{"id":"c835cabb-a505-4e49-93b4-fed8963b08f7","arxiv_id":"2508.20507","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":8,"one_line_summary":"HERA 21-cm power spectrum forecasts could constrain dark matter-baryon scattering cross-sections 5 to 10 times better than global 21-cm forecasts and far better than current CMB and satellite bounds.","lead":"This paper predicts how well the HERA radio telescope could detect dark matter that scatters off ordinary gas, by measuring fluctuations in the 21-cm hydrogen signal from the early universe. It finds that HERA power spectrum measurements could outperform earlier global signal forecasts by 5 to over 10 times, and beat current cosmic microwave background and satellite limits.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Omitted frictional heating in 21cmFirstCLASS is load-bearing for the n=-4 headline: Section 4 concedes it could counteract the drag-induced cooling that drives the projected limits, and the Fisher derivatives at sigma0=0 are computed without it.","rationale":"The paper is a forecast that rests on the simulation code 21cmFirstCLASS. The authors identify the omission of frictional heating as a known limitation of that code, but they do not quantify its impact. My reading of the physics is that this omission is the weakest link in the chain connecting the simulation output to the headline numbers. For the n=-4 case, the signal is produced by enhanced cooling in low-relative-velocity patches; frictional heating adds energy in exactly those patches, potentially canceling part of the cooling signature. Since the Fisher derivatives are evaluated at the fiducial CDM point, the omitted term contributes to the derivative at first order in sigma0, so it affects even the tiny cross-sections at the projected upper limits. The paper's suggestion that the effect is 'mild' is not supported by a calculation; the ratio of frictional to drag terms depends on m_chi and the local velocity-temperature balance and can be O(1) for m_chi around a few GeV. A dedicated rerun with the term included would settle whether the projected limits shift by a factor of about two or more. I agree with the reader that this is the load-bearing assumption, and I would keep the conditional verdict pending that check. Other limitations (fixed Population-III parameters, Fisher-to-upper-limit conversion) are secondary and do not change this assessment.","tokens_in":14783,"tokens_out":15195,"duration_ms":138630,"concrete_test":"Modify 21cmFirstCLASS to include the frictional heating term in the baryon temperature equation following Munoz et al. (2015), rerun the n=-4 Fisher forecasts at m_chi = 100 keV, 10 MeV, 1 GeV, and 100 GeV for both HERA noise scenarios, and recompute the 95% upper limits in Table 2 and the improvement factors over the global-signal forecasts (Rahimieh et al. 2025). If any upper limit degrades by more than ~0.3 dex, or the order-of-magnitude improvement over global forecasts at a given mass is lost, the central claim needs revision.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 4 explicitly states that 21cmFirstCLASS omits frictional heating from damping of the DM-baryon relative bulk velocity (the term proportional to V_chi_b^2 in the baryon temperature equation), while retaining the drag-induced coupling proportional to (T_chi - T_b). The paper concedes this could 'partially counteract the drag-induced cooling in the n=-4 case, reducing the amplitude of temperature fluctuations and mildly broadening the contours,' and that 'the projected upper limits may become less stringent in the regions where cooling is currently dominant.' That region is precisely the one responsible for the headline claim: the n=-4 improvement over global-signal forecasts and current CMB/MW-satellite bounds is driven by the cooling-generated power-spectrum enhancement at low DM masses (Figures 1, 3, 5; Table 2). Because the Fisher matrix is evaluated at the fiducial sigma0=0, the derivatives with respect to sigma0 include the first-order response of the omitted frictional term; it is not merely a small-sigma0 correction. For m_chi around or below a few GeV, the ratio of the frictional to drag terms can be of order unity or larger depending on the relative-velocity distribution and temperature, so the forecasted 95% upper limits (e.g., log10(sigma0/cm^2) = -45.41 at 10 MeV optimistic) could shift. The paper's 'mildly' is an assumption, not a result. Consequently, the claimed factor-of-five and order-of-magnitude improvements over global 21-cm forecasts are not yet robust.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript presents Fisher-matrix forecasts of the sensitivity of HERA 21-cm power-spectrum measurements to dark matter-baryon scattering, for cross sections parametrized as σ(v)=σ0 v^n with n=0 and n=-4. The authors use the 21cmFirstCLASS simulation pipeline to generate 21-cm power spectra, a 21cmSense-based noise model for two foreground scenarios (moderate and optimistic), and marginalize over six Population-II astrophysical parameters while fixing Population-III parameters. They report 95% upper-limit forecasts on log10 σ0 as a function of dark-matter mass (Figs. 5 and 6; Table 2), and compare them with CMB bounds, Milky Way satellite bounds, and their own global 21-cm forecasts (Rahimieh et al. 2025). The main quantitative claims are that HERA can improve on global-signal forecasts by at least a factor of five for n=0 and by more than an order of magnitude for n=-4, and can surpass current CMB and satellite constraints. An important caveat is acknowledged in Section 4: the simulation omits frictional heating from the damping of the DM-baryon relative bulk velocity.","tokens_in":15220,"tokens_out":5690,"duration_ms":53900,"significance":"If the forecasts are correct, the paper would make a useful case for 21-cm power-spectrum observations as a probe of DM-baryon interactions, extending the companion global-signal analysis and providing quantitative target constraints for HERA. The work uses public simulation and noise codes, and it is honest in flagging the main modelling limitation. However, the central quantitative claims rest on two issues that are not fully resolved: the Fisher parameterization at the σ0=0 fiducial is ambiguous, and the omitted frictional heating term is acknowledged to be relevant precisely in the n=-4 regime that drives the headline improvement. Because these issues affect the reported numbers themselves, rather than only their interpretation, the paper is not yet ready for publication in its present form.","major_comments":[{"comment":"The fiducial model sets σ0=0 (CDM), yet the results are presented as 95% upper limits on log10(σ0). Please state explicitly whether the Fisher derivatives in Eq. (6) are taken with respect to σ0 or log10 σ0. If the latter, the Fisher information at σ0=0 vanishes identically because ∂O/∂(log10 σ0) = σ0 ∂O/∂σ0 = 0 at the fiducial point, so the quoted limits cannot come from that parameterization. If the former, the conversion from a linear Gaussian error on σ0 to a 95% upper limit on log10(σ0) is a non-Gaussian transformation that must be described, and the entries of Table 2 should be checked under that transformation. This ambiguity is load-bearing because the headline improvement factors over global-signal forecasts and CMB/satellite bounds are derived from these numbers.","section":"§3.2, Table 1, Table 2"},{"comment":"The paper concedes that 21cmFirstCLASS omits frictional heating from the damping of the DM-baryon relative bulk velocity (the term proportional to V_χb^2), and states that this could 'partially counteract the drag-induced cooling in the n=-4 case' and 'mildly broadening the contours', but no quantitative support is given. Because the Fisher derivatives are evaluated at σ0=0, the linear response of this omitted heating term to σ0 contributes to the derivative that sets the projected upper limits; the omission is therefore a first-order effect on the forecast, not a subleading correction that only matters at large σ0. The n=-4 forecasts are driven by cooling-generated power-spectrum enhancement at low DM masses (Figs. 1 and 5, Table 2), which is exactly the regime (m_χ of order or below a few GeV, σ0 ≳ 10^-41 cm^2) where the paper itself says frictional heating becomes important. I ask the authors to either include the term, quantify its impact with an approximate calculation, or explicitly restrict the headline claims to the model without frictional heating. As it stands, the claimed factor-of-five and order-of-magnitude improvements are not robust.","section":"§4, final paragraph"},{"comment":"All Population-III parameters are fixed at their fiducial values, and the paper itself states that marginalizing over them 'could potentially broaden the uncertainties and relax our constraints'. Since the 21-cm power spectrum at Cosmic Dawn is sensitive to Population-III star formation (Hirano & Bromm 2018 is cited in this context), the abstract's unconditional claim that HERA 'can significantly improve sensitivity' is stronger than the evidence presented. The authors should either include Population-III marginalization, quantify the broadening with a sensitivity test, or qualify the abstract and conclusions accordingly.","section":"§3.2, §5"}],"minor_comments":[{"comment":"The text says the simulation is 'discretized into grid cells with a spatial resolution of 128 Mpc', which contradicts the later statement of 'approximately 1 Mpc' resolution and would not support convergence at k≈0.5 Mpc^-1; presumably '128^3 grid cells' is meant, and the wording should be corrected.","section":"§3.1"},{"comment":"The sign of the correlation between σ0 and X-ray luminosity in the n=-4 case is inconsistent: the abstract and Section 4 describe a positive correlation, while Section 5 says 'the X-ray luminosity ... is negatively correlated with the cross-section of interaction only in the Coulomb-like case'.","section":"§4 versus §5, Abstract"},{"comment":"Dark-matter mass entries use inconsistent capitalization ('kev', 'Mev', 'Gev'); these should be keV, MeV, GeV.","section":"Table 2"},{"comment":"The abstract's phrase 'improvement over global signal forecasts' overgeneralizes the comparison: Section 4 notes that the HERA n=0 forecasts are comparable to the 'Future 1' and 'Future 2' global-signal scenarios, and the factor-of-five claim is specifically relative to the EDGES-like scenario. The baseline should be named in the abstract and conclusions.","section":"Abstract, §4"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is built on public codes and the authors' companion global-signal paper; the central scientific question is timely and the comparison framework is appropriate. My main concern is that the abstract and conclusions state the sensitivity improvements without the qualifications that the authors themselves attach to the frictional-heating and Population-III assumptions in Section 4. A revision that quantifies or explicitly conditions those effects would be a much stronger paper."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a solid, transparent forecast paper from a group that knows the 21-cm/IDM literature. The genuinely new piece is the HERA-specific Fisher forecast for the power spectrum, not the global signal, with explicit marginalization over five Population-II astrophysical parameters and two foreground scenarios. It uses public, standard codes (21cmFirstCLASS, 21cmSense), does convergence checks, and gives a clear table of projected 95% limits across mass and interaction models. The comparison to CMB and Milky Way satellite bounds is useful. The correlation structure—especially the sigma0–L_X degeneracy in the Coulomb-like case and its absence for n=0—is cleanly presented.\n\nThe soft spots are real but localized. The main one is frictional heating. As the paper admits in Section 4, 21cmFirstCLASS includes the drag-induced cooling term but not the V^2 frictional heating from damping of the DM-baryon relative bulk velocity. The regime where that omission matters is precisely the low-mass, high-sigma0 corner where the n=-4 forecast is most aggressive. The Fisher derivatives are computed at the CDM fiducial, so an omitted term that is first-order in sigma0 affects the forecast at leading order, not as a small correction. The phrase 'mildly broadening the contours' is an assumption. It could be right, but the paper gives no quantitative argument. For a forecast whose headline is an order-of-magnitude enhancement for n=-4, this needs either a calculation or a much stronger caveat.\n\nSecond, the upper limits are quoted as 95% bounds on log10(sigma0), but the fiducial sigma0 is zero. A Fisher matrix at a physical boundary gives a symmetric error bar; converting that to a one-sided limit requires some extra step, and the paper doesn't say what it is. That's an easy fix but it should be stated.\n\nThe fixed Population-III parameters are a milder concern. The authors acknowledge it, and it mostly makes the absolute numbers optimistic, not the relative comparison within the paper.\n\nThe n=0 results are less affected by the frictional-heating issue; the factor-of-five improvement claim for n=0 should hold up better.\n\nCitation pattern is fine; the self-citation to the companion global-signal paper is natural since they compare against those forecasts.\n\nBottom line: send it to peer review. A good referee should push on the frictional heating and the upper-limit derivation. If the authors can quantify the frictional effect or soften the n=-4 claims accordingly, this will be a useful contribution. If not, the central n=-4 claim should be treated as provisional.","headline":"Clean, well-scoped Fisher forecast of HERA's 21-cm power-spectrum sensitivity to dark matter-baryon scattering, but the n=-4 headline numbers rest on an omitted frictional-heating term whose effect the authors concede but do not quantify.","tokens_in":15724,"tokens_out":3820,"would_cite":true,"duration_ms":35881,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["95.35.+d","98.80.-k"],"model":"deepseek-v4-flash","headline":"The 21-cm power spectrum can improve dark matter-baryon scattering limits by more than an order of magnitude.","keywords":["21-cm cosmology","dark matter-baryon scattering","interacting dark matter","velocity-dependent cross-section","HERA","Fisher matrix forecast","Cosmic Dawn","power spectrum"],"falsifier":"Recompute the $n=-4$ forecast with the frictional heating term (proportional to the square of the dark matter-baryon relative bulk velocity) included in the simulation's thermal evolution. If the 95% upper limit at $m_\\chi=10$ MeV moves from the optimistic value of $\\log_{10}(\\sigma_0/\\mathrm{cm}^2)\\approx -45.19$ to above $\\approx -44$, the core claim of an order-of-magnitude improvement in that cooling-dominated regime would fail.","tokens_in":14592,"feed_emoji":"📡","tokens_out":12391,"duration_ms":91525,"temperature":0.7,"pith_summary":"This paper forecasts how well the upcoming HERA radio interferometer could detect dark matter that scatters off ordinary baryons. Using simulated 21-cm power spectra and a Fisher matrix analysis, the authors claim that HERA measurements of spatial fluctuations in the cosmic hydrogen signal would constrain the interaction cross-section $\\sigma_0$ at least five times better than global 21-cm signal experiments for velocity-independent scattering, and more than an order of magnitude better for Coulomb-like scattering. These projected limits would also improve on existing bounds from the cosmic microwave background and from the abundance of Milky Way satellite galaxies. The result matters because 21-cm interferometry could become a leading probe of non-gravitational dark matter interactions during Cosmic Dawn, provided key astrophysical uncertainties are accounted for.","feed_headline":"21-cm maps could probe dark matter scattering 10x deeper","feed_subtitle":"HERA power-spectrum forecasts beat global-signal and CMB bounds, reaching cross-sections near log10 sigma0 = -45.","key_machinery":"The load-bearing object is the spherically averaged 21-cm power spectrum, $\\Delta^2_{21}(k,z)$, with per-bin uncertainties from a model of HERA's thermal noise and foregrounds. The Fisher information matrix takes derivatives of that spectrum with respect to the dark matter mass $m_\\chi$, cross-section normalization $\\sigma_0$, and five astrophysical parameters, and inverts them to get projected error contours. Physically, the chain runs from modified initial conditions (altered baryon temperature and dark matter-baryon bulk velocity) through a semi-numerical simulation of Ly-$\\alpha$ coupling, X-ray heating, and reionization to the brightness-temperature fluctuations. The input cross-section is a power law in relative velocity, $\\sigma(v)=\\sigma_0 v^n$, with $n=-4$ (Coulomb-like) and $n=0$ (velocity-independent) as the two benchmark cases.","core_discovery":"The central claim is that the 21-cm power spectrum, not just its sky-averaged global signal, carries a sharp imprint of dark matter-baryon elastic scattering, and HERA can exploit it. For a velocity-independent cross-section ($n=0$) the scattering delays structure formation and shifts the power-spectrum peak to higher frequencies, while for a Coulomb-like cross-section ($n=-4$) it cools gas in low-relative-velocity patches and amplifies fluctuations during Cosmic Dawn. The authors forecast 95% upper limits on $\\log_{10}(\\sigma_0/\\mathrm{cm}^2)$ as low as $-45.41$ for $n=-4$ at a dark matter mass of 10 MeV under optimistic HERA assumptions, improving over current CMB bounds by at least two orders of magnitude for $n=-4$ and over Milky Way satellite bounds by more than an order of magnitude for $n=0$. They also find that the cross-section is essentially uncorrelated with star formation efficiency and ionizing escape fraction, but is positively correlated with X-ray luminosity in the $n=-4$ case.","pith_inferences":["If frictional heating from bulk-velocity damping were included, the $n=-4$ low-mass sensitivity could weaken in exactly the regime where the largest claimed improvement lies; the paper's own discussion flags this possibility.","Marginalizing over Population-III star parameters, which the paper fixes, could introduce additional degeneracies and broaden the projected contours, potentially reducing the stated improvement factors.","The same Fisher pipeline could be applied to other upcoming interferometers or to combined datasets, where extra $k$-coverage and redshift range might push sensitivity further.","A joint fit to both the velocity index $n$ and $\\sigma_0$, rather than fixing $n$, would test whether the power spectrum can actually distinguish Coulomb-like from velocity-independent scattering."],"forward_implications":["Projected HERA 95% limits reach $\\log_{10}(\\sigma_0/\\mathrm{cm}^2)\\approx -45.4$ for $n=-4$ at 10 MeV in the optimistic foreground scenario, below current CMB and satellite bounds.","For $n=0$, HERA power-spectrum forecasts beat Milky Way satellite abundance constraints by over an order of magnitude and global-signal forecasts by at least a factor of five.","The $n=-4$ cross-section is positively correlated with X-ray luminosity, so accurate X-ray modeling is needed to avoid biasing dark matter inferences.","Star formation efficiency, escape fraction, and their power-law slopes do not correlate strongly with the cross-section, so the dark matter constraint is relatively robust to those astrophysical uncertainties.","The improvement comes from using the full redshift and scale information in the power spectrum rather than the sky-averaged signal alone."],"supporting_citations":[{"why":"Supplies the simulation pipeline used to compute 21-cm power spectra with dark matter-baryon scattering.","marker":"Flitter & Kovetz 2024a,b"},{"why":"Companion global 21-cm signal forecast whose sensitivity the power-spectrum forecasts are compared against to define the reported improvement factors.","marker":"Rahimieh et al. 2025"},{"why":"Provides the CMB bounds that the projected n=-4 limits are compared with and improve upon.","marker":"Nguyen et al. 2021"},{"why":"Provides Milky Way satellite abundance bounds that the projected n=0 limits are compared with and improve upon.","marker":"Maamari et al. 2021"},{"why":"Identifies the drag-induced cooling and frictional heating terms in dark matter-baryon scattering, including the frictional heating the simulation omits.","marker":"Muñoz et al. 2015"},{"why":"Basis of the semi-numerical simulation approach underlying the power-spectrum computation.","marker":"Mesinger et al. 2011"},{"why":"Supplies the noise and foreground model used to set per-k, per-z error bars for HERA.","marker":"Pober et al. 2013, 2014"},{"why":"Source of the Fisher matrix formulation tailored to the 21-cm power spectrum that the forecast is built on.","marker":"Mason et al. 2023"},{"why":"Describes the HERA instrument configuration whose frequency range, antenna layout, and observing parameters enter the noise model.","marker":"DeBoer et al. 2017"},{"why":"Provides the modified initial conditions for dark matter-baryon interactions used as input to the simulations.","marker":"Boddy & Gluscevic 2018; Gluscevic & Boddy 2018"}],"fun_headline_variants":["21-cm power spectrum can probe dark matter scattering 10x deeper than global signal","HERA 21-cm forecasts improve dark matter scattering sensitivity 10x or more","21-cm power spectrum beats CMB and satellite bounds on dark matter scattering","HERA 21-cm power spectrum forecasts improve dark matter scattering limits 10x"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The forecast rests on the simulation's omission of frictional heating from the damping of the relative bulk velocity between dark matter and baryons; including that heating could counteract the drag-induced cooling that drives the claimed $n=-4$ sensitivity, especially at dark matter masses below a few GeV, and would likely make the projected limits less stringent.","fun_headline_variants_meta":{"raw":{"variants":["21-cm power spectrum can probe dark matter scattering 10x deeper than global signal","HERA 21-cm forecasts improve dark matter scattering sensitivity 10x or more","21-cm power spectrum beats CMB and satellite bounds on dark matter scattering","HERA 21-cm power spectrum forecasts improve dark matter scattering limits 10x"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000973,"raw_usage":{"total_tokens":4213,"prompt_tokens":1103,"completion_tokens":3110,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":719,"completion_tokens_details":{"reasoning_tokens":3021}},"tokens_in":719,"tokens_out":3110,"duration_ms":18847,"temperature":1.0,"reasoning_tokens":3021,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T16:44:43.818830+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute the $n=-4$ forecast with the frictional heating term (proportional to the square of the dark matter-baryon relative bulk velocity) included in the simulation's thermal evolution. If the 95% upper limit at $m_\\chi=10$ MeV moves from the optimistic value of $\\log_{10}(\\sigma_0/\\mathrm{cm}^2)\\approx -45.19$ to above $\\approx -44$, the core claim of an order-of-magnitude improvement in that cooling-dominated regime would fail.","supporting_citations":[],"review_version":1}