The Belgian Wave

The Night view is 3D because the subject is an engagement in space: two fighters two nautical miles apart, a ghost that jumps thousands of metres between radar sweeps, and the geometry of one radar finding the other. One scene unit is one metre: the pair flies its racetrack over a plain glowing with Belgium's lit motorway net, the star field overhead is computed for that exact night, and the ghost is drawn as a wireframe on purpose, because it is a radar track and not a thing anyone saw. Press Run and the 23:00 to 01:10 clock plays in a minute: the first attempt, the 00:13 jump, the wingman lock, the last lock, the quiet.

The Tape view is the recorded record, drawn as recorded: the 00:13 lock's speed and altitude channels with their published points, the 22.5-second lock that flew straight and then 'evaded', and the thirteen lock durations from 45.9 seconds down to one tenth. The dashed zero line matters: some altitude readings fall below it, and no craft flies under Belgium. Toggle the sceptical annotations and each episode wears its artifact reading.

The Court view is three shelves of closed-form arithmetic on the tape's own numbers, with no fitted constant anywhere: the Mach test and the boom carpet nobody heard, the jump priced in g and gigawatts with the step time as a dial instead of a secret, and the vertical channel read as motion, a supersonic dive over towns in silence.

The Sky view rebuilds the star field over Ramillies for that night, hour by hour, exactly: the sceptics' whole identification set, Sirius flashing colours low in the south-west, Jupiter at magnitude -2.1 sinking slowly toward the horizon, each object's dotted arc showing its stately all-night drift. What stood where the lights were reported is a computation; whether a given witness watched a given star stays a reading, and the bench says so.

The Ledger view is the wave on nine cards: Eupen and its thirteen gendarmes, the counts, the night, the tape, the Air Force's own analysis, the ground reports and the 143 letters, the press conference, the confessed photograph, and the stealth denial. It closes on two verdicts at equal size: the tape, and the wave.

This bench was called into existence by the casework line itself. The File view of INST-52, the Stealth Bench, carries a Belgium card that kills the black triangles' most convenient attribution on the spot: the United States denied, on the record, flying over Belgium, and a faceted penetrator cannot do the low, slow, silent hover the reports describe. With the attribution dead, the real question is exposed: does the radar evidence that made this wave stand up on the world's evening news survive arithmetic. This station's wiki entry gives the wave one paragraph; this instrument puts its hardest night on the table whole.

It also picks up two older threads. INST-30, Washington 1952, taught how a radar sees things that are not there inside a temperature inversion; Belgium's night had a documented inversion at 3,000 ft, the same machine running again under a different flag. And INST-53, the Shootdown Bench, is this instrument's sister: the same F-16s, the same lock-ons, except that in 2023 the missiles actually flew, while Belgium in 1990 never produced anything to shoot at. Read side by side, the phrase "confirmed on radar" weighs what it honestly weighs.

Three exact shelves and zero fitted constants run everything. The panel tags every number exact, modelled, reported or reading.

M = v / √(γRT(h)) · w = 2h·√(M²−1) · a = Δv/Δt · P = m·a·v · az/el = f(GMST(JD), α, δ, φ, λ)

The atmosphere, exactly. The International Standard Atmosphere gives temperature, and with it the local speed of sound, in closed form at any altitude on the dial. Every recorded speed becomes a Mach number the reader can check, at altitude and at the ground, and the ground number is the one no refraction profile can argue with.

The boom, geometrically. A supersonic aircraft drags a Mach cone whose ground intersection is the boom carpet, width 2h√(M²−1) in still air. At the tape's 970 knots and 5,000 ft that is about 3.4 km of carpet laid along every supersonic kilometre of track, across Walloon Brabant, at midnight. The silence of that night is carried as reported, and it is the court's loudest exhibit.

The jump, priced. The tapes' public record gives endpoints, not step times, so the bench puts the step time on a dial and computes dv/dt live: 2 s reads 21.5 g, 1.1 s reads the cited 40 g. P = mav prices the top of the step for any dialled mass; a false echo, gate transfer or interference product pays none of it.

The sky, solved. GMST from the IAU 1982 expression, verified in the tune script against the canonical J2000 anchors; the bright-star catalogue rotated to alt-az over Ramillies; Jupiter from Standish mean elements, verified to stand in Gemini at magnitude -2.1 that night. The scintillation that turns low bright stars into colour-changing lights is animated over exact positions.

The chase, staged honestly. The reported structure is replayed: two ships 2 NM apart, nine attempts, six locks, three simultaneous, one of them the wingman. The coordinates are invented staging, because the real tracks were never published, and the instrument prints that split on its face.

What is deliberately not computed. No propagation solution for the angels (INST-30 holds that machinery), no claimed identification of any witness's light, no verdict on the wave's unexplained residue. Where the record is silent or contested, the bench carries a label instead of a guess.

The Mach numbers, the boom geometry, the jump kinematics and the star positions are exact and auditable; the chase staging, the tape interpolation and the scintillation are labelled models; the chronology, the tape's numbers, the studies, the counts, the confession and the denials are reported as documented; and what the wave's surviving testimony amounts to is a reading, for which the bench hangs two verdicts at equal size and adds no third.

A timeline, a dt dial, a mass, a speed, an altitude and an hour: between them they hold the whole public argument about the Belgian wave's hardest night.

• The timeline, with Run. The 23:00 to 01:10 clock compressed to a minute, with the record's fixed moments marked: 00:07 first attempt, 00:13 the jump, 00:29 the wingman lock, 00:40 the last lock. Scrub anywhere.
• The dt dial. The single most argued-about number of the wave, 40 g, is not on the tape: it is a reading of the step time. This bench refuses to type a dt in secret. Slide it and watch the g follow, exactly.
• Speed and altitude. The court's Mach test, fed the tape's own band, 150 to 1,010 knots, 11,000 ft down to the deck. Above roughly 660 knots at low level the ground-Mach test starts owing Wallonia a boom.
• The mass. P = mav needs one, so it is a dial, not an assumption: any real mass pays gigawatts at the top of the step; an artifact pays nothing. That asymmetry is the whole shelf.
• The hour. 21:00 to 04:00 CEST over Ramillies. The witness set rises, scintillates and sinks west on rails the tune script verifies; the dotted arcs are what 'it moved slowly all night' looks like when stars do it.
• The sceptical annotations. Off, the bench shows the tape as the 1990 press conference showed it. On, each episode wears the artifact reading the later analyses gave it: gate transfers, angels, mirrors, the wingman.

Six claims that orbit the Belgian wave, from the Air Force's own study to the reading that circled the world, with who made each and where it lands once the arithmetic is done.

1. Press Run and watch the whole night once. Note what never happens: the pilots never see anything. Six locks, nine attempts, zero visuals. That asymmetry is the entire case, and the bench draws the ghost as a wireframe to keep it in view.
2. At the 00:13 lock, watch the streak. The track crosses kilometres between sweeps, dives at the ground, and prints readings below it. Then open the Tape view and find the same seconds as a stepped line on paper.
3. In the Court, slide dt from 2.0 s down to 1.1 s. The g readout walks from 21.5 to 40: both famous numbers, one pair of endpoints, and the dial is the only difference. Nothing about the wave is a cleaner lesson in how a headline number gets made.
4. Slide the speed to 660 kt and watch the ground Mach cross 1. Everything the tape records above that line owes Belgium a boom. The tape spends its most famous seconds well above that line, and the night stayed silent.
5. In the Sky, drag the hour from 23:00 to 03:00. The whole witness set sinks slowly west, scintillating, exactly the slow stately drift the ground reports describe. Then remember not one call came in during the chase, and 143 letters came after the headlines.
6. Toggle the sceptical annotations during the wingman phase. The Air Force's own study concluded one triumphant lock was pointed at the lead F-16. Watch the cone swing between the jets and decide how much of the rest of the tape you would now bet on.
7. End on the Ledger's two verdicts. They do not fight. The tape refutes itself; the wave's testimony stands unexplained and unattributed. Holding both at once is what this file actually supports, and it is rarer than either alone.

The honest line between what is exact, what is modelled here, what the record reports, and what is a reading:

In one line: the ISA atmosphere, every Mach number, the boom-carpet geometry, the jump's dv/dt and P = mav, and the star field over Ramillies are EXACT and machine-checked, with no fitted constant anywhere; the chase staging, the tape interpolation and the scintillation are MODELLED and labelled; the chronology, the tape's numbers, the Gilmard-Salmon findings, Meessen's analyses, the 143 letters, the confession and the denials are REPORTED as documented; and what two thousand sincere reports amount to once the exhibits fall is a READING, for which this bench hangs two verdicts at equal size and adds no third. Fly the night, dial the dt yourself, and decide.

• The Belgian Air Force summary report on the night of 30-31 March 1990 (English translation at ufologie.patrickgross.org): the chronology from the 23:00 gendarmerie call to the 01:10 recovery, nine interception attempts and six lock-ons at 5-8 NM, recorded speeds 150 to 1,010 kt, the 3,000 ft temperature inversion, and the exclusion of balloons, microlights, RPVs, aircraft and laser projections.
• Maj. Gen. (ret.) Wilfried De Brouwer's chapter in Leslie Kean, UFOs: Generals, Pilots and Government Officials Go On the Record (2010): the Eupen chronology and the thirteen gendarmes at eight locations, the 00:13 lock (150 to 970 kt, 9,000 to 5,000 to 11,000 ft), the pilots' names, the wave counts (over 2,000 reports, ~540 unexplained), the DIA memo quote and the 1993 Delcroix letter.
• The Gilmard-Salmon electronic-warfare analysis (Belgian Air Force, 1992), as summarised in the public record: anomalous-propagation angels in the inversion, the 00:29-00:31 lock identified as the wingman locking the lead, and below-zero altitude readings as ground mirror images.
• Auguste Meessen, "Analyse approfondie des mystérieux enregistrements radar des F-16", Inforespace (December 1998), and his Bertem ground-radar study: acceptance of the angel hypothesis for many contacts, two-radar interference proposed as a mechanism for the extreme Doppler speeds, some 180 hours of ground tape analysed, and the one trace he retained as unexplained.
• Tim Printy, "Belgium 1990: A case for radar-visual UFOs?" (astronomyufo.com): the tabulated lock durations (45.9 s down to 0.1 s), the winds-aloft and drift analysis, and the star identifications for the night's ground reports (Sirius, Procyon, Rigel, Betelgeuse, Jupiter, Castor, Pollux, Aldebaran, Capella).
• Jean-Michel Abrassart (2019) on the simultaneous locks as mutual locks with Bragg scattering for the remainder, carried on this bench as the sceptical extension of the BAF finding; Marc Hallet, La Vague OVNI Belge (1992); Pierre Magain and Marc Remy, Physicalia Magazine 15(4), 1993 (bibliographic record).
• The COBEPS statistics paper (Franck Boitte): the independent COB catalogue of 1,282 notifications and 2,531 witnesses across 1989-1993, and the wave historiography after 2011.
• The Petit-Rechain confession: Patrick Maréchal on RTL-TVI, 26 July 2011 (VRT News coverage), a painted polystyrene triangle with corner lamps suspended on threads; the preceding two decades of optical analyses (Louange, Acheroy, Marion); and COBEPS's 27 July 2011 response that one faked photograph does not downgrade the wave.
• The stealth denials: the declassified DIA memo "Belgium and the UFO Issue" recording the USAF's confirmation that no stealth aircraft operated in the Ardennes area, and Defence Minister Leo Delcroix's 1993 letter; contemporary reporting on the wave and the 11 July 1990 Evere press conference.
• Radar hardware: the Westinghouse AN/APG-66 pulse-Doppler radar of the Belgian F-16As (IEEE, "F-16 Pulse Doppler Radar (AN/APG-66) Performance", 1983; radartutorial.eu): medium-PRF downlook, 64-point FFT filter bank, CFAR thresholding, single-target-track gates, the artifact modes (gate capture and transfer, multipath, mutual interference, anomalous propagation) that the sceptical readings invoke.
• Physics: the International Standard Atmosphere; Mach and sonic-boom carpet geometry (any gas-dynamics text); GMST per the IAU 1982 expression (Meeus, Astronomical Algorithms); the J2000 bright-star catalogue; Standish mean Keplerian elements for Jupiter. All constants are solved and checked in scripts/belgian-wave-tune.mjs, which is re-runnable and prints every number the instrument is allowed to carry. This bench fits no constant anywhere.
• Cross-referenced instruments on this site: INST-52, the Stealth Bench, whose Belgium card sent us here; INST-30, Washington 1952, for what an inversion does to a radar picture; INST-53, the Shootdown Bench, for the same fighters with live missiles; and the station's wiki entry on the Belgian wave.