate. This makes the reported 7% (AIA) and ~5% average circular. An out-of-sample or hold-out validation, or at least a systematic sensitivity analysis over a range of thresholds, is required. In addition, the threshold used to exclude pairs from the same active region is never stated in Section 2.3; it must be specified and its influence on the detected peak and on the candidate list must be quantified.","section":"Section 2.3, Figures 4-5"},{"comment":"The control analysis in Appendix B compares the distribution of angular separations between nearest active regions (or nearest transequatorial active regions), not the distribution of consecutive flare pairs expected under a null of independent flaring. A null in which flares occur independently within each active region with rates proportional to the observed flaring rates, but with no inter-AR coupling, could produce a surplus at ~30 degrees simply from the longitudinal clustering of flare-productive active regions. The current control therefore does not support the claim in Section 2.3 that 'this peak does originate from the flares and not the underlying spatial distribution of active regions.' The authors should construct the relevant null for the consecutive-flare-pair statistic.","section":"Appendix B, Figures B.2-B.3"},{"comment":"The transequatorial deficit ('unsympathetic flares') is a central new claim, but it is based on a very small sample (the text itself notes 'despite the small sample size') and has no significance quantification beyond √N error bars. The deficit is then used in Section 6 to propose a characteristic length scale of ~30 degrees. The authors should provide a significance test for this deficit, ideally using the same null model as for the hemispheric excess, and should report the sample size and confidence interval. Without this, the deficit is not established as a robust phenomenon.","section":"Section 3.2, Figures 7-8"},{"comment":"The background model in Eq. (3) is ad hoc, and the paper does not report goodness-of-fit, parameter uncertainties, or a comparison with alternative background parameterizations (e.g., a power law or a nonparametric estimate). The detection of the Gaussian excess is therefore conditional on the arbitrary functional form chosen for the background. The sensitivity of the peak location, width, and amplitude to the background model should be assessed, and the fit uncertainties should be propagated into the significance estimate.","section":"Section 2.3, Eq. (3), background model"}],"minor_comments":[{"comment":"Please specify whether 'log' denotes the natural logarithm, and clarify that the fit is performed on binned probability densities; the choice of bin width and its effect on the fitted parameters should be stated.","section":"Equation (3)"},{"comment":"The claim that 'these results are not significantly impacted by splitting the cycle at a slightly different point in time' is not demonstrated; either provide the sensitivity analysis or remove the claim.","section":"Section 3.1"},{"comment":"The statement that the RHESSI decaying-phase peak is 'neither representative nor statistically significant' is qualitative; provide quantitative support, such as a p-value or confidence interval, or state the sample size.","section":"Section 4.1"},{"comment":"The conclusion that sympathetic flares are 'unambiguously present' is stronger than what the current statistical analysis supports; please temper the abstract and conclusion until the significance analysis is completed.","section":"Section 6"},{"comment":"Several references have corrupted author lists, for example 'Démoulin, L. G. B. . C. H. M. ... 2000' and 'Harrison, G. P. . R. A. & Harrison, G. P. . R. A. 1990'; these need to be corrected.","section":"References"},{"comment":"The PFSS-derived peak of 21 +17/-15 degrees is only marginally consistent with the observed 30-degree excess; the large uncertainty should be acknowledged more explicitly when proposing the magnetic-connectivity interpretation.","section":"Section 6, PFSS interpretation"}],"recommendation":"major_revision","confidential_remarks":"The paper addresses a long-standing question and the multi-instrument approach is commendable, but the central detection is currently not supported by rigorous statistics. The fit-ratio significance and circular candidate selection are the main problems, and Appendix B does not provide the needed null. I would encourage the editor to seek a revision that includes a proper null-hypothesis test; if the authors cannot provide one, the paper should not be accepted. The transequatorial deficit needs separate attention because it is a new and weaker claim."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's the short version: this paper likely identifies the same ~30° longitudinal excess in short-waiting-time flare pairs that earlier work had hinted at, now seen in three independent datasets, plus a new transequatorial deficit. But the abstract's 'strong statistical evidence' overstates what is shown: the significance estimate is a fit ratio with no null distribution, and the candidate selection is fitted to the same histogram used to count events.\n\nWhat's good: three instruments covering two cycles; careful coordinate work; a public catalog; the cycle-phase and hemisphere dependence is a genuine addition; the PFSS footpoint distribution is a testable suggestion for the 30° scale. The authors cite the earlier ~30° detections honestly, so the novelty claim is appropriately moderate.\n\nThe soft spots are all in the statistics. Section 2.3 quantifies the peak by the ratio of two fits to the same data (Eq. 3 with and without a Gaussian). That ratio has no null distribution, no p-value, and no multiple-testing correction. The Gaussian parameters and the 1.5-hour threshold are read off the same plot, and the candidate sample is then used to compute the occurrence rate, making the 5–7% claim partly circular. The same-active-region exclusion threshold is never stated, which matters because the peak sits near the angular size of large active regions.\n\nAppendix B is meant to rule out an active-region clustering origin, but it compares the distribution of nearest active-region separations, not the distribution of consecutive flare pairs you would get under a null with clustered flare rates. That control is suggestive but not decisive. The transequatorial deficit is a genuinely new claim, but it rests on a small sample with no significance quantification, so it cannot carry much weight yet.\n\nNone of these are fatal. The peak is visually present in all three instruments and fades with increasing waiting time, which is what you'd expect from a real interaction. I suspect the detection is real. But the paper needs a proper null test—shuffle flare times or locations while preserving the active-region and flare-rate map, and see whether a 30° peak appears in synthetic pairs—before the significance claim is credible.\n\nWho should read it: solar physicists working on flare statistics and space weather. I'd send it to a referee, but the referee should demand that null test before publication. I'd hold off citing it as an established result until then.","headline":"Plausible multi-instrument confirmation of the ~30° sympathetic-flare excess, but the significance is not established: no calibrated null test, and the selection criteria are read off the same data.","tokens_in":19524,"tokens_out":5207,"would_cite":false,"duration_ms":47703,"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":"Sympathetic solar flares are real: flares pair up at about 30 degrees of longitude within 1.5 hours.","keywords":["Sun: flares","Sun: activity","sympathetic flares","unsympathetic flares","flare waiting times","magnetic connectivity","PFSS extrapolation"],"falsifier":"Reshuffle the start times of the observed flares (or, separately, their longitudes) while keeping the flare locations and the active-region distribution fixed, and rebuild the waiting-time versus separation histograms; if a peak near 30 degrees at waiting times under 1.5 hours still appears in the scrambled data, the claimed sympathetic-flare signal is an artifact of active-region clustering rather than a flare-triggering interaction.","tokens_in":18496,"feed_emoji":"☀️","tokens_out":10253,"duration_ms":93376,"temperature":0.7,"pith_summary":"This paper sets out to establish that solar flares can trigger other flares at a preferred angular separation: same-hemisphere flares separated by roughly $30^\\circ$ in Carrington longitude and starting less than 1.5 hours apart occur more often than expected. Using consecutive-flare pairs from SDO/AIA, RHESSI, and Solar Orbiter/STIX over cycles 24 and 25, the authors find the excess in all three datasets, at an average occurrence rate near 5%, and show that its position shifts with solar cycle phase and hemisphere. They also report a complementary deficit, fewer transequatorial flare pairs at $25^\\circ$--$30^\\circ$ latitude separation when longitudes nearly coincide, which they name unsympathetic flares. Control distributions of active-region positions and PFSS-derived magnetic footpoint separations are used to argue that the angular scale comes from magnetic connectivity between active regions rather than from where active regions simply sit. If the detection holds up, it would make sympathetic flaring a quantitative, reproducible feature of solar activity rather than a collection of case studies.","feed_headline":"Flares on the Sun trigger each other at a 30-degree spacing","feed_subtitle":"Three instruments show ~5% of flares trigger a second flare about 30 degrees away within 1.5 hours.","key_machinery":"The analysis is carried by pair statistics of consecutive flares: for each pair the waiting time $w = T_{i+1} - T_i$ and the Haversine angular separation $\\lambda$, with its longitudinal and latitudinal components $\\Delta\\phi$ and $\\Delta\\theta$, are computed in Carrington coordinates so that solar rotation does not inject spurious correlations. The signal is isolated by fitting Equation 3, a Gaussian peak added to an exponential-decay background plus constant, to the longitudinal-separation histogram of same-hemisphere pairs in short waiting-time bins; the Gaussian selects candidate sympathetic pairs, and a fit without the Gaussian supplies the background ratio used as the significance estimate. Two separate control constructions carry the argument that the signal is not geometric: the distribution of separations between nearest simultaneously present NOAA active regions (Appendix B), and the distribution of footpoint separations of closed coronal field lines traced with potential-field source-surface (PFSS) extrapolations from GONG magnetograms, which peaks at a comparable scale. The PFSS machinery is what converts the observed $30^\\circ$ angular scale into a statement about magnetic connectivity between active regions.","core_discovery":"The paper's central claim is that consecutive flares in the same hemisphere show a statistically significant excess at a longitudinal separation $\\Delta\\phi \\approx 31^\\circ \\pm 10^\\circ$ when the waiting time is less than about 1.5 hours. The excess is modeled by adding a Gaussian component to an exponential background (their Equation 3), and the ratio between the two fits implies about three times as many events in the peak as the background alone would give. The effect appears in the cumulative distribution only at short waiting times and is reproduced by RHESSI and STIX data, with peak positions of roughly $22^\\circ$--$34^\\circ$ depending on instrument and cycle phase. For transequatorial pairs with $\\Delta\\phi \\lesssim 5^\\circ$, the paper finds a deficit around $\\Delta\\theta \\approx 25^\\circ$--$30^\\circ$ that is absent from the underlying active-region distribution; this is the proposed unsympathetic-flare effect. Finally, PFSS extrapolations give a footpoint longitudinal-separation distribution peaking at about $21^{+17}_{-15}$ degrees, broadly consistent with the flare-pair peak, which the authors offer as the structural reason why 30 degrees is special. Across instruments, candidate sympathetic pairs involve about 7%, 3.5%, and 3.9% of flares, for a mean occurrence rate near 5%.","pith_inferences":["Editorial extension: the same consecutive-pair histogram method could be applied to coronal mass ejections and filament eruptions; a ~30-degree excess there would show that the coupling scale is generic to eruptive events, not specific to X-ray flares.","Editorial extension: the unsympathetic-flare deficit makes a concrete prediction for magnetohydrodynamic simulations, namely that two active regions separated by 25--30 degrees in latitude and nearly aligned in longitude should show a reduced probability that the second region erupts after the first one flares.","Editorial extension: because the peak position shifts with cycle phase and hemisphere, the ~30-degree scale may track the latitude separation of the two activity belts rather than a fixed magnetic-connectivity length; comparing the fitted peak against the instantaneous width of the sunspot butterfly diagram would separate those explanations."],"forward_implications":["If the excess is real, roughly 5% of solar flares occur as part of a sympathetic pair, so flare statistics and space-weather forecasting models must include inter-active-region triggering rather than treating flares as independent events.","The peak position changes with solar cycle phase and hemisphere, so the coupling mechanism is modulated by the Sun's large-scale magnetic configuration, not fixed at a universal 30-degree constant.","The implied propagation speeds for candidate sympathetic pairs are cut off below about 80 km/s, with a maximum around 104 km/s, well above the 45 km/s geometric minimum; this leaves slow EUV waves or mass surges as plausible mediators but excludes very slow mechanisms.","The reported transequatorial deficit at 25--30 degrees of latitude for nearly aligned pairs implies that a flare can also suppress later flaring at a comparable angular scale, a phenomenon the authors name unsympathetic flares.","The PFSS footpoint-separation peak around 21 degrees, with a wide spread, is broadly consistent with the flare-pair peak, supporting the proposal that magnetic connectivity between active regions sets the preferred angular scale."],"supporting_citations":[{"why":"Supplies the primary SDO/AIA 131 Å flare catalog used for the main distributions.","marker":"van der Sande et al. (2022)"},{"why":"Provides the RHESSI flare list and instrument description used as the second dataset.","marker":"Lin et al. (2002)"},{"why":"Provides the STIX instrument and flare-list reference used as the third dataset.","marker":"Krucker et al. (2020)"},{"why":"Earlier statistical excess of pairs of active regions closer than 30 degrees; this paper reproduces and extends that result.","marker":"Fritzova-Svestkova et al. (1976)"},{"why":"Earlier deviation from random flares for pairs separated by less than 35 degrees; supports the angular scale.","marker":"Harrison & Harrison (1990)"},{"why":"Earlier SDO/AIA-based finding of increased flare rate within 4 hours beyond 20 degrees; comparison baseline for the new detection.","marker":"Schrijver et al. (2015)"},{"why":"pfsspy code used to trace PFSS field lines for the footpoint-separation interpretation.","marker":"Stansby et al. (2020)"},{"why":"GONG magnetograms provide the photospheric boundary conditions for the PFSS extrapolations.","marker":"Harvey et al. (1996)"},{"why":"Heliophysics Events Knowledgebase used to build the active-region control distributions in Appendix B.","marker":"Hurlburt et al. (2012)"}],"fun_headline_variants":["Sun's flares trigger partners 30 degrees away","Sympathetic flares: 5% of solar flares spark a twin 30° away","Flares pair up at 30-degree separation on Sun","Study finds solar flares often trigger each other at 30° apart","Sun's flaring twins: 30-degree spacing revealed"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the excess of flare pairs at about 30 degrees of longitude within 1.5 hours is caused by flare-to-flare interaction rather than by the spatial and temporal clustering of active regions; the significance estimate compares two fits to the same histogram, and no reshuffled-null test that randomizes flare times or locations while preserving the active-region pattern is presented.","fun_headline_variants_meta":{"raw":{"variants":["Sun's flares trigger partners 30 degrees away","Sympathetic flares: 5% of solar flares spark a twin 30° away","Flares pair up at 30-degree separation on Sun","Study finds solar flares often trigger each other at 30° apart","Sun's flaring twins: 30-degree spacing revealed"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000231,"raw_usage":{"total_tokens":1581,"prompt_tokens":1133,"completion_tokens":448,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":749,"completion_tokens_details":{"reasoning_tokens":361}},"tokens_in":749,"tokens_out":448,"duration_ms":4756,"temperature":1.0,"reasoning_tokens":361,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T16:17:22.746996+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Reshuffle the start times of the observed flares (or, separately, their longitudes) while keeping the flare locations and the active-region distribution fixed, and rebuild the waiting-time versus separation histograms; if a peak near 30 degrees at waiting times under 1.5 hours still appears in the scrambled data, the claimed sympathetic-flare signal is an artifact of active-region clustering rather than a flare-triggering interaction.","supporting_citations":[{"cited_title":"E., & Gagnon , R","cited_arxiv_id":null,"evidence_quote":"Supplies the primary SDO/AIA 131 Å flare catalog used for the main distributions."},{"cited_title":"P., Dennis , B","cited_arxiv_id":null,"evidence_quote":"Provides the RHESSI flare list and instrument description used as the second dataset."},{"cited_title":"C., Svestka, Z., et al","cited_arxiv_id":null,"evidence_quote":"Earlier statistical excess of pairs of active regions closer than 30 degrees; this paper reproduces and extends that result."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Earlier deviation from random flares for pairs separated by less than 35 degrees; supports the angular scale."},{"cited_title":"J., Higgins, P","cited_arxiv_id":null,"evidence_quote":"Earlier SDO/AIA-based finding of increased flare rate within 4 hours beyond 20 degrees; comparison baseline for the new detection."}],"review_version":1}