# Battle of the Beams: Critique Responses Systematic point-by-point response to external arguments advanced against the [[Knickebein_Propagation_Null|Knickebein globe-vs-flat null hypothesis]]. Each section contains (1) a steel-manned restatement of the argument (2) verbatim quotes with source timestamps so the restatement can be audited against the original, (3) identification of any logical fallacies or factual errors, and (4) a refutation grounded in math, primary-source citations, or the ITU-R P.526-16 analysis already in the main null doc. The first dossier covers the ~2h45m *Drunk Debunk Show ep. 58 | D C Dan Presents The Battle Of The Beams* Link: https://youtu.be/H8xiAPDFK1o --- ## How to read this document Each argument is numbered. Within an argument: - **Quoted.** Direct, literal words from the source transcript with the video-clock timestamp. The argument opens with the speaker's own words so the reader hears the position before any paraphrase. - **Claim.** The argument restated in the strongest, most charitable, most impersonal form available, built from the quotes above rather than a hostile paraphrase. - **Fallacies.** Named logical errors or category mistakes present in the claim. - **Refutation.** The counter-argument with math, citation, or graph, referencing the main null doc for derivations already worked out there. > [!important] Argument 6 is handled in its own full companion document > The single most significant technical argument in the presentation is **Argument 6 — "the beam refracts and diffracts around the curve, ground waves hug the Earth, the optical horizon is not the RF horizon"**. This is the one that invokes Sommerfeld ground-wave propagation, 4/3 Earth refraction, and smooth-sphere diffraction as rescue mechanisms for a beyond-line-of-sight 31.5 MHz path on a globe. Because it is the load-bearing falsification target and because the rebuttal requires walking through the ITU's own ground-wave propagation standards (P.368-10, the 1991 CCIR Handbook R-HDB-13 *Curves for Radio Wave Propagation over the Surface of the Earth*, P.526-16 §3 smooth-sphere diffraction, and the ITU reference LFMF-SmoothEarth C++ source), the full technical response lives in its own companion document rather than in the inline Refutation section of Argument 6 below. > > See [[GRWAVE_P368_BotB]] for the complete breakdown: the per-model sweeps for all thirteen Knickebein and Telefunken paths, the β ≈ 0.81 Eq. 16 correction that had to be applied to P.526 for sea paths at 31.5 MHz, the Sommerfeld-Norton plane-Earth cross-check from Handbook Part 1 §3.2.1, and the side-by-side comparison of all four propagation models (Friis flat-Earth, Sommerfeld-Norton flat-Earth, ITU-R P.526-16 Fock, ITU-R P.368 GRWAVE) against the 31.5 MHz receiver noise floor at every confirmed target distance. The short version of the conclusion is reproduced in the Refutation section of Argument 6 below; the full derivation, ITU citations, primary-source screenshots, and model cross-validation all live in the companion doc. --- # Dossier 1: DC Dan Dano (13 Feb 2025) Source: *Drunk Debunk Show* livestream, 13 Feb 2025, guest DC Dan Dano. YouTube: <https://youtu.be/H8xiAPDFK1o>. Local working copy of the auto-generated YouTube transcript at `/tmp/dan_transcript.txt`. Quotes below retain the auto-caption typos and include the timestamp displayed in the transcript file. The presentation's position is that the flat-Earth interpretation of Knickebein does not survive when the British RAF source data are re-evaluated for engineering usability and when the globe-model propagation is computed with the full set of VHF mechanisms used in real-world radio planning. **Nine** distinct arguments are offered in support of that position. Ordered from least technical to most technical (each expounded in full below from verbatim transcript quotes): - **Argument 1:** the British Spalding reading is an observation, not a measurement. - **Argument 2:** the Anson cockpit and civilian Hallicrafters S-27 made the British reading a miracle at best. - **Argument 9:** only about six Rolls-Royce bombings out of ~33,000 recorded incidents implies the beam was 60 to 70% reliable, not 100%. - **Argument 8:** the real Kleve tower is 84 to 91 m, not the 200 m flat-Earth analyses use. - **Argument 3:** Spalding sits outside the Luftwaffe's own usable-region chart. - **Argument 4:** the Luftwaffe 0.3° beam-width spec projects to about 3 km at 440 km, not 400 to 500 yards. - **Argument 7:** a vertically polarised array is narrow horizontally and tall vertically, so altitude is not a problem. - **Argument 5:** an amateur-radio VHF link-budget tool gives a 45 dB fade margin at Kleve → Derby, enough for 1,000 km on a globe. - **Argument 6:** atmospheric refraction, grazing-ray ground-wave propagation, and diffraction around the Earth's curvature extend the signal past the optical line of sight. (This is the load-bearing technical argument; the full rebuttal lives in [[GRWAVE_P368_BotB]].) --- ## Argument 1: "Bufton's reading is an observation, not a measurement" ### Quoted > [!quote] DC Dan Dano — *Drunk Debunk Show* 13 Feb 2025 > "Should we really call these measurements I mean to be honest let's think about this how was the signal measured it was measured by some guy with ear you know with headphones on and a pilot who's guesstimating his ground speed on a cloudy night what variables did they count for what did they have a device that actually measured signal strength none of that is known" > — **35:04** > > "Considering all these complications would any reasonable person consider these measurements or observations no right these would be observations not measurements there's your two choices right" > — **42:01** ### Claim The British June 1940 detection of the Knickebein equisignal at Spalding does not qualify as a scientific measurement and should not be used as the factual anchor for any propagation analysis of the system. Specifically: (1) the reading was a single auditory tone heard once by a radio operator on a single flight, not a repeatable measurement programme; (2) the aircraft was an Avro Anson reconnaissance patrol plane with multiple glazed panels, exposed twin engines, no EM shielding, and ad-hoc modifications, rather than a purpose-built VHF platform; (3) the receiver was a civilian Hallicrafters S-27 purchased from a consumer retailer (Webb's Radio, Soho), converted from AC to DC mains, and spliced into antennas not designed for VHF; (4) the operator had no specific training on the target signal and was told only to look for Lorenz-type signals in a given frequency range; (5) no device aboard the aircraft measured signal strength in absolute units, so the "400 to 500 yard" width is derived from pilot ground-speed guesstimation and stopwatch timing. Under any reasonable engineering standard a reading produced under those conditions is an observation, not a measurement. ### The category error at the core of Argument 1 The whole of Argument 1 rests on a single sentence from 42:01 of the *Drunk Debunk Show* livestream: *"these would be observations not measurements there's your two choices right"*. That one sentence is doing all of the load-bearing work, because if observation and measurement are mutually exclusive categories and the Bufton reading falls on the observation side, then the factual anchor for the entire flat-Earth interpretation of Knickebein is gone before the globe-vs-flat calculation even starts. The sentence is not true. Observation and measurement are not alternatives. They are stages of the same scientific process. Observation is the broader act of using the senses (or an instrument that substitutes for the senses) to notice and describe what is happening. Measurement is the subset of observation that assigns a number to a property through a prior-set rule. A measurement is an observation with a dial on it. The word "measurement" does not exclude the word "observation"; it names one specific *kind* of observation, the kind that returns a number. This is not a subtle epistemology point. It is the standard definitional framing used in statistical practice, in international geospatial standards, and in elementary-school science curricula. Three sources, one for each audience: **1. Undergraduate statistical practice.** The SJSU StatPrimer operationally defines the two terms in exactly this order: > "Observation, in which the scientist observes what is happening, collects information, and studies facts relevant to the problem." > > "Measurement is the assigning of numbers or codes according to prior-set rules. It is how we get the numbers upon which we perform statistical operations." > > — [SJSU StatPrimer, "Measurement"](https://www2.sjsu.edu/faculty/gerstman/StatPrimer/measure.htm) Observation gathers the facts. Measurement assigns numbers to those facts. The second is a subset of the first, and the two are *conjoined* in every real scientific protocol. **2. International standards.** The Open Geospatial Consortium and ISO 19156 "Observations and Measurements" (O&M) is a formal international standard that defines a single unified conceptual schema covering both words at once, explicitly designed to refute the idea that an observation and a measurement are two different things that need two different schemas: > "Observations and Measurements (O&M) is an international standard which defines a conceptual schema encoding for observations, and for features involved in sampling when making observations." > > — [Wikipedia: Observations and Measurements](https://en.wikipedia.org/wiki/Observations_and_Measurements) One schema. One model. One act. Not two choices. **3. NGSS-aligned grade 4 to 8 science curriculum.** The Quantum for All lesson plan *Measurement and Observation: A "taste" of Quantum* (Matsler 2019, distributed under CC BY-NC-SA) is pitched at nine-year-olds and is the classroom implementation of **NGSS DCI 5-PS1-3**, whose canonical wording is *"Make observations **and** measurements to identify materials based on their properties."* The standard conjoins the two with *and*, not with *or*. The Matsler lesson plan operationalises this by treating "measurement" as one of *"different ways to measure"* and observation as the broader sense-based act, then using the two terms **interchangeably** in its coin-toss example: > "The process of observing, or trying to measure what the coin was, changed the coin." > > — [[2019_Matsler_QuantumForAll_Measurement_Observation|Matsler 2019]], p. 2 (screenshots: [[2019_Matsler_QuantumForAll_Measurement_Observation-p1.png|page 1]], [[2019_Matsler_QuantumForAll_Measurement_Observation-p2.png|page 2]]) When a grade-4 NGSS lesson uses *observing* and *trying to measure* as synonyms inside the same sentence, and when the US K-12 science framework explicitly conjoins the two with *and* in the title of the standard it is aligned to, the claim that adult science requires us to choose between them collapses. Argument 1 is built on a false binary. The binary is false because the premise that observation and measurement are mutually exclusive is, by the nine-year-old classroom definition, the opposite of what both words mean. Note also that Argument 1 nowhere offers its own definition of either term. The presentation displays the word "measurements" in sarcastic quotes on its title slide and uses the phrase "would any reasonable person consider these measurements or observations" as a rhetorical question, but it never says what the speaker understands those words to mean. The definitional work is skipped, and the audience is invited to agree with the conclusion *"there's your two choices right"* without ever being shown a definition on which the two choices could rest. ### What the British actually did, in the correct two-stage sequence If the two words are used correctly, the British 21 June 1940 detection of the Knickebein equisignal was *both* an observation and a measurement, in the order those two stages are supposed to happen in a disciplined scientific investigation. **Stage 1 — observation (all the circumstantial evidence collected before the Anson ever flew).** The observation stage of the investigation started months before Bufton took off. It included: the Farnborough analysis of the captured tail antenna from a Heinkel He 111 shot down in Scotland (which showed a Lorenz-class VHF receiver tuned to a band the RAF did not monitor); POW interrogations of captured Luftwaffe navigators mentioning a *Knickebein* beam system; an Enigma intercept on 5 June 1940 that gave the approximate geographical alignment of the Kleve and Bredstedt/Stollberg beams; and R.V. Jones's own correlation of bomber wrecks along apparent beam axes. This is the observation stage in the SJSU sense: the scientists *observed what was happening, collected information, and studied facts relevant to the problem*. It produced a working hypothesis: a German VHF navigation beam existed at a specific frequency range and was being aimed at the English Midlands. No numbers yet. Just observation. **Stage 2 — measurement (the Bufton Anson flight on 21 June 1940).** Once the observation stage had produced a specific enough hypothesis to test, the RAF instrumented it and flew through it. Flight Lieutenant H.E. Bufton and Corporal Mackey took an Avro Anson from RAF Wyton, flew the expected Kleve beam axis at a known altitude (5,800 m), and used a Hallicrafters S-27 receiver tuned to the 30 to 33.3 MHz band documented by the Farnborough tail-antenna analysis. Crossing the beam at approximately right angles, the operator called the transitions from *dashes heard only* (inside the dash lobe) to *steady tone* (inside the equisignal) to *dots heard only* (inside the dot lobe), and the pilot timed the transitions on the aircraft's clock while holding a known ground speed. This produced a single number with units: **the equisignal corridor at 5,800 m altitude above Spalding was 400 to 500 yards wide.** That is a measurement by the SJSU definition, by the ISO 19156 definition, and by Matsler's classroom definition. A number was assigned to a physical property (the width of a corridor) through a prior-set rule (width = ground-speed × transit-time). Source: [[1978_Jones_Most_Secret_War|Jones 1978]] pp. 181 to 182, quoted verbatim in the Jones source note under the heading "Bufton's Anson Flight Report". The British did what every disciplined investigation is supposed to do. They observed first, formed a hypothesis, instrumented the hypothesis, and measured. Argument 1 is asking us to discard the measurement stage by re-labelling it as "only an observation", which is a move that only works if the two words are treated as alternatives. They are not alternatives. They are consecutive stages of the same workflow. ### The "guesstimated ground speed" objection, and why it collapses The argument that specifically targets the pilot's ground-speed figure as unreliable: *"a pilot who's guesstimating his ground speed on a cloudy night"*. The implication is that if the ground-speed figure is unreliable, the measured width is unreliable, because width = speed × transit-time. The objection collapses on arithmetic. The Knickebein antenna geometry is an independent primary-source parameter, not a British figure, and it predicts a specific equisignal width at Spalding. The Telefunken Large Knickebein array is a rectangular aperture $L_H = 99\text{ m}$ wide by $H_V = 29\text{ m}$ tall at the 31.5 MHz transmit frequency ([[2004_Bauer_German_Radio_Navigation_1907_1945|Bauer 2004, p. 12]] from Telefunken archives). The beam is the superposition of two squinted sub-beams offset by 5° from the array centreline, one keyed with dots and the other with dashes, and the equisignal is the narrow corridor where the two sub-beam amplitudes are equal. The angular width of that equisignal crossover on a rectangular-aperture Lorenz array is set by the squint angle and the lobe slope near crossover, not by the main-lobe half-width. For the 99 m × 29 m Telefunken geometry at 31.5 MHz with 5° squint, the equisignal 3 dB corridor is approximately $\theta_{eq} = 0.066°$. Projected rectilinearly from Kleve over the 440 km path to Spalding, this angle gives: $W = d \cdot \tan(\theta_{eq}) = 439{,}541 \text{ m} \times \tan(0.066°) \approx 506 \text{ m} \approx 554 \text{ yd}$ Bufton's flight report: **400 to 500 yards**, at 440 km, 5,800 m altitude, Spalding. The arithmetic from the German antenna geometry predicts **554 yards** at the same range. The discrepancy is ~10%, which is about what you would expect from aircraft ground-speed noise on a cloudy night plus stopwatch timing resolution. If Bufton had been off by the factor-of-six amount Argument 4 wants us to believe ([2:07:53 to 2:08:09 on the same stream](#argument-4-the-luftwaffe-beam-width-spec-is-03-not-0066-so-the-british-measurement-contradicts-the-source)), the measured width should have been about 2,300 metres. It was not. The stopwatch reading landed within 10% of the number that the Telefunken antenna physically produces at that range on any Earth shape. The ground-speed estimate therefore could not have been more than ~10% off, because if it had been more off than that, the British measured width and the German antenna geometry would not have agreed. ### The worked Kleve → Spalding numbers on a globe (post β-fix library) For completeness, this is what the BotB propagation library gives for the Kleve → Spalding geometry that Bufton measured. All four models are in agreement at this distance, as expected for a path inside the Kleve director-beam's operational envelope. These numbers are generated by the same code that produces the master bar chart and the Kleve sweep in [[GRWAVE_P368_BotB]], using the post-β-fix `botb_itu_analysis.py` with the correct ITU-R P.526-16 Eq. 16 β parameter (β = 1 for vertical polarisation over land above the 20 MHz ITU cut; Kleve at 31.5 MHz over avg land is on the β = 1 side of that cut, so the β fix does not change the Kleve numbers): | Quantity | Value | Source | |---|---|---| | Path length | 439.5 km | haversine, Kleve (Kn-4) 51.7886 N 6.1031 E → Spalding 52.7870 N 0.1530 W | | TX height | 111 m | [[1979_Trenkle_Deutsche_Funk_Navigation\|Trenkle 1979 p. 67]] (83 m terrain + 28 m antenna frame) | | RX altitude | 6,000 m | He 111 operational cruise altitude (Jones 1978, Price 2017) | | Frequency | 31.5 MHz | 5 May 1940 Luftwaffe memo + BArch RL 19-6/40 | | TX power | 3,000 W | Telefunken Large Knickebein spec, BArch 230Q8 | | TX gain | 26.0 dBi | $G = 4\pi A/\lambda^2$, $A = 99 \times 29 = 2{,}871\text{ m}^2$ | | RX gain | 3.0 dBi | EBl 3F hanging-wire antenna approximation | | FSPL | 115.3 dB | $20\log_{10}(4\pi d/\lambda)$, Friis | | ITU-R P.526-16 Fock diffraction loss (β = 1) | 38.1 dB | Eq. 13 to 18, `botb_itu_analysis.py` | | Total path loss (globe) | 153.4 dB | FSPL + Fock | | Total path loss (flat) | 115.3 dB | FSPL only | | Globe peak SNR | **+61.6 dB** | post-β-fix library, matches master bar chart | | Globe equisignal SNR (peak − 19 dB) | **+42.6 dB** | 5° squint crossover | | Detection floor | +10 dB | ITU-R P.372 bare-detection rule of thumb | The globe model clears the detection floor by **32.6 dB** at the equisignal. The Kleve director beam reaches Spalding on both world models at 31.5 MHz with the 111 m / 6,000 m geometry. The Bufton reading is consistent with the globe model at this specific path, and it is consistent with the flat model at the same path. Kleve → Spalding is not where the flat-vs-globe test discriminates (for that, see Stollberg → Beeston and the Telefunken range tests in [[GRWAVE_P368_BotB]]). But what Kleve → Spalding *does* show is that Bufton measured a width consistent with the German antenna geometry, on a globe, at a range that the globe model says is well inside the operational envelope. He measured what was there. ### The "reasonable person" framing is self-defeating > "Considering all these complications would any reasonable person consider these measurements or observations no right these would be observations not measurements there's your two choices right" > > — **42:01** The reasonable person who took the reading was Flight Lieutenant H.E. Bufton of the Blind Approach Development Unit, a trained radio operator and navigator whose written flight report gave the British their first direct numerical measurement of the Kleve beam over England: a 400 to 500 yard equisignal corridor passing 1 mile south of Spalding, on a 104°-284° true bearing, with 31.5 MHz carrier and 1,150 Hz modulation. All of that is quoted verbatim in [[1978_Jones_Most_Secret_War|Jones 1978]] pp. 181 to 182. The entire British counter-Knickebein programme was built on the proposition that Bufton was a reasonable person whose reading was reliable. "Would any reasonable person consider this a measurement" is a rhetorical question whose answer is *"yes, the British War Cabinet did in June 1940, and the Merlin engine factory attack was thrwarted becasue of it"* ### The real question is not about British audibility, it is about whether the signal can exist on a globe The most important move to make about Argument 1 is to note what it is trying to change the subject away from. The Knickebein null hypothesis is not "the British heard the beam, therefore flat Earth". It is *"could a 31.5 MHz signal from Kleve at 111 m TX have existed at Spalding at 6,000 m RX with a 400 to 500 yard equisignal corridor, on a sphere of radius 6,371 km?"*. The British audibility chain is a secondary confirmation. The load-bearing primary source is the Telefunken July 1939 over-sea range campaign, which is a **controlled German-internal measurement programme** run by the system's own manufacturer at documented ranges from 400 km to 1,000 km and altitudes of 4,000 m ([[1939_BArch_RL19-6-40_230Q8_App2_Telefunken_Range_Tests|BArch RL 19-6/40 ref. 230Q8 App. 2]]). Dan himself characterises that Telefunken campaign on-air as *"actual measurements done by the Luftwaffe, repeatable, done multiple times, done with multiple pieces of equipment, done with multiple systems"* (2:08:52 to 2:09:00). He treats the German measurements as authoritative. Those German measurements are what the globe model has to explain, and the globe model cannot explain them past 500 km on either Fock or GRWAVE. Whatever is said about Bufton's cockpit acoustic environment or his Hallicrafters receiver changes nothing about the Telefunken range table, because the Telefunken range table is not a British reading. It is the reading Dan himself accepts. In short: Argument 1 is the *"let us discard the British primary source"* argument. Even if one accepts the premise entirely and throws the Bufton flight report in the bin, the null hypothesis is still constrained by the Telefunken primary source Dan accepts in the same presentation, and the globe model still cannot deliver the Telefunken ranges. The argument does not reach the conclusion it aims for, because the conclusion depends on the load-bearing primary source (the Telefunken range table) which the argument leaves untouched. --- ## Argument 2: "The Anson cockpit was too loud, and the S-27 was a civilian receiver. The fact they heard anything is a miracle." ### Quoted > [!quote] DC Dan Dano — *Drunk Debunk Show* 13 Feb 2025 > "Ad hoc design of the receiver in the airplane let me just tell you it's a \[fucking\] miracle that they were able to hear that to begin with... this guy was had his eyes closed he was concentrating he was probably climbing underneath the desk anywhere he can to get away from the noise" > — **2:07:02 to 2:07:18** > > "So considering all these complications would any reasonable person consider these measurements or observations no... it's talking about an audible tone a guy with headphones is sitting in this right here with all these windows with all these the exhaust from two twin engines the wind noise and everything else versus two armored bulkheads where the German radio operator sat" > — **2:39:28 to 2:39:35** ### Claim Separate and additional to the measurement-versus-observation argument, the environment in which the British detection occurred was acoustically and electromagnetically hostile to the point of making any reading through it unreliable as evidence. The Avro Anson used for the 21 June 1940 detection flight was a non-armoured, multi-glazed reconnaissance airframe with exposed twin engines, high ambient noise from wind and exhaust, no compartmentalisation, and no EM shielding on the improvised receiver installation. The receiver itself, a Hallicrafters S-27, was a consumer product designed for controlled quiet environments like a home radio shack, not for airborne operation at 12,000 feet in a noisy, vibrating, cold-soaked platform. The contrast with the German operational setup is severe: the He 111 bomber placed the radio operator between two armoured bulkheads in a shielded compartment, using a purpose-built FuBl 2 superheterodyne with advanced circuit design, multi-stage amplification, integrated antennas, extensive EM shielding, and a crew trained specifically on the Lorenz-style dot/dash beam-riding procedure. Given those differences, the strongest statement that can honestly be made about the British detection is that "the fact they heard it at all is nothing short of a miracle"; it is not a statement about what the signal actually looked like at the antenna. ### The Argument 2 question is almost entirely moot The whole of Argument 2 is about whether the British cockpit environment on 21 June 1940 was acoustically clean enough for Bufton to hear a tone over the Anson's twin Cheetah exhausts. That question is secondary at best and irrelevant at worst, because the null hypothesis does not depend on whether Bufton's ears were in a quiet room. It depends on whether a 31.5 MHz signal from Kleve at 111 m TX can reach Spalding at 6,000 m RX on a globe of radius 6,371 km with the 400 to 500 yard equisignal corridor the system is supposed to produce. That is a Maxwell-equations question about RF propagation across a curved dielectric surface, not a human-factors question about the decibel level of an open-frame radial engine cockpit. Cabin noise is not in the link budget for the same reason that the operator's body temperature is not in the link budget: it is on the wrong side of the receiver, downstream of the audio amplifier and the earphone cups. ### Bufton's measured width matches the Telefunken antenna geometry within ~10%, so "he couldn't hear it" contradicts arithmetic Even if one is willing to entertain Argument 2 on its own terms, the objection collapses the moment it meets the geometry I already walked through in the reply to Argument 1 above. The Telefunken Large Knickebein array is a 99 m × 29 m rectangular aperture at 31.5 MHz with a 5° squint, and the equisignal crossover corridor for that geometry is $\theta_{eq} \approx 0.066°$. Projected rectilinearly from Kleve to Spalding at 439.5 km, that angle gives a 506 m (~554 yard) corridor. Bufton's flight report in [[1978_Jones_Most_Secret_War|Jones 1978]] pp. 181 to 182 gives 400 to 500 yards. That is a ~10% match. If Bufton had been straining through cockpit noise and hearing a phantom that was not really there, his number would not have landed on the German antenna's actual physical equisignal width. The fact that it does is the strongest possible falsification of Argument 2 on its own terms: the arithmetic says the signal was heard cleanly enough to report a number within 10% of the number the Germans' own antenna mathematically produces at that range. ### Even if the British detection is discarded entirely, Telefunken still measured it and the Luftwaffe still flew it Suppose Argument 2 is accepted in full. Throw Bufton's flight report in the bin. Stipulate that the S-27 was a civilian receiver and the Anson cockpit was too loud for any British reading to be reliable. The null hypothesis is still not falsified, because the load-bearing primary source was never the British audibility chain in the first place. It is the German-internal Telefunken July 1939 over-sea range campaign that the company's own engineers ran in their own aircraft at 4,000 m altitude over open water, documented in [[1939_BArch_RL19-6-40_230Q8_App2_Telefunken_Range_Tests|BArch RL 19-6/40 ref. 230Q8 App. 2]], the source note for which includes the embedded photograph of the 10 September 1939 appendix with the full measured range table. Those measurements were taken by Telefunken engineers in a German aircraft with the full FuBl 1 receiver chain and the proper antennas, not by a British intelligence officer in a Anson with a Hallicrafters. The Anson was not on those flights. The S-27 was not on those flights. The Telefunken campaign reports operational ranges from 400 km to 1,000 km for three transmitter configurations, and the ranges on the 3,000 W FFuGt (the large Knickebein transmitter) extend to 1,000 km on every antenna combination tested. And importantly, the transcript at 2:08:52 to 2:09:00 has the speaker characterising exactly that Telefunken campaign on-air as *"actual measurements done by the Luftwaffe, repeatable, done multiple times, done with multiple pieces of equipment, done with multiple systems"*. Argument 2 does not touch it. Argument 2 attacks the British cockpit. The Telefunken range table survives untouched. Beyond Telefunken, there is also the operational fact that the Luftwaffe actually flew Knickebein missions. Between June 1940 and May 1941, He 111 crews from KG 26, KG 27, KG 55 and others flew hundreds of Knickebein-guided missions against Birmingham, Coventry, Derby, Bristol and Liverpool. Those missions happened. The bombs fell. The Coventry raid of 14 November 1940 was executed on a Knickebein fix from Kleve plus Bredstedt ([[1978_Jones_Most_Secret_War|Jones 1978]] Knickebein timeline). None of that was the British listening to the beam in an Anson. That was German aircrews listening to their own navigation system in their own bombers, flying real paths to real cities, with real ordnance. If the signal did not physically exist at the receiver antenna of the He 111 at those ranges and altitudes, those missions would have flown blind and the navigation would have failed. The navigation did not fail. ### Argument 2 undercuts itself: praising the He 111 cockpit is an admission that the signal was audible Argument 2's own internal contrast is the quickest refutation. The transcript at 2:39:28 to 2:39:35 argues that the German radio operator sat *"between two armored bulkheads"* in a quieter, shielded, purpose-built compartment with the operational FuBl 1/2 receiver, while the British operator was in an open-frame Anson with wind, exhaust and glazing noise. The speaker offers this contrast to explain why the Germans might have heard the beam when the British could not. This is an admission that the beam was audible in the He 111. The Germans sat in the quiet, optimal, purpose-built environment. The Germans flew the missions. The Germans used the receiver the manual actually documents. If the speaker's own argument is that the German operator's setup was better than the British one, then by his own argument the German operator heard the signal. And if the German operator heard the signal, then the signal was present at the He 111 antenna at 31.5 MHz at the operational altitudes and ranges the Luftwaffe flew, which is the only thing the null hypothesis needs the signal to have done. The argument that was intended to undermine Bufton ends up vouching for the Germans. Argument 2 cannot have it both ways. Either the He 111 was a better platform to hear Knickebein on, in which case the Germans heard it operationally and the signal exists on a globe, or the signal was never audible anywhere, in which case the praise for the He 111 cockpit is irrelevant because it is praising the ability to hear an inaudible signal. One of those two positions has to be given up, and as soon as either one is, Argument 2 loses. ### What Argument 2 is avoiding: the paired cross-beam geometry The central fact Argument 2 never addresses, and never mentions, is that Knickebein is a **cross-beam** system. It is not one beam from one transmitter. It is two independent VHF beams from two different transmitter sites that intersect over the target, and the aircraft navigates by crossing the dash lobe of the director beam from (say) Kleve while listening for the intersection zone of the marker beam from (say) Stollberg. The system cannot function unless **both** beams, from **both** transmitter sites, arrive at the same point in the sky with clean enough equisignal structure to define the intersection. That means two simultaneous VHF paths across the North Sea of comparable length, from two different Continental locations, both successfully reaching the same aircraft with dot/dash structure intact. A globe model has to explain both beams. It cannot explain even one past the Stollberg → Beeston range on the Fock or GRWAVE calculation in [[GRWAVE_P368_BotB]]. It cannot explain two. The load-bearing hard problem in the null hypothesis is *"where in the sky over the English Midlands do these two beams intersect, on a sphere where both of them are in geometric shadow?"*, and that problem is what any honest defence of the globe model would have to solve. Argument 2 does not try to solve it. Argument 2 picks at the British receiver chain instead. That is a visible rhetorical substitution. It is easier to cast doubt on an ageing Hallicrafters in a noisy Anson than to defend a two-transmitter cross-beam geometry against Fock diffraction on a 6,371 km sphere, so Argument 2 chooses the easier target and never once steps onto the harder one. This entire line of argumentation is a red-herring and a weak-man. There's stronger, more important to address. Nothing mutually exclusive comes from this. --- ## Argument 3: "Spalding is outside the Luftwaffe's own 'usable region'" ### Quoted > [!quote] DC Dan Dano — *Drunk Debunk Show* 13 Feb 2025 > "If we look at Spalding Spalding was around 440 meters right... it was around 400 about 430 kilometers it was about at 5600 meters... it's a little outside that lower audibility limit but all that means is that it is in a sea of of uh static" > — **2:06:38 to 2:06:54** > > "This is outside the usable region the fact that you would consider this an accurate measurement when it's not even in this region right come on" > — **2:08:34** ### Claim A Luftwaffe memo dated 5 May 1940, titled *Navigatorische Verwendung der Fernsender* (Navigation Use of Long-Distance Transmitters), contains a distance-versus-altitude graph of the Knickebein "usable region" based on actual receiver-value averages from field testing. When the Spalding detection point (approximately 440 km and 5,600 to 5,800 m altitude) is overlaid on this graph, it sits just outside the Luftwaffe's own stated lower audibility limit for the usable zone. This does not imply that the signal at Spalding was identically zero (because an analog radio signal does not hard-cut at a threshold, it degrades into noise progressively), and it does not contradict the idea that a trained military operator would still be able to recover a weak tone under favourable conditions. What it does imply is that Spalding is in "a sea of static" rather than in the clear operational envelope the Luftwaffe was designing for, and that any British reading taken at that range is therefore at the edge of the Germans' own expected reliability. The Luftwaffe's reason for drawing the region conservatively is explicitly acknowledged on-stream: military operational envelopes are built for 100% reliability under wartime stress, so an average-case audibility limit is drawn inside the maximum technical reach of the signal. Even granting that conservatism, Spalding falls outside the line, and the British detection occurred at a point the Luftwaffe's own document did not guarantee as usable. ### The Nutzbereich chart is a primary source that contradicts the operational record, which means one of the two has to be wrong ![[DC_Dan.png]] The chart shown on the livestream is the *Geheim!* Nutzbereich diagram from BArch RL 19-6/40 ref. 230Q7, May 1940. It is real. It is a Luftwaffen-Führungsstab document. It is classified secret and the header says *"Darf nicht mit ins Flugzeug genommen werden!"* (not to be taken into the aircraft). Those facts are accepted. The question is what the chart actually *represents*, and the answer is that the chart cannot represent the true operational capability of the Large Knickebein, because if it did the Germans would never have built the Stollberg station in the first place. ### The reductio from Stollberg's existence Knickebein is a **cross-beam** system. It produces a position fix by crossing a director beam (from one transmitter) with a marker beam (from a second transmitter at a different site). The Luftwaffe operated two primary Large Knickebein stations for the 1940-41 campaign: **Kleve** (Kn-4, 51.7886 N 6.1031 E) on the lower Rhine, and **Stollberg/Bredstedt** (Kn-5, 54.6481 N 8.9381 E) in Schleswig-Holstein near the Danish border. Both were built by the Reichsluftfahrtministerium during 1939-1940 specifically as complementary Knickebein sites to provide the two-beam fix. Now apply the Nutzbereich chart to Stollberg's geometry. Stollberg sits ~200 km north of Kleve. For the Stollberg beam to **cross** the Kleve beam over any target in the English Midlands that Kleve can reach at 440 km, Stollberg's own beam has to travel a geometrically longer path, because the Stollberg-to-Midlands distance is larger than the Kleve-to-Midlands distance at almost every Midlands target: | Target | Kleve → target | Stollberg → target | | ------------------------------------------ | -------------- | ------------------ | | Spalding | 440 km | 632 km | | Beeston (Derby / Rolls-Royce Merlin works) | 530 km | 710 km | | Sheffield | 541 km | 694 km | | Birmingham | 551 km | 754 km | | Manchester | 593 km | 740 km | | Bristol | 600 km | 847 km | | Southampton | 530 km | 810 km | | Liverpool | 640 km | 790 km | | Plymouth | 732 km | 1,003 km | Every Stollberg path to a meaningful cross-fix over England is at least 600 km, and most are 700 km or more. If the Nutzbereich chart is taken at face value, with its maximum drawn distance of 500 km, then *every single one of those Stollberg paths is off the right edge of the chart and outside the useable region*. That is not "just outside the line". That is *off the chart entirely*, by 100 to 300 km, before the beam has even crossed the Dutch coast. Every Stollberg path to a British target that the Kleve beam can also reach is, by the chart's own drawing, unusable. Here is the reductio. If the chart accurately represented Stollberg's operational range, the Luftwaffe signals command spent 1939 building a 3,000 W transmitter, a 99 m × 29 m antenna array, and a complete operational beacon site at Stollberg to service zero cross-fix targets in England. That is not a plausible reading of how the Wehrmacht procurement process worked. Signals command knew the geometry of the North Sea, they knew where Kleve was, they knew where Stollberg was, and they chose Stollberg for its specific location because its beam could cross Kleve's beam at operational ranges. They would not have built the second transmitter if the chart's envelope were real. So either the chart is wrong about the operational envelope, or the existence of Stollberg is wrong. The existence of Stollberg is not in dispute: it is documented in the same Bundesarchiv file (RL 19-6/40), it is photographed in Bauer 2004 and Dörenberg's reference pages, its concrete turntable survives in Schleswig-Holstein today, and British intelligence identified its beam on 21 June 1940 (*"the bearings were consistent with transmitters at Cleves and Bredstedt"*, [[1978_Jones_Most_Secret_War|Jones 1978]] p. 182). The chart is the thing that has to give. (Bredstedt = Stollberg) ### Why the chart understates the true envelope There are several plausible reasons why a primary-source document can understate operational capability and they all apply here. The chart is labelled *Geheim!* and *"not to be taken into the aircraft"*. That is the wording of a conservative crew-reference document meant to be memorised on the ground and left behind before flight, precisely because the actual operational ranges the Luftwaffe was using were classified at a higher level than the cockpit document. Nothing about that labelling implies the chart represents the system's true maximum. Everything about it implies the opposite. - **Military secrecy / classification layering.** The chart is explicitly marked "not to be taken into the aircraft", which tells you the Luftwaffe did not want the document to be captured by the British if an aircraft went down. Aircraft are the things most likely to be shot down. The reason for keeping it on the ground is that the true range information should not reach British intelligence in the event of a crash. It follows that the range information on the chart is *closer to the minimum the Luftwaffe wanted the British to know* than it is to the actual operational envelope. A conservative chart kept deliberately off-aircraft is consistent with intentional under-statement of capability. This is standard military practice. Regardless of the document says the system was used at distances are beyond the conservative specs. ### The operational record is decisive The Luftwaffe flew Knickebein-guided missions against British targets from the start of the Night Blitz in September 1940 through to May 1941. Knickebein was the **mass-bomber navigation system**: every He 111 and Ju 88 in the main force units (KG 26, KG 27, KG 55 and others) carried a FuBl 2 and navigated on the Kleve+Stollberg cross-fix during the night bombing campaign. The more precise X-Gerät system used by Kampfgruppe 100 Pfadfinder for pathfinding is a separate beam chain and is not the system Argument 3 is about, so X-Gerät target raids (Coventry on 14 November 1940, for example) are not counted in the following list. Blitz-period targets hit by the main force while navigating on Knickebein include: - **Birmingham** (multiple raids, August through December 1940 into spring 1941) — Midlands industrial centre, munitions and aero-engine production - **Derby / Rolls-Royce Merlin works at Beeston** — the iconic Knickebein target, subject of the Aspirin deflection story in [[1978_Jones_Most_Secret_War|Jones 1978]] - **Liverpool and Merseyside** (August 1940 onwards) — port and docks, sustained heavy raids through May 1941 - **Manchester** (including the Christmas Blitz of 22–24 December 1940) — industrial and rail - **Sheffield** (12–15 December 1940, "Sheffield Blitz") — steel and armament works - **Bristol** (multiple raids, November 1940 onwards) — port and aircraft production at Filton - **Southampton** (November–December 1940 raids) — port and Supermarine Spitfire works at Woolston - **Plymouth** (March–April 1941, "Plymouth Blitz") — naval base and dockyards - **Hull** (multiple raids, March–May 1941) — port Every one of those targets is more than 530 km from Kleve, every one is more than 690 km from Stollberg, and the Plymouth paths exceed 1,000 km on the Stollberg side. The Knickebein cross-fix system reached every one of them, because the bombs fell on every one of them. The British counter-Knickebein jamming campaign (Aspirin, Bromide, Benjamin) was built specifically because the beams did arrive over the Midlands, the North-West, and the South coast. There is no interpretation of the historical record in which the Nutzbereich chart's drawn maximum of 500 km represents the system's true reach. ### Spalding itself, on the chart as drawn Even granting the chart on its own terms, Spalding at 440 km / 5,800 m does not sit where Argument 3 says it sits. Reading the heavy lower-audibility curve at the 440 km column gives an altitude of roughly 5,000 to 5,500 m, meaning Bufton at 5,800 m was **at or slightly above** the drawn curve, not below it and not outside the useable region. Calling Spalding "a little bit outside" requires bisecting a hand-drawn pencil curve with millimetre precision, which is exactly the move Whitehead described in *Science and the Modern World* (1925) as the **fallacy of misplaced concreteness**: treating an abstraction (a smoothed continuous-curve crew-reference diagram) as if it were a hard concrete binary cutoff at arbitrary sub-pixel precision. The chart was drawn with km-scale uncertainty. The "just outside" claim rests on gap smaller than the thickness of the pencil line. ### The globe-model Fock calculation says Spalding is detectable, and puts the chart in context Putting the Nutzbereich chart to one side, the ITU-R P.526-16 Fock globe-model calculation in `botb_itu_analysis.py` gives Kleve → Spalding at 31.5 MHz, 111 m TX, 5,800 m RX, 3 kW, vertical polarisation: FSPL 115.3 dB, Fock diffraction loss 38.1 dB, total path loss 153.4 dB, and equisignal SNR of **+42.6 dB** above the ITU-R P.372 galactic noise floor. That is 33 dB above the +10 dB bare-detection rule of thumb. The signal at Spalding is not "a sea of static". It is comfortably audible on the globe model at the geometry Bufton flew. The two graphs below are per-station distance sweeps produced by `make_p526_vs_p368_graphs.py` using the same post-β-fix `botb_itu_analysis.py` library that generates the master bar chart in [[GRWAVE_P368_BotB]]. Each graph plots four curves as a function of distance from the transmitter, with the dashed horizontal lines marking the 31.5 MHz ITU-R P.372 galactic+thermal noise floor (0 dB reference) and the +10 dB bare-detection rule of thumb. Vertical lines mark every Blitz-period Knickebein target city along the station's beam corridor. **The four curves per graph are:** 1. **Friis + Sommerfeld-Norton flat-Earth, peak** — rectilinear path with full surface-reflection reflection coefficient and Norton attenuation function. The upper envelope, corresponding to a bomber centred on the beam axis. 2. **Friis + Sommerfeld-Norton flat-Earth, equisignal** — same flat-Earth path, minus the ~19 dB fall-off that occurs at the 5° squint crossover between the dot and dash lobes. This is the level the pilot actually hears when sitting on the equisignal centreline. 3. **ITU-R P.526-16 Fock globe, peak** — full smooth-Earth diffraction with Fock height-gain functions, 4/3 Earth radius, β per Eq. 16. The upper envelope on a spherical Earth. 4. **ITU-R P.526-16 Fock globe, equisignal** — same Fock globe path at the 5° squint crossover, this is what the pilot hears at the equisignal on a globe. **Kleve → Midlands sweep** ([[GRWAVE_P368_BotB|per-station graph]]): ![[ITU_Calc_sn_vs_grwave_uv_kleve.png]] Reading the graph at 440 km (Spalding) gives approximately +61.6 dB globe peak SNR and +42.6 dB globe equisignal SNR, matching the canonical library values to the tenth of a dB. Both flat and globe models place Spalding comfortably above the +10 dB detection floor, by 32.6 dB on the globe-model equisignal curve. Following the Fock globe equisignal curve further out across the rest of the Kleve target list (Beeston, Sheffield, Birmingham, Manchester, Liverpool), every Kleve → Midlands city stays above the detection floor out to ~700 km on the globe model. The Kleve director beam reaches the full Blitz-period target list on *both* Earth models. The implication for the Nutzbereich chart is that the chart's drawn maximum of 500 km is not only smaller than the operational record (Night Blitz target list above), it is also smaller than what the globe-propagation physics itself predicts for the Kleve site. A propagation model that predicts audibility out to 700 km is incompatible with a chart that stops at 500 km unless the chart is a conservative crew-reference envelope, which is exactly the reading supported by the document's *Geheim!* and *"not to be taken into the aircraft"* markings. **Stollberg → Midlands sweep** ([[GRWAVE_P368_BotB|per-station graph]]): ![[ITU_Calc_sn_vs_grwave_uv_stollberg.png]] The Stollberg graph is where the flat-Earth and globe-Earth models sharply diverge, and it is the graph that exposes the deepest problem with using the Nutzbereich chart as a physical ceiling. On the flat-Earth curves, Stollberg's equisignal stays comfortably above +10 dB detection right across the Blitz target list to 1,000 km and beyond. On the Fock globe-Earth curves, the equisignal falls through the noise floor around 500 to 550 km and drops into deep shadow loss by 700 km, meaning on the globe model Stollberg cannot reach Beeston, Sheffield, Birmingham, Manchester, Liverpool or Plymouth. Every single operational Stollberg cross-fix path during the Night Blitz falls below the globe-model detection floor. That leaves two mutually exclusive readings for anyone using the Nutzbereich chart to argue for a globe: 1. **The globe model is correct and the Nutzbereich chart is conservative crew guidance.** In this reading, the chart's 500 km maximum is a safety envelope, Spalding at 440 km is inside the envelope (the globe model gives +42.6 dB at that point), and the Stollberg operational record has to be explained by accepting that the globe model cannot reproduce it (which is the core BotB null hypothesis finding discussed in [[GRWAVE_P368_BotB]] and [[Knickebein_Propagation_Null]]). 2. **The Nutzbereich chart is the physical ceiling.** In this reading, the chart's 500 km cutoff is hard, nothing past 500 km works, and therefore Stollberg could never have guided any bomber to any British target. The Night Blitz Knickebein raids on Birmingham, Derby, Liverpool, Manchester, Sheffield, Bristol, Southampton, Plymouth and Hull did not happen under the guidance of the Knickebein system. That is not a reading the historical record supports. Argument 3 is effectively asking the audience to choose reading (2), which requires the historical record of the Night Blitz to be wrong. That is a lot of contradictory evidence to set aside in order to place Spalding a few hundred metres outside the drawn lower audibility curve on a pencil-drawn crew-reference chart. ![[ITU_Calc_sweep_telefunken 1.png]] ![[ITU_Calc_sn_vs_grwave_uv_master_bargraph.png]] ### Treating a conservative crew-reference chart as a physical ceiling Argument 3 picks up a primary-source crew-reference document, treats its pencil curve as a precise operational cutoff, and uses that cutoff to discard the British reading at 440 km while ignoring the fact that the same chart's drawn envelope excludes every operational Stollberg path the Luftwaffe flew for the entire Night Blitz. If the chart were a hard cutoff, Stollberg would be useless and the Knickebein-guided raids on Birmingham, Derby, Liverpool, Manchester, Sheffield, Bristol, Southampton, Plymouth, and Hull would have been impossible. The bombs fell on every one of those cities. The chart therefore cannot be a hard cutoff. At most it is a conservative teaching aid drawn deliberately tight for crew-reference safety margins, kept off the aircraft so the British could not learn the true numbers from a downed Heinkel. Argument 3 reads the tightest, most-classified, most-conservative document in the folder as if it were the physical ceiling of what the beam could do, and then uses that misreading to evict the British detection from a point the chart itself approximately *includes*. Both halves of the move fail. The chart does not exclude Spalding. And even if it did, the chart's own max distance of 500 km would also exclude Stollberg from reaching any British target, which the operational record says it demonstrably did. --- ## Argument 4: "The Luftwaffe beam-width spec is 0.3°, not 0.066°, so the British measurement contradicts the source" ### Quoted > [!quote] DC Dan Dano — *Drunk Debunk Show* 13 Feb 2025 > "Those 400 to 500 yard wide mhm well according to this document this thing should have been roughly 1 and a half kilomet from Center so one and a half kilom here roughly one and a half kilometers here roughly okay that's a three kilometer wide signal not 400 yards" > — **2:07:53 to 2:08:09** ### Claim The 5 May 1940 Luftwaffe memo lists a beam divergence of 0.3° for the Large Knickebein in the header of its navigation-use graph. At 440 km, a 0.3° full angular width projects geometrically to approximately 2,300 m (roughly 1,150 m each side of the centreline), i.e. a 3 km wide signal at Spalding, not the 400 to 500 yards (~ 400 m) reported in the British flight report. The British 400 to 500 yard width figure is therefore either wrong by a factor of six or it is not describing the same quantity the Luftwaffe source is describing. Either way, it cannot be taken as a corroboration of the Luftwaffe specification. This cross-check failure is offered as a second independent reason (on top of Argument 1) to discount the British reading as physically unreliable. ### Three angular quantities, three different parts of the same radiation pattern Argument 4 is a single geometric move: take the 0.3° figure from the 5 May 1940 Luftwaffe navigation-use memo, project it rectilinearly to 440 km, get a ~2,300 m wide strip at Spalding, and declare that Bufton's 400 to 500 yard reading contradicts the source by a factor of six. The move only works if 0.3° and 0.066° are two different values of *the same* physical quantity. They are not. They are values of three different physical quantities, and the 5° squint is a fourth. Each describes a different part of the Knickebein radiation pattern, and it is a specific feature of how a Lorenz-type dot-and-dash beam works that multiple distinct angular widths coexist in the same antenna system at the same time. Before walking through what Bufton measured and why the comparison fails, it helps to lay the three quantities out in order. #### 1. The 5° total squint This is the angular distance from the **dot sub-beam peak** to the **dash sub-beam peak**. It is an antenna design parameter set by how the Telefunken engineers arranged the feed elements on the 99 m × 29 m Large Knickebein array, and it is quoted explicitly in [[2004_Bauer_German_Radio_Navigation_1907_1945|Bauer 2004 p. 12]] and [[2024_Doerenberg_Knickebein_Reference|Dörenberg's reference pages]] citing BArch RL 19-6/40. The two sub-beams are squinted by ±2.5° from the array centreline, giving 5° total from dot peak to dash peak. It is **not** the width of any single lobe. It is the gap between two lobes. Projected to 440 km, $2 \cdot 440{,}000 \cdot \tan(2.5°) \approx 38{,}400$ m: that is the full horizontal extent of the merged pattern from the dot peak direction to the dash peak direction. No pilot flies this width. No British operator measures this width. It is the outer envelope within which the two squinted sub-beams are centred. ![[DC_Dan-3.png]] > [!note] R.V. Jones, *Most Secret War* (1978), Figure 1 — the Lorenz beam principle > This is the fundamental Lorenz navigation pattern. Three regions are visible radiating outward from the transmitter at the bottom left: the **dot zone** (the left sub-beam, whose main lobe is marked by the left dashed-line envelope), the **dash zone** (the right sub-beam, marked by the right dashed-line envelope), and the **equisignal beam** (the narrow centreline corridor between them, labelled at the top). Each zone carries its own rhythmic tone: inside the dot zone the pilot hears dot-keyed Morse, inside the dash zone the pilot hears dash-keyed Morse, and on the equisignal the dots and dashes exactly fill each other's gaps so the pilot hears a continuous steady note. **The 0.3° Luftwaffe main-lobe spec refers to the width of one of the dashed-line envelopes (one sub-beam).** Bufton's 400 to 500 yard reading refers to the width of the narrow "equisignal beam" strip running up the middle. These are different parts of the same diagram. Argument 4 treats them as the same thing. ![[DC_Dan-4.png]] > [!note] Lorenz beam for blind landing — operational view from above > The aerial view makes the distinction vivid. The shaded **dot zone** and **dash zone** are the two wide sub-beam lobes, occupying most of the horizontal extent of the beam pattern. Between them runs the narrow **equi-signal** corridor, drawn as the thin central strip where the dots and dashes interlock into a steady tone. The sinusoidal **track of aircraft** shows a bomber weaving left and right across the equisignal: each time the pilot drifts into the dot zone he hears dots and steers right, each time he drifts into the dash zone he hears dashes and steers left, and each time he is on the equisignal he hears a continuous tone and holds course. **The pilot's useable corridor is the equisignal (0.066°), not the full dot-or-dash zone (0.3°).** Dan's 0.3° projection of "1.5 km each side" at 440 km describes the width of one of the two wide outer zones, not the inner strip the pilot actually navigates in. ![[DC_Dan-5.png]] > [!note] Dörenberg (2020) — Knickebein Großanlage and Kleinanlage antenna structures, from BArch RL 19-6/40 > This is the physical antenna that produces the three angular quantities described above. Both the Large Knickebein (*Großanlage*, Kn-2, Kn-4, Kn-12; ~85 m wide × ~20 m tall) and the Small Knickebein (*Kleinanlage*, Kn-1, Kn-3, Kn-5 through Kn-13; ~45 m × ~9.5 m) are built on the same principle: a horizontal array of 1 λ dipoles spaced at ½ λ, with the array **bent at 165° in plan view** (visible in both top-view insets). The 165° bend is the eponymous "Knickebein" (bent leg) and is the physical feature that creates the squinted dot/dash sub-beams. Each half of the bent array radiates in a slightly different direction; when the two halves are alternately keyed with dots and dashes, the result is the two overlapping sub-beam lobes shown in the Jones and Lorenz diagrams above. The Kleve transmitter (Kn-4) is the Large variant on the left, and it is the antenna whose 5° squint, 0.3° main lobe, and 0.066° equisignal corridor are the three angular quantities at issue in Argument 4. The wavelength band (30 to 33.3 MHz, λ ≈ 9.5 m ± 5%) is printed in the upper right corner. ![[lorenz_equisignal_diagram 1.png]] > [!note] BotB analysis — Large Knickebein radiation pattern with all three angular quantities and their ground projections at 440 km > **Left panel (schematic radiation pattern):** The **dot zone** (magenta, left) and **dash zone** (cyan, right) are the two sub-beam wedges, meeting flush at the yellow **equisignal corridor** (hatched centreline strip). The **purple arc at the top** traces the full 5° peak-to-peak squint from the outer edge of the dot zone to the outer edge of the dash zone, end to end across the merged pattern. The **orange arc** below it traces the 0.3° Luftwaffe main-lobe spec (labelled "Dan's reading — total spec width") at a shorter range inside the pattern, spanning the full width of both sub-beam zones through the equisignal in the middle, because the 0.3° figure is the total beam width from the Luftwaffe document, not a half-width of one sub-beam. The **light-blue annotation** at the top of the yellow strip marks where Bufton measured the 400 to 500 yd equisignal width at Spalding (440 km). All three annotations are drawn at exaggerated angular scale for visibility (the true 5° squint and 0.066° equisignal are far narrower), but every label gives the true degree value. The critical visual: the orange arc (what Dan is comparing against) spans the entire width of the merged pattern at its radius, while Bufton's measurement spans only the thin yellow equisignal slice. These are not the same geometric quantity, and the factor-of-six ratio between them is the real geometric difference between sub-beam main-lobe width and equisignal crossover width, not an error in either number. > **Right panel (projected ground widths at 440 km, log scale):** The four horizontal bars are the ground footprints of the three angular quantities plus Bufton's actual measurement, all projected at the same 440 km distance. The purple bar (5° squint) is 38,495 m. The orange bar (0.3° main lobe) is 2,304 m. The yellow bar (0.066° equisignal) is 507 m. The cyan bar at the bottom is Bufton's 400 to 500 yd (366 to 457 m) measurement. On the log scale, Bufton's cyan bar sits directly adjacent to the yellow equisignal bar and is separated from the orange main-lobe bar by nearly half a decade. Bufton measured what the equisignal predicts, not what the main lobe predicts. #### 2. The 0.3° Luftwaffe main lobe spec This is the angular width of **one sub-beam's main lobe**, at the operational signal-strength level the Luftwaffe documented in the 5 May 1940 memo. It is a property of each individual dot or dash sub-beam taken on its own, and it tells you how wide the signal-present region is for that one lobe. Projected to 440 km, $2 \cdot 440{,}000 \cdot \tan(0.15°) \approx 2{,}300$ m. This is the region inside which a single dot (or single dash) sub-beam is detectable against noise — the region where the **signal exists** at the operational level the memo is drawn for. It is not the region in which a pilot holds course. The 0.3° figure is the number Dan is pointing at on his slide when he computes the 1.5 km per side / 3 km total strip at Spalding. #### 3. The 0.066° equisignal corridor This is the narrow strip at the centreline where the dot sub-beam amplitude exactly equals the dash sub-beam amplitude, to within a 3 dB tolerance. It is **derived** from the 5° squint *plus* the sub-beam shape from the 99 m × 29 m aperture at 31.5 MHz, through the amplitude slope of each sub-beam near the crossover. If any one of those three parameters were off, the equisignal corridor would be off. Projected to 440 km, $2 \cdot 440{,}000 \cdot \tan(0.033°) \approx 507$ m. This is the region inside which a pilot **holds course**. Outside the equisignal but still inside the 0.3° main lobe, the pilot still hears signal, but as rhythmic dots on one side or rhythmic dashes on the other, and steers back toward the centreline until the rhythm merges into a steady 1,150 Hz tone. ### Where each number lives in the pattern The three widths are concentric. The 5° total squint is the outermost angular quantity, describing the horizontal spread of the two peaks. The 0.3° main lobe sits inside that, describing the width of each individual sub-beam around its own peak. The 0.066° equisignal corridor sits inside that, describing the narrow slice where the two amplitudes cross at parity. At 440 km, these project to a 38.4 km peak-to-peak spread, with each sub-beam's main lobe covering 2.3 km of ground around its peak, and a 507 m wide strip at the centreline where the dot and dash cancel to a steady tone. All three are real, all three are documented in the primary sources, and none of them contradict each other. ### What Bufton actually measured, and what it confirms Bufton's flight report, quoted verbatim on pages 181 to 182 of [[1978_Jones_Most_Secret_War|Jones 1978]], gives a corridor width of 400 to 500 yards at the Spalding crossing of the Kleve beam on 21 June 1940. At the 440 km Kleve-to-Spalding range, that width of 366 to 457 m corresponds to an angular width of 0.048° to 0.060°. The predicted equisignal from the Telefunken geometry (5° squint, 99 m × 29 m aperture, 31.5 MHz) is **0.066°**. Bufton's reading lands within ~10 percent of that prediction. What his measurement confirms, strictly: the *combined* geometry — 5° squint, 99 m × 29 m aperture, 31.5 MHz — is correct. If the squint were meaningfully off, or the aperture dimensions were meaningfully off, the equisignal corridor would be a different width and Bufton's reading would not match. The whole beam system as Telefunken documented it is cross-checked by the 21 June 1940 flight report. What his measurement does not directly isolate: the squint angle on its own. Bufton did not point a theodolite at the two sub-beam peaks from the ground. He flew through the equisignal crossover corridor in an Anson and timed the transit. His number is a cross-check on the whole system, not a standalone measurement of any one antenna parameter. ### Where Argument 4 fails Argument 4 takes the 0.3° figure (the width of one sub-beam's main lobe, meaning 2) and projects it rectilinearly to 440 km, arriving at the ~2,300 m "1.5 km each side" strip at Spalding. It then compares that number to Bufton's 400 to 500 yards (meaning 3, the equisignal corridor) and declares a factor-of-six contradiction. The two numbers are not describing the same thing. One is the width of the region in which one of the two sub-beams is detectable against noise at the operational level the Luftwaffe documented. The other is the width of the narrow slice at the centreline where the dot and dash amplitudes are balanced within 3 dB and the pilot hears a steady tone. They are different parts of the same radiation pattern, they are both documented in the primary sources, they are both correct, and the ratio $0.3°/0.066° \approx 4.5$ — which is close to Argument 4's "factor of six" — is simply the real geometric ratio between the main lobe width and the equisignal corridor, not an error in either number. ### The fallacy: equivocation on the phrase "beam width" Argument 4 is a textbook **equivocation** (Fowler / Copi). The English phrase "beam width" carries several distinct technical meanings in antenna engineering, each tied to a different measurable quantity: 1. **First null to first null of one sub-beam** — the outermost angular extent of any signal from a single squinted lobe. 2. **Half-power main lobe width of one sub-beam** — the 3 dB angular width of one lobe around its peak. Several degrees for the Telefunken aperture. 3. **Operational main lobe spec at a chosen signal-strength threshold** — what the 5 May 1940 Luftwaffe memo calls the 0.3° *Leitstrahlschärfe* figure. This is the region in which the sub-beam signal is usable. 4. **Peak-to-peak squint between the two sub-beams** — the 5° figure quoted in Bauer 2004 and the Telefunken archives. 5. **Equisignal crossover corridor** — the narrow slice at the centreline where the dot and dash amplitudes match within some threshold. The 0.066° figure that projects to Bufton's measured 400 to 500 yards at 440 km. Argument 4 uses meaning 3 on Dan's side of the comparison (the Luftwaffe 0.3° spec) and implicitly expects meaning 5 on Bufton's side (the equisignal corridor the pilot flies in). A factor-of-six ratio appears between them, and Argument 4 interprets that ratio as an error in one of the two measurements. But the factor of six is not an error. It is the correct geometric ratio between two distinct features of the same radiation pattern, and both features are documented in the same primary sources. The equivocation is in the English phrase "beam width", not in the data. The usual fix for an equivocation is to subscript the definitions. Once they are subscripted — 0.3° = main lobe operational spec, 0.066° = equisignal crossover corridor — the apparent contradiction disappears, and the two numbers slot into their respective places in the same Telefunken antenna pattern. ### Closing Bufton never claimed to measure the Luftwaffe's 0.3° main lobe width. He measured the equisignal corridor, which is the thing a Lorenz-type blind-approach receiver is designed to resolve: the narrow strip of "steady tone" that distinguishes the pilot's course line from the dot and dash regions on either side. The predicted equisignal width from the primary-source Telefunken antenna geometry is 507 m at 440 km. Bufton reported 366 to 457 m at 440 km. The ratio is 0.72 to 0.90, a match inside measurement tolerance. Argument 4's "factor of six contradiction" collapses as soon as each number is matched to the angular quantity it actually describes. ![[ITU_Calc_knickebein_beam_map 1.png]] --- ## Argument 5: "Link-budget analysis gives a 45 dB fade margin, therefore 1,000 km is achievable on the globe" ### Quoted > [!quote] DC Dan Dano — *Drunk Debunk Show* 13 Feb 2025 > "I used a tool uh to actually create the link budget... antenna height uh of 91 meters... transmit power 3,000 kows translates to about 65 B... we also include the uh receive antenna gain for the hanging wire and we also include the receiver sensitivity which is actually one to two microv volts for the fu1 and we calculate all that and does it for us we see that there is a loss of about 117 DB across free space that's just atmospherics loss obstruction loss is 15 DB forest and the urban loss statistical loss that's let's call it overhead just to be sure so the total path loss is 152 DB that means the fade margin... is almost 45 DB which means that is the equivalent of yes you can go a th000 km with this" > — **2:36:29 to 2:37:57** ### Claim An independent numerical check of the Kleve-to-Derby (Marble Hall) geometry at 528 km using an amateur-radio VHF link-budget tool produces a net fade margin of approximately 45 dB that demonstrates the signal is operationally available well past any confirmed Knickebein target. The inputs used: Kleve transmitter at 91 m effective antenna-centre height, receive aircraft at 6,000 m altitude, transmit power 3,000 W (about 65 dBW), transmitter antenna gain approximately 25 dBi (inferred from the directional array geometry, could be in the 20 to 30 dB range with 25 chosen as a reasonable midpoint), receiver antenna gain from the hanging-wire trailing antenna on the He 111, and receiver sensitivity 1 to 2 μV for the FuBl 1. The tool computes approximately 117 dB of free-space path loss, 15 dB of obstruction loss (a forest / urban / statistical overhead term), for a total path loss of 152 dB. Subtracting from the transmitter EIRP plus receiver sensitivity gives a fade margin of almost 45 dB. A 45 dB fade margin at 528 km directly implies, by the definition of fade margin as "the extra buffer you have before it doesn't become actually audible anymore", that the path can be extended by roughly a factor of two or more without losing the signal. The argument is therefore that the 45 dB margin is the equivalent of being able to "go a thousand km with this" on the globe, and the Stollberg cross-beam geometry is inside that range. ![[DC_Dan-6.png]] ![[knickebein_derby_linkbudget.png]] (Recreated for clarity) ### The link-budget tool omits the dominant loss mechanism on a globe The screenshot from 2:36:29 shows a standard amateur-radio VHF link-budget calculator. Dan enters the Kleve geometry (91 m TX, 6,000 m RX, 3 kW, 31.5 MHz), and the tool returns 117 dB of free-space path loss, 15 dB of "obstruction loss" (a forest/urban/statistical factor), for a total path loss of ~152 dB and a fade margin of ~45 dB. From that 45 dB margin, Dan infers the signal can reach 1,000 km. The problem is what the tool does not include. The 117 dB free-space path loss is the Friis formula, $20\log_{10}(4\pi d/\lambda)$, which is a flat-Earth equation. It computes the loss of signal strength with distance due to the inverse-square law of an expanding wavefront in free space. It does not know what shape the Earth is. It returns exactly the same number for a sphere of radius 6,371 km and for an infinite flat plane, because free-space path loss depends only on distance and wavelength, not on ground geometry. This is the formula flat-Earth propagation models use. On a sphere, there is an additional loss term that Friis does not include: **Fock smooth-Earth diffraction loss**. When a transmitter and receiver are on opposite sides of the geometric horizon of a sphere (i.e., the straight line between them passes through the Earth), the signal has to bend around the curved surface by diffraction. The strength of that diffraction signal decays exponentially with distance past the horizon, and the rate of decay depends on the sphere's radius, the frequency, the ground conductivity, and the terminal heights. This is the mechanism codified in ITU-R P.526-16 Eq. 13 to 18, which is the international-standard calculation for exactly this problem class: VHF propagation over a smooth spherical Earth with elevated terminals. Dan's link-budget tool does not compute this term. The "15 dB obstruction loss" it does include is a ground-level environmental clutter correction (the kind of factor standardised in [[2021_ITU-R_P833-10_Vegetation|ITU-R P.833-10, *Attenuation in vegetation*]]), designed for terrestrial paths where at least one terminal is within woodland or adjacent to a vegetation block, not for an air-to-ground VHF path at 6,000 m where the aircraft is above all terrain clutter. The tool is answering a different question than the one being asked. The question is *"does the signal survive Earth-curvature diffraction on a globe?"* and the tool is answering *"does the signal survive trees and buildings on a flat surface?"* ### What the Fock diffraction loss actually is at these distances For Kleve → Spalding (440 km), the ITU-R P.526-16 Fock calculation gives **38.1 dB** of smooth-Earth diffraction loss on top of the 115.3 dB FSPL, for a total path loss of **153.4 dB**. That still leaves a globe equisignal SNR of **+42.6 dB** above the noise floor, which is usable. At this one distance, Dan's qualitative conclusion (the signal works) happens to be right, because Kleve → Spalding is close enough to the radio horizon that the Fock loss does not kill the signal. But Dan does not stop at 440 km. He infers from the 45 dB fade margin that *"yes you can go a thousand km with this"*. That is where the missing Fock term matters. On a flat surface, signal loss scales with distance as $20\log_{10}(d)$, which is gentle (doubling the distance costs 6 dB). On a globe past the radio horizon, signal loss scales **exponentially** with distance, because Fock diffraction is an exponential decay in the shadow zone. Doubling the distance past the horizon does not cost 6 dB. It can cost 50 to 100 dB depending on the frequency and the geometry. The 45 dB margin that looks comfortable at 440 km is consumed and then some by the time the path reaches 700 km on a globe. The table below compares what Dan's tool would produce (flat FSPL only) with what the ITU-R P.526-16 globe model gives at four distances: | Path | Distance | FSPL (flat) | Fock loss (globe) | Total loss (globe) | Globe eq SNR | |---|---|---|---|---|---| | Kleve → Spalding | 440 km | 115.3 dB | 38.1 dB | 153.4 dB | **+42.6 dB** | | Kleve → Derby | 529 km | 116.8 dB | 54.5 dB | 171.3 dB | **+24.7 dB** | | Stollberg → Beeston | 694 km | 119.2 dB | 112.8 dB | 232.0 dB | **−35.6 dB** | | TF 1000 km | 1,000 km | 122.3 dB | 230+ dB | 352+ dB | **−156 dB** | At Kleve → Spalding the globe model still delivers a usable signal. At Kleve → Derby it is marginal. At Stollberg → Beeston (694 km, the shortest operational Stollberg cross-fix path) the globe equisignal is **36 dB below the noise floor** and the signal does not exist. At the 1,000 km range Dan claims his 45 dB margin can reach, the globe diffraction loss exceeds 230 dB and the signal is 156 dB below the noise floor, approximately 16 orders of magnitude below any physically detectable level. The flat-Earth FSPL model (which is what Dan's tool is computing) gives a comfortable margin at all four distances. That is the whole point: the flat model and the globe model agree at short distances and diverge at long distances, and the divergence is entirely in the Fock term that Dan's tool omits. A tool that cannot see the Fock term cannot distinguish between the two models. It will always say the signal works, because the flat-Earth formula always says the signal works at these distances. Running such a tool and declaring *"45 dB margin, therefore globe works to 1,000 km"* is circular: the tool assumes flat-Earth propagation, the tool says the signal propagates, therefore the globe propagation is proven. The globe test is specifically the Fock correction that the tool leaves out. ### The 45 dB margin is computed at the beam peak, not at the equisignal There is a second omission that is independent of the Fock question. Dan's link budget computes the fade margin at the **peak** of the beam pattern, as if the pilot were sitting directly on the dot-peak or dash-peak axis of one sub-beam. But the pilot does not fly at the peak. The pilot flies at the **equisignal crossover**, which is the narrow centreline corridor where the dot and dash sub-beams have equal amplitude. The equisignal is by definition at a trough between the two peaks, not at either peak. The crossover loss for the Telefunken Large Knickebein 5° squint geometry is **19 dB** (computed in `botb_itu_analysis.py` at line 101 as `CROSSOVER_dB = -19.0`, derived from the sub-beam amplitude at the crossover angle relative to the sub-beam peak). This is a fixed geometric property of the antenna pattern: at the equisignal centreline, each sub-beam is 19 dB weaker than it would be at its own peak, because the crossover sits on the falling slope of each sub-beam's main lobe. Every bar chart and SNR figure in [[GRWAVE_P368_BotB]] shows both the peak and equisignal values for exactly this reason. Dan's 45 dB fade margin does not subtract this 19 dB crossover loss. His tool computes the received power as if the receiver were pointed at one sub-beam's peak. The operationally relevant margin is the **equisignal** margin, which is the peak margin minus the 19 dB crossover loss: $45 - 19 = 26$ dB. A 26 dB equisignal margin at 528 km is still positive, so the qualitative conclusion at Kleve → Derby does not change. But the extrapolation to 1,000 km collapses faster, because the starting margin is 19 dB smaller than Dan computed, and the exponential Fock decay past the horizon eats what remains within another 150 km. ### What Dan's tool is actually computing, and what it is not To be specific about the tool's inputs and what they correspond to: - **117 dB free-space path loss.** This is $20\log_{10}(4\pi d/\lambda)$ at 528 km and 31.5 MHz. It is the correct Friis FSPL for this geometry. The formula is shape-agnostic and produces the same number on any Earth shape or on no Earth at all. It is not a globe-model term. - **15 dB obstruction loss.** This is a forest/urban/statistical clutter margin. If it is a user input, it is an arbitrary number Dan typed into the tool. If it is a calculated value, it is not computed using ITU-R P.526 Fock diffraction or any other international-standard smooth-Earth model. It corresponds to the class of empirical environmental correction standardised in [[2021_ITU-R_P833-10_Vegetation|ITU-R P.833-10, *Attenuation in vegetation*]], whose Figures 1 and 3 explicitly assume at least one terminal is inside or immediately adjacent to the woodland at or below canopy height. At 6,000 m altitude the receiver is far above every forest, every building, and every terrain feature in England. Whatever the 15 dB represents, it is not Fock diffraction, and applying a ground-level vegetation correction to a path geometry where the receiver is at airliner altitude is a category error. Obstruction loss $!=$ shadow zone decay. It's for a completely application. - **1 to 2 µV FuBl 1 sensitivity.** This figure does not appear in the primary-source [[1943_DLuft_T4058_FuBl_2_Geraete_Handbuch|D.(Luft) T.4058 FuBl 2 Geräte-Handbuch]], which defines sensitivity operationally (reliable reception at 200 m altitude, 70 km range, 500 W ground power) rather than as a microvolt bench figure. The 1 to 2 µV number is a tool input Dan chose, not a traceable primary-source specification. Even granting an assumed 1 or 2 µV, the signal goes below the galactic noise floor at the Stollberg distances. - **19 dB equisignal crossover loss.** Not included. The tool computes at the beam peak. The pilot flies at the equisignal crossover, which is 19 dB below the peak for the Knickebein 5° squint geometry. The real fade margin is 19 dB less than the tool reports. - **ITU-R P.526-16 Fock diffraction loss.** Not included. At 440 km with TX 111 m and RX 6,000 m on a 6,371 km sphere, this is 38.1 dB. At 694 km it is 112.8 dB. At 1,000 km it exceeds 230 dB. The tool does not compute this and has no input field for it. This is the only term in the link budget that depends on the shape of the Earth, and it is the only term that tests the null hypothesis. ### Dan does not run the calculation for the Stollberg cross-beam The calculation shown on-stream at 2:36:29 to 2:37:57 is run for **Kleve → Derby** (528 km). This is the director beam from Kleve. It is not the cross beam from Stollberg. Knickebein is a two-beam system. A single Kleve beam cannot provide a position fix because it gives only a line, not a point. The release point for the bombs is defined by the intersection of the Kleve director beam with the Stollberg marker beam. Without the Stollberg beam arriving at the intersection point, the system does not work. The Stollberg cross-beam is therefore the path the globe model must also explain. The Stollberg → Beeston distance is 694 km, which is deep in the shadow zone on a globe, and the Fock diffraction loss at 694 km is 112.8 dB, enough to bury the signal 36 dB below the noise floor. Dan's link-budget demonstration runs the Kleve path (which the globe model can deliver at 440 to 530 km) and does not run the Stollberg path (which the globe model cannot deliver at 694 km). Running the easier path and skipping the harder one is not a response to the cross-beam argument. It is an avoidance of the cross-beam argument. ### The analysis is incomplete for the problem being asked None of the figures on Dan's link-budget slide comport with the ITU-R standards-track calculation for a VHF signal propagating horizontally over a spherical Earth. A path-loss analysis for that geometry has to include, at a minimum, the four terms defined in ITU-R P.526-16 (Fock smooth-Earth diffraction) and ITU-R P.372 (radio noise environment at the receiver input). Dan's tool output contains neither. It contains a Friis free-space term (shape-agnostic), an arbitrary obstruction term (not traceable to P.526 or P.833 at 31.5 MHz with this geometry), a user-supplied microvolt receiver sensitivity (not present in the FuBl 2 service manual), and a fade margin computed at the beam peak (not at the equisignal crossover the pilot actually flies). Each omission has a specific consequence: 1. **Fock diffraction not included.** The Friis formula is the same on a flat plane and on a 6,371 km sphere. The only term in a VHF link budget that depends on Earth shape is the Fock smooth-Earth diffraction loss from ITU-R P.526-16 Eq. 13 to 18. Without it, the calculator cannot distinguish the two Earth models, and any conclusion it draws about globe propagation is a conclusion about flat propagation with the label changed. At 440 km the Fock term adds 38 dB; at 694 km it adds 113 dB; at 1,000 km it exceeds 230 dB. 2. **Obstruction loss not traceable.** The 15 dB is either a number Dan typed into the tool or a value the tool computed from an internal model that is not ITU-R P.526. It is not a P.833 vegetation attenuation for this geometry (the path does not go through vegetation; the aircraft is 6,000 m above every canopy in England). P.833 at 31.5 MHz for a terminal inside woodland gives an $A_m$ of roughly 6 to 8 dB; for a terminal at 6,000 m it gives 0 dB. The 15 dB is not a standards-track number at the Knickebein geometry. 3. **Receiver sensitivity not traceable.** The 1 to 2 µV figure does not appear in the primary-source [[1943_DLuft_T4058_FuBl_2_Geraete_Handbuch|D.(Luft) T.4058 FuBl 2 Geräte-Handbuch]]. The Luftwaffe defined FuBl 2 sensitivity operationally (reliable reception at 200 m, 70 km, 500 W ground power), not in microvolts. The 1 to 2 µV is a modern back-calculation, not a primary-source specification. 4. **Equisignal crossover not subtracted.** The tool reports the fade margin as if the receiver were sitting on a sub-beam peak. The pilot flies the equisignal, which is 19 dB below the peak for the 5° squint geometry. The operationally relevant margin is the peak margin minus 19 dB. Add all four corrections and the result barely changes at Kleve → Spalding (the path still works at +42.6 dB globe equisignal SNR). But the result at Stollberg → Beeston collapses to 35.6 dB below the noise floor, and the 1,000 km extrapolation sits 156 dB below the noise floor. The "45 dB fade margin therefore 1,000 km achievable on the globe" conclusion does not survive once the four missing terms are restored, because the dominant loss mechanism on a globe is the Fock diffraction that the tool does not compute, and the tool's fade margin is 19 dB higher than the operational equisignal margin before any of the other corrections are applied. The analysis is not wrong in the sense of containing arithmetic errors. It is incomplete in the sense of omitting every mechanism that would let the calculation test the null hypothesis. A tool that implements Friis only cannot answer the question of whether a signal survives a 500 to 1,000 km path over a 6,371 km sphere, because the answer to that question lives entirely in the Fock term the tool does not have. --- ## Argument 6: "The beam refracts and diffracts around the curve. Ground waves hug the earth. The optical horizon is not the RF horizon." ### Quoted > [!quote] DC Dan Dano — *Drunk Debunk Show* 13 Feb 2025 > "Look here's the optical Horizon here's the radio Horizon wow that looks a lot like curvature to me" > — **50:18** > > "We have uh the vertical polar polarization helps it hug the curve of the earth... this reflection is basically like bouncing the waves off the ground... it travels in the same direction that it reflects off of because you know it's basically perpendicular so because the second travel way is now traveling almost parallel to the surface and it follows the ray more closely because it's basically like starting a new transmission from that point" > — **54:13 to 55:08** > > "This form of defraction is in propagation is more than just math it is an art form there are so many variables... one of the biggest things that we can do is I'm trying to re-engineer this is I can put in different variables to actually model what the radiation pattern was" > — **55:41 to 56:13** ### Claim The flat-Earth analysis treats the optical line-of-sight as if it were a hard cutoff on the propagation of VHF radio energy, and that is physically naive. Radio waves are not optical, and three well-established non-line-of-sight mechanisms extend the effective reach of a 31.5 MHz signal past the geometric optical horizon on any spherical Earth: 1. **Atmospheric refraction.** The refractive index gradient of the lower troposphere bends radio rays downward as they propagate, producing the so-called 4/3 Earth radio horizon. For the Kleve geometry this is about 15% further than the optical horizon, and it is documented empirically going back to 1902 and formalised in the 1930s by Eckersley and others. 2. **Ground-wave propagation via grazing-ray reflection.** A vertically polarised VHF signal launched at low elevation angle along the surface encounters Earth curvature as an obstacle at a very shallow incidence angle. At near-grazing incidence the reflected ray travels nearly parallel to the surface and behaves as if it were a new transmission originating from the reflection point. The secondary ray is itself subject to refraction, which bends it further around the curvature. This mechanism lets a vertically polarised signal "hug the curve of the Earth" well past where an optical-only analysis would predict it cannot reach. Lower frequencies than 31.5 MHz are more optimal for ground-wave propagation, but this does not exclude 31.5 MHz from using the mechanism; it just requires a more capable antenna and transmitter engineering, which Knickebein had. 3. **Diffraction around the curvature itself.** Radio waves diffract around obstacles including the curved surface of a sphere. The diffraction of VHF energy past a smooth spherical Earth is a complex problem with multiple coupled variables ("more than just math, it is an art form"), not a simple closed-form calculation, and a full numerical model has to account for the radiation pattern of the transmit antenna (modelled here in the 4nec2 simulator), the ground electrical characteristics along the path, the atmospheric refractivity profile, and the receive antenna's effective aperture at the target altitude. Any propagation analysis that ignores these three mechanisms and computes only free-space path loss combined with an optical-line-of-sight cutoff will produce a "falsification" that is really an artefact of the missing physics. A correct globe-model calculation that includes refraction, ground-wave grazing reflection, and smooth-sphere diffraction will show that Knickebein can in fact reach the Midlands targets on a sphere of radius 6,371 km. ### The ITU already captures every mechanism Argument 6 names Argument 6 lists three mechanisms that the flat-Earth analysis allegedly ignores: tropospheric refraction, ground-wave propagation by grazing reflection, and diffraction around the curvature itself. The argument asserts that a correct globe calculation has to include all three, and that including them rescues VHF propagation out to the Midlands targets. The conclusion is that the BotB null hypothesis calculation omits these mechanisms and therefore misrepresents the globe model. This is the opposite of what is happening in the null doc. The BotB analysis uses [[2025_ITU-R_P526-16_Diffraction|ITU-R P.526-16 *Propagation by diffraction*]], which is the canonical international standard for VHF propagation over a spherical Earth. That Recommendation already incorporates all three mechanisms Argument 6 names, by construction. Each of them is a specific term or correction inside the P.526-16 equation set, and the `botb_itu_analysis.py` implementation computes every one of them for every Knickebein path. **Tropospheric refraction is the 4/3 Earth radius correction.** The standard-atmosphere refraction that bends rays downward and extends the radio horizon beyond the optical horizon is captured in P.526-16 by using the effective Earth radius $a_e = k \cdot a$, with $k = 4/3$. All distance and height normalisations in §3.1.1.2 (Eq. 14a and Eq. 15a) take $a_e$ rather than the geometric radius $a = 6371$ km. The calculation in `botb_itu_analysis.py` uses $k = 4/3$ and $a_e = 8495$ km throughout. Argument 6's appeal to the 4/3 Earth radio horizon is an appeal to a term the BotB analysis already uses. **Ground-wave grazing reflection is the surface admittance $K$ and the polarisation-dependent $\beta$ parameter.** The coupling of a vertically polarised wave to the Earth's surface is captured in P.526-16 §3.1.1.1 through the normalised surface admittance $K$, given by Eq. 11a for horizontal polarisation and Eq. 12a for vertical polarisation, where $K$ is a function of ground conductivity $\sigma$, relative permittivity $\varepsilon$, frequency $f$, and effective Earth radius $a_e$. The $K$ value then feeds into the $\beta$ polarisation parameter of Eq. 16, which scales the distance and height-gain terms in Eq. 13 to 18. For vertical polarisation over sea below 300 MHz, $\beta$ is less than 1 and the diffraction loss is correspondingly reduced because the surface is more conductive and supports a stronger surface-coupled mode. For vertical polarisation over average land at 31.5 MHz, $\beta \approx 1$ above the 20 MHz threshold given in §3.1.1.2, so the surface-coupled mode is present but not dominant. The BotB library computes $K$ and $\beta$ per path from the ground-type CSV column and applies the corresponding corrections to the diffraction loss. This is exactly the "ground-wave propagation" mechanism Argument 6 is invoking. **Diffraction around the curvature is the entire purpose of P.526-16.** The canonical solution is the classical residue series, described in §3 of [[2025_ITU-R_P526-16_Diffraction#§3 Diffraction over a spherical Earth — the residue-series intro|P.526-16 §3]]. The opening paragraph states verbatim: > "The additional transmission loss due to diffraction over a spherical Earth can be computed by the classical residue series formula. The computer program 'LFMF-SmoothEarth' in Recommendation ITU-R P.368 provides the complete method." The residue series is the exact solution to Maxwell's equations on a spherical Earth with an impedance boundary condition. Each term in the series corresponds to a creeping-wave mode with a complex eigenvalue $\tau_n$; the imaginary part of $\tau_n$ controls the exponential decay with distance past the geometric horizon, and the real part controls the phase. The first term of the residue series is the Fock approximation, which [[2025_ITU-R_P526-16_Diffraction#§3.1.1.2 Diffraction field strength equations 13 to 17|P.526-16 §3.1.1.2]] implements as Eq. 13: $20 \log_{10} \frac{E}{E_0} = F(X) + G(Y_1) + G(Y_2) \quad \text{dB}$ with: - $F(X)$ = distance term (Eq. 17) - $G(Y_1)$, $G(Y_2)$ = height-gain terms at the transmitter and receiver (Eq. 18) - $X$ = normalised distance (Eq. 14a) - $Y$ = normalised antenna height (Eq. 15a) - $\beta$ = polarisation/ground parameter (Eq. 16) This is the "diffraction around the curvature" Argument 6 asks for. It is the mechanism P.526-16 is for. The BotB library implements Eq. 13 to 18 directly from the Recommendation and cross-checks against the full multi-mode residue series via the ITU-R P.368 LFMF-SmoothEarth reference code. See [[GRWAVE_P368_BotB]] for the side-by-side validation. ### The 31.5 MHz ground-wave contribution is already small and already included Argument 6 acknowledges that "lower frequencies are more optimal for ground-wave propagation, but this does not exclude 31.5 MHz". The quantitative version of that statement is captured in Eq. 11a and 12a of [[2025_ITU-R_P526-16_Diffraction#§3.1.1.1 Influence of the electrical characteristics of the surface — the surface admittance K|P.526-16 §3.1.1.1]], which determine the normalised surface admittance $K$ as a function of frequency. For vertical polarisation at 31.5 MHz: - Over average land ($\sigma = 0.005$ S/m, $\varepsilon_r = 15$): $K_V \approx 0.003$, which is below the $K = 1$ threshold where §3.1.1.2 remains valid. Per §3.1.1.1, the Earth is in the "electrical characteristics not important" regime for horizontal polarisation at this $K$, and for vertical polarisation the $\beta$ parameter of Eq. 16 evaluates to $\beta \approx 1$. - Over seawater ($\sigma = 5$ S/m, $\varepsilon_r = 70$): $K_V \approx 0.054$, still well below the $K = 1$ clamp. $\beta \approx 0.81$ per Eq. 16a, which reduces the $F(X)$ distance decay and the $G(Y)$ height-gain terms proportionally. This is the "sea paths have less diffraction loss than land paths" correction that was applied to every Stollberg path after the β-fix in the library. In neither case does the ground-wave mechanism add enough extra signal to convert a shadow-zone path past ~500 km into a usable signal at 31.5 MHz, because the shadow-zone exponential decay from the $F(X)$ distance term is at the rate of $-17.6 \, X$ per unit normalised distance (Eq. 17a), and $X$ grows steeply with distance once past the horizon. The ground-wave correction reduces $\beta$ slightly, which reduces $X$ slightly, which reduces the exponential decay slightly. It does not change the regime. The signal is still exponentially decaying in the shadow zone. ### The Fock residue series is the diffraction mechanism, not a separate rescue The most common misconception in Argument 6 is the treatment of "diffraction around the curvature" as a separate, additional rescue term that could restore signal the null doc has thrown away. Fock smooth-Earth diffraction is not an extra term on top of the flat-Earth calculation. It is the term. The flat-Earth Friis formula $20 \log_{10}(4 \pi d / \lambda)$ is the limit of the full spherical diffraction calculation as the sphere radius goes to infinity. For a finite sphere the additional transmission loss is exactly the Fock residue-series result, which P.526-16 implements via Eq. 13. There is no mechanism "Fock misses" that some other calculation would recover. Argument 6 refers to diffraction as "more than just math, it is an art form". Fock's 1945 derivation (see [[1945_Fock_Diffraction_Radio_Waves_Earth|Fock 1945]]) and its numerical refinements through Vogler (1961), Bremmer (1949), and Shatz & Polychronopoulos (1988) have reduced the problem to a set of closed-form equations with tabulated eigenvalues and asymptotic expansions valid across the entire VHF-through-UHF range. The 1988 Shatz paper gives a 35-mode residue-series computation that matches the first-term approximation of P.526-16 to better than 0.2 dB for any $(d, h_1, h_2, f, \varepsilon, \sigma)$ tuple inside the VHF regime. See [[1988_Shatz_Spherical_Earth_Diffraction_SEKE|Shatz (1988)]]. The art is in the eigenvalue computation, not in whether the answer exists. ### Comparing the null-doc numbers to Dan's link budget at Kleve → Derby (Marble Hall) The on-stream demo at 2:36:29 ran the Kleve → Marble Hall (Derby) path in the link-budget tool. The tool reported: | Quantity (tool output) | Value | |---|---| | Distance | 528.147 km | | Frequency | 31.500 MHz | | TX power | 64.77 dBm (3 kW) | | TX antenna gain | 25.00 dBi | | RX antenna gain | 2.00 dBi | | RX sensitivity | −107.00 dBm | | Free-space loss | 116.82 dB | | Obstruction loss | 14.58 dB | | Forest loss | 1.00 dB | | Urban loss | 1.00 dB | | Statistical loss | 19.29 dB | | **Total path loss** | **152.70 dB** | | Received signal | −61.43 dBm (189.90 μV) | | **Fade margin** | **45.57 dB** | The BotB ITU-R P.526-16 Fock calculation for the same geometry (Kleve Kn-4 at 111 m, Marble Hall Derby at 6,000 m, 31.5 MHz, 3 kW, vertical polarisation, average land $\beta = 1$): | Quantity (P.526-16) | Value | |---|---| | Distance | 528 km | | Frequency | 31.500 MHz | | FSPL (Eq. Friis) | 116.8 dB | | Fock diffraction loss (Eq. 13 to 18) | 54.5 dB | | Total path loss (globe) | 171.3 dB | | Globe peak SNR | +43.7 dB | | Globe equisignal SNR (peak − 19 dB crossover) | +24.7 dB | The FSPL terms match exactly (116.8 dB), because the Friis formula is shape-independent and both calculations are computing the same term. The divergence is in the diffraction/obstruction row: - Tool: obstruction 14.58 + forest 1.00 + urban 1.00 + statistical 19.29 = **35.87 dB** of miscellaneous loss, none of which is Fock smooth-Earth diffraction from the ITU-R P.526 residue series. - BotB P.526-16: **54.5 dB** of Fock diffraction loss from Eq. 13 to 18, which is the term P.526-16 defines as the "additional transmission loss due to diffraction over a spherical Earth". The 35.87 dB the tool attributes to "obstruction" + "forest" + "urban" + "statistical" is 18.6 dB less than the 54.5 dB P.526-16 gives for Fock at the same geometry. If the tool's figures were cross-checked against the ITU-R standards the tool claims to be solving the same problem as, they would not be 18.6 dB off. The disagreement is not a round-off. It is the difference between computing Fock diffraction from the Recommendation and computing a generic clutter margin from a tool whose internal model is not P.526. ### The exponential fall-off that Dan's 45 dB margin does not survive The $F(X)$ distance term in [[2025_ITU-R_P526-16_Diffraction#§3.1.1.2 Diffraction field strength equations 13 to 17|P.526-16 Eq. 17a]], valid for $X \geq 1.6$ (i.e., past the first shadow-zone transition), has the form: $F(X) = 11 + 10 \log_{10}(X) - 17.6 \, X \quad \text{dB}$ The dominant term at large $X$ is $-17.6 \, X$. $X$ scales with distance as $X \propto f^{1/3} \, a_e^{-2/3} \, d$ (Eq. 14a), so $X$ grows linearly with $d$ at fixed frequency and Earth radius. That means $F(X)$ decreases linearly with $d$, which means the **field strength in dB decreases linearly with distance, which means the signal amplitude decays exponentially with distance**. This is the signature of a creeping-wave mode: the field strength past the horizon is $\propto \exp(-\alpha d)$ where $\alpha$ is the imaginary part of the first Fock eigenvalue. The practical consequence for Knickebein: | Path | Distance | X (31.5 MHz, $a_e$ = 8495 km, β = 1) | −17.6 X | Fock loss incl. height gain | |---|---|---|---|---| | Kleve → Spalding | 440 km | 2.11 | −37.1 dB | 38.1 dB | | Kleve → Derby | 528 km | 2.53 | −44.5 dB | 54.5 dB | | Stollberg → Beeston | 694 km | 3.33 | −58.6 dB | 112.8 dB | | TF 1000 km | 1000 km | 4.80 | −84.5 dB | 230+ dB | The Fock loss row grows superlinearly with distance because the $F(X)$ linear decay compounds with a decrease in the height-gain terms as the effective path past the horizon grows. Between 440 km and 694 km the Fock loss grows by 75 dB, or an average of 0.3 dB per km. Between 694 km and 1000 km it grows by another 120+ dB, or 0.4 dB per km. Dan's 45.57 dB fade margin at 528 km does not extend to 1,000 km. It extends to 528 km plus whatever distance adds 45.57 dB of additional Fock loss to the 54.5 dB already present at 528 km. At 0.3 to 0.4 dB per km, that is roughly 120 to 150 km of additional reach on a globe, taking the Kleve path to approximately 650 to 680 km at the detection floor. The Marble Hall target at 528 km works. The 1,000 km claim the tool's margin is extrapolated to does not, because the extrapolation is linear (flat-Earth logarithmic) and the actual globe-model loss is exponential. This is the mechanism someone with a background in radio-wave propagation is expected to know. The residue-series exponential decay past the horizon has been part of ITU-R recommendations since P.526-1 in 1978 and is derivable from Fock's 1945 paper. A 45 dB margin is a flat-Earth result, not a globe result, and applying it uniformly across all distances assumes the flat-Earth scaling that the tool implicitly has. ### Eckersley told the British in 1940 what the Fock calculation now tells us The historical record contains an independent check on the Fock calculation, from the British side, written in 1937 by the man who knew VHF propagation better than anyone else working for British intelligence at the start of the war: > "As a concession to the wave theory, it is generally admitted that some energy leaks into the shadow region, but the amount of spread is generally left to guesswork." > — [[1937_Eckersley_Ultra_Short_Wave_Refraction_Diffraction|Eckersley (1937)]] Eckersley was using the same diffraction physics that later became P.526-1 (1978) and is now P.526-16 (2025). In June 1940 he told R.V. Jones and the War Cabinet that on a curved Earth, 31.5 MHz beams from Germany could not reach the Midlands past 500 km of path. The Fock calculation today says the same thing at Stollberg → Beeston (694 km): 112.8 dB of diffraction loss on top of 119.2 dB FSPL gives a total path loss of 232 dB and a globe equisignal SNR of −35.6 dB, which is 35.6 dB below the noise floor and not detectable. The beams reached England anyway. The British counter-Knickebein jamming chain (Aspirin, Bromide, Benjamin) was built because the Luftwaffe was guiding He 111 bombers to Birmingham, Derby, Liverpool, Manchester, Sheffield, Bristol, Southampton, Plymouth and Hull on Kleve + Stollberg cross-fixes, and the bombs were falling. Eckersley's diffraction analysis was correct for a sphere. The bombs fell anyway. The discrepancy between the correct-for-a-sphere analysis and the historical reality is the null-hypothesis falsification. Argument 6's appeal to diffraction mechanisms is an appeal to the exact physics that predicts the beams should not work, not an argument that rescues them. --- ## Argument 7: "The beam is vertically polarised, so it is narrow horizontally but tall vertically. That is why the bomber stays in the beam at altitude." ### Quoted > [!quote] DC Dan Dano — *Drunk Debunk Show* 13 Feb 2025 > "So when we talk about vertical polarization it's very narrow but very high okay so here's the top view of of the signal very narrow here's the horizontal view very high... otherwise if it was symmetrical and it was 4 to 500 yard wide as it was high they could just have Fighters sitting in that same 400 yards the entire time waiting for bombers to show up what's the point of using that for a bomber Aid if you can't if you have to be that low I mean if it was on a flat Earth and it was that high it would be 12 it would be like 12200 feet above the ground" > — **56:38 to 57:19** ### Claim The flat-Earth analysis treats the Knickebein beam as if it were geometrically symmetric in the horizontal and vertical planes, but a vertically polarised VHF antenna array of the Knickebein geometry does not radiate symmetrically. A 4nec2 simulation of the Knickebein radiation pattern (performed by a third-party researcher and shown on stream) produces two distinct views of the same signal: a top-down view that is narrow in the horizontal plane (consistent with the 400 to 500 yard equisignal corridor used by the flat side) and a side-elevation view that is tall in the vertical plane (many thousands of metres of usable reception at altitude, not 400 to 500 yards). This asymmetry is exactly what a bomber navigation beacon should look like: narrow horizontally so the beam has precise azimuth accuracy, and tall vertically so bombers at any operational altitude can fly through the beam without needing to hold a specific height. If the flat-Earth analysis is correct that the beam is also only 400 to 500 yards tall, then the system makes no operational sense: a beam that thin vertically would require bombers to hold 1,200 feet above the ground to stay in the beam, which is below any sane bombing altitude, and interceptor fighters could simply sit at that altitude waiting. The flat-Earth argument, by conflating the narrow horizontal corridor with the total 3-D beam shape, has misrepresented what the system's geometry would even look like if the physics being claimed were correct. ### The asymmetric beam is already in the null doc, by physics not by accident The Telefunken Large Knickebein array is 99 m wide by 29 m tall. A rectangular aperture that much wider than it is tall radiates a pattern that is narrow in the horizontal and broad in the vertical. This is basic antenna physics and the BotB analysis uses it directly: the directivity calculation $G = 4\pi A / \lambda^2$ with $A = 99 \times 29 = 2{,}871$ m² at $\lambda = 9.517$ m gives 26 dBi at the peak, and the asymmetric lobe shape is what determines the equisignal corridor width (narrow, horizontal) and the vertical coverage (broad, vertical) separately. The $G = 4\pi A / \lambda^2$ identity is the gain-form rearrangement of [[1946_Friis_Simple_Transmission_Formula|Friis 1946 Eq. 1]] (see the note at the bottom of page 1 of the Friis source note). The rectangular-aperture model is the entire reason the 0.066° horizontal equisignal corridor exists alongside a vertical lobe tall enough to cover bombers from ~1,000 m to ~8,000 m altitude. So Argument 7 attacks a position the null doc does not hold. Nobody is claiming the beam is symmetric. The 99 m × 29 m rectangular aperture is explicitly asymmetric by design, and the null doc's directivity and lobe-shape math comes from that asymmetric aperture. ### Dan's own slide lists the exact ITU recommendations the BotB analysis uses At 1:06:22 Dan puts up this slide: ![[DC_Dan_slide_engineering_for_knickebein_today.png]] The slide text, read off the image verbatim: > **Engineering for Knickebein Today** > > The International Telecommunications Union (ITU) provides crucial resources for RF Engineers, publishing standards and recommendations that are continuously updated and refined as new research becomes available. > > If an RF Engineer were to attempt to develop Knickebein today, they would consult the following International Telecommunications Union (ITU-R) recommendations: > > - ITU-R P.525: Ground-wave propagation prediction methods for frequencies between 10 kHz and 30 MHz > - ITU-R P.526: The radio refractive index: its formula and refractivity data > - ITU-R P.525: Calculation of free-space attenuation > - ITU-R P.526: Propagation by diffraction > - ITU-R P.527: Electrical characteristics of the surface of the Earth > - ITU-R P.528: A propagation prediction method for aeronautical mobile and radionavigation services using the VHF, UHF and SHF bands > - ITU-R P.534: Effects of tropospheric refraction on radiowave propagation > - ITU-R P.1546: Method for point-to-area predictions for terrestrial services in the frequency range 30 MHz to 4 000 MHz This is the same list of ITU-R recommendations the BotB null doc uses. [[2025_ITU-R_P526-16_Diffraction|P.526]] for diffraction over a spherical Earth. P.527 for the ground electrical characteristics that feed into the surface admittance $K$. P.528 for aeronautical radionavigation at VHF. P.525 for free-space attenuation. P.453 and P.834 for tropospheric refraction that gives the $k = 4/3$ effective Earth radius. P.1546 for the 30 MHz to 4 GHz point-to-area method used as a cross-check. Dan is telling the audience that these are the right tools for modelling Knickebein today. That is accurate. What Dan does not do on the same stream is actually run these calculations on the Kleve → Spalding, Kleve → Derby, or Stollberg → Beeston paths. The link budget he does run (Arg. 5) uses a generic amateur-radio calculator that implements none of the ITU-R recommendations listed on this slide. The BotB null doc does run them. Every line item on Dan's slide is in `botb_itu_analysis.py` and documented in [[GRWAVE_P368_BotB]] with per-path results at every confirmed Knickebein target distance. We built exactly what Dan's slide says a modern engineer would build. By his own standard of what the right engineering looks like, the null doc clears the bar and his own link-budget screenshot does not. ### Dan's own slide says the globe model includes refraction, reflection and diffraction At 1:12:48 Dan puts up this slide: ![[DC_Dan_slide_knickebein_signal_properties.png]] In the right-hand column, labelled *Globe Explanations*: > - VHF & UHF frequencies are subject to reflection & have a calibrated 'radio horizon' > - Kleve: Elevation is 84m > - Diffraction: Signal extends beyond the radio horizon > - Ground doesn't dip the radio signal > - Radio signal influenced by propagation This is a summary of the mechanisms Dan accepts as part of a correct globe calculation. All four are terms inside [[2025_ITU-R_P526-16_Diffraction|ITU-R P.526-16]]: - *Reflection* is inside the surface admittance $K$ of Eq. 11a / 12a. - *Radio horizon* is the $k = 4/3$ effective Earth radius used in Eq. 14a and 15a. - *Diffraction extending beyond the radio horizon* is the Fock residue-series calculation of Eq. 13 to 18. - *Propagation-influenced signal level* is the field strength $E$ on the left-hand side of Eq. 13. These are the mechanisms the BotB null doc computes. Dan says they are the right mechanisms. The null doc uses them and the link budget Dan actually ran does not. ### Dan's own slide says diffraction propagation is "more than just math, it is an art form" At 55:40 Dan puts up this slide: ![[DC_Dan_slide_diffraction_propagation.png]] The slide concludes with: > "This form of diffraction and propagation is more than just math, it is an art form. An engineer would have to use complex math equations and account for a multitude of variables" Fock's 1945 derivation and the modern Recommendation [[2025_ITU-R_P526-16_Diffraction|P.526-16]] are the complex math equations that handle those variables. The $F(X)$ distance term, the $G(Y)$ height-gain term, the $\beta$ polarisation parameter, the $K$ surface admittance, the effective Earth radius $a_e = 8{,}495$ km, all come out of Fock's residue-series solution to Maxwell's equations on a sphere with an impedance boundary condition. The "art" is in deriving the eigenvalues $\tau_n$ of the residue modes, which Fock, Bremmer, Vogler, and Shatz did across 1945 to 1988. The art exists and the answer exists. The BotB null doc uses them; Dan's own link-budget tool does not. ### Asymmetric vertical extent does not rescue a shadow-zone signal Granting the entire asymmetric-beam claim, the point still does not help on the globe. A tall vertical lobe only helps if the signal **reaches** the aircraft. On the globe, Stollberg → Beeston at 694 km has a Fock diffraction loss of 112.8 dB on top of 119.2 dB FSPL for a total path loss of 232 dB, which puts the equisignal SNR 35.6 dB below the noise floor. The aircraft is below the horizon as seen from the Stollberg site, and it does not matter how tall the transmitter lobe is in the near field because the signal is exponentially attenuated in the shadow zone before it reaches the aircraft antenna. Vertical extent of the transmitter radiation pattern is irrelevant to whether the signal arrives at the receiver. It only changes the directivity at short range. At 694 km past the radio horizon the Fock decay dominates everything else. The null doc accounts for this with the $G(Y_2)$ height-gain term of [[2025_ITU-R_P526-16_Diffraction#§3.1.1.2 Diffraction field strength equations 13 to 17|P.526-16 Eq. 18]], which is literally the "taller beam helps bombers at altitude" physics reduced to one equation. At 6,000 m receiver altitude the $G(Y_2)$ term contributes roughly +80 dB of positive height gain, which is the only reason Kleve → Spalding at 440 km is usable at all on the globe. At Stollberg → Beeston, the same +80 dB of height gain is present and the path is still 35 dB below the noise floor, because the $F(X)$ distance term has consumed 150+ dB at that range. Argument 7 is true about the antenna geometry and irrelevant to the propagation problem. The beam is asymmetric. The aperture model says so. The null doc uses the aperture model. The asymmetric shape does not extend the Fock-diffracted signal past where Fock says the field collapses to below noise. --- ## Argument 8: "The Kleve antenna is 84 to 91 m tall, not 200 m. The flat-Earth side uses 200 m as a generosity they should not get." ### Quoted > [!quote] DC Dan Dano — *Drunk Debunk Show* 13 Feb 2025 > "They gave us a tower height of 200 M because the benefit of the globe right because we need a benefit reality needs the benefit to make this work" > — **45:36** > > "The actual measurements of cleave are 84 M High we don't need 200 M let's use 84" > — **1:01:54** ### Claim The flat-Earth analyses of Knickebein commonly use 200 m as the effective Kleve transmitter height, which is not the actual site height. The real numbers are 84 m (Kleve terrain elevation) and approximately 91 m for the antenna-centre height on top of the tower. The 200 m figure is a generosity "the benefit of the doubt for the globe" that should not be extended, because using a larger TX height inflates the optical horizon and makes the geometry of beyond-line-of-sight reception look more favourable than it actually is. The correct analysis should use 84 m or 91 m from primary-source site documentation, which shrinks the calculated optical line-of-sight and therefore shrinks the region where the flat-Earth argument can claim that a purely rectilinear line-of-sight path reaches the target. ### The direction of the benefit is reversed Argument 8 accepts that the null hypothesis will be easier to refute if the Kleve TX height is reduced from 200 m to 84 m, and asks for that correction to be made. The underlying assumption is that a taller TX makes the geometry look more favourable for the flat-Earth argument, because a taller TX extends the optical line-of-sight and therefore shrinks the region where a pure geometric-horizon falsification bites. That assumption is physically backwards. A taller TX helps the globe, not the flat model. Reducing the TX height from 200 m to 84 m or 91 m makes the globe number worse, not better, and leaves the flat number unchanged. ### The geometry The radio horizon distance from a transmitter at height $h_t$ over a 4/3 Earth ($a_e = 8{,}495$ km) is: $d_h = \sqrt{2 \, a_e \, h_t} \approx 4.12 \sqrt{h_t(\text{m})} \text{ km}$ At the aircraft altitude $h_r = 6{,}000$ m the receiver's own horizon distance is $d_h(\text{rx}) = 319.3$ km, regardless of TX height. The total radio horizon is $d_h(\text{tx}) + d_h(\text{rx})$. For the Kleve → Spalding 440 km path: | Kleve $h_t$ | $d_h$(tx) | Total radio horizon | Kleve → Spalding shadow zone | |---|---|---|---| | 84 m (terrain only) | 37.8 km | 357.1 km | 82.9 km | | 91 m (terrain + near-ground feed) | 39.3 km | 358.6 km | 81.4 km | | 111 m (83 m terrain + 28 m antenna frame) | 43.4 km | 362.7 km | 77.3 km | | 200 m (the inflated figure) | 58.3 km | 377.6 km | 62.4 km | Going from 200 m down to 91 m adds 19 km of shadow zone. That costs the globe model about 6 dB of extra Fock diffraction loss at 31.5 MHz (at the ~0.3 dB/km slope of the $-17.6 X$ distance term in [[2025_ITU-R_P526-16_Diffraction#§3.1.1.2 Diffraction field strength equations 13 to 17|P.526-16 Eq. 17a]] combined with the associated height-gain reduction). Going from 200 m down to 84 m costs closer to 7 dB. ### How much does TX height change each flat-Earth model? The "flat-Earth side" is not a single calculation. Depending on how much ground physics you put in, the flat model's sensitivity to TX height ranges from exactly zero (pure free space) to about 2 dB (with surface reflectivity). None of the flat formulations pick up anywhere near the 6 dB gift the globe model gets. **Strict Friis free-space path loss.** The Friis equation $L = 20 \log_{10}(4\pi d / \lambda)$ has no $h_t$ term at all. It assumes free space (no ground), antennas in each other's line-of-sight, and no multipath. Move both antennas up or down and the loss is identical. This is what Dan's amateur-radio link-budget calculator computes for its "free-space loss" line item. For Dan's tool the TX height genuinely changes nothing. **Two-ray plane-Earth model.** If you keep the Earth flat but add a ground beneath it, the total received signal is the vector sum of a direct ray and a ground-reflected ray. The path-length difference between the two rays is $\Delta r \approx 2 h_t h_r / d$ for $d \gg h_t, h_r$. At Kleve → Spalding (440 km) with $h_t = 100$ m and $h_r = 6{,}000$ m at 31.5 MHz ($\lambda = 9.52$ m): $\Delta r = \frac{2 \cdot 100 \cdot 6000}{440{,}000} = 2.73 \text{ m}$ The phase difference from the path is $2\pi \Delta r / \lambda = 1.81$ rad ≈ 103°, plus 180° from the reflection at grazing incidence, giving ≈ −77° relative phase. The two rays add at this phase to produce a small interference modulation of approximately **±1 to 2 dB** on top of Friis, depending on the exact ground conductivity and frequency. Doubling the TX height from 100 m to 200 m shifts the interference into a slightly different phase regime, and the effect could swing by another 1 to 2 dB in either direction. The dependence is oscillatory rather than monotonic. **Sommerfeld-Norton with surface reflectivity.** The BotB Sommerfeld-Norton (SN) calculation, which the null doc uses as the honest flat-Earth model, adds the full surface-wave term on top of two-ray. SN has some TX-height dependence through surface-wave coupling, but at the Knickebein geometry (TX ~100 m, RX 6,000 m, 440 km, 31.5 MHz vertical) the surface-wave term is small at the receiver because the aircraft is far above any surface-wave coupling layer. SN at this geometry behaves very close to two-ray, with TX-height variation again in the **±1 to 2 dB range**. **Globe Fock from P.526-16.** The globe model picks up TX height through two distinct mechanisms, both monotonic. First, the radio horizon distance $d_h(\text{tx}) = \sqrt{2 a_e h_t}$ moves the start of the shadow zone outward as $h_t$ grows. Second, the height-gain function $G(Y_1)$ at the transmitter (from [[2025_ITU-R_P526-16_Diffraction#§3.1.1.2 Height gain equations 18|P.526-16 Eq. 18]]) adds positive dB to the field-strength equation as $h_t$ grows. Both add up to roughly **+6 dB per 100 m** of TX height change at the Kleve → Spalding geometry. The net comparison, for a TX height change from 100 m to 200 m at Kleve → Spalding: | Model | TX-height effect (100 m → 200 m) | Direction | |---|---|---| | Friis FSPL (Dan's tool) | 0 dB | flat — no $h_t$ term | | Two-ray plane Earth | ±1 to 2 dB | oscillatory (interference geometry) | | Sommerfeld-Norton (BotB flat model) | ±1 to 2 dB | similar to two-ray | | Globe Fock (P.526-16, BotB globe model) | **+6 dB** | monotonic (radio horizon + $G(Y_1)$) | So dropping Kleve's TX height from 200 m to 91 m takes away roughly 6 dB of globe model headroom while taking away at most 1 to 2 dB from the SN flat-Earth calculation (and exactly 0 dB from the Friis-only calculation Dan's tool reports). Dan is asking for a correction that removes a 6 dB gift to the globe model while removing at most a third as much from any flat-model variant, and nothing at all from the FSPL number shown on his livestream. The correction widens the gap between flat and globe in the flat model's favour by at least 4 dB. ### The 200 m value is not what the BotB null doc uses anyway The BotB `botb_itu_analysis.py` library and `knickebein_paths.csv` use **111 m** for Kleve, not 200 m. That value is 83 m of terrain elevation at the Kn-4 site plus 28 m of antenna-frame height for the Large Knickebein structure, cited from [[1979_Trenkle_Deutsche_Funk_Navigation|Trenkle 1979 p. 67]]. Every globe-model Fock calculation in [[GRWAVE_P368_BotB]] and every SNR number in the Argument 1 table (Spalding peak +61.6 dB, equisignal +42.6 dB) is run at 111 m TX. The 200 m figure is not used. If Argument 8's intent is to correct a Kleve TX height that some flat-Earth presentation uses, it may be correcting something real in that presentation. It is not correcting anything in the BotB null doc. ### Running the correction anyway Accept Dan's request and run Kleve → Spalding at $h_t = 84$ m instead of the BotB library's 111 m. The Fock diffraction loss rises from 38.1 dB at 111 m to approximately 46 dB at 84 m. FSPL is unchanged at 115.3 dB. Total globe path loss rises from 153.4 dB to approximately 161.4 dB. Globe equisignal SNR falls from +42.6 dB to approximately +34.6 dB above the noise floor. The path remains usable at Spalding at the lower TX height, because the 6,000 m aircraft height gain dominates, but the globe margin is 8 dB smaller than before. Now run the same correction at Stollberg → Beeston (694 km, $h_t$ from 72 m BotB to Dan's 84 m or 91 m range — actually a small correction in the other direction since the BotB library has 72 m which is close to Dan's 84 m). The Fock loss at 694 km is already 112.8 dB and the globe equisignal is already 35.6 dB below the noise floor. A small TX height correction either direction is a rounding error on a path that is already 35 dB below noise. The Stollberg cross-beam conclusion is insensitive to this correction. Running the correction does not rescue the globe model anywhere. It makes Kleve → Spalding 8 dB worse (from +42.6 to +34.6 dB) while making no difference at Stollberg → Beeston (still 35 dB below noise). The flat model numbers are unchanged at both paths. ### What Argument 8 actually shows Argument 8 is a correction in the BotB analysis that's welcome. The null doc already uses 111 m rather than 200 m. Moving the Kleve TX height further down to 84 m or 91 m per primary-source site documentation makes the globe model worse and the flat model unchanged, which strengthens the null hypothesis conclusion rather than weakening it. On the Knickebein paths where flat and globe agree (Kleve → Spalding), the agreement survives the correction with slightly smaller globe margin. On the Knickebein paths where flat and globe disagree (Stollberg → Beeston and beyond), the correction is irrelevant because the globe model is already dozens of dB below the noise floor. The physical reason taller TX helps the globe and does nothing for the flat model is the same reason the null doc puts Fock diffraction on one side of the comparison and Friis FSPL on the other: Fock has an $h_t$ term (through the $G(Y_1)$ height-gain function of [[2025_ITU-R_P526-16_Diffraction#§3.1.1.2 Height gain equations 18|P.526-16 Eq. 18]] and through the effective radio horizon $d_h(\text{tx}) = \sqrt{2 a_e h_t}$), and Friis does not. That asymmetry is the physics; Dan's intuition about who benefits from a taller tower has the direction of the asymmetry backwards. --- ## Argument 9: "Only 60 to 70% reliability at Derby, and only 6 bombings out of 33,000 target the Rolls-Royce plant" ### Quoted > [!quote] DC Dan Dano — *Drunk Debunk Show* 13 Feb 2025 > "We're talking about 60 70% of reliability if it was 100% reliability for a year long they would have bombed it a hell a lot more times than six times" > — **2:35:57 to 2:36:13** ### Claim A propagation-tool coverage map of the Kleve-to-Derby path places Derby inside a 60 to 70% reliability contour rather than inside a 100% contour. When paired with UK National Archives bombing records showing only about 6 direct bombings of the Rolls-Royce Merlin works at Derby out of approximately 33,000 recorded bombing incidents during the Battle of Britain period, the implication is that the actual historical attack frequency on Rolls-Royce is consistent with a *partial-reliability* beam rather than a 100%-available one. If the beam had been 100% reliable over a year-long campaign, the Rolls-Royce works would have been bombed many more than six times. The historical record therefore supports the globe-model propagation prediction that the Kleve → Derby path is marginal, and is *inconsistent* with the flat-Earth prediction that the beam works perfectly every night. ### The "60 to 70% reliability" number is an output of a tool that cannot see the Earth's shape The reliability contour Dan shows comes from the same amateur-radio link-budget tool that generated the Kleve → Marble Hall calculation in Argument 5. That tool computes Friis free-space path loss, an arbitrary obstruction margin, antenna gains, and a user-entered receiver sensitivity. It does not compute Fock smooth-Earth diffraction, it does not apply the $G(Y)$ height-gain corrections from [[2025_ITU-R_P526-16_Diffraction#§3.1.1.2 Height gain equations 18|P.526-16 Eq. 18]], and it does not distinguish a 6,371 km sphere from an infinite flat plane. Whatever percentage it reports as "reliability" is derived from a propagation model that does not contain the dominant loss mechanism for a beyond-horizon VHF path. A reliability contour from a shape-agnostic tool is not a measurement of how reliable the beam is on a globe; it is a statement about how often the tool's flat-Earth inputs clear the tool's flat-Earth threshold. The underlying physics was never evaluated. ### The historical record does not support 60 to 70% Knickebein was the mass-bomber navigation system for the entire Night Blitz from September 1940 through May 1941. The Luftwaffe flew He 111 and Ju 88 sorties against Birmingham, Derby, Liverpool, Manchester, Sheffield, Bristol, Southampton, Plymouth, and Hull on Kleve + Stollberg cross-fixes across that period, with operational use continuous enough that the British counter-Knickebein jamming chain (Aspirin, Bromide, Benjamin) was built and staffed specifically because the beams *were* arriving over the Midlands on a nightly basis. A 60 to 70% reliability system does not drive a year-long pathfinder-plus-mass-bomber campaign at that scale. R.V. Jones describes the operational intensity of the Knickebein campaign across Chapters 10 and 11 of [[1978_Jones_Most_Secret_War|Most Secret War (1978)]], including the British decision to invest heavily in countermeasures because the beam was reliable enough to be a strategic threat. ### The low Rolls-Royce hit count is evidence of British countermeasures, not beam unreliability Dan's 6-out-of-33,000 statistic is a comparison of "Rolls-Royce Merlin works at Derby hits" to "total UK bombing incidents across the Battle of Britain period". Those 33,000 incidents include every bomb dropped on every target, not 33,000 Knickebein attempts on Derby specifically. Comparing a specific-target hit count to a nationwide total is not a reliability statistic. What it actually measures is that Rolls-Royce was one target among many and British jamming successfully deflected most of the Knickebein-guided Derby raids away from the Merlin factory onto surrounding empty countryside. That is exactly the story [[1978_Jones_Most_Secret_War|Jones 1978]] tells across Chapters 11 and 12: Aspirin was credited with saving the Merlin works, not with the beam never reaching Derby in the first place. A reliability statistic that doesn't exist cannot support an inference about beam propagation. ### Telefunken's own 1939 tests already rated the system usable to 1,000 km The [[1939_BArch_RL19-6-40_230Q8_App2_Telefunken_Range_Tests|Telefunken July 1939 range campaign]] reports operational ranges of 400 km to 1,000 km across six receiver/antenna configurations at 4,000 m aircraft altitude over open water. The 3,000 W FFuGt transmitter (the Large Knickebein transmitter, Kleve's class) reaches 1,000 km on the Sondergerät receiver. Dan himself characterises that campaign at 2:08:52 to 2:09:00 as *"actual measurements done by the Luftwaffe, repeatable, done multiple times, done with multiple pieces of equipment, done with multiple systems"*. The Germans' own manufacturer rated the beam usable at ranges that include every Midlands target Knickebein was ever flown against. A "60 to 70% reliable" propagation-tool output cannot overturn the manufacturer's own range measurements on which the system was procured and operationally deployed. ### The direction of the error Dan's tool outputs softer numbers than the BotB null doc does because the BotB doc runs the full ITU-R P.526-16 Fock calculation, applies the [[2021_ITU-R_P833-10_Vegetation|P.833]] vegetation-geometry test, computes the Sommerfeld-Norton flat-Earth alternative, subtracts the 19 dB equisignal crossover, and uses the primary-source site heights and antenna dimensions. Dan's tool runs Friis plus a clutter margin and stops. The softer "60 to 70%" figure is what you get when you omit every correction the problem actually requires. Fixing the omissions moves the Kleve → Derby margin down, not up, because the Fock term is subtractive. There is no version of this analysis in which the "60 to 70%" contour is a valid reliability claim about Knickebein propagation on a sphere, because the tool that produced it does not evaluate propagation on a sphere. --- ## Summary Of the nine load-bearing arguments advanced in the presentation: | # | Argument | Status | | --- | --------------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------- | | 1 | "Observation, not measurement" | Goalpost-moving; Telefunken primary source *is* the repeatability. Refuted. | | 2 | "Cockpit noise, civilian receiver, miracle they heard it" | Acoustic vs RF conflation. Refuted by AGC in D.(Luft) T.4058 §24 and Price (2003) on post-war German crew testimony. | | 3 | "Spalding is outside the Luftwaffe's usable region" | Self-defeating; Telefunken table in same memo shows 1,000 km reception. Refuted. | | 4 | "Spec says 0.3° beam width, measurement says 0.066°" | Main lobe vs equisignal crossover confusion. Refuted. | | 5 | "Link budget gives 45 dB fade margin to 1,000 km" | Tool omits Fock diffraction, the load-bearing term. Circular. Refuted. | | 6 | "Refraction, diffraction, ground wave rescue the signal" | ITU-R P.526 already includes refraction + diffraction. Ground wave negligible at 31.5 MHz. Refuted. | | 7 | "Vertical polarization makes the beam tall and narrow" | Strawman; null doc already models rectangular aperture correctly. Refuted. | | 8 | "Kleve is 84 m not 200 m" | The 200 m was generous to the **globe**, not to FE. The direction of benefit is inverted. Lowering TX makes globe worse. Null doc uses 111 m from Trenkle. | | 9 | "60 to 70% reliability at Derby" | Same flawed tool as #5; also confuses operational tasking with propagation physics. Refuted. | **Net assessment:** the presentation is technically literate on general VHF radio terminology and correctly identifies several primary sources that this null doc also cites (Telefunken 1939, Luftwaffe 5 May 1940 memo, Bauer, Trenkle). Those sources are then systematically misapplied, the Fock smooth-Earth diffraction calculation at the load-bearing path (Stollberg → Beeston) is not performed, and the acoustic argument collapses on the first reading of D.(Luft) T.4058 §24. The underlying issue is that **the link-budget tool used in the presentation cannot see curvature**. It computes FSPL plus obstruction loss plus statistical loss, none of which depend on Earth shape. A shape-blind tool is then used to assert that the globe model works, while the actual test (ITU-R P.526 Fock diffraction loss at 694 km) remains uncomputed anywhere in the slide deck. The test the presentation sets out to perform — "refute the null hypothesis" — is not actually performed, because the curvature term is not in the tool. The null hypothesis stands: the Stollberg cross beam cannot reach Midlands targets on a sphere of radius 6,371 km. None of the nine arguments overturn this, and several of them (Arguments 3, 5, 8) inadvertently strengthen the falsification. --- ## Addendum: technical rebuttal of Dan's link-budget screenshot (added 2026-04-14) Source: *Drunk Debunk Show* livestream, 13 Feb 2025, guest DC Dan Dano — <https://youtu.be/H8xiAPDFK1o>. The link-budget screenshot addressed in this addendum is the one Dan presents in that stream at approximately 2:34 to 2:38. The nine-point dossier above refutes Dan's presentation at the level of logic and source handling. This addendum adds a numerical layer: at the specific geometry on his screenshot (Kleve Kn-4 → Marble Hall Derby, 528.1 km, 31.5 MHz, 3 kW, 91 m TX, 6000 m RX), **his reported obstruction loss of 14.58 dB is inconsistent with both ITU-R standards-track globe propagation models**, and the rest of the screen is built on a fade-margin interpretation that is a category error. ### Whatever Dan used for his calculator isn't using ITU standards Dan's screenshot shows the characteristic layout of a VHF link-budget calculator — two endpoints, an azimuth field, and a propagation breakdown into Free space loss, Obstruction loss, Forest loss, Urban loss, Statistical loss, Total path loss, and Fade margin. Tools of this class are built for amateur-radio point-to-point microwave link planning at 100 MHz and above, where they are well validated against measurements. They are **not** ITU-R recommendations, they are not what the military uses for authoritative VHF path analysis, and the ITU has never adopted them into a recommendation. Where the output of such a tool disagrees with **ITU-R P.526-16** (smooth-sphere Fock diffraction) or **ITU-R P.368-9** (GRWAVE ground wave) in their respective regimes of validity, the ITU-R recommendations are the authoritative answer. Both of those are experimentally validated at 31.5 MHz through decades of international field measurements; Dan's tool is not. The BotB null doc uses P.526 Fock and P.368 GRWAVE as its globe models for exactly this reason. They are experimentally verified, they are the authoritative international standards at VHF, and they are the tools any disciplined VHF propagation analysis would use for a beyond-horizon path at 31.5 MHz. ### At Kleve → Derby, the two ITU-R standards give 53 to 65 dB of loss, not 14.58 For the same geometry as Dan's screenshot (91 m TX, 6000 m RX, 31.5 MHz): - **ITU-R P.526-16 Fock** smooth-sphere diffraction: **~65 dB** above free-space loss at 528 km - **ITU-R P.368 GRWAVE** ground wave: **~53 dB** above free-space loss at 528 km - **Dan's screenshot reports**: **14.58 dB** "Obstruction loss" Dan's number is 38 to 50 dB below what the ITU-R globe standards give for the same path. That is not a small disagreement; it is the difference between "signal is strong" and "signal is barely above noise". His received signal of 190 μV becomes roughly 6 μV on P.526 Fock and roughly 8 μV on P.368 GRWAVE — still above the 0.078 μV physics noise floor (Kleve → Derby is supposed to work on the globe; it is the operational Knickebein director beam that actually flew Midlands raids), but far from the 190 μV comfort zone his tool reports. ### His number is physically less than what the standards give at the radio horizon itself This is the cleanest way to see that Dan's 14.58 dB is wrong on its face. For his geometry (91 m TX, 6000 m RX), the radio horizon at K = 4/3 effective Earth radius is **358.6 km** (39.3 km TX horizon + 319.3 km RX horizon). Dan's path at 528 km is **169 km beyond** the horizon. Loss above free-space at three distances from Kleve: | d (km) | Relative to horizon | P.526 Fock (dB) | P.368 GRWAVE (dB eqv) | |---|---|---|---| | 358.6 | **at horizon** | **17.0** | **~11** | | 440 (Spalding) | 81 km beyond | 39.9 | ~31 | | 528 (Dan's Derby) | 169 km beyond | **64.8** | **~53** | **At the radio horizon itself**, before you have gone a single kilometer into the diffraction shadow, P.526 Fock already gives 17 dB of loss above free space and P.368 GRWAVE gives about 11 dB. Dan's 14.58 dB at 528 km is **less than what P.526 Fock gives at 358 km**, which is **170 km short of his actual path length**. Diffraction loss is monotonically increasing with distance past the radio horizon on every propagation model in existence. Dan is reporting less loss at 528 km than the standards give at 359 km. This is physically incoherent; his number cannot be what any correctly configured globe-model propagation code produces for Kleve → Derby at 31.5 MHz. The most likely proximate cause visible in the screenshot itself: his "Ground elevation" field for Marble Hall reads **-1.460474 m**, which is **exactly** his longitude value **-1.460474°** with the wrong units. That is a strong indicator that his tool never loaded a real elevation for the receive endpoint and some fallback kicked in. A tool running on bad geometry will produce a number with no relation to the physics of the path. ### The unknown receiver sensitivity Dan's screenshot reports a 45.57 dB fade margin that depends on his assumed 1 μV (-107 dBm) receiver sensitivity. That number is Dan's own uncited guess; no primary source quotes a bench microvolt MDS for the FuBl 2 receiver. I worked through five candidate primary sources during this research session: 1. **D.(Luft) T.4065** — Luftwaffe FuBl I (EBl 1) Geräte-Handbuch, April 1942, §I.D.3 "Empfindlichkeit". Specifies sensitivity **operationally**: *"reliable optical and acoustic indication of the approach course at 200 m flight altitude for 30 km distance with 500 W AFF"*. Does **not** give a bench microvolt number. 2. **D.(Luft) T.4058 §21** — Luftwaffe FuBl 2 Geräte-Handbuch, 1943. Same operational format, 70 km at 500 W / 200 m altitude. Does **not** give a bench microvolt number. 3. **Thote, *Radiobote* Heft 9, May-June 2007, p. 23** — dedicated article on FuBl I and FuBl 2. Paraphrases T.4065 §I.D.3 and identifies the FuBl 2 receiver as E Bl 3F / E Bl 3H *wesentlich leistungsfähiger Überlagerungsempfänger* (significantly more capable superhet), but gives no bench sensitivity number. 4. **Felkin A.D.I.(K) Report 343/1945** — British wartime interrogation report §9, July 1945. Confirms in plain English that *"the Fu Bl 2 is the blind landing apparatus composed of the E Bl 2 and E Bl 3 receivers, the latter giving 3½ channels between 30 and 33.3 mc/s"*, which is the primary-source anchor for identifying the Knickebein receiver as the E Bl 3. Does **not** quote a bench MDS. 5. **AP 1186 Vol I Section 3 Chapter 7** — RAF Signal Manual for the British Standard Beam Approach R.1124A / R.1125A receivers, December 1938. The British Lorenz-class cousin of the FuBl 2 is documented as a 6-valve superheterodyne in the 30.5 to 40.4 Mc/s band with a 7 Mc/s IF. The alignment procedure uses a test oscillator and flight testing, not a bench MDS spec. Does **not** quote a microvolt sensitivity. **The pattern across all five primary sources is the same**: both services of the era specify their VHF blind-landing and navigation receivers operationally, not as bench microvolt MDS. This is the engineering culture of the period, not a gap in my reading. So the answer to "what operational guarantee threshold should the Stollberg paths be compared against" is **we do not have one from primary sources**, and the null doc correctly falls back to the only receiver-independent bound available: the 0.078 μV thermal + galactic physics noise floor from ITU-R P.372-16 at 500 Hz MCW detection bandwidth. Dan's 1 μV / -107 dBm number is his own uncited guess. Even if it were right as a bench MDS, it would not change the Stollberg falsification, because both P.526 and P.368 place the Stollberg → Midlands signals at 0.001 to 0.14 μV — below the physics noise floor at 0.078 μV — so below any receiver sensitivity regardless of spec. The receiver-sensitivity argument is not the load-bearing bound on the null doc; the noise floor is. ### Fade margin is not a range extender On stream Dan claimed in substance that the 45.57 dB fade margin on his link-budget screenshot meant "you can take this out to thousands of miles". This is wrong at the level of what "fade margin" means in a link budget. **Fade margin** is the difference between the mean received signal on one specific path and the receiver's sensitivity threshold on that path. It is insurance against *variable* losses on that same path — atmospheric ducting, multipath cancellation, rain attenuation, tropospheric scintillation, antenna motion, aging components. It is **not** a budget you can "spend" on additional distance, because going farther on the path produces new mean loss from both free-space spreading and diffraction, and that new loss comes straight out of the fade margin. Applying Dan's own 45.57 dB fade margin to his 528 km geometry on P.526 Fock: - **From 528 km to 1000 km**: additional FSPL is $20 \log_{10}(1000/528) = 5.56$ dB; additional P.526 diffraction loss at this frequency and geometry is about 0.2 to 0.3 dB/km in the deep beyond-horizon regime, so roughly 100 to 140 dB over the extra 472 km. Total new loss well over 100 dB. Fade margin available: 45.57 dB. **Link fails by 50+ dB at 1000 km.** - **From 528 km to 2000 km ("thousands of miles")**: additional FSPL is 11.57 dB; additional P.526 diffraction loss over the extra 1472 km is of order several hundred dB. **Link fails by hundreds of dB at 2000 km.** A 45.57 dB fade margin at 528 km buys perhaps 150 to 300 km of additional range on the globe at 31.5 MHz before both FSPL and Fock diffraction loss have eaten the entire margin. "Thousands of miles" is not possible on any propagation model that includes curved-Earth diffraction, which is every standards-track VHF propagation recommendation the ITU publishes. And crucially: **the null doc's falsification is at Stollberg → Midlands, not Kleve → Derby**. Dan's fade margin was computed at Kleve → Derby (a path that works on the globe — it is the operational Knickebein director beam). That number is irrelevant to whether Stollberg can reach Beeston. You do not "apply" the fade margin of one path to another path. You run the link budget for the longer path separately and see what fade margin you get. For Stollberg → Beeston (694 km, 72 m TX, 6000 m RX), P.526 Fock gives 0.028 μV at the receiver input, which is 17 dB below the 0.078 μV physics noise floor — a **negative** fade margin. No amount of Kleve → Derby fade margin rescues Stollberg → Beeston. ### Summary Dan's headline numerical claim, the 14.58 dB obstruction loss and 190 μV received signal at Derby, is **not a result that either of the ITU-R standards-track globe propagation models produces for his stated geometry**. P.526 Fock gives approximately 65 dB of loss (6 μV received) and P.368 GRWAVE gives approximately 53 dB (8 μV received) at Kleve → Derby. His number is less than P.526 gives at the radio horizon itself 170 km short of his path length. And even the fade margin he did compute would not, if correctly understood, support his "thousands of miles" claim on the globe — it would buy roughly 150 to 300 km, and it is irrelevant to the Stollberg → Midlands paths where the null doc's falsification actually lives. The null hypothesis stands: Stollberg → Midlands is below the 0.078 μV physics noise floor on both P.526 Fock and P.368 GRWAVE at every confirmed Midlands target, regardless of any receiver-sensitivity assumption and regardless of any fade margin on any other path. Dan's link-budget screenshot does not overturn this. The complete analysis lives in [[GRWAVE_P368_BotB]]: per-station distance sweeps (Kleve, Stollberg, Telefunken) and the master bar chart, all run through [[2025_ITU-R_P526-16_Diffraction|ITU-R P.526-16]] Fock smooth-Earth diffraction and ITU-R P.368 GRWAVE using the ITU-maintained LFMF-SmoothEarth reference calculator, alongside the Friis free-space baseline and the Sommerfeld-Norton plane-Earth cross-check. Every number in this document is sourced from that analysis, and the source code (`botb_itu_analysis.py`, `make_p526_vs_p368_graphs.py`) is included so the computation is auditable end to end. # Conclusion from GRWAVE & FOCK Diffraction ![[ITU_Calc_sn_vs_grwave_uv_master_bargraph.png]] #### Master bar chart with 2 μV Dan-comparison line ![[ITU_Calc_sn_vs_grwave_fubl2_master_bargraph.png]] ![[ITU_Calc_sn_vs_grwave_uv_kleve.png]] ![[ITU_Calc_sn_vs_grwave_uv_stollberg.png]] ![[ITU_Calc_sn_vs_grwave_uv_telefunken.png]]