{"id":"6078cfe5-81c6-49ba-91fd-f0f76989da5d","arxiv_id":"2608.07290","paper_version":1,"verdict":"UNVERDICTED","confidence":"HIGH","novelty_score":2.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"A SKA community chapter surveys how HI rotation curves could test dark matter and modified gravity, with proposed surveys and software needs.","lead":"This chapter reviews how future SKA radio observations of hydrogen gas in galaxies could distinguish dark matter models from modified gravity theories. It argues that sharper, wider, and deeper HI surveys will turn galaxy rotation curves into a large statistical test of competing theories.","discovery_kind":"review","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The z≃1 rotation-curve forecast in Sec. 3.3 assumes the local HI mass-size relation holds at z≃1; if high-z HI disks are more compact, the 10,000-hour survey yields fewer than 5 resolution elements and the DM/MOND test fails.","rationale":"The reader's weakest_assumption identifies exactly the load-bearing extrapolation: the local HI mass-size relation is used in Sec. 3.3 to define the z≃1 survey, and no evidence is provided that the relation holds at that redshift. I agree with that identification. The chapter is a planning document rather than a research paper, so the appropriate verdict remains UNVERDICTED: the concern does not make the chapter internally inconsistent, and there is no new result to accept or reject. However, the concern is real and should be flagged. A second concern, that even with 5 beams the kinematic fit at low SNR may be unreliable, is acknowledged in the chapter's own discussion of barely resolved systems and automated acceptance criteria (Secs. 4.2 and 4.2.2), so the reader's choice of the size-extrapolation issue as the primary weak point is appropriate. The concrete test via simulations would settle whether the z≃1 forecast survives, and the verdict would need no change if the test succeeded.","tokens_in":31247,"tokens_out":5819,"duration_ms":58713,"concrete_test":"Recompute the Sec. 3.3 survey yield replacing the local Wang et al. (2016) size relation with z≃1 sizes from cosmological hydro simulations (TNG50/EAGLE/SIMBA): select M_HI ≃ 10^10 M⊙ galaxies at z≃1, convolve mock HI cubes to 1″/3.1 km/s at the 10,000-h depth, run 3DBarolo or TiRiFiC with the ≥5-beam and inclination acceptance criteria, and compare the number of reliable rotation curves to the paper's implied expectation. If the recovered sample falls below the tens needed for the dynamical-scaling-law test, the z≃1 component of the central claim is unsupported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central forecast that SKA-Mid AA4 will deliver HI rotation curves at z≃1 (Sec. 3.3) is load-bearing for the paper's 'closing the loop' statement in Sec. 5, and it rests on an unmeasured extrapolation: the local Wang et al. (2016) HI mass-size relation is assumed to predict the angular sizes of M_HI ≃ 10^10 M⊙ disks at z≃1. High-redshift stellar disks are smaller at fixed stellar mass, and if HI disks shrink correspondingly, a 1″ beam at z≃1 (≈8 kpc) will place fewer than the required ≥5 resolution elements on the major axis; the outer, low-acceleration regions that distinguish MOND from CDM would remain unresolved. The chapter cites HI-stacking constraints on the evolution of M_HI–M* (Chowdhury et al. 2020; Bera et al. 2023) but provides no empirical or simulation-based anchor for size evolution, so the 10,000-hour Band-1 survey design is not currently supported by the quoted evidence. This does not make the chapter internally inconsistent; it identifies the specific place where the planning forecast would fail if the extrapolation is wrong.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This chapter, written for the 'Advancing Astrophysics with the SKA – II' volume, reviews how SKA-Mid Array Assembly 4 (AA4) will advance the study of galactic dark matter and modified-gravity theories through spatially resolved HI 21-cm observations. It outlines three observing programs: high-resolution (roughly 1-2 arcsec) HI mapping of nearby galaxies to address the cusp-core problem and test MOND, wide and pointed surveys to build samples of tens of thousands of galaxies with resolved kinematics, and a very deep (about 10,000 hours) Band-1 pencil-beam survey to push resolved HI rotation curves to z about 1. The chapter also reviews the state of kinematic modeling tools and identifies key technical challenges for automated analysis. Its central claim is that these programs will allow tests that discriminate between particle dark matter and MOND-type theories, potentially closing the loop on the dark matter problem.","tokens_in":31559,"tokens_out":5664,"duration_ms":50021,"significance":"If the forecasts are correct, the proposed SKA-Mid AA4 programs would be transformative: they would provide the first large-sample, high-resolution HI kinematic data at z about 1, where dark matter and modified-gravity models make divergent predictions for the evolution of dynamical scaling laws. The chapter's survey of the literature is accurate and balanced, and it is careful to label its quantitative estimates as rough rather than precise. It also correctly identifies the need for automated, bias-tested kinematic modeling pipelines and for modeling in the uv-plane. No part of the chapter is internally circular; its claims are extrapolations from established local relations and instrument specifications. The main risk is that the z about 1 rotation-curve forecast rests on an unverified extrapolation of the HI mass-size relation, and that the 'tens of thousands of resolved galaxies' yield estimate lacks an explicit resolution threshold. Both issues are fixable in revision and do not invalidate the review's overall scientific case.","major_comments":[{"comment":"The claim that a roughly 10,000-hour Band-1 survey with SKA-Mid AA4 will map M_HI about 10^10 M_sun galaxies at z about 1 with at least five independent resolution elements along the major axis assumes that the local HI mass-size relation of Wang et al. (2016) holds at z about 1. The manuscript cites HI-stacking results for the evolution of the M_star-M_HI relation (e.g., Chowdhury et al. 2020; Bera et al. 2023), but it provides no empirical or simulation-based constraint on the size evolution of HI disks. Because high-redshift stellar disks are observed to be more compact at fixed stellar mass, a plausible shrinking of HI disks would directly reduce the number of resolution elements and undermine the stated ability to reach the low-acceleration, DM-dominated outskirts at z about 1. This assumption is load-bearing for the 'close the loop' statement in Sec. 5. I recommend that the authors either provide a quantitative assessment of the impact of size evolution (for example, by considering a range of size-mass relations from simulations) or explicitly frame the 10,000-hour survey as an optimistic upper limit contingent on no significant size evolution.","section":"Sec. 3.3 and Sec. 5"},{"comment":"The estimate that a 900-hour, 150 deg^2 survey will 'spatially resolve about 20,000 galaxies' is not reproducible because the text does not define the minimum number of resolved elements needed to classify a galaxy as spatially resolved for kinematic purposes. Given that Sec. 4.2.2 and the cited Deg et al. (2025) work show that at least about five beams across the major axis are required for reliable inclination and rotation-curve extraction, the yield should be recomputed (or the threshold explicitly stated) before the 'tens of thousands' figure is used in the abstract. Without this, the headline sample-size forecast cannot be independently checked.","section":"Sec. 3.2"}],"minor_comments":[{"comment":"The text refers to 'SKA-Mid AA*' in the discussion of the deep Band-2 survey; this should presumably read 'AA4', as the asterisk appears to be a typographical artifact.","section":"Sec. 3.3"},{"comment":"The statement that 20 hours of Band-2 observation detects an HI column density of about 5 M_sun pc^-2 over 16 km/s at resolutions of 1.3 arcsec and 3.1 km/s lacks the assumed system temperature, bandwidth, or a reference to the sensitivity calculator and therefore cannot be independently verified.","section":"Sec. 3.1"},{"comment":"The text alternates between a 'Band-2 survey' for the range z about 0-0.5 and a 'Band-1 survey' for z about 1; please clarify which band and frequency range is meant for each redshift interval.","section":"Sec. 3.3"},{"comment":"Several references are future-dated relative to the arXiv submission (e.g., Haubner et al. 2026, Bianchetti et al. 2025, Deg et al. 2025); if this is a book chapter, please identify these as in press or accepted to avoid ambiguity for readers.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"This is a review and prospects chapter for the SKA science book rather than a research paper with new derivations. The main technical concern is that the z about 1 rotation-curve forecast is presented with more confidence than the underlying size-evolution assumption supports. I recommend that the editors encourage the authors to add a dedicated caveat paragraph on the extrapolation of the HI mass-size relation, or to soften the 'closing the loop' language. The chapter is otherwise authoritative, well-written, and appropriately referenced, and the issues raised are local and fixable."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a review and prospects chapter, not a research paper. It introduces no new data, equations, or measurements, and it does not claim to. What it does—synthesising the current state of HI galaxy dynamics and laying out realistic survey strategies for SKA-Mid AA4—is done well. The authors are clearly expert, the descriptive parts are accurate, and the survey-time estimates are explicitly flagged as rough.\n\nThe best parts are the discussion of the dynamical scaling laws (BTFR, CDR, RAR) and the small-scale CDM problems (cusp-core, diversity, too-big-to-fail). That section is balanced and up to date. Section 3's distinction between a wide-field medium-shallow survey and a pointed snapshot survey of ~20,000 galaxies is practical, and Section 4 gives a solid state-of-the-art of tilted-ring, 3D, Bayesian, and uv-plane modelling tools. I learned a few things from the modelling section.\n\nThe soft spot is exactly what the reader flagged: the z~1 rotation-curve forecast in Sec 3.3. The 10,000-hour Band-1 survey is designed assuming the local HI mass-size relation (Wang et al. 2016) and local HI mass function (Jones et al. 2018) can be extrapolated to z~1. The chapter cites HI-stacking work that shows M_HI-M* evolves, but it cites nothing that anchors the size evolution. If high-z HI disks are more compact at fixed mass, the 1\" beam won't put ≥5 independent resolution elements across the major axis, and the outermost low-acceleration regions—the ones that distinguish MOND from CDM—would stay unresolved. This is a genuine gap. It's not a deal-breaker, because the chapter is a planning document rather than a measurement, but the 'closing the loop' language in Sec 5 leans on a forecast that would need that size-evolution anchor.\n\nMinor quibble: 'close the loop' is a bit grand for a chapter that itself stresses the remaining technical challenges. The body text is appropriately cautious.\n\nBottom line: this is a useful, competent chapter for anyone working in galaxy dynamics, HI surveys, or SKA planning. A brief caveat in Sec 3.3 about the size-evolution uncertainty would make it fully solid. It deserves to go through normal peer review and be published.","headline":"A competent, well-written SKA review chapter whose only real soft spot is the unanchored z~1 size-evolution assumption behind the 10,000-hour survey forecast.","tokens_in":32064,"tokens_out":3765,"would_cite":true,"duration_ms":32247,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"This chapter forecasts that the SKA-Mid telescope in its fourth array assembly (AA4) will deliver HI rotation curves at 1–2 arcsec resolution, for tens of thousands of galaxies, and out to redshift $z\\simeq1$, with enough statistical…","keywords":["21-cm HI line","galaxy rotation curves","dark matter models","modified gravity","SKA-Mid","dynamical scaling laws","cusp-core problem","high-redshift galaxy kinematics"],"falsifier":"A concrete check would be to carry out a deep, 10,000-hour-class Band-1 observation in a well-studied cosmological field and measure angular sizes of galaxies with HI masses near $10^{10}$ solar masses at $z\\simeq1$. If fewer than five independent beams fit across the major axis of the typical such disk, the paper's redshift-one rotation-curve forecast fails; if five or more beams are available and the outer velocity fields are mapped, the forecast is confirmed.","tokens_in":31098,"feed_emoji":"📡","tokens_out":9049,"duration_ms":85271,"temperature":0.7,"pith_summary":"Seeing the atomic-hydrogen (HI) gas in galaxies at 21 cm has long been one of the cleanest ways to map their outer gravitational pull, because HI disks reach well beyond their stars. This chapter argues that SKA-Mid AA4 will turn that technique into a large-scale test of dark matter: it should reach 1–2 arcsec resolution in inner galaxy regions, produce rotation curves for tens of thousands of galaxies instead of hundreds, and, for the first time, extend reliable HI rotation curves out to redshift $z\\simeq1$. If those forecasts hold, astronomers could compare how particle dark matter models and modified-gravity theories predict the tight scaling relations between baryons and galaxy dynamics should scatter, evolve with cosmic time, and depend on environment. The paper also argues that the data volume will force a shift from visual inspection to automated, Bayesian, and machine-learning kinematic modelling.","feed_headline":"SKA-Mid could map tens of thousands of galaxy rotation curves","feed_subtitle":"That would also extend HI rotation curves to redshift 1, testing particle dark matter against modified gravity.","key_machinery":"The load-bearing object is the spatially resolved HI rotation curve, measured from the 21-cm line and modeled as a rotating disk through the standard relation $V_{\\rm los}(x,y)=V_{\\rm sys}+V_{\\rm rot}(R)\\sin i\\cos\\theta$, with the inclination $i$ and position angle set by the projected ellipse. The forecast for what SKA-Mid AA4 can do rests on a quantitative resolution metric: a usable rotation curve requires a column-density sensitivity near $1\\,M_\\odot\\,{\\rm pc}^{-2}$ and at least five independent beams across the major axis, applied to the local HI mass function and HI mass–size relation together with cosmological redshift dimming and broadening. On the modelling side, the machinery is 3D forward modelling in datacubes, and eventually fitting in the $uv$-plane, to handle beam smearing in barely resolved galaxies, plus automated acceptance and quality-control criteria for samples of tens of thousands.","core_discovery":"On its own terms, the paper's central claim is that SKA-Mid AA4 will change HI rotation-curve studies from a small, heterogeneous enterprise into a statistical and cosmic-time machine. Three capabilities are forecast: targeted observations at about 1–2 arcsec resolution can resolve the inner rising parts of rotation curves and measure dark-matter core properties in dwarf galaxies; a wide or pointed survey can spatially resolve the HI kinematics of roughly twenty thousand galaxies across environments from clusters to voids; and a very deep, roughly 10,000-hour, Band-1 pencil-beam survey can place five or more independent resolution elements across the major axes of galaxies with HI mass near $10^{10}$ solar masses at $z\\simeq1$. The paper connects these capabilities to the empirical dynamical laws of galaxies—the baryonic Tully-Fisher relation, the central density relation, and the radial acceleration relation—and notes that particle dark matter and modified gravity predict opposite behaviors for the intrinsic scatter, redshift evolution, and environmental residuals of these laws. It concludes that discriminating between those predictions is what would allow the dark matter problem to be closed.","pith_inferences":["An implication the authors leave implicit: if the 10,000-hour Band-1 survey resolves only galaxies near HI mass $10^{10}$ solar masses at $z\\simeq1$, tests of the dynamical scaling laws at cosmic noon would be confined to the high-mass end, leaving the low-acceleration regime that most strongly separates dark matter from modified gravity unprobed beyond $z\\simeq0.1$.","The authors describe the $0.1\\lesssim z\\lesssim0.2$ radio-frequency-interference gap but do not state what follows: unless mitigation closes it, any claimed measurement of redshift evolution in the dynamical laws will need to interpolate across a substantial part of cosmic time, weakening the test.","Before SKA data arrive, the proposed five-beam resolution metric could be checked by running the same automated 3D fitting software on mock observations built from very high-resolution local HI datacubes, asking what fraction of realistic disk geometries actually pass the acceptance criteria at low signal-to-noise ratios."],"forward_implications":["At 1–2 arcsec resolution, dark-matter core sizes and inner density slopes in low-mass galaxies can be measured, testing cusp–core expectations and distinguishing feedback-driven core formation from self-interacting or fuzzy dark matter.","A sample of tens of thousands of resolved galaxies lets the baryonic Tully-Fisher, central density, and radial acceleration relations be measured with known selection functions, and their scatter compared with the tiny scatter expected from modified gravity versus the larger emergent scatter of galaxy-formation models.","HI rotation curves at $z\\simeq1$ would probe the outer, low-acceleration regions of massive galaxies, avoiding the disk–halo degeneracy that limits H$\\alpha$ and CO kinematics at cosmic noon.","Testing whether the dynamical laws evolve over roughly 8 Gyr and differ with environment would be one of the few observations capable of separating particle dark matter from modified gravity.","The automated 3D and $uv$-plane fitting tools required for this data volume would also make the selection function reproducible when comparing observations to cosmological simulations."],"supporting_citations":[{"why":"Provides the local HI mass–size relation used to convert predicted HI masses into disk angular sizes at all redshifts in the survey-yield estimates.","marker":"Wang et al. (2016)"},{"why":"Provides the HI mass function used to estimate how many galaxies a wide or pointed survey would spatially resolve.","marker":"Jones et al. (2018)"},{"why":"Supplies the cosmological redshift broadening and dimming corrections that set sensitivity requirements for $z>0$ HI surveys.","marker":"Meyer et al. (2017)"},{"why":"Supplies the simulated HI-selected galaxy cone used to illustrate what a deep survey could detect out to $z\\simeq1$.","marker":"Obreschkow et al. (2009)"},{"why":"Defines the current benchmark sample of local rotation curves with matched near-infrared photometry that SKA surveys would extend by two orders of magnitude.","marker":"Lelli et al. (2016b)"},{"why":"Establishes the radial acceleration relation from local rotation curves, one of the dynamical laws whose scatter, evolution, and environmental dependence would carry the dark-matter versus modified-gravity test.","marker":"Lelli et al. (2017b)"},{"why":"Provides the 3D tilted-ring fitting approach used to model poorly resolved HI datacubes while accounting for beam smearing.","marker":"Di Teodoro and Fraternali (2015)"},{"why":"Shows from pathfinder data how reliable kinematic models are in barely resolved galaxies and informs the automated acceptance criteria needed for large SKA samples.","marker":"Deg et al. (2022)"},{"why":"Describes the current WALLABY survey and its spatial resolution, setting the pathfinder benchmark that SKA-Mid AA4 would exceed.","marker":"Westmeier et al. (2022)"},{"why":"Holds the current record for a well-measured HI rotation curve at $z\\simeq0.083$, defining the pre-SKA limit that a $z\\simeq1$ survey would break.","marker":"Pickering et al. (1997)"}],"fun_headline_variants":["SKA-Mid to map 20,000 galaxy rotation curves","Testing dark matter with 20,000 HI rotation curves","SKA-Mid: HI rotation curves from z=0 to z=1","Probing dark matter and gravity with 20k galaxy curves","SKA-Mid will test dark matter with 20,000 rotation curves"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"For the $z\\simeq1$ part of the forecast, the load-bearing premise is that the local HI mass–size relation and HI mass function still describe galaxies at $z\\simeq1$; if high-redshift HI disks are smaller or more diffuse than that extrapolation, a 10,000-hour survey would not put five resolution elements across them and the outermost, dark-matter-dominated regions would remain unresolved.","fun_headline_variants_meta":{"raw":{"variants":["SKA-Mid to map 20,000 galaxy rotation curves","Testing dark matter with 20,000 HI rotation curves","SKA-Mid: HI rotation curves from z=0 to z=1","Probing dark matter and gravity with 20k galaxy curves","SKA-Mid will test dark matter with 20,000 rotation curves"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000479,"raw_usage":{"total_tokens":2417,"prompt_tokens":1034,"completion_tokens":1383,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":650,"completion_tokens_details":{"reasoning_tokens":1290}},"tokens_in":650,"tokens_out":1383,"duration_ms":9753,"temperature":1.0,"reasoning_tokens":1290,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T10:55:48.885700+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete check would be to carry out a deep, 10,000-hour-class Band-1 observation in a well-studied cosmological field and measure angular sizes of galaxies with HI masses near $10^{10}$ solar masses at $z\\simeq1$. If fewer than five independent beams fit across the major axis of the typical such disk, the paper's redshift-one rotation-curve forecast fails; if five or more beams are available and the outer velocity fields are mapped, the forecast is confirmed.","supporting_citations":[],"review_version":1}