{"id":"55e4d75d-6b60-4b80-a8e0-4009c6c650e4","arxiv_id":"2411.10124","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Neglecting frequency-dependent beams in a Simons Observatory-like analysis biases foreground parameters by more than 2 sigma and cosmological parameters by up to 0.3 sigma, making per-component color-corrected beams necessary.","lead":"This paper shows that ignoring the frequency dependence of a telescope's beam (beam chromaticity) can bias foreground and cosmological parameters in future high-resolution CMB surveys. It presents a formalism and a public-software implementation for including per-component color-corrected beams in power spectrum likelihood analyses.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Most load-bearing issue is model dependence of the quantitative forecast: the >2σ foreground and 0.3σ cosmology biases are obtained with Gaussian beams and ~30 GHz top-hat passbands (Eqs. 17-18), and are not yet tested against realistic SO-LAT beam shapes or passbands.","rationale":"The reader's weakest-assumption analysis already identified the Gaussian beam/passband idealization, and I agree that this is the correct weak spot. The paper's formalism is internally consistent: Eq. 9 follows from Eq. 2 when the per-frequency beam window function is known, and Eq. 13 is the correct achromatic limit; the validation against bplike (Appendix A) is genuine independent support. The remaining question is purely quantitative: do the bias magnitudes survive realistic beam shapes and scalings? Because the paper shows α=1 still produces >2σ foreground biases, and real diffraction-limited systems typically have α close to 2, the central conclusion is probably robust, but not proven for actual instruments. The CVL claim (one sentence, no setup) is a further extrapolation and would benefit from being made reproducible. Overall, this does not change the reader's ACCEPT: the paper delivers a sound formalism, a public implementation, and an illustrative forecast whose assumptions are transparent. The robustness test I propose is a natural next step rather than a blocker. Hence verdict remains UNCHANGED.","tokens_in":11604,"tokens_out":16500,"duration_ms":183725,"concrete_test":"Using the public LAT MFLike and fgspectra code, re-run the Section III forecasts replacing the Gaussian beams of Eqs. 17-18 with realistic SO-LAT beam models (from optical/GRASP simulations or planet measurements) and the actual measured passband shapes. Then recompute the Appendix B parameter shifts and compare the >2σ foreground and 0.3σ cosmological bias claims. If the biases persist, the concern is settled; if they drop below the quoted thresholds, the central claim must be qualified as model-dependent.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing point is that the quantitative necessity claim rests on a specific chromaticity model: Gaussian beams with FWHM(ν) = FWHM(ν0)(ν/ν0)^(-α/2), α ∈ {1, 1.5, 2}, and ~30 GHz top-hat passbands in surface-brightness units (Eqs. 17-18, Section III). The authors never probe α < 1 or non-power-law frequency dependence, nor realistic non-top-hat band shapes. If a real SO-like telescope has effectively weaker main-beam chromaticity (e.g., because feed illumination and edge taper make the beam FWHM scale more weakly than ν^-1, or because sidelobe structure changes the high-ℓ window function differently), the reported >2σ foreground biases and the unquantified 'several standard deviations' CVL statement could shrink. This is a model-dependence concern, not an internal inconsistency: the formalism (Eq. 9 vs Eq. 13) is correct and the bplike cross-check supports the implementation. But the abstract's claim that 'it is necessary' goes beyond the formalism and depends on the assumed beam/passband model, which is acknowledged in the text ('for simplicity, we can assume diffraction-limited Gaussian beams'). A dedicated robustness test is needed before applying the conclusion to a concrete instrument.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a formalism for including beam chromaticity in CMB power-spectrum likelihood analyses. The key idea is that the effective beam window function becomes component-dependent once foregrounds with different spectral energy distributions are observed through broad passbands: the foreground SED must be integrated against a frequency-dependent beam (Eqs. 9-11), rather than against a simple passband average (Eqs. 13-14). The authors implement this in the public SO likelihood stack (LAT MFLike and fgspectra), validate the implementation against the ACT bplike code (Appendix A), and forecast the bias incurred when chromatic beams are present in simulated spectra but omitted in the analysis. For an SO LAT-like setup with Gaussian beams whose FWHM scales as (nu/nu0)^(-alpha/2) with alpha = 1, 1.5, 2 and ~30 GHz top-hat passbands, they find >2 sigma biases on extragalactic foreground parameters and up to 0.3 sigma biases on Neff and H0, with larger cosmological biases in a cosmic-variance-limited case.","tokens_in":11904,"tokens_out":2188,"duration_ms":25509,"significance":"If the quantitative results hold, the paper makes a timely and useful point: next-generation high-resolution CMB experiments cannot treat the beam as a single achromatic window function when modeling foregrounds. The formal derivation in Section II is clean and appears correct: Eq. 9 follows from beam-weighted passband integrals, and the achromatic limit of Eq. 13 is recovered consistently. The implementation in public software, the numerical cross-check against bplike, and the forward-simulation setup (no circular fitting of the beam model) are concrete strengths. The main caveat is that the headline bias numbers are tied to an idealized beam model; the paper itself acknowledges this simplification. The formal framework is likely to be of lasting use regardless of the specific forecast values.","major_comments":[{"comment":"The quantitative claims in the abstract and conclusions---foreground biases larger than 2 sigma and cosmological biases up to 0.3 sigma (plus 'several standard deviations' in the CVL limit)---rest entirely on the assumed beam model: Gaussian profiles with FWHM(nu) = FWHM(nu0)(nu/nu0)^(-alpha/2) for alpha in {1, 1.5, 2}, combined with ~30 GHz top-hat passbands. The authors state 'for simplicity, we can assume diffraction-limited Gaussian beams,' and this is a legitimate first demonstration. However, the abstract's 'it is necessary' conclusion is stronger than what the simulation actually establishes, because no robustness test is given for weaker chromaticity (alpha < 1), non-power-law frequency dependence, realistic non-Gaussian beam shapes with sidelobes, or non-top-hat passband shapes. Since these assumptions are load-bearing for the headline bias magnitudes, I ask the authors either to add a sensitivity/robustness test (e.g., varying the chromaticity scaling and passband shape) or to temper the abstract and conclusions to present the numbers as an illustrative worst-case demonstration under an explicitly idealized beam model.","section":"Section III, Eqs. (17)-(18)"},{"comment":"The statement that in a cosmic-variance-limited survey 'the distortion of the foreground spectra can induce biases on cosmological parameters up to several standard deviations' is not accompanied by any experimental setup, noise level, multipole range, or quantitative figure. Given that this is one of the strongest motivations in the paper, the claim needs either a reference to a specific calculation or a direct forecast; otherwise it should be removed or explicitly labeled as a qualitative expectation.","section":"Section III, paragraph on the cosmic-variance-limited survey"}],"minor_comments":[{"comment":"There are typos in the captions and text: '1-dimentional' and '2-dimentional' should be 'one-dimensional' and 'two-dimensional', and 'baises' should be 'biases'.","section":"Appendix B, Figures 5 and 6"},{"comment":"The phrase 'mesurement errors' should read 'measurement errors'.","section":"Section III, text after Fig. 3"},{"comment":"The sentence describing the ACT DR4 treatment of color correction is informative but slightly awkward: 'Color-corrections for the other sky components were neglected' is followed by a long parenthetical. Consider splitting this into two sentences for readability.","section":"Section I, introduction"},{"comment":"It may help the reader to explicitly note that r^c_{\\ell,\\nu} is normalized to unit integral over the passband when multiplied by the beam, since this normalization is central to the interpretation of Eq. (11) as a beam-weighted SED average.","section":"Eq. (10)"}],"recommendation":"major_revision","confidential_remarks":"The formal framework and public implementation are solid, and the paper will likely be useful to the CMB community. The main revision needed is to align the strength of the abstract's claim with the idealized beam/passband assumptions, or to add a robustness test. This is a model-dependence issue, not an internal inconsistency, and it is fixable within the manuscript's scope. The paper is not an official SO collaboration paper, which is disclosed, and the self-citations to in-preparation works are clearly marked."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThis paper is a solid contribution. It generalizes the standard color-correction idea to a per-component chromatic beam in a full multi-frequency CMB likelihood, implements it in the public SO stack (LAT MFLike and fgspectra), validates the implementation against ACT's bplike at numerical precision, and quantifies the bias from ignoring beam chromaticity in an end-to-end forecast. The derivation in Section II is clean; Eq. 9 follows from beam-weighted bandpass integrals, and it reduces correctly to the achromatic limit. The bplike validation is real evidence, not just a claim.\n\nThe new result is the forecast: for an SO LAT-like configuration with 30 GHz top-hat passbands and Gaussian beams whose FWHM scales as ν^{-α/2}, neglecting per-component chromaticity biases extragalactic foreground parameters by more than 2σ and cosmological parameters by up to ~0.3σ (Neff, H0), with larger biases for a CVL survey. That is a meaningful within-field correction for next-generation analyses, even though it does not change cosmology by itself.\n\nThe main soft spot is model dependence of the headline numbers. The beams are assumed Gaussian and diffraction-limited (α up to 2), and the passbands are top-hat; real SO LAT beams have non-Gaussian wings, sidelobes, and more complex frequency dependence. If the true chromaticity is weaker than the assumed scaling, the 2σ foreground and 0.3σ cosmological biases shrink. The authors acknowledge this simplification explicitly and promise a more thorough SO analysis, but the abstract's \"necessary\" claim is a bit stronger than what the idealized model can support on its own. A robustness test with a non-Gaussian beam or at least α<1 would make the forecast much more convincing. The CVL \"several standard deviations\" statement is also unquantified, and the beam scaling parameterization rests on an in-preparation reference [48].\n\nNone of this undermines the formalism or the implementation, both of which check out. The paper is useful to anyone building or running high-resolution multi-frequency CMB likelihoods, and it deserves a serious referee. I would send it to review with a request for a robustness test and a slightly more cautious abstract.\n\nRecommendation: accept after minor revision, with the robustness caveat addressed.","headline":"A clean formalism and a validated implementation for per-component beam chromaticity in CMB likelihoods, with a forecast whose headline numbers rest on idealized Gaussian beams and deserve a robustness check.","tokens_in":12453,"tokens_out":3058,"would_cite":true,"duration_ms":30718,"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 paper shows that high-resolution CMB analyses must assign each sky component its own frequency-dependent beam, because neglecting this effect biases foreground parameters by more than 2 sigma.","keywords":["cosmic microwave background","beam chromaticity","beam window function","foreground spectral energy distributions","power spectrum likelihood","systematic bias","Gaussian beam model","damping tail"],"falsifier":"Measure the effective beam window function as a function of frequency inside a real 30 GHz-wide channel—for instance, from planet observations across the passband—and compare the predicted ~20% foreground distortion at ℓ = 9000 with the actual distortion; alternatively, run the same likelihood analysis on real data from a high-resolution survey with known passbands and see whether foreground parameters shift by the forecast >2σ when chromaticity is added.","tokens_in":11416,"feed_emoji":"🔭","tokens_out":6511,"duration_ms":60912,"temperature":0.7,"pith_summary":"This paper argues that beam chromaticity—the variation of a telescope's beam profile across its frequency passband—cannot be ignored in high-resolution cosmic microwave background analyses. It develops a formalism that folds a frequency-dependent beam into each sky component's spectral energy distribution separately, and tests what happens when data simulated with such chromatic beams are analyzed with the usual achromatic-beam likelihood. For passbands about 20% wide, neglecting the effect biases astrophysical foreground parameters by more than 2 standard deviations and shifts cosmological parameters such as the effective number of relativistic species and the Hubble constant by up to 0.3 sigma. The paper concludes that next-generation surveys must use a color-corrected beam for every sky component with a distinct spectrum, not a single beam per channel.","feed_headline":"Chromatic beams can bias CMB fits past 2 sigma","feed_subtitle":"New formalism requires a color-corrected beam for each sky component in high-resolution surveys.","key_machinery":"The load-bearing object is the geometric factor $r^c_{\\ell,\\nu} = \\tau^c_\\nu F_\\nu b^c_{\\ell,\\nu} / \\int d\\nu\\, \\tau^c_\\nu F_\\nu b^c_{\\ell,\\nu}$, the normalized product of the channel passband, the temperature-to-surface-brightness conversion, and the frequency-dependent beam window function. It converts the ordinary bandpass integral of a foreground SED into a scale-dependent effective SED, $\\hat{f}^{\\rm FG,c}_\\ell = \\int d\\nu\\, r^c_{\\ell,\\nu} f^{\\rm FG}_\\nu$, so that a beam that changes across the passband distorts each foreground component's angular power spectrum differently. The paper implements this in a public likelihood framework and adopts Gaussian beams with a diffraction-limited scaling of the full width at half maximum, $\\mathrm{FWHM}(\\nu) = \\mathrm{FWHM}(\\nu_0)(\\nu/\\nu_0)^{-\\alpha/2}$, with $\\alpha$ ranging from 0 to 2, to quantify the effect.","core_discovery":"The central claim is that for high-resolution CMB power-spectrum analyses, the beam window function must be treated as frequency-dependent and applied separately to each sky component with its own spectral energy distribution. The paper shows that the observed cross-spectrum between frequency channels factorizes into a CMB beam term and a foreground term whose effective SED is a frequency integral of the physical SED weighted by a normalized beam-passband product. When this chromatic correction is present in simulated data but omitted from the likelihood, the recovered foreground amplitudes shift by more than 2σ for the modeled experiment, and parameters measured from the small-scale damping tail, notably the effective number of relativistic species and the Hubble constant, shift by up to 0.3σ; in a cosmic-variance-limited survey the cosmological biases grow to several standard deviations. The paper also validates its implementation by matching the spectral shapes computed by an independent likelihood code.","pith_inferences":["If real beams have non-Gaussian wings or sidelobes that vary with frequency more than the Gaussian scaling assumed here, the foreground distortion could be even larger at high multipoles than the ~20% level quoted, since sidelobe power tends to grow with angular scale mismatch.","The same chromatic formalism implies that passband uncertainties and beam uncertainties will not be separable in future likelihoods; marginalizing over the frequency scaling of the beam may absorb part of the foreground signal, so joint constraints on passbands and beams from planet observations will be needed.","A testable extension is to apply the formalism to existing high-resolution survey data with measured passbands and beam models; if the predicted >2σ foreground shifts appear in real data, it would corroborate the paper's forecast, whereas null results would point to narrower effective passbands or weaker in-band beam variation."],"forward_implications":["Future high-resolution CMB experiments will need to measure and model the beam's in-band frequency dependence, not just its azimuthally averaged profile, for each sky component.","Foreground parameter estimates from analyses of the damping-tail region will be biased beyond statistical errors if chromaticity is ignored, complicating component separation and astrophysical interpretation.","Cosmological parameters sensitive to small angular scales, such as $N_{\\rm eff}$ and $H_0$, will inherit a systematic shift at the level of a few tenths of a sigma even when the CMB beam itself is perfectly calibrated.","In cosmic-variance-limited surveys, the same omission produces cosmological biases of several standard deviations, making the correction mandatory rather than optional.","The formalism generalizes to temperature and polarization and to any number of foreground components, so existing pipelines can adopt it without rebuilding their spectral models."],"supporting_citations":[{"why":"Supplies the Gaussian beam window function expression adopted for each spectral element.","marker":"[42]"},{"why":"The previous analysis that neglected chromaticity in the baseline likelihood, providing the baseline this paper moves beyond.","marker":"[6]"},{"why":"Provides the benchmark smooth simulation, analysis settings, and foreground parameter values used in the forecasts.","marker":"[40]"},{"why":"Supplies the foreground SED decomposition used to separate astrophysical components.","marker":"[25]"},{"why":"Gives the bandpass-integrated foreground power spectra treatment that the paper's chromatic formalism generalizes.","marker":"[37]"},{"why":"Used for the Legendre-transform relation between beam radial profile and window function.","marker":"[44]"},{"why":"Underpins the SED plus angular template factorization of foreground emission.","marker":"[47]"},{"why":"Source of the extra frequency scaling parameter alpha used in the beam model.","marker":"[48]"}],"fun_headline_variants":["Chromatic beams skew CMB fits past 2σ","Color-corrected beams key for high-res CMB","Beam chromaticity shifts CMB parameters","Ignoring beam chromaticity distorts CMB spectra"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The bias magnitudes rest on assuming Gaussian beams whose width scales as a power law across the passband and top-hat passbands about 30 GHz wide; if the true in-band beam variation is weaker or the effective passbands narrower, the reported shifts shrink.","fun_headline_variants_meta":{"raw":{"variants":["Chromatic beams skew CMB fits past 2σ","Color-corrected beams key for high-res CMB","Beam chromaticity shifts CMB parameters","Ignoring beam chromaticity distorts CMB spectra"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001419,"raw_usage":{"total_tokens":5696,"prompt_tokens":878,"completion_tokens":4818,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":494,"completion_tokens_details":{"reasoning_tokens":4755}},"tokens_in":494,"tokens_out":4818,"duration_ms":36801,"temperature":1.0,"reasoning_tokens":4755,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T19:56:21.151627+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the effective beam window function as a function of frequency inside a real 30 GHz-wide channel—for instance, from planet observations across the passband—and compare the predicted ~20% foreground distortion at ℓ = 9000 with the actual distortion; alternatively, run the same likelihood analysis on real data from a high-resolution survey with known passbands and see whether foreground parameters shift by the forecast >2σ when chromaticity is added.","supporting_citations":[{"cited_title":"Challinor, P","cited_arxiv_id":null,"evidence_quote":"Supplies the Gaussian beam window function expression adopted for each spectral element."},{"cited_title":"The Simons Observatory: impact of bandpass, polarization angle and calibration uncertainties on small-scale power spectrum analysis","cited_arxiv_id":"2403.05242","evidence_quote":"Provides the benchmark smooth simulation, analysis settings, and foreground parameter values used in the forecasts."},{"cited_title":"The Atacama Cosmology Telescope: Measurement and Analysis of 1D Beams for DR4","cited_arxiv_id":"2112.12226","evidence_quote":"Used for the Legendre-transform relation between beam radial profile and window function."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Source of the extra frequency scaling parameter alpha used in the beam model."}],"review_version":1}