{"id":"5bd991f8-c409-4800-8290-6ef3b646c6db","arxiv_id":"2607.24527","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":8,"one_line_summary":"Planned lunar satellites will increasingly contaminate the farside radio band; under this model, ≥30 dB uniform UEMR shielding is required to keep peak RFI below LFT3-like sensitivity.","lead":"Simulations of planned lunar satellites show the Moon’s farside will grow steadily noisier in radio, mainly from unintended electronics leakage rather than deliberate transmitters. Roughly 30 dB of uniform shielding would be needed to keep peak interference under a LFT3-like telescope’s sensitivity, and that clean window is closing.","discovery_kind":"new_application","skeptic_critique":{"model":"moonshotai/kimi-k3","headline":"The headline 31.40 dB shielding figure is set at 410.9 MHz — inside the regime where the UEMR catalogue has no measurements and content is synthesized from sinc²-sidelobe harmonic rules; the number is also a Monte-Carlo maximum with no reported seed sensitivity.","rationale":"I agree with the reader's identification of the UEMR high-frequency extrapolation as the load-bearing weakness, and with the CONDITIONAL verdict, which my analysis does not move in either direction. My contribution is to sharpen the concern in three ways. (1) The coincidence is exact: the headline number's anchor frequency, 410.9 MHz, lies in the extrapolated regime, so the most-cited quantitative result inherits the full uncertainty of the least-evidenced part of the model — it is not a peripheral output. (2) The specific harmonic amplitude rule (sinc² sidelobe maxima) is an arbitrary physical proxy with no stated validation against measured spacecraft leakage; this makes the model dependence concrete and points at exactly which line of the pipeline to perturb. (3) Two additional checkable internals the reader did not flag: the maximum-RFI spectrum is a Monte-Carlo object with no reported convergence or seed analysis, and the bandwidth used in the effective-sensitivity curve is never specified, creating a possible per-channel vs full-band normalization mismatch in the very comparison that yields 31.40 dB. These are calibration issues, not flaws in reasoning: the pipeline is open-source, the geometry-driven results (visibility growth, sky distribution, IEMR localisation) are independent of the UEMR library and look sound, and the authors themselves flag the 230 MHz limitation prominently and propose the measurement campaign that would resolve it. The paper's policy-relevant conclusions survive any plausible outcome of my proposed test; only the precision of the shielding figure is at stake. Hence CONDITIONAL stands, with the condition now precisely specified: the shielding number should be presented as model-dependent (ideally a range over harmonic-envelope choices and seeds) rather than a point estimate.","tokens_in":16997,"tokens_out":2262,"duration_ms":83688,"concrete_test":"Using the released FEARLESS code (github.com/ScienceMoonshot/Lunar-RFI-Study), reproduce Fig. 7 three ways: (a) harmonics truncated above 230 MHz entirely (measured UEMR only), (b) sinc²-sidelobe envelope replaced by a flat −10 dB/harmonic rule, (c) baseline. For each, run 5 random seeds and record the required uniform shielding and its frequency. If the baseline 31.40 dB moves by more than ~6 dB or the worst frequency changes regime, report the shielding figure as a range with stated harmonic-model dependence rather than a point value; also confirm in the code that the 1-s sensitivity curve uses Δν = 0.1 MHz per channel.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The reader's weakest-assumption identification is correct and can be sharpened. The strongest_claim's quantitative core — \"uniform UEMR shielding of at least 31.40 dB, set by the worst frequency 410.9 MHz\" — depends entirely on modelled power at 410.9 MHz, which §5 concedes is above the ~230 MHz limit of directly measured UEMR. Everything above 230 MHz is generated by the harmonic machinery of §3.3: harmonics to order n=8, widths scaled by n, and peak heights taken from \"the relative maxima of the sidelobes of a sinc-squared response.\" That last rule is the soft spot. Sinc² sidelobe envelopes describe a specific idealized waveform (rectangular pulse with negligible rise time); real digital/clock leakage harmonic envelopes are set by slew rates, spread-spectrum clocking, and filtering, and can fall off much faster or slower than 1/n². The choice directly sets the power at 410.9 MHz (which for a ~51 MHz fundamental would be n=8, the most suppressed harmonic under sinc²; for a ~103 MHz fundamental, n=4). So the single frequency that pins the headline number is one whose modelled amplitude is essentially a convention, not a measurement. Second, the maximum-RFI spectrum of Fig. 5/7 is a stochastic object: occurrence-fraction draws, Gaussian EIRP sampling, and uniform draws of narrowband centre frequencies (§3.3) all feed it, and the paper reports no seed-to-seed spread or convergence of the maximum over the ~3×10⁴ timesteps. A different realisation could move the worst bin or its level by several dB. A secondary, checkable worry: §3 never states what Δν enters the \"effective sensitivity, t=1 s\" curve of Figs. 5 and 7. If the full sub-band width was used rather than the 0.1 MHz channel width in which RFI power is evaluated, the shielding requirement shifts systematically (likely upward, since full-band sensitivity is deeper). None of this threatens the qualitative claims (visibility growth, UEMR dominance, clean-window closure), which rest on geometry and catalogue counts","agreement_with_reader":"agree"},"referee_report":{"model":"moonshotai/kimi-k3","summary":"The paper presents FEARLESS, an open-source forward simulation of the radio-frequency interference environment at the planned LFT3 landing site on the lunar farside. A catalogue of ~50 planned lunar/cislunar satellites (Table A1) is propagated over 2025–2030; intended (IEMR, mostly 2.4–2.6 GHz Gaussian features) and unintended (UEMR, catalogue-based up to ~230 MHz, extended by harmonics to order n=8 with sinc²-sidelobe-derived amplitudes) emissions are converted to received power via Friis-style flux and the frequency- and direction-dependent effective areas of the three LFT3 antenna systems. Main findings: visible-satellite counts grow from ~1–2 mean daily (2025–2027) to maxima of ~16 by 2029; UEMR contaminates the HF/VHF/UHF bands broadly while IEMR is spectrally localised; and a uniform UEMR shielding of 31.40 dB — set by the worst frequency, 410.9 MHz — is required to keep maximum UEMR below the 1 s LFT3 sensitivity. The authors conclude the farside remains usable but increasingly contaminated, and that satellite-level shielding is the only effective broad-band mitigation.","tokens_in":17508,"tokens_out":3333,"duration_ms":110837,"significance":"If the results hold, this is a timely and policy-relevant contribution: the lunar farside is currently the only radio-quiet site in the inner solar system, and a credible forward model of its degradation directly informs mission shielding requirements, ITU-style protection discussions, and the scheduling of instruments such as LFT3, LuSEE-Night, and FARSIDE. The work has genuine strengths worth naming: the FEARLESS code is released open source with a documented mission catalogue (Table A1) so individual entries can be revised; the modelling is a forward simulation rather than a fit dressed as prediction; the instrument-response treatment (Eq. 2, beam-weighted off-axis gains across three antenna types) is more careful than typical EIRP-to-flux estimates; and the paper is commendably candid in §5 about the UEMR catalogue limitations, proposing a concrete measurement campaign to remedy them. The qualitative conclusions — visible-satellite counts rising sharply after 2028, UEMR dominating over spectrally localised IEMR, and shielding being the only broad-band mitigation — are robust to the catalogue uncertainties and constitute a useful, falsifiable baseline for the community.","major_comments":[{"comment":"The headline quantitative result — 'uniform UEMR shielding of at least 31.40 dB', set at the worst frequency 410.9 MHz (Fig. 7) — is pinned by modelled power in a regime where the UEMR catalogue contains no measurements. Section 5 concedes directly measured UEMR extends only to ~230 MHz, with higher-frequency content synthesized from harmonics to order n=8 whose peak heights are taken from 'the relative maxima of the sidelobes of a sinc-squared response' (§3.3). That envelope corresponds to an idealised rectangular-pulse waveform; real digital/clock leakage envelopes are set by slew rates, spread-spectrum clocking, and filtering, and can fall off considerably faster or slower. For a ~51 MHz fundamental, 410.9 MHz is the n=8 harmonic — the most suppressed under a sinc² rule — so the single frequency that fixes the shielding number is one whose modelled amplitude is a convention. I do not","section":"§3.3, §4.4, Fig. 7"},{"comment":"The maximum-RFI spectrum of Figs. 5 and 7 is a stochastic object: occurrence-fraction on/off draws, Gaussian EIRP sampling, and uniform draws of narrowband centre frequencies (§3.3) all feed it, over ~3×10^4 timesteps per year. No seed-to-seed spread or convergence diagnostic is reported anywhere. Because the shielding requirement is set by the single largest value in the maximum spectrum, it is precisely the statistic most sensitive to realisation noise. The authors should run multiple realisations (the code is open source, so this is a modest computational request) and report the dispersion of both the worst-bin power and the derived 31.40 dB figure. If the spread is small, stating so strengthens the result considerably; if it is not, an uncertainty band is required.","section":"§4.3–4.4, Figs. 5 and 7"},{"comment":"The effective sensitivity curve in Figs. 5 and 7 (purple, '1 s integration') is load-bearing for the shielding criterion, yet the manuscript never states the numerical inputs to Eq. (1) used to produce it: T_sys (and its frequency dependence across 0.1–2700 MHz), η_s, n_pol, and the assumed usable bandwidth Δν per channel. Eq. (1) gives the radiometer equation but no values appear in §2 or §3.1, and Table 1 lists only gains. A 0.1 MHz channel width can be inferred from §3.3, but T_sys at 0.1 MHz versus 2.5 GHz for an LFT3-like system differs by orders of magnitude and directly shifts the 31.40 dB requirement dB-for-dB. The assumed values (or a reference to DeBoer et al. 2026 with explicit citation of which numbers were taken) must be stated.","section":"§2 (Eq. 1), §4.3–4.4, Figs. 5 and 7"}],"minor_comments":[{"comment":"The Appendix heading 'APPENDIX A:' appears with no title text; Table A1 itself is present and useful. Please also clarify the 'Luna26 End' entry (2029) versus the 'Luna26' 12-month lifetime starting 2028 — presumably a mission-extension placeholder, but this should be explained in the table caption.","section":"Appendix A, Table A1"},{"comment":"The shielding analysis (Fig. 7) applies only to UEMR, while Fig. 5 shows IEMR maxima near 2.4–2.6 GHz also exceeding the sensitivity curve from 2028 onward. The text is technically clear on this, but the abstract phrase 'place the maximum interference below the sensitivity' could be read as covering all RFI; suggest 'maximum unintended interference' in the abstract for precision.","section":"Abstract, §4.4"},{"comment":"Table 1 leaves gaps in LFT3 coverage (50–60 MHz, 260–300 MHz). It is unclear how received power is treated in these gaps in Figs. 4–7 — zero gain, nearest antenna, or interpolation? One sentence in §3.1 would resolve this.","section":"§3.1, Table 1"},{"comment":"The 500 s timestep is justified only by computational cost. For a ~100 km orbit (period ~2 h), 500 s is a substantial fraction of a farside transit; please comment on whether the instantaneous-maximum statistics (Figs. 2, 5) are sensitive to this cadence, even if only via a spot check at finer resolution for one month.","section":"§3 (timestep choice)"},{"comment":"Eq. (6): 'PFD ii(ν, t)' appears to be a typo for F_i(ν, t) from Eq. (5); the subscript is duplicated. Also 'Moon-fixed coordination system' should read 'coordinate system' (§3.4).","section":"§3.4, Eq. (6)"},{"comment":"Several typos and grammar slips: 'such fast radio bursts' → 'such as fast radio bursts' (abstract and §1); 'noticably' → 'noticeably' (§4.3); 'attentuation' → 'attenuation' (Fig. 7 caption); 'decline of UEMR about 1.5 GHz' → 'above 1.5 GHz' (§6); 'a LFT3-like' → 'an LFT3-like' throughout.","section":"Various"},{"comment":"Fig. 3 shows a colourbar only on the 2028 panel; since the three panels span very different pass densities, please state whether the colour scale is common across panels or per-panel, and if per-panel, add colourbars to all three.","section":"Fig. 3"},{"comment":"Reference list inconsistencies: 'Grigg, D. et al. 2025' has a malformed author list; the entries 'Aerospace F.,', 'Lavochkin N.,', 'Technologies V. S.,', 'Ltd S. S. T.' are alphabetised and formatted as personal names — please regularise these corporate/website citations.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The receiver under test (LFT3 gains, beams, site) is taken from the coauthored DeBoer et al. 2026 concept paper, which is listed as \"MNRAS, submitted\". This is disclosed and does not drive the RFI-growth conclusion, but the editor may wish to note that a key input reference is not yet citable in final form. The author list includes the LFT3 concept lead; I see no evidence this biases the analysis — the shielding result, if anything, argues for constraints on the authors' own mission's neighbours — but the connection is worth noting. The work is a good fit for the journal's instrumentation/methods scope."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"The useful core here is not a new link-budget formula. It is FEARLESS plus a compiled lunar/cislunar mission list, run as a time-resolved IEMR/UEMR forecast at an LFT3-like farside site through ~2030, with open code. That package is actually new in this niche and is what people planning timing, shielding, or spectrum policy will want.\n\nGeometry, horizon cut, free-space flux, and A_eff from gain are standard and cleanly stated. Visibility growth, low-elevation pass density, and the qualitative split (IEMR stuck near 2.4–2.6 GHz; UEMR the broad problem) follow directly from the catalogue and do not depend on the softest spectral assumptions. The limitations section is honest: measured UEMR only to ~230 MHz, higher frequencies filled by harmonics to n=8 with sinc²-style amplitudes. Code availability is a real plus.\n\nThe soft spot is exactly where the stress note points. The headline 31.40 dB is set by the worst bin at 410.9 MHz—inside the synthesised regime—and the max spectrum is a Monte-Carlo object (occurrence draws, EIRP scatter, random narrowband centres) with no seed or convergence check reported. Change the harmonic envelope or the realisation and that single number moves by several dB. Minor secondary gap: the 1 s sensitivity curve never states which Δν was used relative to the 0.1 MHz grid. None of that sinks the paper; it means treat 31.4 dB as an order-of-magnitude, model-dependent threshold, not a regulatory constant. The qualitative claim—usable for LFT3-like work now, increasingly and more persistently contaminated late this decade—still holds.\n\nThis is for lunar radio, RFI mitigation, and cislunar spectrum people, not a general methods audience. Math and citation pattern look fine; free parameters are listed in the model, not hidden. I would send it to referees. Engage with the pipeline and the clean-window argument; quote the shielding figure only with the UEMR caveat attached.","headline":"Open multi-year lunar-farside RFI forecast with a real catalogue and code; the ~31 dB number is soft because it sits on unmeasured harmonic UEMR, but the clean-window warning is solid and useful.","tokens_in":18356,"tokens_out":562,"would_cite":true,"duration_ms":18146,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.5","headline":"Planned lunar satellites will flood the farside with unintended radio leakage unless they are shielded by at least about 30 dB.","keywords":["lunar farside","radio frequency interference","unintended electromagnetic radiation","satellite shielding","LFT3","radio astronomy","cislunar satellites","RFI modelling"],"falsifier":"Wide-band, calibrated measurements of actual lunar or cislunar spacecraft unintended emission spectra above a few hundred megahertz—compared against the paper’s harmonic-filled catalogue and against the predicted received power at a farside site—would confirm or overturn the 31.40 dB uniform-shielding requirement and the claimed contamination density by 2029–2030.","tokens_in":17891,"feed_emoji":"🌕","tokens_out":1172,"duration_ms":28435,"temperature":0.7,"pith_summary":"The lunar farside is still the cleanest radio site in the inner solar system, which is why instruments like an LFT3-style telescope want to work there. This paper models how that quietness erodes as planned lunar and cislunar satellites appear over the next few years. It finds that deliberate communications stay mostly in narrow bands near 2.4–2.6 GHz, but unintended leakage spreads across the HF, VHF, and UHF bands and becomes nearly continuous as more craft rise above the horizon. Under the model’s assumptions, only uniform shielding of roughly 30 dB or more brings the worst leakage below a one-second LFT3-like sensitivity floor. The practical message is that the farside remains usable for such instruments for now, but the clean window is closing and mission design and policy must act while it still exists.","feed_headline":"Moon's quiet farside needs ~30 dB satellite shielding","feed_subtitle":"Unintended leakage from planned lunar craft will blanket HF–UHF bands unless missions harden designs soon","key_machinery":"FEARLESS — a time-stepped simulation that propagates each catalogue satellite relative to a fixed farside site, keeps only sources above the local horizon, applies frequency-dependent IEMR and UEMR emission models, folds in the LFT3 antenna effective area and beam, and sums the received power spectrum.","core_discovery":"Using a catalogue of planned lunar missions and separate models for intended and unintended emissions, the authors show that an LFT3-like receiver on the farside will see rising, often broadband contamination as the visible satellite count grows through 2028–2030. Intended links concentrate near communications bands, especially 2.4–2.6 GHz; unintended electromagnetic radiation contaminates wide stretches of the 0.1–2700 MHz band. A uniform UEMR attenuation of at least about 30 dB (31.40 dB at the worst frequency in their run) is required to push the maximum received interference below the instrument’s one-second sensitivity; more sensitive telescopes would need still more shielding. Without","pith_inferences":["If operators treat ~30 dB UEMR shielding as a design requirement for cislunar craft, the same standard would also reduce leakage seen by Earth-based low-frequency arrays that already struggle with mega-constellations.","The sky maps’ concentration of passes near the horizon suggests that future farside arrays with steerable high-gain beams could still carve out usable high-elevation windows longer than fixed broad-beam dipoles can.","A public, regularly updated ‘farside RFI budget’ tied to real launch manifests would turn this static catalogue study into an operational planning tool for every proposed lunar radio experiment.","Without binding leakage limits, commercial relay constellations could lock in a permanently elevated noise floor before the first generation of farside science payloads finishes its baseline surveys."],"forward_implications":["LFT3-class observations remain feasible in the near term, but the fraction of clean sky and clean spectrum shrinks quickly after about 2028 as more satellites become simultaneously visible.","Preserving the farside as radio-quiet requires satellite-level UEMR shielding, stricter out-of-band limits, and pre-launch emission tests, not only observing-strategy tricks such as avoiding low elevations.","More sensitive future farside arrays would need higher than ~30 dB uniform UEMR suppression to meet the same ‘below sensitivity’ criterion.","There is a time-limited window in which the natural radio quietness of the farside can still be measured before persistent satellite RFI sets in, which should drive mission timing and policy.","Intended communications near 2.4–2.6 GHz are a localized bandwidth loss that may be inevitable for data relay; unintended broadband leakage is the harder, spectrum-wide threat."],"fun_headline_variants":["Farside RFI rises with lunar satellites; 30 dB UEMR shield needed","Unintended lunar craft emissions blanket HF–UHF unless hardened","LFT3-like farside needs ≥30 dB shielding as satellites multiply","Intended links cluster at 2.4 GHz; leakage spans 0.1–2700 MHz","Lunar farside stays usable now but grows contaminated by 2030"],"cache_read_input_tokens":128,"weakest_assumption_plain":"The model treats future satellites’ unintended leakage as well represented by a catalogue that is only directly measured up to about 230 MHz and is filled at higher frequencies by harmonics and representative power levels; if real leakage is much weaker, stronger, or differently shaped, both the contamination maps and the 30 dB shielding number can move by large factors.","fun_headline_variants_meta":{"raw":{"variants":["Farside RFI rises with lunar satellites; 30 dB UEMR shield needed","Unintended lunar craft emissions blanket HF–UHF unless hardened","LFT3-like farside needs ≥30 dB shielding as satellites multiply","Intended links cluster at 2.4 GHz; leakage spans 0.1–2700 MHz","Lunar farside stays usable now but grows contaminated by 2030"]},"model":"grok-4.5","effort":"low","cost_usd":0.00527,"raw_usage":{"total_tokens":1562,"prompt_tokens":971,"num_sources_used":0,"completion_tokens":95,"cost_in_usd_ticks":52704000,"prompt_tokens_details":{"text_tokens":971,"audio_tokens":0,"image_tokens":0,"cached_tokens":128},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":496,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":971,"tokens_out":95,"duration_ms":6479,"temperature":1.0,"reasoning_tokens":496,"cache_read_input_tokens":128,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-31T12:20:00.412646+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Wide-band, calibrated measurements of actual lunar or cislunar spacecraft unintended emission spectra above a few hundred megahertz—compared against the paper’s harmonic-filled catalogue and against the predicted received power at a farside site—would confirm or overturn the 31.40 dB uniform-shielding requirement and the claimed contamination density by 2029–2030.","supporting_citations":[],"review_version":1}