{"id":"487e2caf-c89f-40f7-bc0d-a53dc6a17238","arxiv_id":"2411.18945","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"A frequency-dependent ('chromatic') antenna can in principle recover the cosmological recombination signal, provided its beam changes smoothly and modestly, within site- and time-dependent tolerances.","lead":"This simulation study asks whether the extremely faint cosmological recombination radiation (CRR) can still be detected when the observing antenna's beam changes with frequency. Using toy-model beam wobbles and stretches, it finds that certain smooth types of chromaticity still permit a CRR detection, with tolerances that depend on observing site and local sidereal time.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The paper's detection metrics and tolerance table are computed from noiseless simulations, with no specified noise amplitude or sensitivity model; since the CRR is ~10 nK, realistic receiver noise would dominate and erase the claimed distinguishability.","rationale":"The reader's weakest assumption is precisely the missing noise specification, and I agree it is the load-bearing concern. The paper is transparent about its toy-model treatment and explicitly calls itself a 'first step,' so it does not overclaim maturity; however, the abstract's statement that detection is 'indeed possible' with a chromatic antenna goes beyond what the noiseless simulations show. The metrics γ and ϱ are well-defined and the pipeline is reproducible in principle, but without a noise budget they measure only systematics. Realistic noise would add a common term to both ED_null and ED_sig, shrinking γ and reducing ϱ, and Table 1 gives no indication of how the tolerances degrade. The appropriate outcome is therefore the same as the reader's: conditional acceptance pending a noise and sensitivity model plus uncertainty quantification on the metrics. My concern does not move the verdict, so I recommend UNCHANGED.","tokens_in":12064,"tokens_out":9246,"duration_ms":83102,"concrete_test":"Using APSERa-like parameters (Tsys ≈ 50 K, Δν = 2 GHz, integration time t = 10^4 hr), compute the per-channel noise σ = Tsys/√(Δν·t) and add independent Gaussian noise to both signal-present and signal-absent test spectra at a Table 1 tolerance (e.g., 1-D wobble α = 4.8° at M9). Repeat the MS fit and metric computation ~100 times; if the 0.8 threshold on γ or ϱ is not met for the signal-present case, or if the null-case metric is not clearly separated, the detection claim fails. Equivalently, check analytically that Nσ² is small compared with the noiseless ED_sig0² and ED_null0² at the quoted tolerances.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim ('we demonstrate that it is indeed possible to detect the signal using a chromatic antenna') rests on the fractional Euclidean distance γ and Pearson ϱ computed from single, deterministic residuals. In Section 2 the only noise mention is that the ideal null hypothesis 'returns the simulated thermal noise' (Figure 5), but no amplitude, Tsys, bandwidth, or integration time is given, and this noise is not propagated into the test cases or the metric definitions. Reference and test residuals are otherwise noiseless; γ and ϱ therefore measure purely systematic similarity, not statistical distinguishability. The CRR brightness temperature is 1-10 nK over 2-4 GHz; a ground-based receiver (Tsys ~ tens of K) needs a noise level well below nK even to see the signal, and any noise component Nσ² entering both ED_null and ED_sig will drive γ = (ED_null - ED_sig)/ED_null toward 0 once it dominates the systematic differences. Consequently the tolerances in Table 1 are upper limits under zero noise, and the paper does not specify whether they survive realistic sensitivity. Without this, the feasibility conclusion is not established.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper investigates whether the predicted cosmological recombination radiation (CRR) lines, an additive nK-level distortion to the CMB over 2-4 GHz, can be distinguished from foreground systematics when the observing antenna is not perfectly achromatic. The authors build a simulation pipeline that generates mock spectra from Galactic foreground maps with a power-law model and a sin^2(theta) beam, optionally injects a CosmoSpec CRR template, fits the spectra with maximally smooth polynomials, and compares residuals using a fractional Euclidean distance (gamma) and the Pearson correlation coefficient (rho). They introduce toy-model beam perturbations (1-D and 2-D wobble, and beam stretching), compute gamma and rho for two observing sites and three LSTs, and report tolerance limits at a threshold of 0.8. The central claim is that, within finite and case-dependent tolerances, the CRR signal can still be detected even with a chromatic antenna.","tokens_in":12277,"tokens_out":6147,"duration_ms":60895,"significance":"The paper is a useful first step toward a practical CRR detection because it moves beyond the ideal-instrument assumption of earlier feasibility studies. Its strengths are the clear, self-contained simulation pipeline, the explicit parameterization of three beam-chromaticity toy models, and the systematic exploration of observing location and LST with a summarized tolerance table. These results can inform antenna design and observing strategy for APSERa. The detection metrics are standard and the zero-perturbation limits behave as expected (gamma=1, rho=1 with signal; rho=0 without). The main weakness is that the entire analysis is deterministic: thermal noise is mentioned but never specified or propagated, and the detection thresholds are arbitrary. Thus the quantitative tolerances are noiseless upper limits, and the abstract's claim that detection is 'demonstrated' is stronger than what the simulations establish. The paper is not circular, since it is a matched-template recovery test against an external theoretical CRR template, but the absence of noise and uncertainty quantification is a load-bearing gap for the practical conclusion.","major_comments":[{"comment":"The simulated thermal noise is mentioned in Section 2 ('returns the simulated thermal noise in the null hypothesis case', Figure 5), but its amplitude, spectral shape, and generation are never specified, and it is not propagated into the test cases or the metrics. Equations (3)-(4) and Figures 13-18 are computed from single deterministic residuals. Since the CRR signal is 1-10 nK, a realistic ground-based receiver would produce per-channel noise many orders of magnitude larger unless the integration time and bandwidth are specified; once a noise component N enters both ED_null and ED_sig, gamma = (ED_null - ED_sig)/ED_null tends to 0 when the noise dominates the systematic residual differences. The tolerances in Table 1 are therefore noiseless upper limits. To support the claim that detection is possible, the authors must specify a noise model (system temperature, bandwidth, integration time, spectral binning), inject noise realizations into both null and signal mocks, and show that gamma and rho remain distinguishable at the quoted perturbation levels with a defined false-alarm rate.","section":"Section 2, Section 3, Table 1"},{"comment":"The detection threshold of 0.8 for both gamma and rho is introduced without justification, and the metrics are evaluated on one deterministic residual per configuration rather than on a distribution over noise realizations. No confidence intervals, false-positive rates, or detection significances are reported, so the 'tolerances' in Table 1 (first crossing below 0.8) have no statistical meaning. The authors should either derive the threshold from the noise model and a target false-alarm probability, or report the full distributions of gamma and rho over many realizations and state the detection criterion in terms of those distributions.","section":"Section 3.1 and 3.2"},{"comment":"The 2-D wobble model uses the absolute value in Eq. (9), which creates a kink (slope discontinuity) in the beam-direction-versus-frequency relation at 3 GHz. The near-zero tolerances for 2-D wobble in Table 1 appear to be driven largely by this non-smooth toy-model feature rather than by a general property of beam direction reversal. The authors should clarify whether this kink is intended to represent a physical antenna response, and they should test a smoothed variant (for example a sinusoidal or low-pass-filtered direction change) before concluding that 2-D wobble is inherently 'detrimental' and requires channel dropping.","section":"Section 4.1, Eq. (9), Section 5.2, Table 1"}],"minor_comments":[{"comment":"There are numerous typographical errors and copyediting issues, including 'transitions form being fully ionized' in the Introduction, 'the the' in Section 5.1, and 'Table 5.3' referring to Table 1; these should be corrected in a final proof.","section":"Throughout"},{"comment":"The notation αν is ambiguous because it could be read as the product alpha times nu; a clearer notation such as alpha(nu) would improve readability.","section":"Eqs. (8)-(9)"},{"comment":"Calling the beam behavior at 3 GHz a 'discontinuity' is imprecise; the absolute value in Eq. (9) produces a continuous beam direction with a discontinuous slope (a kink), and the text should say so.","section":"Section 5.2"},{"comment":"The entry 0 for the 2D-γ tolerance at M9 needs an explanation: does it mean that the metric is already below threshold for the smallest simulated nonzero perturbation, or that no positive perturbation is allowed?","section":"Table 1"},{"comment":"The statement that 'any MS function of order 3 or above (order = infinity) will result in the same residual when fitting the CRR signal' is nontrivial and should be justified or demonstrated, since one might expect higher-order polynomials to absorb more spectral structure.","section":"Section 2"}],"recommendation":"major_revision","confidential_remarks":"The paper fits the scope of the journal and makes a worthwhile first contribution, provided the noise-model gap is addressed. I do not see citation or novelty concerns. I would encourage the authors to make the simulation code public and to frame the abstract more carefully as a noiseless feasibility demonstration rather than a full practical detection proof."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nQuick take: this is a well-scoped simulation study that does something genuinely new—quantifying how much antenna beam chromaticity a CRR detection can tolerate under a toy model—but the central numbers in Table 1 are computed without any noise model. Treat them as optimistic upper limits, not practical detection thresholds.\n\nWhat's new: previous feasibility studies (MSR2015, Sathish et al. 2024) assumed a perfectly achromatic antenna. Here they perturb a sin²(θ) beam with three simple chromatic distortions (1-D wobble, 2-D wobble, stretching) and use two standard similarity metrics (fractional Euclidean distance and Pearson correlation) against a reference CRR template to decide whether signal presence is distinguishable from absence. The pipeline—mock sky spectra, MS-function foreground fitting, residual comparison—is coherent, and the qualitative trends are physically sensible: 1-D wobble is more tolerable than 2-D wobble (which introduces a spectral discontinuity at band centre), stretching is most forgiving, and looking away from the Galactic centre helps. The paper is honest that this is a first step with a toy model.\n\nThe soft spots are real, and one is load-bearing. There is no noise model. The text mentions 'simulated thermal noise' in the null case (Section 2, Figure 5) but never gives a Tsys, bandwidth, integration time, or noise amplitude. The γ and ρ values that produce Table 1 come from deterministic residuals. Since the CRR is ~1–10 nK, any plausible receiver noise is many orders of magnitude larger; once noise dominates the residual, γ collapses toward zero and the quoted wobble/stretch tolerances no longer describe a distinguishable signal. The paper does not claim to have solved this, but it also doesn't flag that the tolerances are noiseless upper limits. That needs to be stated clearly, and ideally a sensitivity calculation should be added showing what integration time is required to reach the relevant noise floor.\n\nSecondary issues: the γ=ρ=0.8 threshold is arbitrary, and no error bars or noise realizations are given on the metrics, so we don't know how sharp the tolerance boundaries actually are. The introduction's 'paradigm shift in cosmology' for a non-detection is overblown. None of these undermine the core logic, but they do reduce the current paper to a proof-of-concept rather than a design-ready result.\n\nWho this is for: anyone planning a CRR experiment or working on global 21-cm foreground systematics. It deserves a serious referee; the gap is fixable. I'd send it to review, with a request for a noise treatment and a dampened conclusion.","headline":"Useful first-pass simulation showing CRR detection can tolerate some beam chromaticity, but the missing noise model means the headline tolerances are noiseless upper limits, not a demonstrated detection.","tokens_in":12810,"tokens_out":2837,"would_cite":true,"duration_ms":25888,"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 simulates whether the faint cosmological recombination radiation (CRR) can still be detected when the observing antenna is not perfectly achromatic, and finds that it can, within limits.","keywords":["cosmological recombination radiation","CMB spectral distortions","antenna chromaticity","Maximally Smooth functions","APSERa","foreground removal","signal detection metrics"],"falsifier":"A calculation that injects realistic radiometer noise (e.g., Tsys about 100 K, a channel bandwidth, and an integration time) into the simulated residuals and shows that γ and ϱ can no longer separate the signal-present from signal-absent cases at the Table 1 tolerances; alternatively, a measured beam pattern of an APSERa candidate antenna that violates the smoothness requirement (e.g., exhibiting 2-D wobble) would falsify the practical recommendation.","tokens_in":11802,"feed_emoji":"📡","tokens_out":5306,"duration_ms":42414,"temperature":0.7,"pith_summary":"The paper asks whether the cosmological recombination radiation (CRR) — a quasi-periodic few-nanoKelvin ripple in the sky spectrum predicted by standard cosmology but never detected — can still be found when the observing antenna is not perfectly achromatic. The authors simulate 2–4 GHz sky spectra with a toy-model beam that wobbles or stretches with frequency, fit out the smooth foreground with Maximally Smooth functions, and compare residuals against a reference template using two similarity metrics. They claim that within finite tolerance ranges — which depend on the type of chromaticity, the observing site, and the local sidereal time — the CRR signal can be distinguished from a signal-absent null case. If true, this gives concrete antenna-design and observing-strategy guidance for the APSERa experiment.","feed_headline":"Recombination signal detectable with a non-ideal antenna","feed_subtitle":"Simulation sets wobble and stretch tolerances for the 2–4 GHz APSERa experiment's antenna.","key_machinery":"The argument is carried by a simulation pipeline that generates mock antenna temperatures from a sky model (power-law Galactic synchrotron maps at 408 MHz, 1420 MHz, and 23 GHz), weights them with the antenna beam pattern $G(\\theta,\\phi,\\nu)$, applies a return-loss window, fits with arbitrarily high-order Maximally Smooth functions to remove smooth foregrounds, and then compares the residuals to a reference template using two detection metrics. The beam chromaticity toy model perturbs the ideal $\\sin^2\\theta$ beam in three ways: 1-D wobble (peak direction tilts linearly with frequency), 2-D wobble (tilt direction reverses at mid-band), and stretching (FWHM grows with frequency, parameterized by $\\beta$). The metrics are the fractional Euclidean distance $\\gamma = (\\mathrm{ED}_{\\mathrm{null}} - \\mathrm{ED}_{\\mathrm{sig}})/\\mathrm{ED}_{\\mathrm{null}}$ and the Pearson correlation coefficient $\\varrho$, with 0.8 taken as a detection threshold for both.","core_discovery":"The central claim is that the CRR lines remain detectable with a non-ideal antenna, using the fractional Euclidean distance and the Pearson correlation coefficient to separate signal-present from signal-absent residuals, provided the beam chromaticity stays within type- and geometry-dependent tolerances (Table 1). For 1-D directional wobble, tolerances reach up to about 11 degrees of total tilt depending on site and local sidereal time; for beam stretching, up to roughly 26 degrees of FWHM change; whereas 2-D wobble (direction reversal at band center) is effectively intolerable because the kink at 3 GHz injects non-smooth structure into the residuals. The paper presents this as a first step beyond the ideal-instrument assumption of the original feasibility study.","pith_inferences":["If real antenna beams can be characterised well enough to certify which chromaticity class they fall in, the same simulation pipeline could compute per-experiment safe limits on beam variation, turning the tolerances into a commissioning specification.","Because the paper's residuals are deterministic (no explicit noise model), the practical detectability claim implicitly assumes a noise floor below the ~1–10 nK signal; realistic radiometer noise would shrink the quoted tolerances, so the next step is to inject system-temperature noise and re-derive the tolerances as a function of integration time.","The metric threshold of 0.8 is arbitrary; a proper detection statistic would also need a false-alarm probability, which the deterministic simulation cannot yet provide.","The 2-D wobble result suggests a design principle: any non-monotonic frequency dependence of the beam peak should be pushed to the band edges or flagged, because smoothness of the beam-versus-frequency function matters more than the total change in pointing."],"forward_implications":["A real antenna may be usable for CRR detection without reaching one-part-per-billion achromaticity, as long as its beam frequency-dependence is of the smooth type (monotonic wobble or stretch) within the quoted tolerances.","Observing strategy matters: pointing away from the Galactic plane (for example Murchison at LST 9h) widens the tolerable wobble range, while having the Galactic centre overhead tightens it sharply.","Abrupt beam-direction reversals (2-D wobble) should be avoided in antenna design, or the affected channels must be dropped, because the mid-band kink destroys detectability.","The two metrics are complementary: fractional Euclidean distance is the conservative detector, while the Pearson coefficient allows larger tolerances and can serve as an early-stage detection indicator."],"supporting_citations":[{"why":"Original feasibility study defining the 2–6 GHz window, the Maximally Smooth function foreground-separation method, and the ideal-antenna assumption this paper relaxes.","marker":"[Rao et al., 2015]"},{"why":"CosmoSpec code providing the predicted CRR template (Figure 1) used as the reference signal.","marker":"[Chluba & Ali-Haimoud, 2016]"},{"why":"Sky-measurement simulation and signal-extraction pipeline that this paper adapts.","marker":"[Sathish et al., 2024]"},{"why":"408 MHz all-sky map used as input to the power-law Galactic synchrotron model.","marker":"[Haslam et al., 1982]"},{"why":"1420 MHz maps used in the foreground model.","marker":"[Reich, 1982; Reich & Reich, 1986]"},{"why":"HEALPix pixelisation used to grid the sky maps at Nside=64.","marker":"[Gorski et al., 2005]"}],"fun_headline_variants":["Recombination lines detectable despite antenna chromaticity","Non-ideal antenna can still catch recombination signal","CRR lines detectable with a chromatic antenna","Antenna flaws don't block cosmic recombination lines","Recombination signal survives antenna imperfections"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The paper never specifies the thermal noise level in the residuals; it assumes the instrument can integrate deeply enough that noise is negligible compared with the ~10 nK CRR signal, and all tolerances are derived from noiseless simulations.","fun_headline_variants_meta":{"raw":{"variants":["Recombination lines detectable despite antenna chromaticity","Non-ideal antenna can still catch recombination signal","CRR lines detectable with a chromatic antenna","Antenna flaws don't block cosmic recombination lines","Recombination signal survives antenna imperfections"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000622,"raw_usage":{"total_tokens":2873,"prompt_tokens":930,"completion_tokens":1943,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":546,"completion_tokens_details":{"reasoning_tokens":1877}},"tokens_in":546,"tokens_out":1943,"duration_ms":14506,"temperature":1.0,"reasoning_tokens":1877,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T10:42:50.649549+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A calculation that injects realistic radiometer noise (e.g., Tsys about 100 K, a channel bandwidth, and an integration time) into the simulated residuals and shows that γ and ϱ can no longer separate the signal-present from signal-absent cases at the Table 1 tolerances; alternatively, a measured beam pattern of an APSERa candidate antenna that violates the smoothness requirement (e.g., exhibiting 2-D wobble) would falsify the practical recommendation.","supporting_citations":[{"cited_title":"2016, Monthly Notices of the Royal Astronomical Society, 456, 3494","cited_arxiv_id":null,"evidence_quote":"CosmoSpec code providing the predicted CRR template (Figure 1) used as the reference signal."},{"cited_title":"S., & Sarkar, D","cited_arxiv_id":null,"evidence_quote":"Sky-measurement simulation and signal-extraction pipeline that this paper adapts."},{"cited_title":"1982, Astronomy and Astrophysics supple- ment series, 48, 219 Rubi˜no-Mart´ın, J., Rubi ˜no-Mart´ın, J., Chluba, J.,","cited_arxiv_id":null,"evidence_quote":"1420 MHz maps used in the foreground model."},{"cited_title":"M., et al","cited_arxiv_id":null,"evidence_quote":"HEALPix pixelisation used to grid the sky maps at Nside=64."}],"review_version":1}