{"id":"eb9deab0-ddbf-41da-b7f4-82e045ee1b5e","arxiv_id":"1908.10689","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"Horizontal radio transmissions through 1 to 5 km of South Polar ice yield an attenuation length near 1.5 km and a polarization-dependent arrival-time asymmetry of about 8 to 9 ns per km.","lead":"An Antarctic research team sent radio pulses through up to 5 kilometers of South Pole ice and measured how far the signals travel before fading. The results confirm that cold polar ice is transparent enough at radio frequencies to host next-generation neutrino telescopes and reveal how ice crystals delay differently polarized signals.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Attenuation-length extraction assumes smooth 1/r plus exponential absorption, while the paper's own Section 3 documents multipath-like amplitude modulation; this unmodeled propagation effect is the main threat to the 1.43 km central value.","rationale":"The reader's weakest assumption focused on the geometric-optics n(z) model for timing and birefringence. My concern is adjacent but more directly load-bearing for the headline attenuation claim: the same unmodeled propagation anomalies that appear in Section 3 enter the amplitude ratios used in Section 6. The paper is honest about these anomalies and the central value is independently consistent with prior vertical-bounce estimates and across multiple days, so I would not reject or downgrade the paper. However, the systematic error budget does not appear to include a bias term from coherent multipath or focusing, and the quoted 0.37 km systematic may therefore be optimistic. The conditional verdict already captures this uncertainty; a targeted data-driven test using the paper's own depth-modulation fit would settle whether the concern is real. The proposed test is feasible with existing data and directly probes bias rather than width of the attenuation histogram.","tokens_in":23084,"tokens_out":4329,"duration_ms":50315,"concrete_test":"Recompute the SPICE attenuation-length histogram after dividing each event's VPol amplitude by the empirical depth-modulation envelope A(z) = A0 cos(kz + φ0) shown in Figure 13, before forming the same-antenna V1/V2 ratios. If the refitted peak shifts by more than roughly 0.25 km relative to 1.43 km, the Section 3 multipath-like modulation is a source of bias, not just width; if the peak is stable, the central attenuation claim is robust to this concern.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing claim is the horizontal attenuation length L_atten = 1.43 ± 0.25 ± 0.37 km (Section 6). The extraction uses V1/V2 = (d2/d1) exp((d2-d1)/L_atten), i.e. it assumes spherical spreading plus exponential absorption along simple geometric paths with no multipath, scattering, or focusing. The paper itself reports signals in the geometric shadow zone, 'unexpectedly large variation' in signal strength with depth, 'apparent modulation of the signal amplitude with depth', and 'considerable deviations from smoothness' (Section 3), attributing them to 'multi-path interference effects'. Those same amplitude anomalies enter directly into the amplitude ratios from which L_atten is derived. The paper propagates event-by-event amplitude scatter into the width of the L_atten histogram (Section 7) but does not propagate a possible depth- or frequency-dependent bias from coherent multipath or focusing into the 0.37 km systematic term. Also, the deep-pulser estimate relies on A4 with a 12 dB attenuator correction; a 1 dB calibration error shifts L by roughly 0.15 km. The n(z) model is not the crucial issue here; the crucial issue is that the observed non-geometric amplitude structure could bias the peak of the histogram rather than merely widen it.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports on long-baseline (1--5 km) horizontal radio-frequency propagation measurements through South Polar ice, using two deep IceCube pulsers (IC1S, IC22S) and a mobile transmitter (SPUNK PVA) lowered into the SPICE borehole, with signals recorded by the ARA stations. Three main results are presented: (i) constraints on the englacial refractive-index profile n(z) from direct/refracted double-pulse time differences; (ii) a birefringence measurement parameterized as δt(H−V)[ns/km] = a cosθ + b, with a = 8.3 ± 1.3 ns/km and b = −8.6 ± 0.9 ns/km, proposed as a range-estimation tool for neutrino vertices; and (iii) the first horizontal-geometry radio-frequency electric-field attenuation length, with the deep-pulser measurement L_atten = 1.43 ± 0.25 ± 0.37 km and supporting SPICE-core daily distributions with peak values between about 1.3 and 2.4 km. The paper also documents anomalous propagation effects, including signals inside the geometric shadow zone and strong depth-dependent amplitude modulation, which it attributes to multipath interference and near-surface index structure.","tokens_in":23304,"tokens_out":4044,"duration_ms":44772,"significance":"If the attenuation-length result holds, it is an important experimental input for in-ice radio neutrino detectors: it directly supports km-scale station spacing for Askaryan detectors and provides the first horizontal-baseline measurement of cold-ice RF attenuation, complementing previous vertical bottom-bounce measurements. The birefringence measurement is also valuable for reconstruction of neutrino interaction vertices, and the n(z) comparison gives a useful cross-check of the AraSim ice model. The paper combines several independent datasets (deep pulsers, SPICE core drops, surface pulser timing checks, and Monte Carlo smearing studies), and it is appropriately candid about observed anomalies that are not captured by smooth geometric-optics models. These strengths make the paper a useful contribution if the central attenuation-length systematics can be brought under control.","major_comments":[{"comment":"The extraction V1/V2 = (d2/d1) exp((d2−d1)/L_atten) assumes spherical spreading with path lengths d1 and d2 plus exponential absorption along simple geometric paths, but Table 1 lists lateral distances rather than curved ray-path lengths, and the paper's own §3 documents multipath-like amplitude modulation, shadow-zone propagation, and depth-dependent focusing (Figures 10--13). These effects can bias the peak of the L_atten histogram rather than merely broaden it. The Monte Carlo smearing study in §7 addresses voltage-resolution broadening only, not a propagation-induced shift. Please quantify an additional systematic from focusing/multipath, or restrict the attenuation estimate to event samples for which single-path geometric propagation is verified.","section":"§6, attenuation-length formula"},{"comment":"The deep-pulser result relies on A4 data taken with a 12 dB input attenuator that is corrected offline. A 1 dB uncertainty in this correction changes L_atten by roughly 0.15 km, which is comparable to the statistical error of ±0.25 km. The paper does not report the calibration accuracy of this attenuator or propagate it into the 0.37 km systematic term. Please add this contribution explicitly.","section":"§6, A4 12 dB attenuator"},{"comment":"The four daily SPICE attenuation-length distributions have peak values 2377 ± 645, 1540 ± 361, 1348 ± 383, and 1302 ± 372 m, spanning a factor of about 1.8. Day 358, which uses the piezo pulser, is described as most prone to systematic uncertainties, yet it is still included in the supporting claim of values 'clustering around 1.5 km'. Please state whether the main conclusion remains unchanged when Day 358 is excluded and quantify a day-to-day common-mode systematic that could explain the spread.","section":"§6 and Figure 25"},{"comment":"The claimed ~15% range-estimation capability for future neutrino measurements is based on a linear fit to the same data used to determine the birefringence parameters a and b. Since the parameters are not held fixed from an independent calibration, the quoted precision is optimistic. A leave-one-out or independent-data validation, or an explicit statement of how much of the 15% comes from the geometric D/R lever arm versus the birefringence fit, would make the claim reliable.","section":"§5.1, Figure 23"}],"minor_comments":[{"comment":"The caption 'r corrected' should be expanded to explain which radial/geometric correction is applied and whether it is the same 1/r factor used in the attenuation-length formula in §6.","section":"§3, Figure 13"},{"comment":"There is a typo: the text says the third axis is 'perpendicular to both ê1 and ê1'; it should read 'ê2' in the second instance.","section":"§5, paragraph 1"},{"comment":"The table header 'Testbed (no data)' is confusing because §5.1 reports testbed birefringence measurements from an earlier campaign; please add a clarifying note about the different epochs.","section":"Table 1"},{"comment":"The summary states attenuation lengths 'clustering around 1.5 km' while also quoting the Day 358 value of 2.38 km; consider rephrasing to 'values in the range 1.3--2.4 km, with deep-pulser mean 1.43 km'.","section":"§8, bullet list"}],"recommendation":"major_revision","confidential_remarks":"This is a solid experimental paper with multiple independent datasets and a refreshingly honest treatment of anomalies, but the headline attenuation-length number needs a more complete propagation of systematic effects, especially multipath/focusing bias and the A4 attenuator calibration. The day-by-day SPICE spread also needs a defensible interpretation. These are fixable with additional analysis, so major revision is appropriate rather than reject."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague—\n\nThe thing to know: this is the first direct measurement of the horizontal RF attenuation length in cold polar ice, and the headline number is ~1.5 km. If it holds, km-scale spacing for in-ice radio neutrino detectors is viable. The paper also gives a clean linear parameterization of birefringence versus angle to ice flow (a=8.3±1.3, b=-8.6±0.9 ns/km) and shows the refractive-index uncertainty contributes only ~2–3% to effective volume. These are real, new results from actual in situ transmission data—deep pulsers in IceCube holes and the 2018 SPICE core antenna drop—and they are consistent with older vertical-bounce estimates. That consistency, plus the multiple independent station pairs, is why the central claim is credible.\n\nWhat the paper does well: it documents its own anomalies. Section 3 reports signals in the geometric shadow zone, strong depth-dependent amplitude modulation, and caustic-like focusing at the shadow boundary. That honesty makes the paper more useful than a clean-but-incomplete analysis would be.\n\nThe soft spot is exactly where the stress-test points. The attenuation-length extraction uses V1/V2 = (d2/d1) exp((d2−d1)/L), which assumes smooth geometric spreading plus exponential decay. But the paper itself shows the amplitudes are modulated with depth in ways that look like multipath and focusing. That structure could bias the amplitude ratio, not just widen the histogram. The Monte Carlo smearing in Section 7 addresses width; it does not set an upper limit on a coherent bias. The quoted 0.37 km systematic does not include such a term. Also, the deep-pulser value relies on the 12 dB attenuator correction; a 1 dB error moves L by roughly 0.15 km. So the central value 1.43 km is less certain than the error bars imply. The broader claim—cold polar ice has >1 km attenuation length over horizontal baselines—is very likely correct, since the day-by-day SPICE values range from 1.3 to 2.4 km and nothing dips below 1.3 km. But the precise number should be treated with caution.\n\nThe n(z) constraints are useful but modest; the model comparison is fit and evaluated on the same data, so it is more interpolation than validation. The authors do not oversell it.\n\nThis paper deserves serious peer review. It is a genuine experimental contribution with new data, cross-checks, and clear limitations. My recommendation: send it to review, and ask the authors to add a systematic term for coherent multipath/focusing bias to the attenuation-length error budget, or explicitly state that such a term is not included. Either way, publish; the community needs this measurement on record.","headline":"First direct horizontal-baseline RF attenuation length in polar ice (~1.5 km) with honest anomaly reporting; the >1 km conclusion is solid, but the exact central value carries a larger uncounted multipath systematic than the quoted errors suggest.","tokens_in":24248,"tokens_out":4843,"would_cite":true,"duration_ms":46717,"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 pulses transmitted horizontally through 1–5 km of cold South Polar ice attenuate with a measured field attenuation length of about 1.5 km, supporting kilometre-spaced in-ice neutrino detectors.","keywords":["radio-frequency attenuation in ice","South Polar ice sheet","in-ice neutrino detection","Askaryan radio emission","ice birefringence","refractive index profile","shadow-zone propagation","horizontal long-baseline transmission"],"falsifier":"A dedicated multi-station run would settle it: broadcast one sharp transmitter pulse and record it at three or more stations at different baselines on roughly the same bearing with independently calibrated channel gains, then check that the inferred $L_{atten}$ is identical for every station pair. If the apparent attenuation length changes with baseline, receiver depth, or pulse sharpness (the slower piezo day already gives ~2.4 km versus ~1.3–1.5 km for the fast pulsers), the amplitude-ratio method is contaminated by multipath or scattering, and 1.43 km is an effective, not intrinsic, value. A laboratory cross-check: measure the complex permittivity of deep South Pole ice cores over 100–800 MHz and ask whether the implied absorption length over 200–1200 m depth is consistent with $1.43 \\pm 0.44$ km.","tokens_in":22815,"feed_emoji":"📡","tokens_out":22126,"duration_ms":189472,"temperature":0.7,"pith_summary":"Cold polar ice is transparent enough to carry radio-frequency pulses across kilometre-scale horizontal baselines, and this paper measures exactly how transparent: an electric-field attenuation length of $1.43 \\pm 0.25 \\pm 0.37$ km in the upper 200–1200 m of the South Polar ice sheet. The measurement is the first made in the horizontal geometry that in-ice neutrino detectors actually use, rather than the vertical 'bed-bounce' geometry of all earlier attenuation estimates. Using two transmitters frozen into 1400-m-deep boreholes and a movable dipole lowered into the 1700-m ice-core hole, the experiment also constrains the depth-dependent refractive index, measures a polarization-dependent arrival-time asymmetry (birefringence) at the 0.15% level, and observes signals arriving from the geometric shadow zone where ray optics predicts none. If the attenuation result is right, kilometre-spaced antenna arrays are viable for ultra-high-energy neutrino astronomy, and the refractive-index uncertainty that had clouded sensitivity estimates shrinks to a few percent. The shadow-zone and amplitude-modulation anomalies, however, indicate that the ice is not the smooth, laterally uniform medium the analysis assumes.","feed_headline":"Antarctic ice attenuates radio over 1.43 km, a horizontal first","feed_subtitle":"Pulses recorded 1–5 km from deep transmitters confirm the radio reach that kilometre-spaced neutrino detectors need.","key_machinery":"The argument rests on three measured quantities carried by the received waveforms. The attenuation length uses amplitude ratios: for the same transmitter pulse seen by the same type of antenna at two stations at distances $d_1$ and $d_2$, the ratio $V_1/V_2 = (d_2/d_1)\\exp((d_2-d_1)/L_{atten})$ isolates the exponential loss, since the $1/r$ spreading factor and all source and bedrock unknowns cancel. The refractive-index constraint uses the direct/refracted 'double pulse': a refracted ray that turns over in the low-density firn arrives tens to hundreds of nanoseconds after the direct ray, and the evolution of that delay $\\delta t(D,R)$ with transmitter and receiver depth, plus the depth at which the signal emerges from the geometric shadow zone, selects among parameterizations of $n(z) = 1.78 + B\\exp(Cz)$, with the paper's best-fit $B=0.61$, $C=0.0172$. The birefringence observable is the per-kilometre HPol-minus-VPol arrival-time difference plotted against the angle between the propagation direction and local ice flow, whose linear fit parameterizes the crystal-orientation fabric.","core_discovery":"On its own terms, the paper establishes that radio-frequency pulses propagate across 1–5 km of cold polar ice with a field attenuation length of $1.43 \\pm 0.25 \\pm 0.37$ km, the first such value measured in the horizontal geometry characteristic of in-ice neutrino signals rather than by vertical bed-bounce. The value comes from comparing the amplitudes of the same transmitter pulse at two differently distant receiver stations, a ratio that cancels unknown transmitter power, bedrock reflectivity, and flux-focusing factors. The paper also reports that the direct/refracted double-pulse timing data allow several refractive-index parameterizations, all of which agree to within 3% in neutrino effective volume; that the HPol-minus-VPol arrival-time asymmetry is fit by $\\delta t(H-V) = (8.3\\pm1.3)\\cos\\theta - (8.6\\pm0.9)$ ns/km, where $\\theta$ is the angle to the local ice-flow direction, enabling ~15% range-to-vertex estimates; and that signals appear inside the geometric shadow zone together with strong depth-dependent amplitude modulation, which a smoothly varying $n(z)$ model cannot reproduce.","pith_inferences":["A testable extension sits in the paper's own data: the attenuation length could be recomputed depth-by-depth from the movable-transmitter descent, and if $L_{atten}$ oscillates with transmitter depth in phase with the observed SNR modulation, then the modulation is contaminating the amplitude ratio and the 1.43 km figure is an average over interference, not a pure absorption length.","The slower piezo transmitter yields a longer apparent attenuation (~2.4 km) than the fast pulsers (~1.3–1.5 km); if pulse sharpness changes the apparent attenuation, an untested prediction is that narrowband measurements at different carrier frequencies would find a frequency-dependent $L_{atten}$, which would indicate scattering losses rather than pure absorption.","The same double-pulse timing technique could serve as a continuous firn monitor: the shadow-boundary depth inferred from the $\\delta t(D,R)$ x-intercept should respond to seasonal surface-density changes, so the calibration infrastructure doubles as a glaciological instrument.","The nonzero intercept in the birefringence fit ($b = -8.6$ ns/km) means a flow-parallel H-V asymmetry exists even at zero angle to flow; if real, it says the fabric model needs more than the simple girdle picture and it would set the zero-point of the range estimator for the most common, roughly flow-parallel, geometries."],"forward_implications":["Kilometre-scale spacing of antennas and stations is viable: with a ~1.5 km field attenuation length, Askaryan radio signals remain detectable across the multi-kilometre baselines that next-generation in-ice arrays contemplate.","Refractive-index uncertainty no longer limits detector sensitivity: the tested $n(z)$ parameterizations agree to roughly 3% in neutrino effective volume, so array layouts and triggers can be fixed without a definitive density profile.","Birefringence becomes a ranging observable: the measured $\\delta t(H-V) = (8.3\\pm1.3)\\cos\\theta - (8.6\\pm0.9)$ ns/km lets an experiment convert a measured polarization arrival-time asymmetry into a distance to the neutrino vertex with ~15% precision, which is needed for neutrino energy estimation.","The shadow-zone anomalies enlarge the effective aperture: signals observable where geometric optics predicts none mean near-surface and shallow stations can still trigger on in-ice sources, as the paper notes may enhance the detector aperture.","The horizontal attenuation value anchors depth-dependent absorption models: unlike the old depth-averaged bed-bounce numbers, this measurement constrains the upper (colder) half of the ice sheet separately, tying down the dominant contribution to the neutrino target volume."],"supporting_citations":[{"why":"Prior estimate of ~1.5 km attenuation length for the upper half of the South Polar ice sheet, which the new horizontal measurement confirms.","marker":"[39]"},{"why":"Earlier South Polar in situ radio-frequency attenuation measurement in vertical geometry, the benchmark that the previous >1 km determinations relied on.","marker":"[15]"},{"why":"Earlier work with the same receiver stations measuring the real dielectric permittivity of glacial ice; it establishes the double-pulse timing and range-estimation techniques the present paper extends.","marker":"[13]"},{"why":"Earlier vertical-transmission birefringence measurements showing a ~0.15% polarization-dependent wavespeed asymmetry, the benchmark for the present HPol-VPol result.","marker":"[14]"},{"why":"Prior observation of radio signals propagating from within the geometric shadow zone, the anomaly the paper corroborates and the motivation for questioning propagation in a smooth refractive index profile.","marker":"[22]"},{"why":"Concurrently recorded data from a separate Antarctic radio array showing similar depth-dependent signal-amplitude modulation, supporting the paper's claim that the modulation is a real ice-propagation effect.","marker":"[32]"},{"why":"South Pole temperature-versus-depth profile used to correct and interpret the depth dependence of attenuation in earlier measurements and in the present data.","marker":"[36]"},{"why":"Ice-core density data from the upper South Polar ice sheet, one of the reference datasets used to validate the fitted refractive index profile n(z).","marker":"[34]"},{"why":"In situ firn index-of-refraction measurements that anchor the density-to-refractive-index relation used in the n(z) model comparisons.","marker":"[8]"},{"why":"Measurements and modeling of near-surface glacial-ice radio propagation, cited jointly with [22] for the unexpected shadow-zone signal propagation.","marker":"[23]"}],"fun_headline_variants":["Radio travels kilometers horizontally through polar ice","First horizontal ice radio attenuation: 1.43 km","Radio spans 5 km in polar ice, attenuation 1.43 km","1.43 km attenuation length in horizontal ice radio path","Ice radio reach extends to 5 km for neutrino detectors"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that received signals follow geometric-optics rays through a smoothly varying, laterally uniform refractive index profile $n(z)$; the paper itself reports shadow-zone signals and depth-dependent amplitude modulations that this model cannot reproduce, so if those anomalies come from volume scattering or internal layers, the derived index profile and the incidence-angle corrections entering the birefringence extraction could be biased — though the central attenuation length, based on amplitude ratios, is the part least exposed to this assumption.","fun_headline_variants_meta":{"raw":{"variants":["Radio travels kilometers horizontally through polar ice","First horizontal ice radio attenuation: 1.43 km","Radio spans 5 km in polar ice, attenuation 1.43 km","1.43 km attenuation length in horizontal ice radio path","Ice radio reach extends to 5 km for neutrino detectors"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000939,"raw_usage":{"total_tokens":4038,"prompt_tokens":991,"completion_tokens":3047,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":607,"completion_tokens_details":{"reasoning_tokens":2966}},"tokens_in":607,"tokens_out":3047,"duration_ms":24420,"temperature":1.0,"reasoning_tokens":2966,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T10:37:12.299209+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A dedicated multi-station run would settle it: broadcast one sharp transmitter pulse and record it at three or more stations at different baselines on roughly the same bearing with independently calibrated channel gains, then check that the inferred $L_{atten}$ is identical for every station pair. If the apparent attenuation length changes with baseline, receiver depth, or pulse sharpness (the slower piezo day already gives ~2.4 km versus ~1.3–1.5 km for the fast pulsers), the amplitude-ratio method is contaminated by multipath or scattering, and 1.43 km is an effective, not intrinsic, value. A laboratory cross-check: measure the complex permittivity of deep South Pole ice cores over 100–800 MHz and ask whether the implied absorption length over 200–1200 m depth is consistent with $1.43 \\pm 0.44$ km.","supporting_citations":[{"cited_title":"Allison and J","cited_arxiv_id":null,"evidence_quote":"Prior estimate of ~1.5 km attenuation length for the upper half of the South Polar ice sheet, which the new horizontal measurement confirms."},{"cited_title":"South Polar in situ radio-frequency ice attenuation","cited_arxiv_id":null,"evidence_quote":"Earlier South Polar in situ radio-frequency attenuation measurement in vertical geometry, the benchmark that the previous >1 km determinations relied on."},{"cited_title":"Measurement of the real dielectric permittivity epsilon’ of glacial ice","cited_arxiv_id":null,"evidence_quote":"Earlier work with the same receiver stations measuring the real dielectric permittivity of glacial ice; it establishes the double-pulse timing and range-estimation techniques the present paper extends."},{"cited_title":"Radio Frequency Birefringence in South Polar Ice and Implications for Neutrino Reconstruction.Astropart","cited_arxiv_id":null,"evidence_quote":"Earlier vertical-transmission birefringence measurements showing a ~0.15% polarization-dependent wavespeed asymmetry, the benchmark for the present HPol-VPol result."},{"cited_title":"Temperature profile for glacial ice at the South Pole: Implications for life in a nearby subglacial lake","cited_arxiv_id":null,"evidence_quote":"South Pole temperature-versus-depth profile used to correct and interpret the depth dependence of attenuation in earlier measurements and in the present data."},{"cited_title":"A 237-meter ice core from south pole station.Antarct","cited_arxiv_id":null,"evidence_quote":"Ice-core density data from the upper South Polar ice sheet, one of the reference datasets used to validate the fitted refractive index profile n(z)."},{"cited_title":"In situ index-of-refraction measurements of the south polar firn with the rice detector.Journal of Glaciology , 50(171):522–532, 2004","cited_arxiv_id":null,"evidence_quote":"In situ firn index-of-refraction measurements that anchor the density-to-refractive-index relation used in the n(z) model comparisons."}],"review_version":1}