{"id":"543a9b53-9ac7-435f-88a8-804a46e838bd","arxiv_id":"2412.14264","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Using two clustering mechanisms, concurrent collapse and gravitational clustering, the authors derive hot dark matter density profiles and recast CDM telescope bounds into axion-photon coupling limits for 1 to 10 eV axions.","lead":"This paper computes how light, fast-moving dark matter particles cluster around galaxies and then uses that clustering to turn existing telescope limits into new constraints on hot axions. The result is a new indirect-detection target for eV-scale particles that decay into photons.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Concurrent-collapse amplitude rests on a Maxwellian Jeans criterion for a species whose actual distribution is Bose-Einstein; an order-unity error in vc or in the EPS merger-tree normalization translates directly into the reported gaγ limits.","rationale":"The reader's weakest_assumption identifies the same load-bearing point: the concurrent-collapse amplitude, and hence the inner D-factor for massive halos, rests on a Maxwellian Jeans criterion applied to a Bose-Einstein relic. I agree that this is the pivotal assumption. The paper is honest about the limitation, and the framework is coherent, but the abstract and conclusions present the resulting constraints as stringent without attaching uncertainty bands to the dominant profile normalization. The check I propose would settle the concern: either a controlled simulation of CDM plus HDM or a linear kinetic-theory computation for the true phase-space distribution. Until one of these is done, CONDITIONAL remains the right verdict; I do not see a reason to reject the paper, since the recasting logic is clear and the limitations are disclosed. My recommendation is therefore to keep the reader's conditional verdict rather than upgrade to accept or downgrade to reject.","tokens_in":19656,"tokens_out":22950,"duration_ms":223284,"concrete_test":"Run a cosmological zoom N-body simulation with collisionless CDM plus a subdominant Bose-Einstein HDM component for representative masses mχ = 1, 4, 10 eV in a 10^15 Msun galaxy cluster and a 10^8–10^9 Msun dwarf, and compare the measured z=0 HDM density profile and line-of-sight D-factor to the sum of Eqs. 2.15 and 2.23. Independently, compute vc^BE from the linearized Vlasov-Poisson dispersion relation for the exact distribution of Eq. (1.1) in a CDM-dominated background and compare it to Eq. (2.6). If either the simulated D-factor or vc^BE differs from the paper's estimate by more than ~30%, the reported gaγ limits need explicit uncertainty bands or downward revision.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The reported limits scale as sqrt(D_CDM/D_a) (Eq. 4.5), and for the heavy halos that provide the most stringent recast (VIMOS clusters, and for ma above a few eV also dwarfs and the Milky Way) the inner D-factor is dominated by the concurrent-collapse term ρccχ = ρCDM Mχ,tot/Mvir (Eq. 2.15). Mχ,tot is built from Eq. 2.11 using a critical velocity vc derived from a Maxwellian Jeans radius (Eqs. 2.4–2.6) and a fraction ε of the unperturbed Bose-Einstein distribution with p < mχ vc (Eq. 2.10). The paper explicitly disclaims a rigorous treatment of gravitational instability for the actual phase-space distribution in Sec. 2.1 and repeats this limitation in Sec. 5. This is not a small perturbation: for a Bose-Einstein distribution the low-momentum phase-space density is enhanced relative to a Maxwellian, so both the unstable-mode boundary and the fraction of particles that participate in concurrent collapse can differ by an O(1) factor from the Maxwellian estimate. Because Mχ,tot enters the flux linearly and the coupling enters as its square root, an O(1) error in Mχ,tot shifts the quoted gaγ bounds by O(1), which is enough to move the claimed exclusion of the CAST-improving region. No kinetic-theory calculation or simulation is provided to calibrate this normalization, so the central claim is only conditionally supported.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper develops a framework for indirect detection of hot dark matter (HDM) that is a subdominant component of the dark matter. The HDM density profile around a CDM halo is modeled as the sum of a concurrent-collapse component, obtained by summing the HDM mass below a Jeans critical velocity over the EPS merger tree (Eqs. (2.9)-(2.15)), and a gravitational-clustering component obtained from a linearized Boltzmann equation (Eq. (2.23)). The authors define a D-factor and use the scaling gaγ = sqrt(D_CDM/D_a) gCDMγ (Eq. (4.5)) to recast published CDM axion-decay constraints from MUSE, WINERED, JWST, and VIMOS into 95% C.L. bounds on thermal axions in the O(1-10) eV mass range. They conclude that the new bounds improve on CAST over most of the range and on globular-cluster constraints for ma around 4.5-7.6 eV.","tokens_in":20011,"tokens_out":17575,"duration_ms":164809,"significance":"The recasting idea is clean, and the paper is careful to state its assumptions. Its strengths include a transparent derivation of the gravitational-clustering kernel with analytic asymptotics in Appendix B, explicit formulas for the D-factors, and the use of external CDM constraints rather than a fit to the axion signal, so there is no circularity in the limit-setting procedure. If the HDM profile model is correct, the resulting constraints are a useful new probe of thermal axions and, more generally, of eV-scale relics. The quantitative claims, however, rest on the uncalibrated Maxwellian Jeans criterion used to set the concurrent-collapse amplitude, which is the dominant contribution for the most constraining halos; an order-unity error in that amplitude translates into an order-unity shift in the reported coupling bounds. The paper also extends a linear clustering solution into the nonlinear regime and presents the final curves without propagated systematic uncertainties. These issues are addressable, but they make the current bounds conditional rather than final.","major_comments":[{"comment":"The central amplitude Mχ,tot is computed using a Jeans critical velocity vc derived for a Maxwellian distribution, while the HDM relic distribution is Bose-Einstein. The paper states explicitly in Sec. 2.1 that a rigorous treatment of gravitational instability for the actual phase-space distribution is beyond its scope, and repeats this limitation in Sec. 5. Because ρccχ(r) is proportional to Mχ,tot (Eq. (2.15)) and the recast bound enters as gaγ = (D_CDM/D_a)^{1/2} gCDMγ (Eq. (4.5)), an O(1) error in vc or in the efficiency factor ε(M,zf) of Eq. (2.10) shifts all reported coupling bounds by O(1). The low-momentum enhancement of the Bose-Einstein distribution relative to a Maxwellian can affect both the location of the unstable-mode boundary and the fraction of particles that participate in concurrent collapse. I ask the authors to calibrate this step with a kinetic-theory calculation or simulation, or to give conservative bracketed limits that cover the plausible range of vc and ε.","section":"Sec. 2.1-2.2, Eqs. (2.4)-(2.15)"},{"comment":"The gravitational-clustering profile is obtained from a linear perturbation of the homogeneous HDM distribution and is then used to compute total profiles and D-factors integrated to large radii. The paper acknowledges in Sec. 2.3 that the solution is formally valid only for δnχ/n̄χ ≪ 1 but is extended into the nonlinear regime. Since the final limits depend on the D-factor, and since for bosons the central clustering density of Eq. (2.27) contains a ln(2Λ)k_FS^2 enhancement relative to the fermionic case, the uncontrolled extrapolation can matter. Please quantify the nonlinear correction for representative halos, for example by comparison with N-body or Vlasov simulations for the HDM component, or demonstrate that the reported bounds are insensitive to it.","section":"Sec. 2.3-2.4, Eq. (2.23) and Fig. 3"},{"comment":"The WINERED curve labeled as a constraint in Fig. 6 is not a well-defined 95% C.L. limit. The text states that the CDM constraint from Ref. [29] combines Leo-V and Tucana II, that the two targets cannot be separated, and that the authors therefore use Leo-V alone and assume the combined constraint applies to it. This makes the teal-dashed curve an estimate whose statistical coverage is not defined. Either perform the full two-target likelihood recast or present this curve as an illustrative projection and exclude it from the set of claimed bounds.","section":"Sec. 3.2 and Sec. 4.3, Fig. 6"},{"comment":"The reported limit curves are presented without propagated uncertainties. The D-factors depend on the target virial masses, the concentration-mass relation of Eq. (3.9), the formation-redshift fitting function of Eq. (2.7) (which is extrapolated below Mvir = 1.5×10^10 Msun), and the Leo-V generalized NFW parameters. These inputs carry substantial empirical scatter, and an order-unity change in the D-factor ratio moves the bounds by order unity in gaγ. Please provide a representative uncertainty band or identify which input dominates, so that the reader can assess the robustness of the claimed improvement over CAST.","section":"Sec. 3.2 and Sec. 4.3, Eqs. (3.7)-(3.9), Fig. 6"}],"minor_comments":[{"comment":"Equation (2.30) does not appear to reproduce the efficiency factor ε defined in Eq. (2.10): for mχvc/Tχ → 0 the bracket tends to unity rather than to the small fraction of particles below vc, and for x = O(1) it differs from ε by a factor of a few. Please correct or remove this approximate expression.","section":"Eq. (2.30)"},{"comment":"The cosmological-stability requirement is mentioned but never written explicitly; please state the condition Γa→γγ ≲ H0 and indicate the corresponding cutoff in Fig. 6.","section":"Sec. 4.3, Fig. 6"},{"comment":"Reference [38] is cited as an unpublished communication; the VIMOS analysis should cite a published paper or the constraint should be omitted.","section":"References [37,38]"},{"comment":"The assumption that HDM and CDM line shapes are identical is justified by the coarse experimental resolution, but for the gravitational-clustering component the velocity dispersion can be comparable to the HDM thermal velocity; a quantitative sentence with the relevant velocity scales would be helpful.","section":"Sec. 3.2"},{"comment":"The caption and the text differ on whether the dot-dashed segment denotes the extrapolated regime; please clarify the line styles.","section":"Fig. 2"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within the scope of JHEP, and the recasting idea is worth publishing after revision. My main concern is the normalization of Mχ,tot; I would want to see either a calibration of the concurrent-collapse prescription or a clearly conservative bracketing of its amplitude before endorsing the quantitative bounds. The WINERED estimate and the unpublished VIMOS reference should also be cleaned up before the paper is accepted."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The thing to know: this is the first paper I have seen that gives hot dark matter a proper halo-profile framework for indirect detection rather than just scaling CDM profiles. The axion bounds it derives are new in the 1–10 eV mass window, but they are not as robust as the abstract implies. The load-bearing part of the profile calculation is the concurrent-collapse amplitude, and that amplitude rests on a qualitative Jeans estimate for a Bose-Einstein species.\n\nWhat is genuinely good: the authors extend the neutrino gravitational-clustering machinery to bosons (Eq. 2.23) and add an EPS merger-tree treatment to accumulate HDM that collapses with CDM across the halo's assembly history. The recasting logic — compare D-factors, rescale CDM limits via Eq. 4.5 — is transparent and honestly done. They also clearly flag the Maxwellian approximation in Sec. 2.1 and again in the conclusion. The relevant literature is cited, including the axion-limits database, and the input halo parameters are stated openly.\n\nThe soft spot is the one the stress-test note identifies. The critical velocity vc in Eqs. (2.4)–(2.6) is derived from a Maxwellian Jeans radius, while the actual phase-space distribution is Bose-Einstein. The low-momentum enhancement in a Bose-Einstein distribution can change both the unstable-mode boundary and the fraction of particles below vc by an O(1) factor. Because Mχ,tot enters the D-factor linearly and gaγ enters as its square root, an O(1) error in the amplitude shifts the quoted limits by O(1). For the halos doing the heavy lifting — the VIMOS clusters and, for higher masses, the Milky Way — concurrent collapse dominates the inner D-factor, so this is not a corner of parameter space. The authors admit the limitation but do not quantify it, and the abstract still says “stringent limits” with no uncertainty band. The nonlinear extension of the linear clustering solution is a real caveat but secondary, since clustering is subdominant in the inner regions that set the D-factor for the most stringent recasts. Halo-model and concentration–mass uncertainties also go unpropagated; those are probably minor compared with the vc issue.\n\nWho is this for: people working on eV-scale axions and on HDM structure formation. The framework is worth reading and reusing; the specific limits should be treated as order-of-magnitude constraints until the concurrent-collapse normalization is checked against simulations or kinetic theory.\n\nRecommendation: send it to review. It is coherent, honestly written, and the central limitation is disclosed even though it is under-weighted. A good referee can push for uncertainty bands or a softened abstract without needing to reject the paper.","headline":"First real HDM indirect-detection framework; new axion limits are plausible but the dominant profile term carries an acknowledged, unquantified O(1) systematic.","tokens_in":20517,"tokens_out":2040,"would_cite":true,"duration_ms":20413,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["95.35.+d"],"model":"deepseek-v4-flash","headline":"Thermal axions in the 1–10 eV range can be indirectly detected through their decay into photon pairs, and existing observations of dwarf galaxies, the Milky Way, and galaxy clusters already constrain the axion–photon coupling more tightly…","keywords":["hot dark matter","thermal axions","axion-photon coupling","indirect detection","D-factor","concurrent collapse","gravitational clustering","extended Press-Schechter"],"falsifier":"A cosmological simulation that follows a Bose–Einstein thermal relic of mass a few eV through structure formation could test the core prediction directly: if the measured inner density profile does not match the sum of Eq. (2.15) and the clustering profile, the recast bounds do not follow. Observationally, a targeted search for the two-photon decay line at wavelength $\\lambda = 2479.68\\,{\\rm nm}\\,({\\rm eV}/m_a)$ toward a galaxy cluster or dwarf galaxy would measure the product of the D-factor and the decay rate; comparing that flux with the predicted D-factor would settle whether the profile calculation is correct.","tokens_in":2115,"feed_emoji":"🔭","tokens_out":3459,"duration_ms":122913,"temperature":0.7,"pith_summary":"The paper proposes a way to detect hot dark matter — light particles that once thermalized with the Standard Model and survive today as a subdominant dark-matter component — by searching for their decay products in the densest parts of dark-matter halos. It argues that a hot relic's density profile around a cold dark matter halo is set by two mechanisms acting together: concurrent collapse, in which the slow tail of the relic's velocity distribution falls in with cold matter during halo formation and mergers, and gravitational clustering, in which the remaining particles accrete onto the finished halo. The two components have different radial shapes, and the paper combines them into a D-factor, the line-of-sight integral that controls the decay flux. Applied to thermal axions with masses in the 1–10 eV range, the framework yields new 95% confidence upper bounds on the axion–photon coupling that are stronger than helioscope limits over most of that range and stronger than globular-cluster limits near 4.5–7.6 eV. If the density-profile calculation holds, existing infrared and optical surveys are already sensitive to a class of particles that is otherwise very hard to probe.","feed_headline":"Telescope data bound hot axions beyond helioscope limits","feed_subtitle":"Recasting existing dark-matter decay searches turns dwarf galaxies, the Milky Way, and galaxy clusters into axion laboratories.","key_machinery":"The load-bearing objects are the two density-profile formulae. Concurrent collapse gives $\\rho^{\\rm cc}_\\chi(r) = \\rho_{\\rm CDM}(r)\\, M_{\\chi,\\rm tot}/M_{\\rm vir}$, with $M_{\\chi,\\rm tot}$ set by the extended Press–Schechter merger-tree integral of Eq. (2.11) over the Jeans-volume relic mass; this component is cuspy like the cold halo. Gravitational clustering gives the perturbation $\\delta\\tilde n_\\chi(k,\\eta)$ in Eq. (2.23), obtained from the linearized collisionless Boltzmann equation and expressible as a convolution of the cold halo overdensity with a window function controlled by the free-streaming scale; this component is flat inside and extended outside. The argument is carried by the recasting identity $g_{a\\gamma} = \\sqrt{D_{\\rm CDM}/D_a}\\, g^{\\rm CDM}_{\\gamma}$ (Eq. 4.5), which turns published cold-dark-matter decay limits into hot-dark-matter limits without requiring new observations.","core_discovery":"The paper's central claim is that the hot dark matter halo profile is the sum of a concurrent-collapse component and a gravitational-clustering component. The concurrent-collapse component is built from the halo's merger history: at each step, the fraction of relic particles with velocity below a Jeans-derived critical velocity $v_c$ is taken to collapse with the cold matter, so its density follows the cold halo profile with an amplitude fixed by an extended Press–Schechter integral over the progenitor mass function. The gravitational-clustering component is derived by solving the linearized collisionless Boltzmann equation for a Bose–Einstein relic in the time-independent potential of an NFW halo, giving the number-density perturbation in Eq. (2.23), which flattens in the inner region and extends farther out. The paper then observes that any existing cold-dark-matter decay constraint can be converted into a hot-dark-matter constraint through the D-factor ratio, $g_{a\\gamma} = \\sqrt{D_{\\rm CDM}/D_a}\\, g^{\\rm CDM}_{\\gamma}$, because the photon flux factors into a decay rate times a line-of-sight density integral. Carrying out this recast with dwarf-galaxy, Milky-Way, and galaxy-cluster data produces 95% C.L. bounds on the axion–photon coupling for thermal axions with masses in the $\\mathcal{O}(1{-}10)\\,{\\rm eV}$ range.","pith_inferences":["If the Maxwellian Jeans treatment overestimates the slow tail that collapses, the inner cusp amplitude — and hence the coupling bounds — would shift down by an order-unity factor; a simulation of Bose–Einstein relics in halos would settle this.","The framework suggests a morphology test: because gravitational clustering gives a flat, extended profile, a resolved decay line from a cluster should appear more extended than a cold-dark-matter decay line, providing an observational handle on which mechanism dominates.","The recast method could be inverted: a future nondetection at the predicted D-factor could be used to bound the relic's effective temperature today, since the profile amplitude depends on $T_{\\chi,0}$.","For fermionic relics, the phase-space bound from Liouville's theorem and Fermi–Dirac statistics caps the inner density, so the same framework predicts a suppressed or saturated inner signal for fermionic hot dark matter, a testable difference from the bosonic case."],"forward_implications":["Thermal axions with $m_a \\sim 1{-}10\\,{\\rm eV}$ and $T_{a,0}\\simeq 0.91\\,{\\rm K}$ are now excluded for couplings above the recast bounds, with the galaxy-cluster analysis giving the strongest limit.","Heavier halos produce larger hot-relic D-factors, so galaxy clusters are the most promising targets for future line searches.","The same D-factor recasting applies to any light relic whose decay into photons produces an unresolved line, including sterile-neutrino-type decays, not only axions.","Because concurrent collapse dominates the inner profile and gravitational clustering dominates the outer profile, the two mechanisms can in principle be separated by the radial shape of a detected signal.","For axion masses where the coupling is large enough that the axion decays within a Hubble time, the assumption of cosmological stability breaks down, so the reported bounds apply only below that maximum coupling."],"supporting_citations":[{"why":"Supplies the excursion-set progenitor mass function used to sum concurrent collapse over the merger tree.","marker":"[21]"},{"why":"Provides the assembly-redshift fitting formula that sets when each progenitor halo forms and hence the critical velocity.","marker":"[12]"},{"why":"Establishes the linearized Boltzmann calculation of relic clustering around cold halos that is extended to bosons.","marker":"[18]"},{"why":"Gives the NFW profile assumed for cold halos and for the concurrent-collapse component.","marker":"[17]"},{"why":"Defines the Maxwellian Jeans radius used to set the critical velocity $v_c$.","marker":"[11]"},{"why":"Galaxy-cluster search for decaying relic axions whose cold-dark-matter limit is recast into a hot-axion bound.","marker":"[10]"},{"why":"Dwarf-galaxy optical search for axion-like dark matter; its limit is rescaled by the D-factor ratio.","marker":"[28]"},{"why":"Ultra-faint dwarf search; the most sensitive dwarf constraint used in the recast.","marker":"[29]"},{"why":"Infrared-background search for decaying axions in the Milky Way; recast as a hot-axion bound.","marker":"[30]"},{"why":"Aggregated database of existing dark-matter axion constraints that provides the input cold-matter limits.","marker":"[32]"}],"fun_headline_variants":["Galaxy data recast as hot axion detectors","Dwarf galaxies and clusters tighten hot axion limits","Recasting dark matter decay to probe eV-scale axions","New axion-photon coupling bounds from telescope archives"],"cache_read_input_tokens":22528,"weakest_assumption_plain":"The amplitude of the concurrent-collapse component assumes that every hot relic particle slower than a Jeans-derived critical velocity, computed from a Maxwellian distribution rather than the true Bose–Einstein one, collapses exactly like cold dark matter at every merger step; since this component dominates the inner density and the D-factor, any order-unity error in that critical velocity or in the merger-tree efficiency shifts all the reported coupling bounds by a comparable factor.","fun_headline_variants_meta":{"raw":{"variants":["Galaxy data recast as hot axion detectors","Dwarf galaxies and clusters tighten hot axion limits","Recasting dark matter decay to probe eV-scale axions","New axion-photon coupling bounds from telescope archives"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00045,"raw_usage":{"total_tokens":2280,"prompt_tokens":972,"completion_tokens":1308,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":588,"completion_tokens_details":{"reasoning_tokens":1244}},"tokens_in":588,"tokens_out":1308,"duration_ms":11438,"temperature":1.0,"reasoning_tokens":1244,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T12:23:32.959027+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A cosmological simulation that follows a Bose–Einstein thermal relic of mass a few eV through structure formation could test the core prediction directly: if the measured inner density profile does not match the sum of Eq. (2.15) and the clustering profile, the recast bounds do not follow. Observationally, a targeted search for the two-photon decay line at wavelength $\\lambda = 2479.68\\,{\\rm nm}\\,({\\rm eV}/m_a)$ toward a galaxy cluster or dwarf galaxy would measure the product of the D-factor and the decay rate; comparing that flux with the predicted D-factor would settle whether the profile calculation is correct.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the excursion-set progenitor mass function used to sum concurrent collapse over the merger tree."},{"cited_title":"Power, A","cited_arxiv_id":null,"evidence_quote":"Provides the assembly-redshift fitting formula that sets when each progenitor halo forms and hence the critical velocity."}],"review_version":1}