{"id":"203b01e8-7ea2-4481-b686-5e09f12bd790","arxiv_id":"2509.02858","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"Wind-formed CV H-alpha profiles can populate the expected outflow region of the excess-EW diagram, but the diagram is so sensitive to wing definition that an FWHM-based mask and a calibrated EW scaling relation are proposed.","lead":"This paper simulates 3,645 H-alpha lines from winds around accreting white dwarfs and stress-tests a common outflow diagnostic. It finds the diagnostic changes with the analyst's wing definition, proposes an FWHM-based fix, and releases an open line library plus a scaling formula.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Central reliability claim lacks a non-wind control: all grid lines are wind-formed, so mask sensitivity and FWHM-mask 'improvement' are never shown to improve wind/non-wind discrimination.","rationale":"The reader's weakest assumption concerned model adequacy: whether the KWD biconical wind parameterization captures real CV wind geometry and ionization. My concern is more direct and does not depend on that assumption: even if the KWD models are accepted as representative winds, the paper's central claim about diagnostic reliability is not tested because the study lacks a null class of non-wind lines. This is an internal inference gap, not an external-validity dispute. The reader's mention of sample-selection circularity is related but does not target the missing non-wind control. This concern is load-bearing because the abstract and conclusions make a normative claim that the FWHM-based method is a 'sensitive and reliable way to detect disc winds'; the data only show that excess values are sensitive to the mask definition and that wind models move around in the diagram. The gap is addressable with a matched control sample and a quantitative discrimination metric. Given the otherwise open, reproducible presentation and the fact that the concern is a validation gap rather than a demonstrated error, I do not change the reader's CONDITIONAL verdict; the condition should be the control-group test.","tokens_in":24785,"tokens_out":6935,"duration_ms":81087,"concrete_test":"Generate a matched no-wind control sample: run SIROCCO with the same WD/disc parameters and wind mass-loss rate set to zero (or, if no line forms, use an analytic pure-disc H-alpha line model) over the same parameter grid and inclinations, restricting to lines with EW and FWHM inside the Gold selection box. Apply both the original fixed ±1000-2500 km/s mask and the proposed 1x/5x FWHM mask to the wind and no-wind samples, then compute a discrimination metric such as ROC AUC or a confusion matrix for the 'blue deficit/red excess' region. If the FWHM mask does not yield a significantly higher discrimination than the fixed mask, the central claim that fixed masks are unreliable and FWHM masks restore diagnostic power does not hold.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The decisive weakness is in the inference from Sections 6 and 7.1-7.3. The headline claim is that fixed-mask EW excesses are unreliable wind diagnostics and that the FWHM-relative mask (1x/5x FWHM) restores interpretability. That claim requires measuring how well the diagnostic separates wind-formed from non-wind lines. No such measurement is made. Every spectrum in the grid is generated with the KWD wind turned on; Silver/Bronze spectra are weak or multi-peaked wind models, not wind-free. There is no pure disc/no-wind control sample. Thus Fig. 8 demonstrates only that measured excess values depend on the mask, not that the original method misclassifies lines as wind-formed. Similarly, Fig. 9/B3 shows only that FWHM masking moves wind models to different parts of the diagram; since all points are true winds, true-positive rate is undefined and the false-positive rate (non-wind lines falling in the 'wind region') is never estimated. The paper itself notes in Section 7.1 that hot spots, eccentric discs, or other non-wind structures can produce asymmetric wings, but no such non-wind model is included. Consequently, the statement in Section 8 that the refined definition is 'more likely to provide a sensitive and reliable way to detect disc winds' is under-supported. The open-source grid is a genuine resource; the issue is the validation design, not the simulations per se. A matched non-wind control is needed before the reliability claim can be accepted.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a systematic grid of 3645 synthetic H-alpha line profiles (729 KWD biconical wind models viewed at five inclinations) computed with SIROCCO, filters them into quality tiers, and defines a 'Gold' subsample whose EW and FWHM match observed high-state CVs. Using this grid, the authors test the excess-EW diagnostic diagram of Mata Sánchez et al. (2018)/Cúneo et al. (2023). They find that wind-formed models can occupy the previously suggested wind regions, but that the result is highly sensitive to the adopted velocity masking window. They therefore propose a FWHM-relative mask (1.0x–5.0x FWHM) and argue it improves diagnostic reliability. They also derive an approximate power-law scaling relation (Eq. 3) between EW and wind parameters, plus an emission-measure-based curve-of-growth model in Appendix A. All models and analysis scripts are open-source, with a web-based browsing tool.","tokens_in":25237,"tokens_out":4027,"duration_ms":48411,"significance":"If the central reliability claim is established, this would be a valuable methodological contribution: it would place the widely used excess-EW diagnostic on a quantitative footing, provide a standardized masking prescription, and give observers a fast way to connect H-alpha EWs to disc-wind parameters. The paper's concrete strengths are the large open-source grid, the explicit quantitative test of mask sensitivity in Fig. 8, and the physically motivated scaling analysis. However, the validation design currently limits the strength of the main conclusions: there is no wind-free control sample, the Gold sample is selected on the same observations used for comparison, and the scaling relations are fit and tested on the same data. These issues are addressable but require additional work before the reliability claims can be accepted.","major_comments":[{"comment":"The claim that the FWHM-based masking restores sensitivity and reliability is under-supported because every spectrum in the grid is wind-formed. Fig. 8 demonstrates that fixed-mask excess EWs depend on the window choice (11/155 points switch quadrants), but that does not show that the original method misclassifies non-wind lines, nor that the revised method correctly separates wind from non-wind lines. The paper itself notes in Section 7.1 that hot spots and eccentric discs can produce asymmetric wings, but no such non-wind profiles are modelled. Without a matched control sample of wind-free line profiles, the true-positive and false-positive rates of the proposed diagnostic are undefined, so the concluding sentence that the refined definition is 'more likely to provide a sensitive and reliable way to detect disc winds' does not follow. Please add a non-wind control set (e.g., rotating-d","section":"Sections 7.1–7.3 and Section 8 (third bullet)"},{"comment":"The Gold sample is selected by drawing a box 'roughly centred on the Cúneo et al. (2023) sample' in the EW–FWHM plane (Fig. 6), and the same Cúneo data are then used as the observational benchmark in the excess-EW diagram (Fig. 7). Selection in EW/FWHM and comparison in excess-EW are not identical observables, so this is not a full circularity, but it is also not an independent validation: the model–data overlap in Fig. 7 is conditioned on matching the same systems. Please validate on an independent sample (e.g., Zhao et al. 2025 or a withheld subset) and quantify how Gold membership and the resulting diagram change with the selection-box boundaries.","section":"Section 5.2.2 and Section 6"},{"comment":"The scaling relation in Eq. (3) is fitted to the Gold sample and its predictive performance is displayed for the same Gold sample (Fig. 10); the Silver/Bronze inset shows degradation but those samples are not used in the fit. Similarly, Appendix A fits K1, K2, and the EM scaling coefficients to the full data set. As presented, the scatter of ~0.17 dex measures in-sample fit quality, not predictive power. To support the statement that the relation lets observers 'assess whether—and what kind of—accretion disc wind might produce the H-alpha line,' please provide out-of-sample validation, for example by cross-validation or by holding out a random subset of the grid. If the relation is intended only as an empirical description of the grid, that limitation should be stated explicitly.","section":"Section 7.4, Eq. (3), Fig. 10; Appendix A, Table A1"},{"comment":"All diagnostic conclusions and the scaling relation are conditioned on the Knigge–Wood–Drew biconical wind parameterization with fixed white-dwarf parameters, a smooth wind, and 80–90% cell convergence treated as steady state. If real CV winds are clumpy, time-dependent, or differently collimated, the Gold-sample fractions and the fitted exponents in Eq. (3) could change. This is not a fatal objection—the grid is explicitly systematic—but the abstract and Section 8 state conclusions about 'disc winds' in general. Please add a limitations paragraph that spells out the model-validity domain and, ideally, a concrete test (e.g., a small comparison set with different wind geometry or with clumping) showing the robustness of the qualitative conclusions.","section":"Section 2 and Table 1"}],"minor_comments":[{"comment":"The text says a subset of 40 spectra, but the caption and text also refer to 155 data points; clarify that each spectrum is shown at five inclinations and state the total number of plotted points.","section":"Fig. 8 caption and Section 7.2"},{"comment":"There is a tension between the original method's step (ii) (Gaussian fit constrained to the core) and the third modification (fitting the overall line profile). Please provide an explicit algorithmic summary of the actual fitting region after the modification.","section":"Section 3.2"},{"comment":"The choice of 1.0x and 5.0x FWHM is still a user-chosen window. Since the paper emphasizes mask sensitivity, please report the sensitivity of Fig. 9/B3 to reasonable changes in these multipliers, or state that this is left for future work.","section":"Section 7.3"},{"comment":"The notation 'h 100.31(α−0.25) i' is confusing; consider writing these factors as 10^{0.31(α−0.25)} and 10^{0.1(β−1.5)}.","section":"Eq. (3)"},{"comment":"Minor wording: 'the vice-versa view applies' should be 'the vice versa view applies' or 'the converse applies'.","section":"Section 5.2.2"},{"comment":"The statement that 'about 20%' of lines are Gold would be more precise if the range across the five inclinations were given, since Gold membership is inclination-dependent.","section":"Abstract and Section 8"}],"recommendation":"major_revision","confidential_remarks":"The open-source grid and web tool are genuine contributions, and the mask-sensitivity experiment in Fig. 8 is a useful quantitative result. The main gap is the absence of a non-wind control sample, which prevents the central reliability claim from being evaluated. The same-sample fitting of Eq. (3) and the Gold selection box also need to be addressed. If the authors add a control experiment and out-of-sample validation, the paper would be suitable for publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The thing to know: this paper gives the field a systematic, openly available grid of SIROCCO H-alpha wind profiles and a well-demonstrated warning that the EW-excess diagnostic diagram is sensitive to the user-defined masking window. The Fig. 8 experiment -- 11 of 155 points switching quadrants under a 200 km/s inner-bound shift -- is the clearest and most actionable result in the paper. That alone is worth a look.\n\nWhat's genuinely new: the 729-model grid (3645 spectra across five inclinations), the open data and browser tool, and the explicit FWHM-based masking proposal. The paper is also honest about the calibrated nature of the EW scaling relation and about the clustering of wind models along the y=x diagonal with fixed masks. The Appendix A curve-of-growth framing for line luminosity vs emission measure is a nice physically motivated bridge.\n\nWhere it goes soft: the central reliability claim is under-supported. Every spectrum in the grid is wind-formed; there is no wind-free control (disc plus hot spot, eccentric disc, etc.). So the paper shows that fixed masks are mask-sensitive and that FWHM masks move wind models around the diagram, but it never shows that either method actually distinguishes wind-formed from non-wind lines. Section 7.1 itself concedes that hot spots and eccentric discs can produce asymmetric wings, yet no such model is included. Consequently, the Section 8 statement that the refined definition is 'more likely to provide a sensitive and reliable way to detect disc winds' is not established. This is a fixable gap, but it is load-bearing for the headline claim. The Gold-sample selection also has circularity -- the box is drawn around Cuneo data and then compared to those same data -- and the scaling relation is a fit, not a prediction, but the authors are candid about both, so I'd treat them as minor.\n\nWho this is for: anyone using EW-excess diagrams in CVs or LMXBs, and anyone building synthetic line-profile libraries. I'd cite the grid and the mask-sensitivity result, but not the reliability claim without a control.\n\nRecommendation: send it to peer review. A good referee should ask for the non-wind control (or a reworded conclusion), but the resource and the sensitivity warning deserve to be published and debated.","headline":"A genuinely useful open grid and a real warning about mask sensitivity, but the strong reliability claim for the FWHM-based fix isn't backed by a non-wind control.","tokens_in":25691,"tokens_out":1599,"would_cite":true,"duration_ms":22203,"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":"Excess-EW wind diagnostic hinges on how the line wing is defined","keywords":["accretion disc winds","cataclysmic variables","H-alpha emission lines","excess equivalent width diagnostic","radiative transfer","SIROCCO","line profile diagnostics","equivalent width scaling relation"],"falsifier":"Take the 61 observed H-alpha spectra of Cúneo et al. (2023) and recompute their excess EWs under both the fixed ±1000–2500 km/s mask and the proposed 1.0x–5.0x FWHM mask; then shift the FWHM estimate by ±30 per cent. If a substantial fraction of spectra change quadrants under the FWHM mask, or if known wind-driving systems do not move toward the blue-deficit/red-excess region, the paper's central reliability claim would be contradicted.","tokens_in":24742,"feed_emoji":"🌀","tokens_out":7489,"duration_ms":77785,"temperature":0.7,"pith_summary":"This paper asks whether the 'excess equivalent width' diagram—a fast way to spot accretion-disc winds from optical H-alpha lines—actually works. To answer it, the authors compute 3,645 synthetic H-alpha profiles from 729 biconical disc-wind models with the radiative-transfer code SIROCCO, and compare them to observed nova-like cataclysmic variables. Their central finding is that the diagram's verdict depends strongly on the user-chosen velocity window that defines the line wing: a small shift of the window can move a data point across the diagram and even switch its quadrant. They argue that a window set at 1.0x to 5.0x the line's full-width at half-maximum makes the diagnostic stable and interpretable, and they supply an approximate scaling relation that predicts H-alpha equivalent width from wind parameters. If right, this means previous outflow classifications based on fixed windows need revisiting, and future surveys should use the FWHM-relative mask.","feed_headline":"Excess-EW wind test too sensitive to line-wing cutoff","feed_subtitle":"A 3,645-profile grid of CV winds shows fixed windows misread the diagram; the fix is a FWHM-based mask.","key_machinery":"The central object is the 'excess equivalent width diagnostic diagram': a best-fit Gaussian is subtracted from a continuum-normalised H-alpha line, and the residual flux is integrated separately in a blue-wing and red-wing velocity mask to give the two coordinates. The paper's refinement replaces the fixed velocity mask with a dynamic one bounded at 1.0x and 5.0x the line's FWHM, so every line is measured over the same fraction of its wings. The supporting machinery is the SIROCCO radiative-transfer grid of 729 Knigge-Wood-Drew biconical wind models (six varied parameters, five inclinations), which supplies the 3,645 self-consistent H-alpha profiles used to test the diagram and calibrate the","core_discovery":"The paper's central claim is that the excess-EW diagnostic diagram, as originally implemented with fixed radial-velocity masking windows, is not a reliable outflow diagnostic for wind-formed H-alpha lines. Using a grid of 729 SIROCCO wind models viewed at five inclinations, the authors show that the position of a line in the diagram depends sensitively on the chosen wing boundaries; shifting the inner edge by approximately 200 km/s can move data points between quadrants. They propose defining the wing masking window as 1.0x to 5.0x the line's FWHM, which avoids core contamination and makes the diagram's regions correspond to actual profile shapes—P-Cygni-like lines land in the blue-deficit/r","pith_inferences":["The mask-sensitivity problem likely extends beyond CVs: any application of the excess-EW method to LMXBs, YSOs, or AGN that uses fixed windows should be re-examined with FWHM-relative masks.","The scaling relation could be inverted into a cheap prior for spectral fitting or emulator-based inference, narrowing the wind-parameter space before expensive radiative-transfer runs.","The paper's finding that most wind models cluster near the diagonal suggests that an off-diagonal excess alone is weak evidence; combining the diagram with blue-shifted absorption or time-variable line shapes would give more robust wind identification."],"forward_implications":["Previous applications of the excess-EW diagram that used fixed velocity windows should be re-checked; their outflow classifications may be mask artefacts rather than wind detections.","Adopting the FWHM-relative mask (1.0x to 5.0x FWHM) makes the diagram stable enough for survey-scale use across thousands of spectra with very different line widths.","The scaling relation lets an observer decide quickly whether a measured H-alpha EW can plausibly be produced by a disc wind, and if so what combination of mass-loss, collimation, and acceleration parameters would do it.","The result that most synthetic wind lines sit near the y=x diagonal under the fixed mask means symmetric non-Gaussian wings—from discs, hot spots, or eccentric discs—can masquerade as outflow features unless the mask is chosen carefully.","The Gold sample's preference for the 'wind region' of the diagram is inclination-dependent, so viewing angle must be accounted for when classifying an observed source."],"supporting_citations":[{"why":"Introduces the excess EW diagnostic diagram and the original Gaussian-subtraction and masking approach that the paper tests.","marker":"Mata Sánchez et al. (2018)"},{"why":"Provides the observational benchmark of 61 H-alpha spectra from four nova-like CVs and the fixed ±1000–2500 km/s masking window used for comparison.","marker":"Cúneo et al. (2023)"},{"why":"Extends the diagnostic for outflow detection and classification; its variant masking choices are part of the sensitivity the paper documents.","marker":"Mata Sánchez et al. (2023)"},{"why":"An application of the excess-EW method to LMXBs that introduced significance contours; shows the method's broad use and motivates reliability testing.","marker":"Muñoz-Darias et al. (2019)"},{"why":"Applies the method with different masking windows; cited as evidence that window definitions vary between studies.","marker":"Panizo-Espinar et al. (2021)"},{"why":"Defines the biconical disc-wind parameterization (KWD model) used to construct the entire simulation grid.","marker":"Knigge et al. (1995)"},{"why":"Describes the SIROCCO radiative-transfer code used to generate the synthetic H-alpha profiles.","marker":"Matthews et al. (2024)"},{"why":"Supplies the LAMOST CV sample used as a secondary observational constraint on the Gold sample's EW/FWHM selection box.","marker":"Zhao et al. (2025)"}],"fun_headline_variants":["FWHM-based wing mask fixes excess-EW wind diagnostic","Wind-line test needs FWHM wings to stay reliable","Refined wing definition sharpens wind-formed line identification","Excess-EW diagram misreads wind lines without FWHM wings","New wing cutoff makes wind-line diagnosis from H-alpha reliable"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The analysis assumes that one smooth, steady, biconical wind model with fixed white-dwarf and disc parameters, treated as converged at 80–90 per cent of cells, adequately represents the geometry, ionization, and velocity structure of real CV disc winds.","fun_headline_variants_meta":{"raw":{"variants":["FWHM-based wing mask fixes excess-EW wind diagnostic","Wind-line test needs FWHM wings to stay reliable","Refined wing definition sharpens wind-formed line identification","Excess-EW diagram misreads wind lines without FWHM wings","New wing cutoff makes wind-line diagnosis from H-alpha reliable"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000264,"raw_usage":{"total_tokens":1483,"prompt_tokens":832,"completion_tokens":651,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":576,"completion_tokens_details":{"reasoning_tokens":567}},"tokens_in":576,"tokens_out":651,"duration_ms":7655,"temperature":1.0,"reasoning_tokens":567,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T11:20:29.371484+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the 61 observed H-alpha spectra of Cúneo et al. (2023) and recompute their excess EWs under both the fixed ±1000–2500 km/s mask and the proposed 1.0x–5.0x FWHM mask; then shift the FWHM estimate by ±30 per cent. If a substantial fraction of spectra change quadrants under the FWHM mask, or if known wind-driving systems do not move toward the blue-deficit/red-excess region, the paper's central reliability claim would be contradicted.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Extends the diagnostic for outflow detection and classification; its variant masking choices are part of the sensitivity the paper documents."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Applies the method with different masking windows; cited as evidence that window definitions vary between studies."},{"cited_title":"SIROCCO: A Publicly Available Monte Carlo Ionization and Radiative Transfer Code for Astrophysical Outflows","cited_arxiv_id":"2410.19908","evidence_quote":"Describes the SIROCCO radiative-transfer code used to generate the synthetic H-alpha profiles."},{"cited_title":"Searching for accreting compact binary systems from spectroscopy and photometry: Application to LAMOST spectra","cited_arxiv_id":"2503.12410","evidence_quote":"Supplies the LAMOST CV sample used as a secondary observational constraint on the Gold sample's EW/FWHM selection box."}],"review_version":1}