{"id":"4b433454-ac7d-4f51-a5df-2674309fd6fc","arxiv_id":"2501.14058","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"A systematic search of ASAS-SN data finds 782 slowly varying stars, including 433 new variables, mostly red subgiants and lower main-sequence stars.","lead":"This paper searches 9.4 million ASAS-SN stars for brightness changes of at least 0.03 magnitudes per year over about a decade, and finds 782 slowly varying systems, of which 433 are newly identified. The result maps a previously unexplored variability regime and gives follow-up targets for stellar activity, dust, and accretion studies.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 782-source catalog lacks a quantified false-positive rate; the brightening/fading balance and visual inspection cannot exclude residual camera/filter systematics, so the central claim is not yet fully secured.","rationale":"The paper is a careful observational search, and its astrophysical interpretation is supported by the recovery of known variable classes and by CMD placement. However, the quantitative headline depends on the assumption that a 0.03 mag/yr trend over a decade is astrophysical rather than residual systematics. The reader's weakest assumption identifies exactly this. I agree: the equal brightening/fading check is far too weak to rule out color- or camera-dependent drifts, and visual inspection is not a reproducible quantitative control. The most concrete mechanism is the V/g intercalibration with constant shifts; slow color evolution or changing seasonal filter coverage could create spurious slopes in the combined light curves. This does not invalidate the paper, but it means the catalog's false-positive rate and completeness are unknown. An injection-recovery or stable-star control would settle the issue. Since the reader already assigned CONDITIONAL and this concern supports that verdict rather than overturning it, no change to the verdict is needed.","tokens_in":15284,"tokens_out":6301,"duration_ms":64999,"concrete_test":"Inject artificial linear trends with amplitudes 0.03-0.10 mag/yr into a random subset of non-candidate ASAS-SN light curves from the same 13<g<14.5 mag range, run the full pipeline blindly, and measure the recovery fraction and the false-positive rate among null injections; additionally, for the 782 candidates, compare slopes fit separately to V-only and g-only seasonal medians and check whether a large fraction show inconsistent band-dependent trends.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim asserts 782 genuine slow variables, 433 of them newly identified. The pipeline's only global systematics check is that 395 sources brighten and 387 fade (Section 2). This balance is insensitive to time-dependent zero-point drifts in individual cameras/filters, color-dependent calibration residuals between the combined V and g bands, and spatially localized systematics, all of which can produce both brightening and fading artifacts. The authors do not run a control sample of known stable stars nor injection-recovery tests to estimate the false-positive rate, and the final acceptance step is subjective visual inspection. A specific concern is the V/g intercalibration: a constant calibration shift per camera/filter is fitted, so if a star's color evolves slowly or the seasonal mix of V and g observations changes over the decade, the combined seasonal medians can acquire a spurious long-term slope. Thus the quoted 782 and 433 could be upper bounds unless the residual systematics are quantified. The conclusion's own caveat that slower changes 'will require significant improvements in false positive rejection' reinforces that the current rejection power is not demonstrated.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript presents a search for long-term (decade-scale) photometric variability in 9,361,613 ASAS-SN sources with 13<g<14.5 mag. Using linear and quadratic fits to seasonal median light curves after intercalibrating cameras and filters, the authors select candidates with slopes ≳0.03 mag/yr and Δg>0.3 mag, then reject false positives via bright-star proximity, south-pole cuts, proper-motion cuts, and visual inspection. They report 782 candidates, 433 of them newly identified variables, cross-match to Gaia, SIMBAD, AAVSO, and WISE, classify the sources into five CMD groups, and analyze periodicities and mid-IR dust indicators. The central claim is that this is the first systematic catalog of slow variability in this magnitude range.","tokens_in":15513,"tokens_out":3406,"duration_ms":30501,"significance":"If the candidate list is robust, this is a valuable resource: it is the first systematic census of slow variability in ASAS-SN at these magnitudes, it identifies a substantial population of previously unclassified variables, and it links them to astrophysical groups (RS CVn, AGB, Be stars, AGN). The paper's strengths include a clearly specified selection pipeline, extensive external cross-matching, a full electronic table, and public availability of the light curves. The main weakness is that the false-positive rate is not quantified: no control sample or injection-recovery test is presented, and the final acceptance step is subjective visual inspection. The equal brightening/fading balance is suggestive but not a substitute for a systematics estimate.","major_comments":[{"comment":"The reduction from 36,705 initial candidates to 782 after false-positive rejection is not accompanied by any estimate of the false-positive rate. The statement that the roughly equal numbers of brightening (395) and fading (387) sources imply the sample is 'not affected by systematic drifts' is insufficient, because time-dependent zero-point drifts, color-dependent intercalibration residuals between V and g, and spatially localized systematics can all produce both brightening and fading artifacts. I request a control sample of known stable stars or injection-recovery tests to estimate the false-positive rate; without this, the quoted 782 sources (and 433 new variables) are upper bounds rather than a measured catalog size.","section":"Section 2"},{"comment":"The intercalibration fits constant camera/filter offsets and then seasonal medians are computed. If a star's color evolves slowly or the seasonal mix of V and g observations changes over the decade, the combined seasonal medians can acquire a spurious long-term slope. The paper does not demonstrate that residual systematics are below 0.03 mag/yr on decade timescales. Please quantify the stability of the photometric system (for example, by measuring the scatter of known stable stars or by comparing with external photometry) and show that the reported slopes are significant relative to this noise floor.","section":"Section 2, intercalibration"},{"comment":"Table 1 reports optical slopes, Δg, periods, and W1−W2 colors without uncertainties. Because the selection thresholds (0.03 mag/yr and 0.3 mag) are defined relative to these quantities, the absence of error bars prevents the reader from assessing whether individual candidates are genuinely above threshold. Please provide uncertainties, or at least the photometric noise floor for the slopes and periods, so that the significance of individual entries can be judged.","section":"Table 1"},{"comment":"The final paragraph's caveat that slower changes 'will require significant improvements in false positive rejection' is appropriate, but it also underscores that the rejection power for the 0.03 mag/yr threshold itself is not demonstrated. Given that the central claim is the existence and classification of 782 genuine slow variables, the paper should either provide a quantitative false-positive estimate for the adopted threshold or soften the claim accordingly.","section":"Section 4"}],"minor_comments":[{"comment":"The Lomb-Scargle period search uses a false alarm probability threshold of 0.1, which is quite loose; with 551 reported periodic variables, many periods may be spurious. Please state the expected number of false positives at this threshold or justify the choice.","section":"Section 2"},{"comment":"The caption contains a grammatical error: 'The format is the same is in Fig. 11' should read 'The format is the same as in Fig. 11'.","section":"Figure 5 caption"},{"comment":"The text 'Roughly, 10 percent of the sources are listed as YSOs (11) or T Tauri stars (5)' includes an unnecessary comma; the sentence would read more cleanly as 'Roughly 10 percent of the sources...'.","section":"Section 3"},{"comment":"The paper states that periods longer than the average observing season are discarded, but it does not define how the average observing season length is computed. Please clarify this definition.","section":"Section 2"}],"recommendation":"major_revision","confidential_remarks":"The paper is within scope for MNRAS and the resulting catalog would be useful to the variable-star community. The main risk is the unquantified false-positive rate; I recommend major revision. In revision, the authors should add a stability control or injection-recovery test, provide error bars for the key quantities, and temper the central claim if those tests cannot be performed. The equal brightening/fading balance is too weak to stand alone as evidence against systematics."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this is a genuinely useful first systematic catalog of slow variability in ASAS-SN's 13<g<14.5 range, with 433 new variables out of 782 candidates. The paper deserves a serious referee, though the headline numbers are softer than the abstract implies.\n\nThe new thing here is the search itself and the catalog, not the techniques. Linear/quadratic fits, seasonal medians, Lomb-Scargle, and WISE cross-matching are all standard, but nobody has systematically combed a decade of ASAS-SN data for 0.03 mag/yr drifts in this magnitude range before. The authors put real work into removing false positives: bright-star bleed trails, south-pole field rotation, high proper motion stars, and failed V/g intercalibration. The equal split of 395 brightening vs 387 fading sources is a sensible first check, and the grouping of candidates on the CMD plus the WISE dust diagnostics adds physical context. The 191 dust-related candidates and the small AGN sample are nice byproducts.\n\nThe soft spots are real but not fatal. The biggest one is the absence of a quantified false-positive rate. The visual inspection is subjective, the brightening/fading balance cannot rule out time-dependent zero-point drifts or color-dependent residuals between V and g, and there is no control sample of known stable stars or injection-recovery test. The seasonal mix of V and g observations changing over a decade could in principle produce spurious slopes, and the paper doesn't demonstrate this is negligible. The periodicity threshold of FAP<0.1 is loose, and Table 1 has no error bars on slopes, periods, or colors. These are all addressable rather than catastrophic. The authors themselves admit that slower changes will require better false-positive rejection, which tells you they know the current rejection is not ironclad.\n\nI'd send this to a referee. The catalog is the product, and for a catalog paper the missing completeness/reliability budget is an important but fixable gap. A revision with a proper systematics estimate and error bars would make the 782/433 numbers trustworthy. I'd probably cite it for the catalog once that's in.","headline":"A genuinely useful first catalog of slow variables in ASAS-SN, with an honest but unquantified contamination risk; deserves refereeing.","tokens_in":16025,"tokens_out":2885,"would_cite":true,"duration_ms":24499,"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 reports that a decade-long automated sky survey of nine million bright stars contains 782 systems changing by more than 0.03 magnitudes per year, 433 of which are newly identified variables.","keywords":["long-term stellar variability","slow variable stars","decade-scale time-domain photometry","stellar activity cycles","circumstellar dust","semi-regular variables","color-magnitude diagram","photometric calibration"],"falsifier":"A direct test would feed thousands of light curves of stars known to be constant through the same intercalibration, seasonal-median binning, and slope-selection pipeline; if a comparable fraction of those stable stars are flagged at more than 0.03 magnitudes per year, the catalog's trends are substantially contaminated by systematics.","tokens_in":15082,"feed_emoji":"🌟","tokens_out":7042,"duration_ms":60165,"temperature":0.7,"pith_summary":"The paper asks whether a ten-year baseline of wide-field automated sky-survey photometry can systematically uncover slow stellar variability, a regime transient surveys usually ignore. It establishes that, among about nine million isolated stars in a narrow bright-sky magnitude range, 782 show a steady brightness trend of more than roughly 0.03 magnitudes per year; 433 of these are new variables and 349 were previously classified under standard classes that do not capture this slow behavior. The sample separates into five behavioral groups, with the largest being red subgiants and lower main-sequence stars, and about 70 percent also show shorter-period periodic variability. If true, this means slowly varying stars are common, have distinct physical drivers such as activity cycles, pulsation, and dust, and can be found systematically in decade-long photometry.","feed_headline":"Ten-year survey finds 782 slowly changing stars","feed_subtitle":"The first systematic slow-variability catalog in this brightness range shows red subgiants and low-mass stars dominate.","key_machinery":"The load-bearing procedure is the construction of decade-long seasonal-median light curves. Each source's camera and filter data are intercalibrated using a damped random-walk Gaussian process, then median-averaged per observing season to remove short-timescale variability, and then fit with linear and quadratic functions of time. The linear slope and the quadratic fit's maximum magnitude excursion select candidates; four classes of false positives are removed, namely bright-star artifacts, south-pole field rotation, high proper motion, and failed intercalibration between two filter bands. This machinery converts raw survey photometry into a slow-variability rate and a maximum magnitude change, and the equal counts of brightening and fading sources serve as the internal check that the trends are not a systematic drift.","core_discovery":"The paper's central claim is that a systematic, false-positive-controlled search of 9,361,613 isolated sources with $13 < g < 14.5$ mag in a ten-year time-domain survey yields 782 genuine slowly variable systems with slopes exceeding roughly $0.03$ mag/yr. It argues these trends are astrophysical rather than instrumental, citing the near balance of brightening and fading sources and the fact that known variable classes such as semi-regular, slow irregular, and spotted stars appear only at the high-amplitude tail of their class distributions. The candidates occupy distinct regions of the observed color–magnitude diagram and are split into main-sequence, subgiant/giant, AGB, luminous blue, and nova-like groups. The paper also shows that 551 candidates are periodic on shorter timescales, mostly longer than 10 days, and that 191 are plausibly linked to circumstellar dust through infrared excess or optical-versus-infrared slope behavior.","pith_inferences":["The paper does not run a control sample of known constant stars, so I infer the absolute size of the catalog is not yet pinned down; an injection-recovery experiment would convert the brightening-versus-fading balance into a direct false-positive rate and a completeness estimate.","I infer the same intercalibration-plus-seasonal-median method transfers directly to other decade-long surveys, so running this selection on fainter sources or independent fields would test whether the subgiant/giant dominance is universal or magnitude-dependent.","The 191 dust-related candidates are the most promising subset for follow-up spectroscopy: measuring temperature and luminosity at high and low states could separate temperature-driven variability from true obscuration by circumstellar dust."],"forward_implications":["A ten-year survey of a single magnitude range already yields hundreds of slowly varying stars, so the slowly varying sky is not rare.","Standard variable-star taxonomies miss this regime: most new systems have no clean standard class, and known SR/L/ROT variables appear as the extreme high-amplitude tail of their populations.","About 70 percent of slow variables also vary periodically on shorter timescales, meaning long-term trends and rotation or pulsation commonly coexist.","Roughly 191 candidates show optical and infrared changes consistent with dust formation or destruction, identifying a substantial dust-linked subset.","The five behavioral groups imply distinct physical drivers: magnetic activity for lower main-sequence stars, spot activity for subgiants, pulsation for AGB stars, eruptive behavior for blue stars, and mass transfer for nova-like systems."],"supporting_citations":[{"why":"Supplies the survey photometry that is the starting point for every light curve in the search.","marker":"Shappee et al. 2014"},{"why":"Describes the survey operations and data release that produced the decade-long baseline.","marker":"Kochanek et al. 2017"},{"why":"Provides the earlier survey variable-star catalog used for cross-matching previously classified sources.","marker":"Jayasinghe et al. 2018"},{"why":"Defines the Sky Patrol V2.0 database from which the 9,361,613 light curves were downloaded.","marker":"Hart et al. 2023"},{"why":"Supplies the damped random-walk Gaussian process used to intercalibrate cameras and filters.","marker":"Kozłowski et al. 2010"},{"why":"Provides the periodogram method used to search for periodic variability in detrended light curves.","marker":"Lomb 1976"},{"why":"Supplies the false-alarm-probability formalism used to select significant periods.","marker":"Scargle 1982"},{"why":"Provides the method for extracting and fitting infrared light curves used to assess dust-related variability.","marker":"Hwang & Zakamska 2020"},{"why":"Supplies the international variable-star catalog used to identify previously classified variables.","marker":"Watson et al. 2006"},{"why":"Supplies the quasar/AGN catalog used to identify the eight active galactic nuclei in the sample.","marker":"Flesch 2023"}],"fun_headline_variants":["782 slow variables found in 10-year survey","Decade-long hunt yields 782 slowly changing stars","9.3M stars scanned, 782 slow variables emerge","433 new variables among 782 slow stars","Slow variability catalog: 782 systems from ASAS-SN"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The premise is that a 0.03-magnitude-per-year drift in the processed decade-long light curve is a real change in the star rather than a small residual calibration artifact, and the paper's checks for this are only a count showing roughly equal numbers of brightening and fading sources plus visual review of each light curve.","fun_headline_variants_meta":{"raw":{"variants":["782 slow variables found in 10-year survey","Decade-long hunt yields 782 slowly changing stars","9.3M stars scanned, 782 slow variables emerge","433 new variables among 782 slow stars","Slow variability catalog: 782 systems from ASAS-SN"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001305,"raw_usage":{"total_tokens":5323,"prompt_tokens":950,"completion_tokens":4373,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":566,"completion_tokens_details":{"reasoning_tokens":4296}},"tokens_in":566,"tokens_out":4373,"duration_ms":25810,"temperature":1.0,"reasoning_tokens":4296,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T15:25:58.708552+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct test would feed thousands of light curves of stars known to be constant through the same intercalibration, seasonal-median binning, and slope-selection pipeline; if a comparable fraction of those stable stars are flagged at more than 0.03 magnitudes per year, the catalog's trends are substantially contaminated by systematics.","supporting_citations":[],"review_version":1}