{"id":"3610611f-e91e-45b1-b2a2-9641c2de22aa","arxiv_id":"2502.08867","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A beamforming plus near-field correction technique estimates RFI source altitude (11.7 km) and speed (792 km/h), identifying an airplane in MWA data and two additional airplane events.","lead":"This paper estimates the altitude and speed of radio-frequency interference sources in MWA telescope data by applying near-field corrections and beamforming. The method identifies three RFI events as airplanes, which could eventually let astronomers subtract instead of flag contaminated data.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Altitude estimate depends on WSClean far-field coordinates that the near-field effect itself biases; no ground truth validates the focus-distance maximum, so the airplane conclusion is not yet established.","rationale":"The paper's stated goal is to estimate an RFI emitter's altitude from the focal distance that maximizes beamformed intensity after near-field corrections. For that estimate to be correct, the RA/Dec used as the phase center must be the true direction to the emitter, and the intensity maximum must correspond to the physical distance. The first condition is the least secure. The source is imaged by WSClean under the far-field assumption even though the paper demonstrates that near-field sources are smeared in such images; the fitted coordinates are therefore a centroid of a blurred source, not a measured geometric direction. The paper acknowledges that coordinate errors are excluded from the quoted uncertainties. The second condition is also untested: no known-altitude source is processed to show that the method recovers the correct distance, and the beamformed intensity as a function of focal distance is not shown to be unimodal or robust to sidelobes. The LEO degeneracy illustrates that the altitude and speed estimates are coupled, so their mutual consistency cannot substitute for an external reference. The reader's conditional verdict is appropriate; the concern is addressable by simulation and by seeking an event with independent flight data, so it does not warrant rejection, but it does require validation before the 'definitive' claim can stand.","tokens_in":9667,"tokens_out":6468,"duration_ms":70176,"concrete_test":"Generate simulated MWA Phase I visibilities for a point source at known altitude (e.g., 11.7 km) using the same near-field delay model, image with WSClean under the far-field assumption to obtain source coordinates, then run the altitude estimator. Repeat for a point source at 400 km with a LEO angular rate. Compare recovered focal distances and altitudes to truth. If the recovered altitude from WSClean coordinates is biased by more than the quoted standard error, or if the 400 km source also produces a beamformed-intensity peak near 12 km, the central inference is not supported without an independent position reference.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central inference in Section 3.1 is that the beamformed-intensity maximum over focal distance f yields the true source altitude. This requires that the far-field RA/Dec coordinates from WSClean, which anchor the near-field corrections in Eq. (1), are unbiased. But the paper itself explains in Section 2.3 that a near-field source appears smeared in far-field images, with different baselines placing it at different positions. WSClean component fitting therefore returns a brightness-weighted centroid of that smear, and nothing in the paper shows this centroid equals the geometric direction to the source. The quoted 11.7 +/- 0.1 km is the standard error of the mean over time-steps (Eq. 4), which measures scatter across samples, not accuracy; Section 4 explicitly excludes WSClean coordinate errors and beamforming errors. Because the speed is computed as angular displacement times the estimated slant distance, any altitude bias propagates directly into the speed, so the agreement of 792 km/h with airplane cruise speed is not independent confirmation. Moreover, a LEO satellite at roughly 400 km altitude moving at 7.5 km/s has a similar apparent angular rate of about 1 deg/s and produces non-negligible wavefront curvature over MWA Phase I baselines; without an external reference the observed consistency cannot distinguish these cases. The two additional observations were selected for their visual similarity to the target and are also unverified. The claim of a 'first definitive detection and localization' therefore rests on an unquantified coordinate step plus a focus-maximum assumption that has not been validated against a source of known position.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a method for estimating the altitude and velocity of near-field radio-frequency interference (RFI) emitters in Murchison Widefield Array (MWA) data by combining far-field phasing, geometric near-field corrections following Prabu et al. (2023), and beamforming. For a two-minute 2013 Phase I observation, the authors identify 59 time-steps containing a bright RFI source, obtain its RA/Dec at each time-step via WSClean component fitting, and then scan a focal-distance parameter, beamforming at each distance to find the maximum intensity. They report an average altitude of 11.7 ± 0.1 km and speed of 792 ± 1 km/h, conclude that the source is an airplane, and apply the technique to two further observations (reported altitudes 11.73 km and 13.9 km; speeds 1050 and 1360 km/h), claiming consistent airplane identification and the first definitive airplane detection and localization in MWA data. The manuscript is explicitly framed as a preliminary study: the authors state that no detailed quantitative assessment is provided, that flight-track validation was unsuccessful, and that the reported errors exclude coordinate, calibration, and beamforming systematics.","tokens_in":9941,"tokens_out":12059,"duration_ms":112892,"significance":"The geometric framework is transparent: the near-field delay corrections (Eqs. 1-3) are explicit, the focal-distance maximization is a clean estimation procedure, and replacing per-distance imaging with beamforming is a genuine computational simplification relative to the imaging-based approach of Prabu et al. (2023). The paper is also admirably candid about its limitations, including the failure to locate the flight, the exclusion of coordinate and beamforming errors from the quoted uncertainties, and the poor performance of the third observation. If the focal-distance maximum can be shown to be an unbiased altitude estimator, the method would be a useful step toward RFI peeling and toward preserving larger fractions of EoR data. The significance is moderate, however: the central quantity (altitude) is never checked against ground truth, the two consistency checks in Section 4 are not independent because the speed is derived from the measured angular rate times the estimated altitude, and the 'definitive detection' claim is stronger than the evidence presented.","major_comments":[{"comment":"The altitude estimate rests on the assumption that the focal distance maximizing beamformed intensity equals the true source distance, and this step inherits a coordinate bias that is neither quantified nor bounded. Section 2.3 states that a near-field source is smeared in far-field images, with different baselines placing it at different angular positions, yet the RA/Dec coordinates used in Eq. (1) come from far-field WSClean component fitting (Section 3). The paper does not demonstrate that the fitted component position equals the geometric direction to the emitter; a brightness-weighted centroid of the smear could differ from it. Any such bias propagates through Eqs. (1)-(3) directly into the fitted focal distance, the altitude, and the derived speed, and Section 4 explicitly excludes this channel from the error budget. A focused test, such as injecting a point-like near-field source at known position and distance into the visibilities and recovering both coordinates and distance, would establish whether the estimator is unbiased; without it, the reported numbers rest on an unverified assumption.","section":"§3.1, §2.3"},{"comment":"The quoted uncertainties (11.7 ± 0.1 km, 792 ± 1 km/h) are standard errors of the mean across time-steps and therefore measure run-to-run scatter, not accuracy; the paper acknowledges this but still uses these values to support the airplane conclusion. Because the speed is obtained from the angular displacement times the estimated slant distance, the altitude and speed checks in Section 4 are not independent: at the measured angular rate, any altitude bias scales directly into the speed. The agreement of 11.7 km and 792 km/h with typical airplane operating ranges is therefore internal consistency of one derived quantity, not two independent confirmations, and it cannot sustain the abstract's 'confidently conclude' unless the focal-distance estimator's accuracy is independently established.","section":"§4, Eq. (4)"},{"comment":"The purported validation observations do not satisfy the paper's own classification criteria. For OBSID 1252945816 the reported altitude (13.9 ± 0.9 km) lies above the 9.4-11.6 km cruising-altitude range quoted in Section 4, and the speed (1360 ± 30 km/h) is transonic to supersonic at that altitude and far above typical civil cruise speeds. The truncation threshold reported for this observation (0.45 σ below the maximum, Section 4.1) indicates a weak intensity peak, and the paper concedes that this observation 'performs significantly worse' than the others, yet it is still counted as an airplane identification and as part of the claim that airplanes are 'consistently' identified. Additionally, the two supporting observations were selected for visual similarity to the target's RFI signature after manual inspection, so the validation is not blind; this should be stated explicitly and the third observation's identification should either be explained or excluded from the consistency claim.","section":"§4.1, Table 1"},{"comment":"The claim of 'the first definitive detection and localization of an airplane in MWA data' is not supported by the evidence presented. No independent validation is provided: the paper reports that flight-track lookup failed, and no ADS-B, radar, orbital, or other corroborating data are used. Alternative near-field emitters are not quantitatively excluded; for example, the measured angular rate (~1 deg/s) is also consistent with a low-Earth-orbit satellite at a much larger distance with a much higher physical speed, and the near-field curvature of such a satellite over MWA Phase I baselines is non-negligible. The introduction itself disclaims a 'detailed quantitative assessment,' which sits in tension with the abstract's confident classification. The claims should be reframed as a proof-of-concept demonstration with a stated accuracy requirement, or the manuscript should be strengthened with external validation (e.g., a known satellite pass or an aircraft track with independent position data).","section":"§5, Abstract"}],"minor_comments":[{"comment":"Equation (4) prints 'SE = σ√n'; the standard error of the mean should be σ/√n.","section":"§3 (Eq. 4)"},{"comment":"The term w_far-field,i,j is used in the definition of Δw_{i,j} but is never defined in the text; define it explicitly, and rewrite 'expi2π' as exp(2πi Δw/λ) for clarity.","section":"§2.3 (Eq. 3)"},{"comment":"The text refers to 'the third observation from the left' and 'all three observations,' but the Figure 4 panels are not labeled with OBSIDs; label the panels so the discussion is unambiguous.","section":"§4.1, Fig. 4"},{"comment":"The measured target altitude of 11.7 km sits marginally above the quoted airplane cruising range of 9.4-11.6 km; clarify whether the values are above sea level or above ground level and note whether this offset is attributable to the unquantified systematics.","section":"§4"},{"comment":"There is a typo in the Figure 1 caption ('where the the object is successfully imaged'), and the paragraph at the start of Section 3 begins with a stray 's' ('sThe data used in our preliminary study...').","section":"§3, Fig. 1 caption"},{"comment":"For reproducibility, state the focal-distance search range and grid spacing used in the beamforming scan, as well as the number of focal distances evaluated per time-step.","section":"§3.1"},{"comment":"The disclaimer that 'This research does not provide a detailed quantitative assessment' should be reconciled with the abstract's 'confidently conclude that the object in question is in fact an airplane'; the current framing makes the paper's evidentiary standard unclear.","section":"§1 (Introduction)"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is in scope for PASA and the method is a plausible incremental contribution over Prabu et al. (2023), but the 'definitive detection' claim is the kind of strong statement that invites scrutiny without external validation. I would urge the editor to require either (a) an external validation test, such as a known satellite pass with well-determined ephemeris or an aircraft observation with independent position data, or (b) a systematic softening of the classification and 'first definitive' claims, plus an expanded error budget that includes the coordinate and beamforming systematics the authors already identify. The authors' candor about limitations is a real strength and should be preserved."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's my read. The genuinely new piece is replacing Prabu et al.'s iterative imaging with a beamforming sweep over focal distance, which is orders of magnitude cheaper, and applying it to airplane reflections in MWA data. That is a real, bounded extension, and the paper is upfront that it's a preliminary demonstration. The geometry in Eqs. (1)-(3) is transparent, the method section is readable, and the authors explicitly say the quoted error bars are standard errors of the mean, not systematics. They also admit they could not find the flight in historical airspace data and that the third observation is much weaker. That honesty earns credit.\n\nThe soft spots are real but not disqualifying for a first demonstration. The main one is that the focus-distance maximum is trusted as the true distance, and the near-field corrections are anchored to RA/Dec coordinates from WSClean component fitting. As the paper itself explains, a near-field source appears smeared in far-field images, so WSClean returns a brightness-weighted centroid of that smear. Nothing in the paper shows that centroid equals the geometric direction to the source. If the centroid is biased, the altitude and speed inherit a systematic error that the reported uncertainties don't cover. The speed check is then not independent confirmation, because speed is derived from angular displacement times the estimated slant distance. The two validation observations were chosen for visual similarity to the target, so they don't break the circularity. The satellite alternative the stress-test raises is less worrying because the estimated altitude is 11.7 km, but that only holds if the focus maximum is unbiased.\n\nThat said, the method is plausible, the paper says the right things about its own limitations, and the weakness is an unvalidated coordinate step rather than an internal contradiction. I'd send it to a good referee, with the expectation that the referee asks for an external validation or a simulation-based test of the coordinate bias. It deserves referee time; it doesn't deserve acceptance as is.","headline":"Useful, honest, bounded extension of Prabu et al. that trades iterative imaging for a beamforming sweep; the airplane claim is plausible but not yet nailed down because the coordinate anchor is unvalidated.","tokens_in":10484,"tokens_out":1976,"would_cite":true,"duration_ms":20244,"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":"The paper shows that an RFI-emitting object crossing an interferometer's field of view can be brought into focus with near-field corrections, yielding an altitude of 11.7 km and a speed of 792 km/h that identify it as an airplane.","keywords":["instrumentation: interferometers","methods: data analysis","radio-frequency interference","near-field corrections","beamforming","Murchison Widefield Array","airplane detection","Epoch of Reionization"],"falsifier":"Run the same focal-distance scan on an object whose range is independently known, such as an aircraft with an ADS-B flight track or a satellite with a precise ephemeris, and compare the beamformed peak distance with the true range; a systematic mismatch would show the peak is not a faithful distance estimator. A simpler check is to scan a bright far-field calibrator: a spurious finite-distance peak would indicate the maximization is responding to sidelobe structure rather than true focus.","tokens_in":9474,"feed_emoji":"✈️","tokens_out":8245,"duration_ms":69941,"temperature":0.7,"pith_summary":"The paper aims to turn radio-frequency interference from a nuisance into a measurable signal. By phasing an interferometer for a source at a finite distance rather than at infinity, a reflecting object can be brought into focus, and the distance that maximizes the beamformed intensity estimates its altitude. Applied to a two-minute Murchison Widefield Array observation, the method gives $11.7 \\pm 0.1$ km altitude and $792 \\pm 1$ km/h speed for an unknown object, which the authors identify as an airplane; two further observations also yield airplane-like altitudes and speeds. The authors claim this is the first definitive detection and localization of an airplane in MWA data, and argue that such localization could replace crude RFI flagging with subtraction or peeling, preserving more of the data needed for Epoch of Reionization science.","feed_headline":"RFI source ID'd as airplane at 11.7 km altitude","feed_subtitle":"Near-field corrections focus a reflected signal, recovering altitude and speed that flagging alone would throw away.","key_machinery":"The load-bearing object is the near-field phase correction: for an assumed focal distance $f$, the geometric delay to each antenna is computed from spherical geometry, and the difference between this near-field delay and the far-field delay is applied as a per-baseline phase. Beamforming then averages all corrected visibilities into a single scalar, and the assumed $f$ that maximizes this scalar is the estimated distance to the emitter. This replaces the earlier iterative imaging-plus-SNR maximization with a one-dimensional scan, making the localization orders of magnitude faster while keeping a similar accuracy.","core_discovery":"The central discovery is that the distance to a near-field RFI emitter can be read off from the focal distance at which the beamformed visibility is maximized. A source close to the array produces spherical wavefronts, so each baseline sees a distance-dependent excess delay over the plane-wave assumption; correcting for that delay at a trial focal distance brings the source into coherence exactly when the trial distance matches the true distance. The paper applies far-field phasing to the object's coordinates, applies per-baseline near-field phase corrections, and averages all visibilities, so no image cube is needed at each trial distance. Repeating the scan over time yields an altitude track, and combining angular displacement with altitude gives a speed; for the target observation the result, $11.7 \\pm 0.1$ km and $792 \\pm 1$ km/h, matches an airplane's cruising altitude and speed.","pith_inferences":["A direct accuracy test would run the same focal-distance scan on a source with an independently known range, such as a calibration drone or a satellite with a precise ephemeris, separating the method's precision from its accuracy; the paper only reports precision from the standard error of the mean.","If the beamformed peak is a faithful distance estimator, the approach could be extended to fainter RFI that escapes current flagging, provided faint-source coordinates can be obtained; the paper explicitly leaves that generalization to future work.","The same scan could be used opportunistically to monitor aircraft over radio-quiet zones from existing telescope data, since the cited study finds aircraft above the MRO horizon at least 13% of the time.","Because the focal scan reduces localization to maximizing a single scalar, replacing the imaging-based coordinate fitting with a joint position-and-distance search could make near-real-time RFI tracking feasible on modest computing hardware."],"forward_implications":["RFI from airplanes could be subtracted or peeled from the visibilities rather than flagged, preserving time-frequency channels that would otherwise be discarded from Epoch of Reionization analyses.","The technique is not restricted to the target observation: two additional MWA observations yield airplane-like altitudes of 11.73 km and 13.9 km and speeds of 1050 km/h and 1360 km/h, showing repeatability across array configurations.","The method's precision is limited more by array configuration and time resolution than by the focal-distance search; for shorter baselines the far-field assumption is valid to closer distances, and 2-second integrations smear fast airplanes into non-point-like images.","The airplane identification, the authors state, is the first definitive detection and localization of an airplane in MWA data, establishing that aircraft-reflected RFI can be identified rather than merely flagged."],"supporting_citations":[{"why":"Supplies the near-field correction formalism and the earlier result that the optimal focal distance maximizes image SNR, which the beamforming shortcut builds on.","marker":"Prabu et al. (2023)"},{"why":"Establishes that near-field corrections bring near-field objects into focus in radio imaging, the conceptual basis for the altitude scan.","marker":"Marr et al. (2015)"},{"why":"WSClean is used to image each time step and fit the right ascension and declination coordinates required for far-field phasing.","marker":"Offringa et al. (2014)"},{"why":"Provided the target observation's OBSID and identified it as containing RFI, making it the known-contaminated dataset for the study.","marker":"Wilensky et al. (2019)"},{"why":"Provides the airplane cruising altitude range (9.4–11.6 km) used to classify the 11.7 km estimate as an airplane.","marker":"Sforza (2014)"},{"why":"Shows that the far-field assumption holds at closer distances for shorter baselines, used to explain the weaker performance of the compact Phase II observation.","marker":"Prabu et al. (2022)"},{"why":"Documents that aircraft are present above the MRO horizon at least 13% of the time, establishing the practical importance of airplane RFI reflections.","marker":"Tingay et al. (2020)"},{"why":"Demonstrates that an effective RFI subtraction strategy can reduce data loss to as little as 1%, motivating the shift from flagging to subtraction.","marker":"Finlay et al. (2023)"}],"fun_headline_variants":["RFI altitude pinpointed: airplane at 11.7 km, flying 792 km/h","Near-field focus finds airplane RFI at 11.7 km altitude","RFI source located: airplane at 11.7 km, speed 792 km/h","Using near-field focus, RFI altitude estimated: plane at 11.7 km","Airplane RFI's altitude and speed from near-field focal scan"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The estimate assumes that the focal distance giving the highest beamformed intensity equals the true distance to a point-like reflector whose sky coordinates come from far-field imaging; if the reflector is extended or the coordinates are biased, the altitude and speed carry a systematic error the paper does not quantify.","fun_headline_variants_meta":{"raw":{"variants":["RFI altitude pinpointed: airplane at 11.7 km, flying 792 km/h","Near-field focus finds airplane RFI at 11.7 km altitude","RFI source located: airplane at 11.7 km, speed 792 km/h","Using near-field focus, RFI altitude estimated: plane at 11.7 km","Airplane RFI's altitude and speed from near-field focal scan"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001126,"raw_usage":{"total_tokens":4694,"prompt_tokens":969,"completion_tokens":3725,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":585,"completion_tokens_details":{"reasoning_tokens":3618}},"tokens_in":585,"tokens_out":3725,"duration_ms":25947,"temperature":1.0,"reasoning_tokens":3618,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T23:24:03.161636+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the same focal-distance scan on an object whose range is independently known, such as an aircraft with an ADS-B flight track or a satellite with a precise ephemeris, and compare the beamformed peak distance with the true range; a systematic mismatch would show the peak is not a faithful distance estimator. A simpler check is to scan a bright far-field calibrator: a spurious finite-distance peak would indicate the maximization is responding to sidelobe structure rather than true focus.","supporting_citations":[{"cited_title":"2023, Publications of the Astronomical Society of Australia, 40, 1","cited_arxiv_id":null,"evidence_quote":"Supplies the near-field correction formalism and the earlier result that the optimal focal distance maximizes image SNR, which the beamforming shortcut builds on."},{"cited_title":"M., Snell, R","cited_arxiv_id":null,"evidence_quote":"Establishes that near-field corrections bring near-field objects into focus in radio imaging, the conceptual basis for the altitude scan."},{"cited_title":"J., Morales, M","cited_arxiv_id":null,"evidence_quote":"Provided the target observation's OBSID and identified it as containing RFI, making it the known-contaminated dataset for the study."},{"cited_title":"2014, in Commercial Airplane Design Principles, ed","cited_arxiv_id":null,"evidence_quote":"Provides the airplane cruising altitude range (9.4–11.6 km) used to classify the 11.7 km estimate as an airplane."},{"cited_title":"2022, Advances in Space Research, 70, 812–824","cited_arxiv_id":null,"evidence_quote":"Shows that the far-field assumption holds at closer distances for shorter baselines, used to explain the weaker performance of the compact Phase II observation."},{"cited_title":"J., Sokolowski, M., Wayth, R., & Ung, D","cited_arxiv_id":null,"evidence_quote":"Documents that aircraft are present above the MRO horizon at least 13% of the time, establishing the practical importance of airplane RFI reflections."},{"cited_title":"A., Kunz, M., & Oozeer, N","cited_arxiv_id":null,"evidence_quote":"Demonstrates that an effective RFI subtraction strategy can reduce data loss to as little as 1%, motivating the shift from flagging to subtraction."}],"review_version":1}