{"id":"ed91b325-eb2a-4343-bacc-1397ad272001","arxiv_id":"2507.17266","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":8,"one_line_summary":"A simulation study demonstrates a radio-only reconstruction chain for inclined air showers, reporting 0.04 degree angular and about 10 percent energy resolution.","lead":"This paper tests a complete pipeline for reconstructing the arrival direction and energy of ultra-high-energy cosmic rays from radio signals of inclined air showers, using simulated antenna arrays. It reports an angular precision around 0.04 degrees and a 10 percent energy resolution, which would be useful for future sparse radio arrays such as GRAND.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Energy resolution is in-sample: S0/γ and f(ρmax) are fit to true Epri on the same library used for the quoted <10% resolution, and a quality cut uses MC truth.","rationale":"The reader's weakest-assumption analysis identifies the same load-bearing issue: the energy calibration and evaluation share the same Monte Carlo events. I agree, and the manuscript text supports it explicitly. Section 5.3 says S0 and γ are obtained by minimizing the residual to true Epri; Section 5.1 fits the density spline to the same library; Section 4.3 includes a cut threshold determined from MC truth; and the energy-resolution plots in Section 5.3 are computed on that same library. The result is that the 10% energy resolution is in-sample calibration quality. A held-out split is the natural, inexpensive check. I do not see an additional concern that is more load-bearing: the electric-field reconstruction and direction pipeline have no explicit calibration to true observables reported, though they still rely on simulated array geometry and galactic-noise-only traces. The correct disposition is unchanged: CONDITIONAL acceptance with the in-sample nature of the energy numbers made explicit and an out-of-sample validation planned.","tokens_in":14286,"tokens_out":9304,"duration_ms":99103,"concrete_test":"Split the 4160-event library into calibration and validation halves (stratified by primary and by energy/zenith bins). Fit S0, γ, and the f(ρmax) spline on the calibration half only; apply the fitted pipeline to the validation half and report the Erec/Epri bias and width. Repeat with swapped halves, and rerun without the §4.3 MC-truth-based fluence-amplitude cut. If the held-out resolution exceeds ~15% or the composition-averaged bias exceeds ~5%, the 10% energy-resolution claim is a calibration artifact rather than predictive accuracy.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central energy claim is not an independent prediction. In §5.3, S0 and γ in Eq. 12 are obtained by minimizing (Erec − Epri) on the analysis library; in §5.1/Fig. 6 the air-density correction f(ρmax) is fit to ideal simulations of the same library; and §4.3 applies a cut requiring the fitted fluence amplitude < 10^9, an explicit MC-truth-determined threshold. Figures 7–8 then report the bias and width of Erec/Epri on those same events. With the calibration and the selection both conditioned on true Epri (or MC-truth fluence), the quoted ≈10% resolution measures how well the model can be tuned to this simulation, not how well it would reconstruct an independent event. The angular and electric-field claims are less exposed, because they do not use true Epri for calibration, but they inherit the same simulated array and noise model. The two fitted parameters alone would be a minor overfitting risk; the density spline and the truth-based cut are the parts that can substantially inflate the reported performance. To support the abstract's claim that the pipeline 'demonstrates the reliability' for future experiments, an out-of-sample evaluation or a clear statement that all quoted numbers are in-sample fit qualities is required.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents an end-to-end reconstruction pipeline for radio detection of inclined ultra-high-energy cosmic-ray air showers, using ZHAireS simulations of proton and iron primaries. The pipeline consists of three stages: an analytical least-squares reconstruction of the electric field from three-polarization antenna voltages (Sec. 3), a direction reconstruction that combines a spherical wavefront model with an angular distribution function (ADF) fit of the radio footprint (Sec. 4), and an energy reconstruction that integrates the geomagnetic energy fluence, applies an air-density and geomagnetic correction, and calibrates a power-law relation to the primary energy (Sec. 5). The reported performance is an unbiased electric-field peak reconstruction (68% interval [-0.02, 0.02]), an angular resolution of about 0.04 degrees, and an energy resolution better than 10% over most of the parameter space, with degradation at low energies and very high zenith angles.","tokens_in":14624,"tokens_out":2100,"duration_ms":23514,"significance":"If the quoted performance is predictive, the paper would demonstrate that sparse radio arrays with three-polarization antennas can reconstruct both arrival direction and primary energy of inclined UHECR showers without particle detectors, which is directly relevant to experiments such as GRAND. The electric-field and direction-reconstruction results are well grounded in comparisons with simulated truth and do not rely on calibrating free parameters against the primary energy, so those parts are valuable. The energy reconstruction, however, is the central new claim, and its current evaluation is in-sample: the calibration parameters S0 and gamma in Eq. (12) are fitted to the true primary energies of the same library used for the quoted resolution, the air-density spline f(rho_max) in Fig. 6 is fitted to the same simulated library, and one of the selection cuts (Sec. 4.3) uses a threshold determined from Monte Carlo truth. As written, the <10% energy resolution is a measure of fit quality, not a predictive accuracy, which limits the strength of the abstract's claim that the pipeline 'demonstrates the reliability' for future experiments.","major_comments":[{"comment":"The quoted 10% energy resolution is measured in-sample. The text states that S0 and gamma are obtained by minimizing the residual between Erec and the true primary energy Epri on the analysis library, and Figs. 7 and 8 then show the distribution of Erec/Epri for those same events. This makes the reported resolution a goodness-of-fit of the calibration curve, not an estimate of how well the pipeline would reconstruct independent events. To support the abstract's claim of reliability, the authors should either (a) calibrate on a training subset and evaluate on a disjoint test subset (e.g., by energy, zenith, or event index), or (b) present the quoted numbers explicitly as in-sample fit qualities and remove the inference of predictive accuracy. Without this change, the central energy-resolution claim is not load-bearing as stated.","section":"Sec. 5.3, Eq. (12), Figs. 7-8"},{"comment":"The selection cut 'The fitted amplitude of the energy fluence must be less than 10^9, a threshold determined from the Monte Carlo truth for this library' uses knowledge of the truth to retain events. Applying a truth-derived cut before reporting the energy resolution on the same sample biases the performance estimate upward and is not directly implementable in real data, where no truth is available. The authors should either remove this cut, justify it as a physically motivated quality cut (e.g., via a fixed absolute fluence limit), or evaluate the energy reconstruction without it and report how often an actual data event would survive.","section":"Sec. 4.3, second bullet"},{"comment":"The air-density correction spline f(rho_max) is fitted to the same simulation library on which the energy resolution is later evaluated. This is a second source of in-sample tuning: the spline absorbs simulation-specific fluctuations of the ratio Egeo/(Eem^2 sin alpha) as a function of rho_max, and applying it to the same events overfits the noise realization of this library. A cross-validation or a spline fit to an independent subset is needed to establish that the correction generalizes. At minimum, the number of degrees of freedom in the spline and the goodness-of-fit should be reported.","section":"Sec. 5.1, Fig. 6"}],"minor_comments":[{"comment":"There is a typo in the abstract: 'reconstruction of the the properties of primary particles' contains a duplicated 'the'.","section":"Abstract"},{"comment":"The word 'incomming' should be 'incoming'.","section":"Sec. 3, first paragraph"},{"comment":"The text 'The minimiation of chi^2' contains a typo: 'minimiation' should be 'minimization'.","section":"Sec. 4.1"},{"comment":"The sentence 'where theta_sim and phi_rec represent true zenith and azimuth angles' should read 'phi_sim' instead of 'phi_rec', since phi_rec is the reconstructed azimuth and phi_sim is the true azimuth used in Eq. (8).","section":"Sec. 4.5, Eq. (8)"},{"comment":"The phrase 'it's origin' should be 'its origin' for the possessive.","section":"Sec. 5.2"}],"recommendation":"major_revision","confidential_remarks":"The paper is a simulation-only study with a clear and well-structured pipeline. The in-sample energy calibration is the main technical obstacle; it is presented as a resolution rather than as a fit quality, which overstates the result. The issue is fixable within the manuscript's scope by adding a cross-validation or train/test split, so I recommend major revision rather than rejection. The direction and electric-field sections are sound and could be published with minor polishing if the energy section is brought to the same standard."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The thing to know: this is a serious simulation pipeline for inclined air showers, and the angular and electric-field results look solid. The energy resolution, however, is an in-sample fit quality, not a predictive accuracy. Don't quote the 10% number without checking whether they've added out-of-sample validation.\n\nWhat's new: the integration. The analytical least-squares E-field method is from their companion paper [21], and the ADF direction fitting builds on Decoene et al. and Schlüter & Huege. This paper adds the application to a 4160-shower ZHAireS library with galactic noise, plus an air-density correction f(rho_max) for the geomagnetic energy. The E-field reconstruction is well validated: peak amplitude unbiased, 68% CI [-0.02, 0.02], and energy fluence at 6% std. The angular resolution of ~0.04 deg is also supported by comparison to simulated truth, with sensible dependence on zenith and triggered antennas. Those parts deserve credit.\n\nThe soft spot is the energy calibration. In Sec. 5.3, S0 and gamma are fitted by minimizing the residual between Erec and true Epri on the same library used for the quoted resolution. The spline f(rho_max) is fitted to ideal simulations of that same library (Fig. 6). And the cut on fitted fluence amplitude < 1e9 is explicitly derived from MC truth. So the reported <10% energy resolution measures how well the model can be tuned to this simulation. The angular and E-field claims don't use true Epri for calibration, so they're less exposed, but they still inherit the idealized star-shaped array and galactic-only noise. The paper honestly notes the star pattern is optimistic, but it never flags the energy numbers as in-sample.\n\nBottom line: useful as a demonstration for sparse radio arrays like GRAND, and the E-field and direction parts are worth reading. The energy claim needs an out-of-sample test or a clear statement that the numbers are fit qualities. I'd send it to peer review, but I'd ask for that before accepting. It's a serious paper with one load-bearing caveat.","headline":"A coherent end-to-end simulation study whose electric-field and angular claims are credible, but whose headline 10% energy resolution is in-sample calibration, not an independent prediction.","tokens_in":15152,"tokens_out":2167,"would_cite":true,"duration_ms":23356,"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":"Radio voltages alone can recover cosmic-ray direction to 0.04 degrees and energy to better than 10 percent.","keywords":["radio detection","extensive air showers","ultra-high-energy cosmic rays","electric field reconstruction","arrival direction reconstruction","energy reconstruction","geosynchrotron radiation","angular distribution function"],"falsifier":"Run the identical pipeline with a strict train/test split of the simulation library (fit $S_0$, $\\gamma$, and the spline on one half, evaluate on the other half), or replace the star-shaped 160-antenna pattern with a realistic sparse layout; a resolution that degrades beyond 10% or develops a bias would show the reported numbers are fit quality rather than predictive accuracy.","tokens_in":14078,"feed_emoji":"📡","tokens_out":7607,"duration_ms":70633,"temperature":0.7,"pith_summary":"This paper tries to establish that the entire chain of ultra-high-energy cosmic-ray observables—electric field at the antenna, arrival direction, and primary energy—can be recovered from radio voltages alone, with no particle detectors in the loop. The authors simulate strongly inclined air showers with full radio emission and galactic noise, then run a closed-form least-squares inversion of the three-polarization antenna response to get an unbiased electric field, with a 68% confidence interval of $[-0.02, 0.02]$ in peak amplitude. On that field they fit a spherical wavefront plus an angular distribution function to locate the emission region and the shower axis, reaching about $0.04^\\circ$ angular resolution. For energy they extract the geomagnetic component, correct it for air density at shower maximum and for the sine of the angle to the magnetic field, and fit a power law to primary energy, achieving better than 10% resolution over most of the tested parameter space. If this holds on real sparse arrays, it removes the need for dense particle-detector arrays in measuring the spectrum and directions of the highest-energy cosmic rays.","feed_headline":"Radio-only pipeline hits 0.04° directions, <10% energies","feed_subtitle":"Closed-form E-field inversion plus air-density correction lets sparse antennas measure ultra-high-energy cosmic rays.","key_machinery":"The argument rides on four coupled objects. The first is the closed-form least-squares electric-field estimator $E = (H^T \\sigma_V^{-1} H)^{-1} H^T \\sigma_V^{-1} V$, which inverts the three-polarization antenna response matrix $H$ given the noise covariance $\\sigma_V$. The second is the spherical wavefront model of Eq. (5), which turns per-antenna peak arrival times into an estimated apparent emission point $R_{X_{max}}$ near shower maximum. The third is the angular distribution function $f_{ADF}$ of Eq. (6), a product of a Cherenkov-ring term $f_{ch}$ and an asymmetry term $f_{asym}$, fitted to the reconstructed field-peak footprint to get the shower axis and the fluence amplitude. The fourth is the energy relation $E_{rec} = (S_{geo}/S_0)^{1/\\gamma}$, where $S_{geo} = E_{geo}/(\\sin\\alpha \\cdot f(\\rho_{max}))$ is the geomagnetic radiation energy corrected for magnetic orientation and for air density at the reconstructed $X_{max}$ through a spline $f(\\rho_{max})$. These pieces are chained: the field inversion feeds the direction fit, the direction fit gives $R_{X_{max}}$, and $R_{X_{max}}$ together with the ADF amplitude gives the corrected radiation energy from which primary energy is read off.","core_discovery":"The central claim is that an end-to-end radio reconstruction pipeline, applied to simulated voltages that include galactic noise, reproduces the true shower properties without bias: the peak electric field is recovered with zero median and a 68% confidence interval of $[-0.02, 0.02]$, the arrival direction within about $0.04^\\circ$, and the primary energy with a resolution below 10% for proton and iron primaries between about $10^{17.1}$ and $10^{18.6}$ eV at zenith angles up to roughly 85 degrees. The reconstruction is analytical rather than iterative: minimizing a chi-square over the three measured polarizations gives $E = (H^T \\sigma_V^{-1} H)^{-1} H^T \\sigma_V^{-1} V$, a direct linear inversion that needs no prior on the signal shape. Direction reconstruction uses peak arrival times to fix the apparent emission point on a spherical wavefront, then fits the footprint of reconstructed field peaks with an angular distribution function encoding the Cherenkov ring and geomagnetic asymmetry. Energy reconstruction isolates the geomagnetic fluence, divides out $\\sin\\alpha$ where $\\alpha$ is the angle between the shower axis and the magnetic field, applies a spline correction for air density at the shower maximum (the geosynchrotron regime), and calibrates $E_{rec} = (S_{geo}/S_0)^{1/\\gamma}$ to true primary energy, yielding an average bias consistent with zero and a composition-dependent offset of opposite sign for protons and iron.","pith_inferences":["Because the calibration constants and the resolution are derived from the same simulated library, the honest reading is that the 10% figure is a statement about algorithmic consistency on ideal footprints; a held-out simulation set would reveal how much of it is overfitting.","The star-shaped footprint with 160 antennas densely samples the Cherenkov ring, which real sparse arrays will not do; replacing it in simulation with a realistic sparse layout is a direct way to see how the angular and energy resolutions degrade.","The reported $S_0$ and $\\gamma$ depend on the local geomagnetic field and atmosphere, so a testable extension is to check whether the same power law holds at other sites or whether $S_0$ must be rescaled with the magnetic-field strength."],"forward_implications":["A sparse radio array could measure the arrival directions of very inclined showers at $0.04^\\circ$ precision, comparable to what particle-detector arrays achieve, without needing a dense ground grid.","Energy assignment at better than 10% over $10^{17.5}$ to $10^{18.6}$ eV would let a radio-only experiment measure the UHECR energy spectrum, provided the calibration transfers to data.","The air-density and $\\sin\\alpha$ corrections make the method applicable in the geosynchrotron regime, which is exactly the regime relevant to Earth-skimming neutrino searches with very inclined showers.","The residual composition-dependent bias (protons overestimated, iron underestimated) means that a composition tag or an $X_{max}$-based correction would be needed for per-species energy estimates."],"supporting_citations":[{"why":"Supplies the air-shower and radio-signal simulation library that generates all simulated voltages used in the pipeline.","marker":"[18]"},{"why":"The companion paper introducing the analytical least-squares electric-field reconstruction that this work applies end to end.","marker":"[21]"},{"why":"Provides the angular distribution function (ADF) reconstruction procedure for direction and emission point.","marker":"[24, 49]"},{"why":"Establishes the spherical wavefront model that relates arrival times to the apparent emission position $R_{X_{max}}$.","marker":"[46]"},{"why":"Documents the suppression of geomagnetic radio emission at low air density, motivating the density correction in the energy reconstruction.","marker":"[15]"},{"why":"Supplies the power-law correlation between corrected geomagnetic radiation energy and electromagnetic energy used in Eq. (12).","marker":"[26]"},{"why":"Parameterizes geomagnetic radiation energy fluence and its air-density dependence, the basis for building $S_{geo}$.","marker":"[37]"}],"fun_headline_variants":["Closed-form radio pipeline: 0.04° directions, <10% energies","Zero-bias E-field inversion for sparse-array UHECR","Radio-only end-to-end: 0.04° and <10% energy","Analytic radio reconstruction: unbiased, 0.04°, <10% energy","Sparse-antenna radio: 0.04° angles, <10% energies"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the energy calibration—$S_0$, $\\gamma$, and the air-density spline $f(\\rho_{max})$—is fitted to the same simulated library on which the 10% resolution is then measured, so the claim assumes these calibrations transfer unchanged to real data with the same footprint sampling.","fun_headline_variants_meta":{"raw":{"variants":["Closed-form radio pipeline: 0.04° directions, <10% energies","Zero-bias E-field inversion for sparse-array UHECR","Radio-only end-to-end: 0.04° and <10% energy","Analytic radio reconstruction: unbiased, 0.04°, <10% energy","Sparse-antenna radio: 0.04° angles, <10% energies"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001087,"raw_usage":{"total_tokens":4599,"prompt_tokens":1054,"completion_tokens":3545,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":670,"completion_tokens_details":{"reasoning_tokens":3439}},"tokens_in":670,"tokens_out":3545,"duration_ms":27414,"temperature":1.0,"reasoning_tokens":3439,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T14:53:14.096256+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the identical pipeline with a strict train/test split of the simulation library (fit $S_0$, $\\gamma$, and the spline on one half, evaluate on the other half), or replace the star-shaped 160-antenna pattern with a realistic sparse layout; a resolution that degrades beyond 10% or develops a bias would show the reported numbers are fit quality rather than predictive accuracy.","supporting_citations":[{"cited_title":"Alvarez-Muñiz, W.R","cited_arxiv_id":null,"evidence_quote":"Supplies the air-shower and radio-signal simulation library that generates all simulated voltages used in the pipeline."},{"cited_title":"Decoene, O","cited_arxiv_id":null,"evidence_quote":"Establishes the spherical wavefront model that relates arrival times to the apparent emission position $R_{X_{max}}$."},{"cited_title":"Chiche, C","cited_arxiv_id":null,"evidence_quote":"Documents the suppression of geomagnetic radio emission at low air density, motivating the density correction in the energy reconstruction."},{"cited_title":"Schlüter and T","cited_arxiv_id":null,"evidence_quote":"Supplies the power-law correlation between corrected geomagnetic radiation energy and electromagnetic energy used in Eq. (12)."},{"cited_title":"Glaser, M","cited_arxiv_id":null,"evidence_quote":"Parameterizes geomagnetic radiation energy fluence and its air-density dependence, the basis for building $S_{geo}$."}],"review_version":1}