{"id":"04b2d3d5-2d4e-4bff-82a6-cda3f6595b55","arxiv_id":"2501.12614","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"A noise-weighted chi-square minimization reconstructs air-shower electric fields from two or three antenna polarizations, with vertical polarization improving peak-amplitude precision by a factor of 3 to 5 in simulations.","lead":"This paper proposes a frequency-by-frequency, noise-weighted least-squares method to reconstruct the full electric field vector of cosmic-ray air showers from antennas with two or three polarizations, and tests it on simulated inclined showers. It claims 3 to 5 times better peak-amplitude precision when a vertical antenna polarization is added, which matters for future radio arrays that aim to measure ultra-high-energy cosmic rays.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The headline 4%/6% numbers are produced under a closed-loop simulation where the noise covariance and antenna response used in Eq.","rationale":"The paper's core estimator is a correct weighted least-squares solution, and the validation against independently simulated ZHAireS showers is a genuine test of the reconstruction pipeline. The reader's weakest-assumption analysis identifies the same load-bearing issue: the quoted resolutions are obtained under a diagonal Galactic-noise model with perfect, exactly known antenna responses and no RF-chain electronics. My read reinforces that concern with the closed-loop observation that the noise model used to generate data is the same model assumed in the chi-square weights, so the simulation measures the estimator's performance on its own generative model rather than its robustness to realistic mismatches. The numerical inconsistency between the abstract/conclusion and the displayed statistics (Fig. 10) is real but secondary; it can be fixed by restating which component and which bias convention the 4%/3% numbers refer to. Because the authors explicitly acknowledge the missing RF-chain and real-data steps, the appropriate verdict remains CONDITIONAL rather than REJECT: the method is promising and internally consistent, but the headline performance should be presented as a simulation-based lower bound pending end-to-end tests with noise measured from a real array and antenna-response calibration.","tokens_in":884,"tokens_out":763,"duration_ms":114303,"concrete_test":"Re-run the Section 4 statistical analysis with a non-ideal data-generation model: add a 10% common-mode correlated noise component across the three arms and independent 5% per-frequency multiplicative errors on the HFSS equivalent lengths, while still reconstructing with the nominal diagonal covariance and nominal response. If the PEA or fluence standard deviations rise above 4%/6%, or if the mean bias moves by more than 1 percentage point, the headline numbers are artifacts of the closed-loop simulation; if the statistics remain essentially unchanged, the concern is answered.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central performance claim rests on a self-consistency test: voltages are generated by convolving the simulated electric field with the HFSS equivalent lengths, and the same equivalent lengths are then inserted into the weighted least-squares solution of Eq. (3.7). Likewise, the Galactic noise is generated from the LFmap spectra that are used as the diagonal covariance in Eq. (3.5). Under this exact generative model, the estimator is near-optimal, so the reported standard deviations are lower bounds for a real detector, not validated resolutions. The paper explicitly excludes RF-chain electronic noise and the Voc-to-VADC calibration step (Sec. 2.4), and it does not include calibration uncertainties in the HFSS antenna response or mutual coupling between arms. Any of these effects can make the true noise covariance non-diagonal and the assumed response inaccurate; the generalized least-squares solution then loses optimality and, with response errors, acquires angle- and frequency-dependent bias. A secondary internal inconsistency also weakens the headline: Fig. 10 shows a 5% mean/median bias for the dipole PEA and a 12% standard deviation for the HORIZON theta-component, while the abstract and conclusion quote 4%/3% without stating that these apply to the phi-dominated total. The mathematical method itself is standard and correctly derived, but the specific numerical claims in the abstract are not yet supported outside the idealized simulation loop.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript presents an analytic chi-square minimization method for reconstructing the radio electric field from two or three orthogonally polarized antenna arms, with the estimator given in Eq. (3.7) as a frequency-domain weighted least-squares solution. The authors validate it with ZHAireS-simulated inclined air showers, HFSS antenna responses for a simple dipole and the HORIZON antenna, and LFmap Galactic noise, reporting standard deviations around 4% for the peak envelope amplitude and 4-6% for the energy fluence, along with an improvement by a factor of 3-5 when the vertical polarization is included. The paper also studies arrival-direction dependence and the effect of 1-degree direction perturbations on the reconstruction.","tokens_in":21554,"tokens_out":7999,"duration_ms":77019,"significance":"The mathematical derivation is clean and the simulation pipeline is well documented. If the stated performance holds in more realistic conditions, the method would be a useful, computationally simple alternative to iterative forward-folding for three-polarization arrays such as GRAND and LOPES-3D, and the demonstration that a vertical arm improves low-SNR reconstruction is a valuable quantitative result. However, the headline numbers are obtained in a self-consistency simulation and are not yet supported as detector-level resolutions; the paper's own figures contain inconsistencies with the abstract and conclusion. The central method itself is sound, but the performance claims need to be scoped and corrected.","major_comments":[{"comment":"Eq. (3.5) uses the same LFmap noise spectra as the diagonal covariance that were used to generate the noise in Sec. 2.4, and Eq. (3.7) uses the same HFSS equivalent lengths that produced the voltages in Eq. (2.2). Under this exact generative model the generalized least-squares estimator is near-optimal, so the quoted 4% and 6% values are lower bounds for an idealized detector, not validated resolutions. The abstract should be qualified accordingly, and the authors should add a robustness test with mismatched noise spectra, non-diagonal covariance, or a perturbed antenna response.","section":"2.4 and 3.2"},{"comment":"The abstract's 'better than 4%' and Sec. 5's 'about 3%' for the PEA are not both consistent with Fig. 10, where the total standard deviations are 0.04 (HORIZON) and 0.03 (dipole), and where the dipole total also shows a mean and median bias of 0.05. The latter bias implies a total RMS relative error of about 6% for the dipole even before considering the HORIZON theta-component, which has std = 0.12. The headline should state explicitly which component the 4% refers to, and the text-figure discrepancy for the dipole std should be corrected.","section":"Abstract, Sec. 4.2, Sec. 5"},{"comment":"A 1-degree Gaussian perturbation of the arrival direction changes the HORIZON PEA relative-error std from 0.04 in Fig. 10 to 0.12 in Fig. 13. The statement in Sec. 5 that this perturbation 'has no significant impact on the overall resolution' is therefore not supported by the paper's own figures, even if the effect is concentrated at zenith angles above 85 degrees. Please quantify the impact conditional on the relevant subpopulation and revise the conclusion accordingly.","section":"4.2.3 and Fig. 13"},{"comment":"For the energy fluence, the text states a 4% standard deviation for both antennas, but Fig. 15 shows std = 0.06 for one antenna in the total panel and std = 0.19 for the HORIZON theta-component. The abstract's 'better than 6%' is only defensible for the total or phi-dominated component, not for the theta component. Please reconcile the text and figure and state the conditional nature of the 6% claim.","section":"4.3 and Fig. 15"}],"minor_comments":[{"comment":"Eq. (3.5): the notation sigma_V = diag(sigma_V1, sigma_V2, sigma_V3) is ambiguous; the diagonal elements are variances, so please denote them as sigma_i^2 and reserve sigma_V for the covariance matrix.","section":"3.2"},{"comment":"Please include a legend or explicitly state in the captions which histogram and color correspond to each antenna, since the current text relies on color, which is not accessible in grayscale.","section":"Figs. 10 and 15"},{"comment":"Please define the exact list of zenith angles used, since the grid in log10(1/cos theta) maps to specific theta values; this would make Figs. 12-14 easier to interpret.","section":"2.1"},{"comment":"The definition of W_psi68 as the '34th percentile above and below the median' should be clarified as the half-width of the central 68% interval, i.e., half the distance between the 16th and 84th percentiles.","section":"4.2.1"},{"comment":"The choice of 100 ns signal and noise windows should be justified or tested for sensitivity, since the fluence estimates depend on the window boundaries.","section":"4.3"}],"recommendation":"major_revision","confidential_remarks":"The paper's novelty is modest, since Eq. (3.7) is a standard generalized least-squares estimator, but the three-polarization application and the simulation study are useful for the radio-detection community. The abstract and conclusion overstate the performance and should be corrected before acceptance. I would not require new experimental data, but the authors should either add robustness tests or explicitly scope the claims to the idealized simulation setup."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper does something genuinely useful: it applies a per-frequency, noise-weighted least-squares estimator to electric-field reconstruction from three polarizations, and it demonstrates on 4,160 simulated inclined showers that adding a vertical arm improves peak-envelope-amplitude precision by a factor of 3 to 5. The estimator itself is textbook generalized least squares, correctly derived in Eq. (3.7), but the application to EAS radio reconstruction, the systematic direction-dependence study, and the comparison of a simple dipole versus the complex HORIZON antenna are new and worth having. The matix-inversion comparison is instructive, especially for the HORIZON antenna, where the three-polarization inversion fails badly while the least-squares version stays stable.\n\nThe simulation campaign is substantial and the paper is honest about its event selection and limitations. The circularity burden is low: no target quantity is fitted, and the validation is against independent ZHAireS electric fields. That said, the headline numbers do not hold up as stated. The abstract claims a standard deviation better than 4% for the peak envelope amplitude and better than 6% for fluence, but the paper's own figures show 4% total for both antennas, 12% for the HORIZON theta-component, and a 5% bias for the dipole. The conclusions quote even more optimistic 'about 3%' and 'about 5%'. The numbers are not falsified, but they are cherry-picked from the phi-dominated total, and the 12% tail is downplayed.\n\nThe bigger soft spot is that the reported resolutions come from a closed-loop simulation. The voltages are generated with the same HFSS equivalent lengths and the same LFmap noise spectra that are then inserted into the reconstruction. That makes the estimator near-optimal by construction. Real detectors have correlated noise, calibration errors, impedance mismatch, and electronic noise, all of which the paper explicitly excludes. So the 4% and 6% numbers are lower bounds, not validated resolutions. The authors acknowledge this in the text, but the abstract does not.\n\nFor peer review, the paper deserves a serious referee. It is a legitimate methods contribution to the radio-detection community. The referee should ask for three things: restate the precision claims to match the figures, add a robustness test where the antenna response or noise model used in reconstruction is perturbed relative to the generation, and ideally benchmark against NuRadioReco's forward-folding or an information-field-theory approach. None of these are fatal; they are revisions. I would send it out.","headline":"A clean, useful methods paper whose headline precision numbers are only supported in a closed-loop simulation; the 3–5x gain from a vertical arm is the real result.","tokens_in":22157,"tokens_out":2117,"would_cite":true,"duration_ms":24903,"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":"A frequency-domain chi-square fit reconstructs cosmic-ray air-shower electric fields to better than 4% peak-amplitude and 6% fluence scatter in simulations.","keywords":["radio detection of air showers","electric field reconstruction","three-polarization antennas","inclined air showers","cosmic rays","energy fluence","Galactic background noise","weighted least squares"],"falsifier":"Re-run the same pipeline on signals with known electric fields but with noise drawn from a covariance that includes off-diagonal correlations between antenna arms, or inject a realistic electronic-noise contribution after the open-circuit voltage; if the relative-error scatter in peak envelope amplitude or energy fluence worsens noticeably beyond 4% or 6%, the diagonal-noise assumption behind the \\(\\$chi^{2}$\\) weights is the failing piece.","tokens_in":21049,"feed_emoji":"📡","tokens_out":12807,"duration_ms":122899,"temperature":0.7,"pith_summary":"This paper claims that the electric field of a cosmic-ray air shower can be recovered from antenna voltages by a frequency-by-frequency weighted least-squares fit, using either two or three orthogonal polarizations. The fit weights each arm's voltage by the Galactic-noise power at that frequency, which suppresses the spurious spikes that direct matrix inversion produces where the antenna response is weak. Tested on simulated inclined showers with a simple dipole and the more complex HORIZON antenna, the method reaches a standard deviation better than 4% for the peak envelope amplitude and better than 6% for the energy fluence, with an antenna-dependent bias. Using the third, vertical polarization improves peak-amplitude precision by a factor of 3 to 5. If the result holds on real arrays, radio detectors can convert voltage traces into accurate electric fields without assumptions about the signal shape, which is the first step toward precise energy and composition measurements.","feed_headline":"Three-polarization fit recovers air-shower electric fields to 4%","feed_subtitle":"Weighted frequency fit removes inversion spikes; with a vertical antenna arm, peak-amplitude errors drop 3-5 times.","key_machinery":"The load-bearing object is the closed-form weighted least-squares estimator \\(E = (H^T \\$sigma_V^{{-1}}$ H)^{-1} H^T \\$sigma_V^{{-1}}$ V\\), applied bin-by-bin in the frequency domain. Here \\(H\\) is the \\(3 \\times 2\\) matrix of vector effective lengths mapping the two transverse field components \\(E_\\$\\theta$, E_\\phi\\) to the three arm voltages, and \\(\\sigma_V = \\operatorname{diag}(\\sigma_{V1}^2, \\sigma_{V2}^2, \\sigma_{V3}^2)\\) holds the noise power in each arm, taken from the Galactic background model. This machinery down-weights low-gain or noisy frequency bins instead of amplifying them, and it lets a weak vertical arm contribute information without destabilizing the estimate. Because the estimator is derived by setting the gradient of the \\(\\$chi^{2}$\\) to zero, it is analytical and assumes no signal shape.","core_discovery":"The paper's central claim is that reconstructing the electric field as a weighted least-squares problem in the frequency domain solves the instability of the standard matrix-inversion approach, especially when a vertical polarization is added. At each frequency bin the estimator\n\\[\nE = (H^T \\$sigma_V^{{-1}}$ H)^{-1} H^T \\$sigma_V^{{-1}}$ V\n\\]\ncombines the three measured voltages \\(V\\) through the \\(3 \\times 2\\) vector-effective-length matrix \\(H\\) and the diagonal noise covariance \\(\\sigma_V\\), with the Galactic background spectrum setting the weights. This keeps the transverse-field estimate stable where one antenna arm has low gain or large noise, whereas direct inversion of the response matrix amplifies those bins into large artifacts. On Monte Carlo proton and iron showers from 63 to 87 degrees zenith, the method gives standard deviations near 3-4% for peak envelope amplitude and 4-6% for energy fluence, with a small systematic underestimation for the dipole and near-zero bias for the HORIZON antenna. The sharpest quantitative result is that the vertical arm improves peak-amplitude resolution by a factor of 3 to 5 across the SNR range.","pith_inferences":["If real arrays supply an estimate of the full noise covariance instead of only the diagonal, the same closed-form estimator generalizes by replacing \\(\\sigma_V^{-1}\\) with the inverse covariance; this is a natural upgrade for stations with correlated or non-Galactic noise.","The factor-3-to-5 gain from the vertical arm is largest at low SNR, which suggests that sparse three-polarization designs could recover otherwise lost events and expand the effective sky coverage of a fixed array.","The HORIZON antenna's 1-degree sensitivity in narrow zenith-angle ranges suggests an iterative workflow: fit the direction from the radio wavefront, reconstruct the field, then re-fit the direction on the clean field; this could shrink the high-zenith tail.","The same estimator, applied per frequency bin, can be reused as a building block for downstream Xmax and composition reconstruction, since it delivers a per-antenna field estimate without assuming a pulse shape."],"forward_implications":["For inclined air showers, a three-polarization antenna plus this fit gives several times smaller peak-amplitude scatter than the same fit with two horizontal polarizations, across the whole SNR range.","The method applies to both a simple dipole and the broadband HORIZON antenna, so it can be adopted by current arrays using two polarizations and by three-polarization designs under development.","Because the estimator weights each frequency by the measured noise, it suppresses the spurious spectral spikes that matrix inversion generates near antenna resonances and low-gain directions.","Energy fluence, the integral of the squared field, can be estimated to better than 6% scatter, which transfers directly to the radiation-energy scale used for cosmic-ray energy reconstruction.","The reconstruction tolerates roughly one-degree errors in the assumed arrival direction for most directions, so it does not demand tighter direction reconstruction than already exists."],"supporting_citations":[{"why":"Supplies the ZHAireS air-shower simulations whose true electric fields define the reconstruction targets.","marker":"[24]"},{"why":"Defines the HORIZON antenna used as the realistic three-polarization test case.","marker":"[27]"},{"why":"Documents the AERA matrix-inversion reconstruction that serves as the conventional baseline.","marker":"[41]"},{"why":"Introduces the forward-folding chi-square reconstruction from which this analytical solution is drawn.","marker":"[42]"},{"why":"Provides the LOPES-3D three-polarization antenna heritage and weighted reconstruction experience motivating the vertical arm.","marker":"[46]"},{"why":"Sets the vector-effective-length and spherical-coordinate response formalism used to convert fields to voltages.","marker":"[52]"},{"why":"Gives the LFmap Galactic-noise spectra that define the diagonal noise weights in the chi-square.","marker":"[57]"},{"why":"Supplies the standard linear least-squares solution used for the closed-form estimator.","marker":"[61]"}],"fun_headline_variants":["Three-polarization chi-squared fit recovers air-shower E-fields to 4%","Vertical antenna arm improves E-field reconstruction by up to 5x","Chi-squared method with three polarizations sharpens cosmic ray E-field estimation","Adding vertical polarization cuts E-field errors for inclined air showers","Three-polarization reconstruction improves cosmic ray E-field accuracy"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The quoted resolutions assume the noise in the three antenna arms is independent, dominated by the Galactic background, perfectly impedance-matched, and free of RF-chain electronic noise; if real noise is correlated or has extra components, the 4% and 6% figures do not automatically transfer to data.","fun_headline_variants_meta":{"raw":{"variants":["Three-polarization chi-squared fit recovers air-shower E-fields to 4%","Vertical antenna arm improves E-field reconstruction by up to 5x","Chi-squared method with three polarizations sharpens cosmic ray E-field estimation","Adding vertical polarization cuts E-field errors for inclined air showers","Three-polarization reconstruction improves cosmic ray E-field accuracy"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001295,"raw_usage":{"total_tokens":5328,"prompt_tokens":1027,"completion_tokens":4301,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":643,"completion_tokens_details":{"reasoning_tokens":4209}},"tokens_in":643,"tokens_out":4301,"duration_ms":29101,"temperature":1.0,"reasoning_tokens":4209,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T16:59:58.046858+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-run the same pipeline on signals with known electric fields but with noise drawn from a covariance that includes off-diagonal correlations between antenna arms, or inject a realistic electronic-noise contribution after the open-circuit voltage; if the relative-error scatter in peak envelope amplitude or energy fluence worsens noticeably beyond 4% or 6%, the diagonal-noise assumption behind the \\(\\$chi^{2}$\\) weights is the failing piece.","supporting_citations":[{"cited_title":"Alvarez-Mu˜ niz, W.R","cited_arxiv_id":null,"evidence_quote":"Supplies the ZHAireS air-shower simulations whose true electric fields define the reconstruction targets."},{"cited_title":"Abreu, M","cited_arxiv_id":null,"evidence_quote":"Documents the AERA matrix-inversion reconstruction that serves as the conventional baseline."},{"cited_title":"NuRadioReco: A reconstruction framework for radio neutrino detectors","cited_arxiv_id":"1903.07023","evidence_quote":"Introduces the forward-folding chi-square reconstruction from which this analytical solution is drawn."},{"cited_title":"Huber, Analysing the electric field vector of air shower radio emission , Ph.D","cited_arxiv_id":null,"evidence_quote":"Provides the LOPES-3D three-polarization antenna heritage and weighted reconstruction experience motivating the vertical arm."},{"cited_title":"Abreu, M","cited_arxiv_id":null,"evidence_quote":"Sets the vector-effective-length and spherical-coordinate response formalism used to convert fields to voltages."},{"cited_title":"Polisensky, Lfmap: A low frequency sky map generating program , Long Wavelength Array Memo Series 111 (2007) 515","cited_arxiv_id":null,"evidence_quote":"Gives the LFmap Galactic-noise spectra that define the diagonal noise weights in the chi-square."},{"cited_title":"Behnke, K","cited_arxiv_id":null,"evidence_quote":"Supplies the standard linear least-squares solution used for the closed-form estimator."}],"review_version":1}