{"id":"132b3c43-1774-4a2b-a1d7-26eb139be3a7","arxiv_id":"2412.14907","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"The half-year 2017 cosmic ray depression was caused by repeated CMEs, SIRs, and CIRs, with the largest contributions from four CMEs originating in one long-lived active region.","lead":"This paper argues that the long 2017 dip in galactic cosmic rays measured by AMS-02 was produced by a series of solar storms and high-speed solar wind streams, not by a change in the Sun's global magnetic field. It identifies four powerful eruptions from one persistent active region as the main drivers.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The paper's largest single impact, a 24.4% decrease attributed to the X9.3 CME in the 1.0-1.16 GV bin, is not directly supported by AMS-02 daily data because that bin has a 09-15 Sep data gap overlapping the CME interval.","rationale":"I read the paper as claiming that the 2017 half-year GCR depression has a solar origin and is produced by repeated transient structures, with the largest contributions coming from four CMEs associated with a recurrent active magnetic source. The evidence for the solar origin (rigidity dependence, Figures 2 and 3) is solid and consistent with modulation physics. The exclusion of a 1974-style minicycle is reasonably supported by the co-temporal increase in eruptive activity once the depression begins. The most fragile part is the event-by-event attribution in Table 1 and Figure 6, because it is entirely qualitative. Within that attribution, the single most concrete weakness is the X9.3 CME's 24.4% decrease: the 1.0-1.16 GV daily data are missing for 09-15 September, exactly the period when this CME is supposed to be acting. The paper does not state how the 24.4% was derived across the gap, so the largest impact in the central claim is not verifiable from the data presented. This is not an accusation of manipulation; it is an unaddressed gap in the evidence trail. A straightforward re-computation from the published daily data would settle it. If the 24.4% value holds, the conclusion is strengthened; if not, the ranking of events changes and the 'enhanced substantially' assertion needs to be softened. I therefore agree with the reader's CONDITIONAL verdict and do not think the paper should be rejected; the missing data point is a condition that should be checked, not a demonstrated error. The reader's weakest_assumption is broader, namely the lack of a transport model or statistical significance test; my concern is a concrete instance of that same causal-attribution problem, so my agreement is partial.","tokens_in":15038,"tokens_out":8430,"duration_ms":68643,"concrete_test":"Obtain the AMS-02 daily proton flux values for the 1.0-1.16 GV bin for September 2017 from Aguilar et al. (2021). Identify the date of the minimum flux and the values on 7-8 September and 16-17 September. Recompute the % decrease using only available data, explicitly stating the baseline and the minimum date. If the minimum falls in the 09-15 September gap, or if the recomputed decrease differs from 24.4% by more than ~3 percentage points, the Table 1 entry is not directly measured and must be revised before the X9.3 CME can be ranked as the largest single impact. As a secondary check, repeat the attribution for the 2.97-3.29 GV bin (where data are available) and compare the rigidity scaling with Figure 8; if the largest event no longer follows the same power-law trend, the event ranking is not robust.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim in Section 4 is that the half-year GCR depression was caused by a combination of CIRs, SIRs and CMEs and was enhanced substantially by strong CMEs from a single active magnetic source, with the X9.3 CME from AR12673 providing the largest impact. However, Table 1 lists this CME as arriving 07-Sep 22:28 to 10-Sep 21:00 with a 24.4% decrease in the 1.0-1.16 GV bin, while the Figure 6 caption states that daily AMS-02 proton measurements below 2.97-3.29 GV are unavailable from 09 to 15 September 2017. Thus the interval of the CME's effect overlaps the data gap exactly in the rigidity bin used for the quantitative attribution. The paper does not explain how the 24.4% value was computed across this gap, for example by using the last pre-gap point, interpolation, or the 2.97-3.29 GV bin. Because this event is one of the four CMEs argued to be the largest impacts from the same magnetic source, an inflated or unverifiable decrease would weaken the conclusion that the depression was enhanced substantially by these CMEs. In addition, the preceding M5.5 CME (06-08 Sep) is listed with no % decrease, showing that the event-by-event decomposition in this period is ambiguous. No uncertainties or statistical significance tests are attached to any of the % decreases in Table 1, so the visual matching in Figure 6 cannot distinguish the claimed CME effects from ordinary daily variability or overlapping CIR/SIR effects.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper investigates the cause of the half-year depressions observed in AMS-02 galactic cosmic ray proton fluxes during the second half of 2017. It compares two hypotheses: (1) a 'minicycle' produced by a global change in the solar magnetic field, of the type argued for the 1974 event, and (2) a combined effect of coronal mass ejections (CMEs), stream interaction regions (SIRs), and corotating interaction regions (CIRs). The authors first establish a solar origin from the rigidity dependence of the depression, then argue that minicycle-like signatures before the event are not accompanied by the absence of eruptive activity required by the 1974 definition. Using catalogued CME, SIR, and CIR arrival times, together with daily AMS-02 proton fluxes in low-rigidity bins and Oulu neutron monitor data, they attribute the depression to a sequence of recurrent CMEs and interaction regions, and conclude that the four largest impacts came from CMEs that all originated from a single persistent, unusually active magnetic source that they identify as AR12665, AR12673, and AR12685 in successive rotations.","tokens_in":15373,"tokens_out":3782,"duration_ms":31700,"significance":"If the conclusion holds, the paper resolves an important puzzle: the 2017 event is shown not to be a 1974-style minicycle but a sustained sequence of solar eruptive activity, with a single active complex as the dominant driver. The analysis is valuable because it brings together publicly available catalogues (Richardson & Cane, HELIO4CAST, STEREO, SolarMonitor) and independent datasets (AMS-02 and neutron monitors), and it makes a falsifiable claim about the role of a recurring active region. The paper does not rely on a fitted transport model, and the rigidity-dependent power-law forms in Figure 8 are presented as descriptive rather than as fitted parameters, which limits the risk of circularity. The main weaknesses are quantitative: several % decreases in Table 1 are read off by eye without uncertainties, and the largest attributed impact overlaps a known AMS-02 data gap. These issues, while locally fixable, currently weaken the quantitative support for the central claim that the depression was 'enhanced substantially' by the four identified CMEs.","major_comments":[{"comment":"The 24.4% decrease attributed to the X9.3 CME (07-09 to 10-09 September, in the 1.0-1.16 GV bin) overlaps the 09-15 September data gap that the Figure 6 caption states exists for daily AMS-02 measurements below the 2.97-3.29 GV bin. The paper does not explain how this value was computed across the gap (e.g., by using the last pre-gap point, interpolation, or a higher rigidity bin). Since this event is one of the four CMEs identified as having the largest impacts, the quantitative attribution is load-bearing. Please state the exact method used and provide an uncertainty estimate, or replace the value with one derived from a rigidity bin that is not affected by the gap.","section":"§3.2, Table 1 and Figure 6 caption"},{"comment":"The % decreases in Table 1 are read from the daily flux by eye and are presented without uncertainties or significance tests, and the M5.5 CME (06-08 September) is listed with no % decrease, which makes the event-by-event decomposition in the same period ambiguous. A reproducible definition of the % decrease (e.g., baseline level minus local minimum, with a stated time window) and error bars from the AMS-02 flux uncertainties are needed so that the reader can judge whether the claimed effects exceed ordinary daily variability or overlap with adjacent SIR/CIR effects.","section":"§3.2, Table 1 and §3.3, Figure 8"},{"comment":"The central conclusion that AR12665, AR12673, and AR12685 are manifestations of 'the same magnetic source' persisting for five Carrington rotations is asserted on the basis of labels and locations, but no quantitative tracking is provided (e.g., Carrington longitudes, magnetic-flux emergence histories, or coronal-hole connectivity). Because the uniqueness of this source is the paper's most distinctive claim, please present the supporting evidence in a table or figure; otherwise the statement that all four CMEs share a single source is an unsupported inference.","section":"§3.2 and §4"}],"minor_comments":[{"comment":"The name is 'Bartels rotation', not 'Bartel rotation'; please correct this typo that appears in the abstract and body.","section":"Throughout"},{"comment":"In the sentence 'Carringtion rotations as the markers of CIR impact', 'Carringtion' should be 'Carrington'.","section":"§3.1"},{"comment":"The phrase 'co-rotating/stream interactions regions' should be 'co-rotating/stream interaction regions' (singular 'interaction').","section":"Abstract"},{"comment":"The sentence 'Further corroboration be found in Figure 4 (b)' is missing 'can'; it should read 'Further corroboration can be found in Figure 4 (b)'.","section":"§2.1, Figure 4 discussion"},{"comment":"The power-law fits in panels (b) and (d) are useful as indicators of trends, but they are shown without uncertainties or goodness-of-fit statistics; please state that they are descriptive fits only, or add a brief caption note.","section":"Figure 8"},{"comment":"The LaTeX artifact 'T able 1' should be corrected to 'Table 1'.","section":"§3.2, Table 1 header"}],"recommendation":"major_revision","confidential_remarks":"The paper is well within the scope of astro-ph.SR and addresses a topic of current interest. The central hypothesis is plausible and the datasets are appropriate. The main risk is the quantitative support for the largest impact and the identification of a single active source; both are fixable with additional analysis and presentation. I do not see evidence of misattribution or circular use of fitted parameters. The manuscript would benefit from a clearer statement of how the Table 1 decreases were defined and from a more rigorous demonstration that the three AR labels indeed belong to one recurring magnetic complex."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a credible catalog-based explanation of a real, well-observed event, and I'd send it out. The new result is the attribution—no one before tied the 2017 half-year GCR depression to a combination of CIRs, SIRs, and CMEs and specifically to one long-lived active region that produced the largest impacts. That is worth having. The paper also does a clean job ruling out a 1974-style minicycle: it shows B and tilt-angle changes before the depression are not associated with flux dips, and that the co-temporal CME activity is too strong to ignore. The rigidity dependence argument is sound and independently confirms a solar origin.\n\nThe soft spots are localized but not trivial. The stress-test finds a real problem: Table 1 assigns a 24.4% decrease in the 1.0-1.16 GV bin to the X9.3 CME, but the AMS-02 daily data below 2.97-3.29 GV are missing from 09 to 15 September, which overlaps the listed arrival interval. The paper never states how that number was computed. Since it is the largest impact and carries the 'substantially enhanced' conclusion, this should be fixed by recomputing with an explicit gap treatment or by removing the number and keeping the qualitative claim. The M5.5 CME with no % decrease shows the event decomposition in that period is genuinely ambiguous. All % decreases in Table 1 are read by eye without uncertainties, and there is no statistical test comparing them to ordinary daily variability.\n\nNone of this sinks the paper's central pattern: a five-rotation depression with CIR dips and CME/SIR interruptions is clearly visible in the plotted data even without the 24.4% point. The causal argument does not rest on fitted parameters, and the use of external catalogues plus independent AMS-02 and neutron monitor data is solid. Citation pattern looks normal; they credit Aguilar et al. 2021 and the relevant SIR/CIR literature.\n\nI would accept this for peer review and ask for a targeted revision: explicit method for the September gap, uncertainties on the % decreases, and a clearer statement of why the four CMEs are from the same magnetic source. I wouldn't cite the per-event numbers until that is done, but I'd bring it to reading group as a useful case study.","headline":"A plausible catalog-based explanation of the 2017 GCR depression, but the largest per-event decrease sits on a data gap and needs an explicit recomputation.","tokens_in":15931,"tokens_out":3129,"would_cite":false,"duration_ms":27383,"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 argues that the half-year 2017 depression in galactic cosmic ray proton flux was produced by a sequence of solar wind interaction regions and coronal mass ejections, with four CMEs from one unusually active magnetic source doing…","keywords":["galactic cosmic rays","solar modulation","Forbush decrease","coronal mass ejection","co-rotating interaction region","stream interaction region","solar active regions","AMS-02"],"falsifier":"Check the in-situ spacecraft data at 1 AU for the times of the four largest flux drops: if there are no corresponding ICME shocks and magnetic ejecta, or if a heliospheric transport simulation including only the catalogued CIRs and SIRs reproduces the half-year depression without the four CMEs, the central claim would be contradicted.","tokens_in":14836,"feed_emoji":"☀️","tokens_out":6622,"duration_ms":46722,"temperature":0.7,"pith_summary":"The paper asks why the galactic cosmic ray proton flux measured by AMS-02 fell by about 17 percent over the second half of 2017, a depression lasting nearly half a year, far longer than any individual Forbush decrease. It argues that the drop was not a “minicycle” of the global solar magnetic field, but the accumulated effect of stream interaction regions and coronal mass ejections, with the four largest drops produced by CMEs from a single unusually active magnetic source that survived five Carrington rotations. The claim matters because it shows that a long cosmic ray modulation event near solar minimum can be built sequentially from ordinary solar eruptions, and it pins the event to specific active regions that could be monitored.","feed_headline":"Four CMEs built the 2017 half-year cosmic ray dip","feed_subtitle":"AMS-02 protons fell half a year because repeated eruptions from one active region blocked cosmic rays.","key_machinery":"The load-bearing object is an event-by-event catalogue table (Table 1) that lists the arrival times at 1 AU of Carrington rotations, stream interaction regions, co-rotating interaction regions and coronal mass ejections during June to December 2017, with flare strength and active-region origin for each CME. The table is matched visually to the daily AMS-02 proton flux (Figure 6) and to Oulu neutron monitor data (Figure 7), so that each measured decrease is assigned to a listed structure. A second piece of machinery is the rigidity-resolved analysis: for the two strongest periods, the percentage decrease follows a power law in rigidity up to about 30 GV and then flattens, which is the paper's diagnostic that the modulation is solar rather than extra-heliospheric.","core_discovery":"On the paper's own terms, the central discovery is that the half-year depression observed in AMS-02 galactic proton flux in the second half of 2017, which initially resembled a very long Forbush decrease, was caused by a combination of co-rotating interaction regions, stream interaction regions and coronal mass ejections, and was substantially enhanced by four strong CMEs from the same magnetic source. That source, labeled AR12665, AR12673 and AR12685 on different rotations, persisted for five Carrington rotations and produced the most active flaring of Solar Cycle 24, including the X9.3 flare of September 2017. Although there were minicycle-like rises in the heliospheric magnetic field magnitude and changes in the heliospheric current sheet tilt before the depression, the authors conclude that the depression itself cannot be attributed to a 1974-style minicycle because enhanced eruptive activity was co-temporal with it. The solar origin is supported by the rigidity dependence of the depression, which is strong below roughly 30 GV and weak above it.","pith_inferences":["A control experiment the paper does not run: remove the four strong CMEs from a transport simulation and ask how much of the half-year depression remains; the paper's claim predicts a much weaker and shorter dip from CIRs and SIRs alone.","The same matching method could be applied to AMS-02 helium, electron and positron fluxes for 2017 to test whether charge-sign dependent modulation supports the CME/IR attribution.","The paper's closing remark implies that a broader notion of “minicycle” may still be needed to explain why the Sun produced its most active region of the cycle near minimum; that would connect the event to dynamo studies rather than purely to weather-like eruptions."],"forward_implications":["The 2017 event should be counted as a solar-activity event, not a recurrence of the 1974 minicycle, so comparisons of solar-cycle minima should treat it differently.","Long, half-year cosmic ray depressions near solar minimum can be produced by a sequence of CMEs and interaction regions from a persistent active region, without requiring a global dynamo change.","Tracking a magnetic source as it crosses the disk over multiple rotations, as AR12665, AR12673 and AR12685 did, could provide early warning for similar prolonged cosmic ray depressions.","Because the depression opposes the general rise of proton flux toward solar minimum, the same activity at another phase of the cycle would be less visible in the flux record."],"supporting_citations":[{"why":"Supplies the AMS-02 daily and Bartel-rotation-averaged proton flux data that define the half-year depression.","marker":"[Aguilar et al. 2021]"},{"why":"Defines the 1974 minicycle by the absence of enhanced CME activity; the paper compares and rejects this explanation for 2017.","marker":"[Wibberenz & Cane 2000]"},{"why":"Provides the catalogue of ICME and interaction-region arrival times at 1 AU used in Table 1.","marker":"[Cane & Richardson 2003]"},{"why":"Extends the catalogue of solar wind stream and ejecta events used to identify SIRs and CMEs.","marker":"[Richardson & Cane 2010]"},{"why":"Supplies STEREO data for CME shock arrival times at 1 AU.","marker":"[Jian et al. 2018]"},{"why":"Supplies STEREO SIR times used to identify stream interaction regions.","marker":"[Jian et al. 2019]"},{"why":"Provides HELIO4CAST catalogue information used to list CME arrival times.","marker":"[Möstl et al. 2020]"},{"why":"Documents the X9.3 flare and AR12673's eruptive character, used to mark the strongest CME impacts.","marker":"[Hou et al. 2018]"}],"fun_headline_variants":["Four CMEs from one active region made the 2017 cosmic ray dip","Repeated CMEs from one source caused the 2017 half-year cosmic ray drop","Half-year cosmic ray dip traced to four CMEs from one active region","2017's long cosmic ray depression came from recurring CMEs and CIRs","One active region's four CMEs drove the 2017 half-year cosmic ray dip"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"Each measured drop in proton flux is attributed to a catalogued CME, SIR, or CIR solely by aligning arrival times with dips in the flux time series by eye, with no transport model or statistical test to rule out coincidence or unlisted modulators.","fun_headline_variants_meta":{"raw":{"variants":["Four CMEs from one active region made the 2017 cosmic ray dip","Repeated CMEs from one source caused the 2017 half-year cosmic ray drop","Half-year cosmic ray dip traced to four CMEs from one active region","2017's long cosmic ray depression came from recurring CMEs and CIRs","One active region's four CMEs drove the 2017 half-year cosmic ray dip"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000481,"raw_usage":{"total_tokens":2436,"prompt_tokens":1057,"completion_tokens":1379,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":673,"completion_tokens_details":{"reasoning_tokens":1269}},"tokens_in":673,"tokens_out":1379,"duration_ms":8609,"temperature":1.0,"reasoning_tokens":1269,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T11:48:05.863889+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Check the in-situ spacecraft data at 1 AU for the times of the four largest flux drops: if there are no corresponding ICME shocks and magnetic ejecta, or if a heliospheric transport simulation including only the catalogued CIRs and SIRs reproduces the half-year depression without the four CMEs, the central claim would be contradicted.","supporting_citations":[{"cited_title":"A., Ambrosi , G., et al","cited_arxiv_id":null,"evidence_quote":"Supplies the AMS-02 daily and Bartel-rotation-averaged proton flux data that define the half-year depression."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the 1974 minicycle by the absence of enhanced CME activity; the paper compares and rejects this explanation for 2017."},{"cited_title":"V., & Richardson , I","cited_arxiv_id":null,"evidence_quote":"Provides the catalogue of ICME and interaction-region arrival times at 1 AU used in Table 1."},{"cited_title":"G., & Cane , H","cited_arxiv_id":null,"evidence_quote":"Extends the catalogue of solar wind stream and ejecta events used to identify SIRs and CMEs."},{"cited_title":"K., Russell , C","cited_arxiv_id":null,"evidence_quote":"Supplies STEREO data for CME shock arrival times at 1 AU."},{"cited_title":"K., Luhmann , J","cited_arxiv_id":null,"evidence_quote":"Supplies STEREO SIR times used to identify stream interaction regions."},{"cited_title":"J., Zhang , J., Li , T., Yang , S","cited_arxiv_id":null,"evidence_quote":"Documents the X9.3 flare and AR12673's eruptive character, used to mark the strongest CME impacts."}],"review_version":1}