{"id":"bd4033c8-3f95-4c2d-923a-aff01820ffa1","arxiv_id":"2502.08913","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A 26.4-hour search with FAST finds no axion gegenschein toward Vela, constraining the axion-photon coupling to below about 2e-10 GeV^-1 for axion masses 8.7-9.44 and 10.85-12.01 micro-eV.","lead":"Using the Five-hundred-meter Aperture Spherical radio Telescope (FAST), astronomers searched for a faint radio counter-image of the Vela supernova remnant, which would appear if dark matter consists of axions. They found no signal, placing new limits on how strongly axions interact with light.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The quoted gaγγ limit hinges on the unmeasured early-time Vela luminosity: Eq. (4) fixes free-expansion luminosity to a constant and no SNR-model systematic is propagated into the 95% bound.","rationale":"The reader's CONDITIONAL verdict is appropriate. The most load-bearing assumption is the Vela luminosity history, because it enters multiplicatively in the predicted signal and hence in the derived exclusion. The paper is transparent about the modeling uncertainty (Sec. 2.2) but does not fold it into the error budget, so the quoted number may overstate the confidence in the constraint. The OFF-weight optimization on the same data is a real but secondary concern; it affects the noise estimate at the tens-of-percent level at most, whereas the luminosity history can change the limit by a factor of order unity or more. I would keep the verdict CONDITIONAL: the null detection itself is credible, and the limit is likely conservative if Vela was indeed brighter in the past (as Bietenholz et al. 2021 suggests), but the published bound should be read as conditional on the SNR model, and the proposed check would quantify that conditionality.","tokens_in":26716,"tokens_out":14800,"duration_ms":167508,"concrete_test":"Recompute the flux in Eq. (4) and the resulting 95% upper limit on gaγγ under three alternative early-light-curve models: (i) no radio emission before t_MFA/2 with a linear rise to the fiducial constant; (ii) free-expansion luminosity reduced by factors of 3 and 10; (iii) the Sedov-Taylor power law continued down to a minimum time t_min=10 yr instead of the constant plateau. Keep all other Table 1 parameters fixed. If any alternative shifts the limit above 3e-10 GeV^-1 in the mass windows, the reported constraint is not robust to the principal astrophysical assumption.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is an upper limit on gaγγ whose normalization is set by the predicted gegenschein flux in Eq. (4). The integral over the line of sight is dominated by large xd (early emission times) because the Sedov-Taylor factor (t/t0)^{-4p/5} grows toward t_MFA, and the separate free-expansion integral contributes a comparable amount at the far end. The luminosity during that early phase is unobserved; Sec. 2.2 sets it to an arbitrary constant, citing 'large modeling uncertainties.' The paper does not propagate this uncertainty into the reported limit. If Vela's early radio luminosity was lower than assumed, the expected signal is smaller, and the true 95% upper limit on gaγγ is weaker than the quoted 2e-10 GeV^-1. This is not a minor calibration effect: it directly scales the particle-physics constraint that is the paper's headline result.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports the first search for an axion gegenschein image of the Vela supernova remnant using 26.4 hours of effective ON-OFF data from the FAST L-band 19-beam receiver. Assuming a Milky Way NFW dark matter halo and a parameterized Vela SNR luminosity evolution model taken from Sun et al. (2022), the authors predict a faint spectral line signal at frequencies corresponding to half the axion mass. They find no convincing candidate after RFI flagging, bandpass and flux calibration, baseline and standing-wave removal, and a two-subgroup cross-check of eight candidate excesses. From the null detection they derive a 95% upper limit g_{aγγ} ≲ 2×10^{-10} GeV^{-1} in the mass ranges 8.7–9.44 μeV and 10.85–12.01 μeV, noting that this is stronger than the FAST galaxy-cluster search limit but about a factor of three weaker than the CAST limit, and that roughly 2000 hours of observation could reach ~10^{-11} GeV^{-1}.","tokens_in":26812,"tokens_out":5136,"duration_ms":57187,"significance":"If the limit is robust, this is the first observational constraint on axion dark matter from the gegenschein effect and demonstrates a new observational probe of axion dark matter in the ~10 μeV mass range. The paper is careful in its data processing: it includes detailed RFI flagging, a mock-signal injection check for the standing-wave removal, a candidate search with a quantitative cross-check between independent data subsets, and an explicit comparison of achieved versus thermal-noise sensitivity. These are real strengths. However, the central limit is model-dependent: the predicted signal flux in Eq. (4) depends on the unobserved early-time luminosity of Vela and on the assumed dark matter halo profile, and the paper does not propagate these systematics into the quoted bound. The reported constraint should therefore be understood as conditional on the adopted SNR and halo models.","major_comments":[{"comment":"The free-expansion phase of the Vela SNR luminosity is unobserved and is set to a constant, with the text noting 'large modeling uncertainties' but no propagation of this uncertainty into the result. Since the predicted gegenschein flux S_g is proportional to the luminosity integral over x_d, and the derived limit scales as g_{aγγ} ∝ S_g^{-1/2}, an unmodeled factor of 2 in the early-time luminosity changes the quoted 2×10^{-10} GeV^{-1} limit by roughly 40%. The 'conservative' choice of a constant is conservative only relative to a brighter past; if Vela was dimmer in the free-expansion phase, the true limit would be weaker. Please propagate the Table 1 parameter uncertainties (t_MFA, p, t0) and explicitly test a dimmer early-luminosity case, as well as the alternative electron model Sν ∝ t^{-2(p+1)/5} listed in Table 1, and show how the limit shifts.","section":"Sec. 2.2, Eq. (4) and Table 1"},{"comment":"The dark matter density is fixed to an NFW profile with r_s = 16 kpc and local density 0.46 GeV/c^2/cm^3, but the line-of-sight integral in Eq. (4) is dominated by large x_d, where halo profile uncertainties are largest. The paper does not quantify the sensitivity of the limit to reasonable alternative halo models (e.g., Einasto or cored profiles) or to the quoted uncertainties on the local density from Sivertsson et al. (2018) and Nitschai et al. (2020). Because the limit again scales as the inverse square root of the integrated column density, this systematic should be evaluated and reported.","section":"Sec. 2.1, Eq. (3)"},{"comment":"The OFF-source weights w_OFF,i are optimized to maximize the S/N of the final spectrum using the same data that are later used to set the upper limit, and the noise RMS entering the likelihood in Eq. (24) is measured from that same weighted spectrum. This introduces data-fitting degrees of freedom that are not accounted for in the statistical interpretation. The authors should demonstrate with signal-injection simulations that this weight optimization does not bias the 95% bound, or alternatively quote the limit for fixed weights, which would make the analysis more conservative and easier to interpret.","section":"Sec. 4.4 and Sec. 5.1"}],"minor_comments":[{"comment":"The title and header contain 'V ela' with a stray space; this should be corrected to 'Vela'.","section":"Title and header"},{"comment":"Equation (4) is typeset with ambiguous integral limits and the notation 'tMF A' should be written as t_MFA; please clarify the integrands and the limits of integration.","section":"Sec. 2.2, Eq. (4)"},{"comment":"The arPLS smoothness parameter is stated as λ = 10^3, but no sensitivity test around this value is given; a brief statement of robustness to λ would strengthen the baseline-subtraction description.","section":"Sec. 4.4"},{"comment":"The sentence 'which had been used in the the Com...' contains a duplicated article; please remove the extra 'the'.","section":"Sec. 3.1"},{"comment":"The term 'galactic HI' should be written as 'galactic H I' to follow standard astronomical nomenclature.","section":"Sec. 4.1"},{"comment":"In Eq. (22), the symbols η_A and A_illu are not defined in the text; please define them explicitly or use the notation introduced in Eq. (12) for consistency.","section":"Sec. 5.1"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within scope for a radio astronomy / astroparticle journal and represents a genuinely new observational probe. The main issue is that the headline limit is conditional on unpropagated SNR luminosity and halo-profile systematics; the weight-optimization issue is also worth a careful response. If the authors can quantify how the limit shifts under reasonable model variations and add the requested robustness checks, the paper would be suitable for publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First, the one thing to know: this is the first real attempt at the axion gegenschein, and the null detection is credible. The quoted limit, g < about 2e-10 GeV^-1, is not robust on its own, because it depends on an unmeasured historical luminosity for Vela that is not propagated into the error.\n\nThe genuinely new content is the data: 26.4 hours of FAST ON-OFF observations, a full reduction pipeline, and candidate screening. The pipeline is careful—RFI flagging, noise-diode calibration, arPLS baseline subtraction, standing-wave removal, and then a sub-group cross-check of any candidate. Eight candidates were found; none survived. That is honest practice.\n\nThe paper also does a good job quantifying its noise floor. The observed RMS is 3–6 times the radiometer expectation, and they track this to residual RFI, baseline, and standing waves. Section 6 is open about calibration error and pointing error, and gives percentage impacts. This is how a pilot search should present systematics.\n\nThe soft spots are two. First, the model dependence of the limit. Equation (4) is dominated by early emission times, where Vela's luminosity is not observed. The authors set a constant for the free-expansion phase, citing large modeling uncertainties, and call it conservative. That direction is plausible—Vela may have been brighter early on—but the paper does not propagate either the luminosity uncertainty or the NFW profile parameters into the final constraint. If Vela was dimmer in the past, the true limit would be weaker. This is the main issue; it does not kill the paper, but it should be quantified before the number is used in a physics comparison.\n\nSecond, the OFF-source weights are tuned on the same data that later set the limit. That is a mild form of parameter fitting on the noise estimate. The authors do check sub-groups, so the effect is probably small, but a sensitivity test would be cheap and should be requested.\n\nWho it is for: people in axion indirect detection, and anyone designing weak spectral-line searches with single-dish telescopes. It is a useful demonstration that gegenschein searches are practical, with a clear path to better sensitivity.\n\nRecommendation: send it to peer review. Ask for a systematic propagation of the SNR luminosity and density profile uncertainties into the limit, and a check that the weight-tuning does not bias the limit. With those, the paper will be solid.","headline":"First gegenschein search is credible but its headline limit rests on an unquantified Vela luminosity assumption.","tokens_in":27448,"tokens_out":4168,"would_cite":false,"duration_ms":42368,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The first search for an axion gegenschein image of Vela finds no signal and limits the axion-photon coupling to below 2e-10 GeV^-1.","keywords":["axion dark matter","axion-photon coupling","axion gegenschein","stimulated decay","Vela supernova remnant","FAST radio telescope","null detection","radio astronomy"],"falsifier":"Recompute the expected flux using the alternative electron model listed in Table 1, $S_\\nu \\propto t^{-2(p+1)/5}$, in place of $S_\\nu \\propto t^{-4p/5}$: if the resulting 95% upper limit on $g_{a\\gamma\\gamma}$ rises above the CAST bound, then the quoted constraint is not robust to the luminosity-evolution assumption, whereas a measurement tying down Vela's early radio light curve would decide which model is correct.","tokens_in":26460,"feed_emoji":"📡","tokens_out":8102,"duration_ms":72943,"temperature":0.7,"pith_summary":"This paper reports the first observational search for an axion gegenschein, a radio counter-image formed when radiation from a bright source stimulates the decay of axion dark matter along the line of sight, using Vela supernova remnant as the primary source and 26.4 hours of effective ON-OFF data from the FAST L-band 19-beam receiver. The search covers axion masses of 8.7 to 9.44 micro-eV and 10.85 to 12.01 micro-eV, corresponding to the two cleanest frequency bands 1050-1129 MHz and 1313-1450 MHz, and finds no convincing line signal. From the null result the paper derives $g_{a\\gamma\\gamma} \\lesssim 2 \\times 10^{-10}\\,\\mathrm{GeV}^{-1}$ at 95% confidence in those mass ranges, a limit stronger than the existing FAST galaxy-cluster decay search in the same mass range and about a factor of three weaker than the CAST helioscope bound. The paper further shows that with roughly 2000 hours of ON-OFF observing the same technique should reach $g_{a\\gamma\\gamma} \\sim 10^{-11}\\,\\mathrm{GeV}^{-1}$, which would go below the CAST limit.","feed_headline":"No axion counter-image of Vela; coupling capped at 2e-10","feed_subtitle":"A 26.4-hour FAST search finds no Vela axion echo in two mass windows, tightening radio limits and pointing to a route past the lab…","key_machinery":"The central object is the axion gegenschein flux formula, $S_g = (\\hbar c^4 g_{a\\gamma\\gamma}^2 / 16) \\int_0^{t_0 c/2} S_\\nu(\\nu_a, x_d) \\rho(x_d) dx_d$, specialized to Vela as Eq. 4 with a luminosity history split into a free-expansion phase of constant specific luminosity and a Sedov-Taylor phase where the luminosity declines as $t^{-4p/5}$. This integral converts a null spectrum into a bound on $g_{a\\gamma\\gamma}$, because the observed flux scales as $g_{a\\gamma\\gamma}^2$ and the primary-source history plus the dark-matter profile fix the proportionality constant. Around this sit a Doppler-Gaussian line profile with width $\\sigma_i = \\nu_i \\sigma_d/c$, a two-dimensional Gaussian beam model for the FAST 19-beam receiver, and a likelihood-ratio test statistic $q_{g_{a\\gamma\\gamma}}$ whose 95% exclusion threshold is 2.71.","core_discovery":"The central claim is that if axions with masses in the two covered windows make up the Milky Way dark matter halo, they do not produce a detectable gegenschein image of Vela at the sensitivity of this pilot observation, so the axion-photon coupling must be below about $2 \\times 10^{-10}\\,\\mathrm{GeV}^{-1}$ there. The predicted signal flux is proportional to $g_{a\\gamma\\gamma}^2$ times an integral over distance of the Vela radio luminosity at the Doppler-broadened frequency $\\nu_a = m_a c^2/2h$ and the dark-matter density. After RFI flagging, bandpass and gain calibration, ON-OFF subtraction, baseline fitting, and standing-wave removal, the nineteen beams and four OFF-source positions are weighted and stacked; eight candidate spectral features are found, but none survives a time-split subgroup cross-check, so the data are treated as null and the noise level is converted into a 95% upper limit using a likelihood-ratio test.","pith_inferences":["If Vela's early free-expansion luminosity were substantially lower than the assumed constant, which the paper itself flags as unknown and conservatively modeled, the quoted limit would weaken roughly as the square root of the flux reduction, so a firmer early light curve would sharpen the result.","The gap in coverage between 9.44 and 10.85 micro-eV is set by radio-frequency interference; targeted RFI mitigation or an observing site with a cleaner spectrum could close that mass window without additional integration time.","The same data set can be re-interpreted as a probe of the dark-matter velocity structure, because a detected line's width would map the velocity dispersion and a null result with a known source could constrain line-of-sight dark-matter clumpiness.","If axions make up only a fraction $f$ of the dark matter, the coupling limit scales roughly as $f^{-1/2}$, so future searches should report limits as a function of $f$ to stay comparable across cosmological and local dark-matter models."],"forward_implications":["If the limit is correct, axions in the mass windows 8.7-9.44 and 10.85-12.01 micro-eV cannot constitute the dark matter with coupling above about $2 \\times 10^{-10}\\,\\mathrm{GeV}^{-1}$, narrowing the allowed parameter space for QCD axion and axion-like-particle models.","The successful end-to-end search demonstrates that a single-dish telescope with ON-OFF observing can place competitive axion constraints, so the same pipeline can be pointed at other bright, nearby radio sources to build a multi-source axion search.","The projected sensitivity of about $10^{-11}\\,\\mathrm{GeV}^{-1}$ with roughly 2000 hours of integration would beat the current CAST limit in these mass ranges, making radio telescopes competitive with laboratory helioscopes for axion dark matter.","The eight candidate features that failed the subgroup cross-check imply that any future claimed axion line must appear stably across time-split data subsets before being treated as a detection."],"supporting_citations":[{"why":"Supplies the gegenschein theory, the choice of Vela as an optimal FAST target, and the luminosity evolution model that becomes Eq. 4.","marker":"Sun et al. 2022"},{"why":"Introduces the stimulated-decay counter-image concept and the back-to-back photon kinematics that define the expected signal.","marker":"Ghosh et al. 2020"},{"why":"Extends the axion counter-image calculation and supports the search for a gegenschein image of a bright radio source.","marker":"Arza & Todarello 2022"},{"why":"Provides the CAST helioscope limit $g_{a\\gamma\\gamma} \\le 6.6 \\times 10^{-11}\\,\\mathrm{GeV}^{-1}$, the benchmark against which the new constraint is compared.","marker":"Anastassopoulos et al. 2017"},{"why":"Gives the earlier FAST galaxy-cluster direct axion decay search whose limit this work improves upon in the same mass range.","marker":"Guo et al. 2024"},{"why":"Supplies the likelihood-ratio and upper-limit test-statistic framework used to set the 95% confidence bound.","marker":"Cowan et al. 2011a"},{"why":"Provides the NFW dark-matter density profile used for the line-of-sight density $\rho(x_d)$ in the signal integral.","marker":"Navarro et al. 1996"},{"why":"Provides the two-phase free-expansion and Sedov-Taylor Vela evolution model and the relation between electron and synchrotron spectral indices.","marker":"Sushch & Hnatyk 2014"},{"why":"Indicates that Vela was probably even brighter during the earliest free-expansion phase, supporting the paper's conservative constant-luminosity treatment.","marker":"Bietenholz et al. 2021"},{"why":"Supplies the FAST L-band beam full-width half-maximum and aperture efficiency used in beam modeling and absolute flux calibration.","marker":"Jiang et al. 2020"}],"fun_headline_variants":["No Vela axion echo; coupling limited to 2e-10","FAST finds no axion counter-image of Vela, caps coupling","Null axion gegenschein search tightens limits on dark matter","26-hour FAST search yields no Vela axion glow, new coupling bound"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The predicted gegenschein flux assumes a particular history for how bright Vela's radio emission was over the past 12,000 years, especially during the unobserved first centuries after the supernova, and if Vela was dimmer then, the derived coupling limit would be looser.","fun_headline_variants_meta":{"raw":{"variants":["No Vela axion echo; coupling limited to 2e-10","FAST finds no axion counter-image of Vela, caps coupling","Null axion gegenschein search tightens limits on dark matter","26-hour FAST search yields no Vela axion glow, new coupling bound"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000539,"raw_usage":{"total_tokens":2678,"prompt_tokens":1132,"completion_tokens":1546,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":748,"completion_tokens_details":{"reasoning_tokens":1466}},"tokens_in":748,"tokens_out":1546,"duration_ms":12027,"temperature":1.0,"reasoning_tokens":1466,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T23:13:18.686345+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute the expected flux using the alternative electron model listed in Table 1, $S_\\nu \\propto t^{-2(p+1)/5}$, in place of $S_\\nu \\propto t^{-4p/5}$: if the resulting 95% upper limit on $g_{a\\gamma\\gamma}$ rises above the CAST bound, then the quoted constraint is not robust to the luminosity-evolution assumption, whereas a measurement tying down Vela's early radio light curve would decide which model is correct.","supporting_citations":[{"cited_title":"2024, Physics Letters B, 852, 138631, doi: 10.1016/j.physletb.2024.138631","cited_arxiv_id":null,"evidence_quote":"Gives the earlier FAST galaxy-cluster direct axion decay search whose limit this work improves upon in the same mass range."},{"cited_title":"2014, A&A, 561, A139, doi: 10.1051/0004-6361/201322569","cited_arxiv_id":null,"evidence_quote":"Provides the two-phase free-expansion and Sedov-Taylor Vela evolution model and the relation between electron and synchrotron spectral indices."}],"review_version":1}