The Fauville-Évreux Radar Sighting

Theoretical Alignment

The case aligns weakly with the Extraterrestrial Hypothesis: it offers a rare instrumented radar-visual pairing, but the radar return itself was faint (1/5 intensity), came from a single uncorroborated station, and carried no shape, size, or acceleration data — nothing in the record describes performance beyond a slow, roughly linear 30–40 knot drift. It aligns more strongly, if still moderately, with Psycho-Social framing once the evidentiary gaps are weighed: a single officer's account stands in for what was reportedly a two-witness event, the corroborating "vigie" testimony was never separately recorded, and GEIPAN's own file flags the radar contact as possibly a false echo. Interdimensional framing finds no support anywhere in the primary record.

Background

The Fauville-Évreux sighting sits at the intersection of two institutional histories that were themselves both barely underway in the spring of 1978: France's newly created state UAP investigation unit, and a French Air Force air base whose primary mission had nothing to do with radar surveillance or air defense. Understanding the case at all — why it was investigated the way it was, why certain questions were asked and others were not, and why the surviving record looks the way it does — requires understanding both institutions in some detail first.

Évreux Air Base 105: From Pilot School to Cold War Airlift Hub

Évreux Air Base 105 (Base aérienne 105 d'Évreux-Fauville, ICAO code LFOE) sits a few kilometers east of the city of Évreux (3 km by English Wikipedia, 7 km by French) in the Eure department of Normandy, adjoining the small commune of Fauville — population in the low hundreds even today — from which the base and this case both take half their name. The airfield's military history runs back to 1912–1913, when it began as a civilian aviation ground, before the Air Ministry formally opened a military pilot training school there in 1937, relocating the airfield itself nearer to Fauville in the process. The base then endured a genuinely violent Second World War: occupied and used by the Luftwaffe from 1940, it was heavily bombed by Allied aircraft in December 1942, April 1943, and again in August 1944, before German forces finally withdrew on August 18, 1944 as Allied ground forces advanced through Normandy.

Postwar, the base's identity shifted entirely. Reconstruction began in 1952 under NATO's basing negotiations, with the airfield rebuilt to the era's standard "marguerite" (daisy) dispersal pattern — hardstands arranged with roughly 150 feet of separation between parked aircraft, a Cold War design specifically meant to limit the damage a single nuclear or conventional strike could inflict on a concentrated air wing. Évreux then operated for the next decade and a half as a U.S. Air Force-tenanted NATO installation: the 465th Troop Carrier Wing arrived in May 1955 flying Fairchild C-119 "Flying Boxcar" transports across three squadrons, earning Air Force Outstanding Unit Awards in 1956 for humanitarian airlift work including relief flights supporting United Nations forces in the Middle East and the evacuation of Hungarian refugees following that year's failed uprising; the 317th Troop Carrier Wing relocated from Germany in April 1958, eventually transitioning to Lockheed C-130A Hercules aircraft; and the 322nd Air Division, established August 12, 1955, coordinated broader European airlift operations from the base until 1964. At the height of this period, more than 9,000 American military personnel were stationed at Évreux — a scale of foreign military presence that is easy to forget when looking at the comparatively modest French transport base that occupied the same runways in 1978.

That change came in 1966, when President Charles de Gaulle announced France's withdrawal from NATO's integrated military command structure (while remaining a NATO member state in the broader political sense). American forces departed Évreux by April 1967, and the base passed back to full French Air Force control, formally redesignated BA 105. From that point on, Évreux became — and remains today — a French tactical airlift hub, never a fighter-interceptor base and never a dedicated air-defense radar station. Initial French operations flew the Nord N2501 Noratlas, the same twin-boom, twin-engine transport type that would still be in frontline service there more than a decade later, on the night of this sighting. In the autumn of 1978, the base's Escadre de transport 64 comprised three Noratlas-equipped squadrons — Escadron de transport 1/64 Béarn, 2/64 Anjou (which had just arrived to reinforce the base on June 13, 1978, two months after this sighting), and 3/64 Bigorre — alongside the mixed Groupement aérien 1/56 Vaucluse, itself flying a diverse fleet of Aérospatiale SA-330 Puma helicopters, de Havilland Canada DHC-6 Twin Otters, and further Noratlas airframes. The base would not receive its now-signature Transall C-160 tactical transports in large numbers until later years; in April 1978, the flight line at Évreux was still overwhelmingly a Noratlas operation.

This detail matters directly for understanding what kind of radar the base's duty officer actually had access to that night: it would have been an airfield approach/control radar serving cargo-transport traffic patterns, not a purpose-built air-defense search radar feeding into France's national STRIDA network the way a dedicated Centre de Détection et de Contrôle (CDC) station like Cinq-Mars-la-Pile would have been. That distinction — a transport base's own local radar picking up an anomalous contact, versus the national air-defense radar chain independently confirming it — sits at the center of why this case was never more thoroughly corroborated, a point developed at length in the Detailed Narrative below.

GEPAN's First Year: A Very New Institution

On the investigative side, the sighting arrived at almost exactly the moment France's Groupe d'études des phénomènes aérospatiaux non identifiés (GEPAN) was completing its first full year of operation. CNES astrophysicist Claude Poher had spent much of the mid-1970s lobbying CNES leadership to create a permanent, scientifically staffed state body for investigating UAP reports, building his case in part on a statistical study of the UFO phenomenon that he presented to the American Institute of Aeronautics and Astronautics in 1975 — an unusual move for a subject more commonly relegated to the fringes of aerospace institutions, and a sign of how deliberately Poher tried to frame the proposal in mainstream scientific terms rather than as fringe research. That advocacy ultimately won the backing of CNES director general Yves Sillard, and GEPAN was created in 1977 (CNES dates it to 1 May 1977), giving France a permanent state body, attached to a recognized scientific agency, specifically to study UAP reports on an ongoing basis.

For its first months GEPAN was, in the most literal sense, Poher and a secretary, doing what statistical and case-file work they could manage with essentially no dedicated staff, though with informal, off-the-books help from other CNES personnel including a young engineer named Jean-Jacques Velasco, who would go on to direct GEPAN's successor unit SEPRA through the 1990s. Even in that skeletal first year, GEPAN moved quickly to establish a real intake pipeline: 1977 records show the unit already processing statistics on more than 300 reports transmitted through the Gendarmerie nationale, which had itself been tasked since 1974 with collecting UFO-related witness statements from the public on the state's behalf. By 1978, the unit had grown to roughly ten members, formally overseen by a seven-person scientific council that included physicist Hubert Curien — already CNES's own president at the time, and eventually France's minister of research under President François Mitterrand — lending the young unit the direct oversight of its parent agency's own leadership even while its day-to-day operational staff remained tiny by any modern standard.

April 1978 was, in fact, an unusually active month for Poher personally, which is worth noting simply to establish how genuinely busy GEPAN's caseload already was barely a year into its existence. That same April, Poher personally led a four-person GEPAN team to Cussac, Cantal, to reconstruct a decade-old 1967 close-encounter case involving child witnesses on a remote plateau — a labor-intensive field investigation running in parallel with whatever attention the Fauville-Évreux report was simultaneously receiving back at CNES's Toulouse offices. The Fauville-Évreux sighting therefore landed on the desk of an organization that was still defining its own working methods in real time, juggling a genuinely active caseload with a handful of staff — a formative-era case in the most literal sense, and one of the earliest instances of GEPAN applying an instrumented radar cross-check to a military-sourced report rather than a purely testimonial civilian sighting.

Complete Timeline

Time (UTC)	Event
1977	GEPAN founded. CNES sets up the Groupe d'études des phénomènes aérospatiaux non identifiés in Toulouse under Claude Poher, with the backing of CNES director general Yves Sillard (CNES dates its creation to 1 May 1977).
Sometime before 20:35, Apr 23, 1978	A telephone call reaches the operational duty officer at Évreux Air Base 105, prompting him to look outside.
20:35 UTC	Visual observation begins. The duty officer sees a red point of light at 10–15° elevation, azimuth approximately 300° from the base. Lookout ("vigie") personnel also see the phenomenon.
~20:35–20:44 UTC	The light's luminous intensity is observed decreasing over several minutes. At some point in this window the officer requests radar assistance.
20:44 UTC	Radar contact. Évreux's own radar detects a weak echo (intensity 1/5) at bearing 300°, 18 nautical miles out, moving toward bearing 330° at an estimated 30–40 knots.
20:44–20:51 UTC	The radar echo is tracked continuously for approximately 7 minutes, in step with the ongoing visual observation.
20:51 UTC	Observation ends. GEIPAN: "20h51 TU marque la fin des observations visuelles et l'arrêt du radar" — the end of the visual observations and the stopping of the radar.
April 24, 1978	The duty officer writes his report the next day. The radar observation report gives the 24th as the date of observation, which GEIPAN treats as a typing error, allowing the visual sighting and the radar trace to be taken as simultaneous.
1978 (year, exact date not on public record)	GEPAN investigation. Analysts test and reject the helicopter and balloon hypotheses, calculate Venus's actual position for the exact time/date/location and find it does not match the reported bearing/elevation, and flag the radar echo itself as possibly, though not confirmedly, a false echo.
December 30, 1978	Claude Poher resigns as GEPAN's director, succeeded by mathematician Alain Esterle — closing out the unit's first, foundational year, of which this case is a part.
1988	GEPAN is reorganized as SEPRA (Service d'expertise des phénomènes de rentrée atmosphérique).
2005	SEPRA is reorganized again as GEIPAN (Groupe d'études et d'informations sur les phénomènes aérospatiaux non identifiés), which digitizes and republishes the historical GEPAN case archive — including this case, catalogued as case ID 1978-04-01879 — for public online access.
Present day	The case is published on GEIPAN's site (case 1978-04-01879, testimony 2096) as class D1, with the record last updated on September 29, 2015; geipan.fr has been offline in 2026, and this file relies on Internet Archive copies.

Detailed Narrative

A Call to the Duty Officer

The available record of this case begins, deliberately, with an institutional detail rather than a dramatic one: the duty officer did not simply notice something odd in the sky on his own initiative. GEIPAN's own case summary specifies that he was "responding to a phone call" (répondant à un appel téléphonique) when he first looked outside and saw the light. The call's source and content are not part of the public record — whether it came from the lookout post itself, from another section of the base, or from a member of the public is not stated — but its presence in GEIPAN's own summary means the sighting did not originate as a single individual's private impression. Someone else, prior to and independent of the officer's own observation, had already noticed something worth telephoning about. This is a small detail, but it is a genuine evidentiary point often missed in secondhand retellings of the case: the officer was, from the very first minute, the second person to notice the phenomenon, not the first, even though he became the only person whose account was ever formally written down.

The witness himself is identified in GEIPAN's structured metadata only by function and gender — "opérateur de permanence," male — consistent with GEIPAN's standing practice of anonymizing individual witnesses in its public case files, especially military personnel reporting in a professional capacity. His role placed him, by definition, on duty and responsible for the base's operational awareness during the shift in question; a "duty officer" at a French Air Force transport base in this era would ordinarily be tasked with monitoring communications, coordinating with air traffic control elements, and responding to any operational anomaly, which is precisely the professional context that led him to request a radar check once the visual observation had continued for several minutes rather than dismissing it immediately.

The Visual Observation, 20:35–20:51 UTC

At 20:35 UTC (22:37 local French time, which in late April uses Central European Summer Time, UTC+2), the officer observed what GEIPAN's structured record describes precisely: a single, point-like ("1D — ponctuelle") object of red color, positioned at 10–15 degrees of elevation above the horizon, on a bearing of approximately 300 degrees from the base — roughly northwest. Its apparent size was recorded as comparable to a planet ("estimée à celle d'une planète"), meaning it presented no resolvable disc, structure, or extended shape; from the base, it looked like an unusually placed bright star or planet, distinguished from the ordinary night sky primarily by its color, its position, and — critically — by the fact that its brightness was observed to fade over the course of the sighting.

That fading intensity is one of the case's more specific and harder-to-dismiss details. A fixed star or planet does not meaningfully change in brightness over a sixteen-minute observation window under normal atmospheric conditions; a conventional aircraft's navigation or landing lights do not typically fade in this way either, though a receding light source, an object banking away from the observer, or a light source being progressively obscured by the evening's "quelques éclaircies ou nuageux" (partial clearing or cloudy) conditions could all plausibly produce a gradual dimming. GEIPAN's environmental metadata for the case explicitly records the sky as partly cloudy with intermittent clearing that evening — a condition that cuts in two directions at once for later analysis: it is compatible with a genuinely fading distant light source of unknown origin, but it also means the viewing conditions were not the pristine, cloudless sky that a fully confident astronomical identification would ideally require.

The observation lasted, by both the visual and radar record, from 20:35 to 20:51 UTC — sixteen minutes total, of which the radar tracked the final seven. Sixteen minutes is a substantial observation window by the standards of most UAP reports, which frequently run to seconds or at most a few minutes; a witness given sixteen minutes to study a light's position, color, and brightness trend has considerably more opportunity to notice genuine behavioral detail (and, weighed the other way, considerably more opportunity for ordinary explanations like a slowly moving aircraft, a descending flare, or an astronomical body sinking toward the horizon to plausibly account for what was seen) than a witness reporting a five-second flash.

The Vigie's Silent Corroboration

GEIPAN's case summary states plainly that lookout personnel — "la vigie," the base's watch post — also observed the phenomenon ("Le personnel de la vigie observe également le phénomène"). This detail deserves its own sustained treatment, because it represents a specific and recurring evidentiary problem in military UAP reporting generally, not just a footnote to this one case: a second, independent set of eyes confirmed the sighting in real time, and yet only one account, the duty officer's, was ever converted into the kind of formal, individually attributed testimony that GEIPAN's investigative and archival process is built around.

The reasons for this gap are not spelled out in the public case file, but the operational logic is not hard to infer. A base's lookout watch exists as a standing function, not a named individual whose statement gets separately taken down each time something is reported; whoever staffed the vigie that night was, in institutional terms, part of the base's routine watch-keeping apparatus rather than a person whom anyone thought to formally debrief once the more senior duty officer's report was already being filed. The effect, whatever the cause, is that this case's real evidentiary base rests on more independent observation than its formal documentation reflects: GEIPAN's own file effectively concedes a second corroborating witness existed while being unable to present that witness's account as separately verified testimony with its own detail, phrasing, or possible discrepancies from the officer's version.

This is a pattern worth naming precisely because it recurs, in different specific forms, across military-sourced UAP reporting more broadly: chains of command and institutional reporting procedures are generally built to produce one authoritative account per incident, funneled through whichever individual holds the relevant reporting responsibility, rather than to independently canvass every person present. A commanding officer's report stands in for a flight crew's collective observation; a single logged radar contact stands in for however many controllers were watching the same scope. When the institutional record consolidates in this way, genuine corroboration — a second, third, or fourth person who saw the same thing — can exist in reality while remaining functionally invisible to the outside researcher relying only on the surviving paperwork. GEIPAN's own file is, in this instance, unusually candid about the gap rather than papering over it: it states the vigie saw the object too, even though it cannot substantiate that claim with the vigie's own words. That kind of institutional honesty about an evidentiary limitation is itself a small positive data point for the file's overall credibility, even as the missing corroborating statement is a real limitation on how much independent weight the sighting can bear.

Requesting Radar Support: Évreux's Scope and the STRIDA Network

At some point during the visual observation — GEIPAN's summary does not specify the exact minute, only that it happened before the 20:44 UTC contact — the duty officer requested that the base's radar be used to check for a corresponding return. This decision, ordinary as it might sound, is the single action that transformed this case from an uncorroborated visual sighting into a radar-visual case, and it is worth being precise about exactly what kind of radar capability Évreux actually had to offer in April 1978, and how that capability related to the wider national system France had built for exactly this kind of airspace-monitoring problem.

France's national air-defense radar network in this era operated under the STRIDA system (Système de Traitement et de Représentation des Informations de Défense Aérienne), a semi-automated architecture developed from the early 1960s and linked with NATO's air defence ground environment. According to EPR 118's history of French air defence, it linked a chain of Centres de Détection et de Contrôle (CDCs) that were modernized in stages: Mont-de-Marsan moved into its "Antigone" room with STRIDA CAPAC II in 1969; Doullens went to STRIDA Visu 3 with the SAFARE radio system in 1971, when the North Zone operations centre moved there from Romilly; and Lyon Mont Verdun went into a STRIDA CAPAC III Visu 3 station at the end of 1973, with a Palmier radar on Mont Thou. From 1972, "Liaison 1" data links connected several of these centres with the air-defence operations centre. Each CDC was a node of national air defence, a different category of system from a single airfield's own radar.

Évreux Air Base 105 was not one of these stations, and was never intended to be. As established above, Évreux in 1978 was a tactical transport hub — home to Noratlas-flying squadrons under the Commandement du Transport Aérien Militaire — and its radar, whatever specific set it operated, would have been an airfield-level asset intended to support approach and departure control of the base's own transport traffic, not a STRIDA-network air-defense search radar with the range, sensitivity, or institutional connectivity of a dedicated CDC. This is not a criticism of the officer's decision to use it; on the contrary, requesting a check from whatever radar capability was actually on hand, immediately, while the visual sighting was still in progress, was exactly the right instinct for capturing corroborating data in real time, and it is precisely the kind of improvised but sound field decision that produced this case's one piece of instrumented evidence at all. But it means the resulting detection, however genuine, came from a radar system never designed or positioned to serve as an authoritative air-defense-grade confirmation of an unidentified contact — which is precisely why GEPAN's investigators would have wanted, and did not get, a cross-check from the nearest actual node of the national air-defense chain, at Cinq-Mars-la-Pile.

The 20:44 UTC Contact

Nine minutes into the visual observation, at 20:44 UTC, Évreux's radar detected an echo at bearing 300° — the same direction as the visually observed light — at a range of 18 nautical miles, moving toward bearing 330° at an estimated 30 to 40 knots. The echo's intensity was recorded as very weak: 1 on a 5-point scale. It was tracked for about seven minutes; GEIPAN records 20:51 UTC as the end of the visual observations and the stopping of the radar.

Several features of this contact cut in different directions. In favor of a genuine correlation: the radar bearing (300°) matches the visually reported azimuth, and the echo was tracked for seven minutes, a duration GEIPAN calls long. Against a confident reading: GEIPAN's phrasing — the end of the visual observations "and the stopping of the radar" — suggests the radar watch was ended when the sighting ended, so the matching end time is not independent confirmation; the 1/5 intensity is about as weak as a logged return can be, and GEIPAN says "on ne peut pas exclure un faux écho (par exemple réflexion secondaire)" — a false echo, for example a secondary reflection, cannot be excluded; the 30–40 knot speed is unremarkable. The contact was not confirmed by the other radar, at Cinq-Mars-la-Pile, discussed in the next section.

No recording or tape of the radar return survived into GEPAN's investigation or the modern GEIPAN archive. The entirety of what is known about the contact's parameters — bearing, range, heading, speed, duration, intensity — comes from the operator's written description in the case file, not from an independently reviewable radarscope photograph, magnetic tape, or printout. This is consistent with how ground radar operations at a 1978-era French transport base would ordinarily have functioned (routine air-traffic radar returns were not typically preserved as a matter of course once a shift ended, absent a specific reason to retain them), but it also means that everything asserted about the radar contact's specific numbers rests on the same single chain of testimony — the duty officer's own account — that the visual sighting itself rests on. The radar corroborates the direction and timing of the visual sighting; it does not independently corroborate itself.

Cinq-Mars-la-Pile and the Absent Second Opinion

Cinq-Mars-la-Pile, a commune in the Indre-et-Loire department near Tours, roughly 190 kilometers south-southwest of Évreux, has hosted a CDC (today CDC 07.927, call sign "Raki," covering north-western France) since 1965, when the centre moved there from the mobile "Ramolo" installation at Tours-Saint-Symphorien. EPR 118's history says it began at Cinq-Mars above ground, in a manual operations room with a vertical filter board and IP7A displays, served by an AN 110 radar and two AN 7B antennas; an AN 226 23-cm radar followed in 1968 and the CESAR system in 1973, and in 1980, two years after this sighting, the CDC moved into its underground facility to run STRIDA Visu 4. In April 1978 it was therefore an established air-defense control centre whose coverage extended over the general airspace around Évreux.

Today the centre, also called Camp Mailloux, monitors the northern half of French territory with a radar and five antennas outside the site and a staff of about 40 controllers and 60 mechanics working in shifts; its controllers train at the military school in Rochefort and the controllers' school at Mont-de-Marsan before a further course on site, according to a 2020s ici (formerly France Bleu) report. That describes the modern facility rather than its 1978 predecessor, but it illustrates the kind of professionally staffed capability the station represents.

GEIPAN's summary says that "la mesure radar d'intensité faible et non confirmée par l'autre radar de Cinq-Mars-La-Pile doit être considérée avec circonspection" — the weak radar measurement, "not confirmed by the other radar at Cinq-Mars-La-Pile," must be treated with caution. (An earlier version of this file quoted it as saying "station radar secondaire"; the summary says "l'autre radar.") It is important to be precise about what this does and does not establish. It does not say that Cinq-Mars-la-Pile looked for the target and positively found nothing, nor that it was never asked; it records only that no confirmation came from it. The summary does not say which of those scenarios applies, a genuine limit on how far this file can interpret the sentence.

What can be said with confidence is why investigators would have wanted that second station's input at all: Cinq-Mars-la-Pile's CDC, whatever its exact 1978 technical configuration, was a proper air-defense-grade STRIDA node with radar coverage over a wide swath of northern France, exactly the kind of independent, higher-capability system that could have either strengthened the case dramatically (an independent detection on a compatible bearing and time, from a totally separate radar site, would be very hard to explain away) or resolved it cleanly in the other direction (a definitive absence of any contact at all, on a properly functioning search radar covering the relevant airspace, would argue strongly that Évreux's own weak 1/5-intensity return was spurious). Neither happened. The case sits, instead, in the specific evidentiary limbo of a single, uncorroborated, low-confidence radar target — not confirmed, but also not affirmatively ruled out by the one system that plausibly could have settled the matter either way. This is, in miniature, exactly the same structural weakness that runs through the visual side of the case: a real, independent second source of confirmation appears to have existed in principle (another radar network covering the same airspace, just as a second human witness watched the same sky) without ever actually producing a usable, separately documented result.

A Clerical Error, Not a Second Date

The duty officer wrote his report the day after the sighting, on April 24, 1978. GEIPAN's summary says it is the radar observation report that gives the 24th as the date of observation, and that this "semble dûe à une erreur de frappe" — seems to be a typing error. On its face, a date discrepancy between an incident and its own paperwork is the kind of small inconsistency that can either be dismissed as trivial or treated as grounds for suspicion; GEIPAN treats it as a slip rather than evidence of a second, distinct event.

The reasoning, as GEIPAN states it, is that reading the radar report's date as a typo "permet à priori de valider la simultanéité de l'observation visuelle et de la trace radar" — allows the visual observation and the radar trace to be accepted, in principle, as simultaneous. Had the 24th been the real date of the radar trace, the radar would not corroborate the April 23 sighting at all. No second visual-plus-radar narrative for April 24 appears in the case file, so the simplest reading is one event on the evening of April 23; but the correction rests on an assumption about the radar report rather than on any surviving radar record.

GEPAN in Its First Year: Founding, Mandate, and Method

Understanding how seriously — and how — this case was actually investigated requires understanding what GEPAN was, structurally and methodologically, in the specific months surrounding April 1978. The unit was, by any standard, still in its institutional infancy. It had existed formally for barely thirteen months by the time of the Fauville-Évreux sighting, having been created in 1977 (on 1 May, by CNES's own history) following Claude Poher's multi-year advocacy campaign within CNES, built substantially on a statistical study of the UFO phenomenon he had presented to the American Institute of Aeronautics and Astronautics in 1975. For its earliest months, GEPAN's entire staff was, functionally, Poher and a secretary, with informal, off-the-books assistance from other CNES personnel including a young engineer named Jean-Jacques Velasco, who would go on to direct GEPAN's successor unit SEPRA through the 1990s and serve as an investigator on the far more physically evidenced Trans-en-Provence case in 1981. By 1978 the unit had grown to roughly ten members, formally overseen by a seven-person scientific council that included Hubert Curien, already CNES's own president at the time and eventually France's minister of research.

Crucially, GEPAN did not generate its own raw case intake independently; its reports arrived primarily through a standing arrangement with the Gendarmerie nationale, which had already been tasked since 1974 with collecting UFO-related witness statements from the public and forwarding them to CNES for scientific evaluation, a pipeline GEPAN inherited and formalized. This gendarmerie-fed, centrally analyzed intake model — a state law-enforcement body serving as first-contact witness interviewer, with a scientific unit doing the subsequent technical evaluation — was, and largely remains today under GEIPAN, structurally distinct from how most other countries' military or civilian UFO-investigation programs of the era operated, including the U.S. Air Force's own Project Blue Book, which folded in 1969 having relied far more on direct military-channel reporting and internal Air Force technical review rather than a national police intake system paired with an independent civilian space-agency evaluation unit. The Fauville-Évreux case, notably, did not arrive through that gendarmerie channel at all: it came directly from a French Air Force base's own internal reporting chain, a reminder that GEPAN's remit, even in its first year, extended beyond civilian gendarme-relayed sightings to include military-sourced reports investigated with the same methodology.

That methodology, as it existed in 1977–1978, is documented well enough — both by GEIPAN's own later technical publications and by contemporaneous French media coverage — to describe with some confidence. An October 1978 French television segment on GEPAN's work, later archived by France's national audiovisual institute (the Institut national de l'audiovisuel, INA), captured the unit's investigative posture at almost exactly the moment of this case: analysts and affiliated amateur investigator networks alike are shown applying site mapping, physical measurement, witness sketching and structured questionnaires, and systematic cross-referencing of witness accounts against known local air traffic and, where relevant, weather station data — the same general cross-referencing logic (checking a report's claimed position and timing against independently knowable facts about what else was actually in the sky or on the ground at that moment) that underlies both the Venus calculation and the Cinq-Mars-la-Pile radar cross-check performed for this specific case. GEPAN's structured, forms-based approach to witness intake — the ancestor of the detailed "questionnaire technique" GEIPAN still requires of every witness submission today — was already taking shape in this period, even though the unit's later, more elaborate astronomical-position calculation tools and multi-laboratory physical-evidence protocols (used to dramatic effect on Trans-en-Provence three years later) had not yet been developed to their mature form. In other words: this case was investigated using a methodology that was itself still being built, by an organization one year into its existence, working a case type — a radar-corroborated military sighting with no physical trace — that was different in kind from both the mostly rural civilian sightings that dominated GEPAN's early gendarme-fed caseload and the elaborate landed-craft investigations, like Trans-en-Provence, that would later define the unit's public reputation for rigor.

Modern GEIPAN, by contrast, operates a considerably more elaborate and standardized version of the same basic logic: a formal multi-tier classification system (A through D, with sub-classifications), a mandatory structured witness questionnaire required before any testimony is formally logged, dedicated astronomical calculation tools and published guidance directing witnesses and investigators alike toward tools like Stellarium or Ciel des Hommes for verifying celestial-object positions, and a fully digitized, publicly searchable online case archive — the very archive, in fact, through which this 1978 case file survives and remains accessible today. The core investigative instincts on display in the Fauville-Évreux file — check the astronomy, check the radar, check the conventional explanations, document what could and could not be confirmed rather than reaching for a premature conclusion in either direction — are recognizably the same instincts modern GEIPAN still applies four and a half decades later; what has changed most is the systematization, the digitization, and the sheer accumulated case-comparison base a modern investigator can draw on that a ten-person unit little more than a year old, in 1978, simply did not yet have.

The Venus Check

Because the observed object was a red, point-like light sitting low on the horizon (10–15° elevation) at dusk, and because Venus is by far the single most common real astronomical cause of exactly this category of UAP report, GEPAN's investigators specifically checked the planet's actual position for the precise date, time, and location of the sighting. This is not a speculative or approximate exercise: an object's azimuth (compass bearing) and elevation (angular height above the horizon) at a given moment, from a given point on Earth's surface, is a matter of deterministic celestial mechanics — the same calculation GEIPAN's own published methodology, and the dedicated "questionnaire astronomique" investigative form it still uses today, direct investigators to perform using ephemeris data or modern planetarium software such as Stellarium, cross-checked against sky charts from tools like Ciel des Hommes.

GEIPAN's file records the result: Venus, "très visible ce soir-là," was at azimuth 296° to 299° and elevation 4° to 1.5° during the sighting, while the phenomenon was reported at azimuth 300° to 330° and elevation 10° to 15°; GEIPAN therefore judges confusion with Venus "extrêmement peu probable." NASA JPL Horizons confirms the planet's position: from Évreux, Venus (magnitude −3.9) was at azimuth 296.2° and 3.4° elevation at 20:35 UTC, 297.7° and 2.3° at 20:43, and 299.2° and 1.1° at 20:51, setting. The elevation difference of roughly 8–13° is real. This file notes, however, that several features of the sighting also fit a setting Venus: the light was first seen at almost exactly Venus's azimuth, it was red and point-like with "the apparent size of a planet," its brightness was seen to decrease, and the visual observation ended at 20:51, when Venus was about one degree above the horizon and would have been fading and reddening through the thick lower atmosphere. Elevation estimates for lights near the horizon are often inaccurate. Against that, GEIPAN places the phenomenon at azimuth 300° to 330° — it moved north, away from Venus — and the radar echo moved the same way. The Venus question is therefore less settled than GEIPAN's wording suggests, though GEIPAN's own judgment stands as the official one.

GEIPAN itself finds it "étonnant" — surprising — that the officer did not mention Venus at all, given how visible it was that evening and how close it lay to the reported bearing. An earlier version of this file presented that silence as corroboration of the Venus rejection; GEIPAN presents it as a puzzle. Either the officer saw both Venus and a separate light and thought Venus not worth mentioning, or the light he reported was Venus itself.

Ruling Out Helicopter and Balloon

GEIPAN's case summary states that the helicopter and balloon hypotheses were both explicitly tested and rejected, without elaborating on the specific reasoning in the portion of the file publicly summarized. Both candidates are, nonetheless, worth taking seriously on their face precisely because of Évreux's own operational profile: the base's Groupement aérien 1/56 Vaucluse flew SA-330 Puma helicopters as part of its mixed transport fleet in this era, meaning a helicopter misidentification is not an arbitrary or generic candidate explanation invented after the fact — it is exactly the kind of aircraft the base itself was flying, at the same location, in the same period.

Against the helicopter hypothesis, several elements of the reported observation cut the other way: a helicopter's navigation lights, unlike the steadily fading single red point described, would ordinarily present a more complex lighting pattern (position lights, anti-collision strobe, and often a distinguishable engine/rotor sound at 18 nautical miles' distance closing or opening, depending on flight path); a helicopter maintaining a slow, level cruise at 30–40 knots for sixteen continuous minutes on a steady compass heading is an unusual, though not impossible, flight profile for a genuine cross-country transport leg, and would ordinarily be a matter of record on the base's own flight schedule for that evening — a record GEPAN's investigators, working directly with Évreux's own personnel, would have been well positioned to check and, evidently, did check closely enough to rule the hypothesis out. Weather balloons, similarly, drift with prevailing wind rather than holding a directionally consistent track toward a specific bearing, and a French meteorological balloon of the era would not typically present as a red point source specifically; a fading intensity is more consistent with a balloon's illumination source (if lit at all) dimming as it either drifted away or lost altitude, which may be part of why GEPAN's investigators considered and then excluded it, though again the file's public summary does not preserve the specific technical reasoning for either rejection in full detail. What can be said with confidence is that both hypotheses represent exactly the kind of mundane, locally-plausible explanations a careful investigation is obligated to test first, given the base's own operational context, and that GEPAN's file states plainly that neither survived that testing.

Searching the French Ufological Record

France's specialist ufology press was, by 1978, a genuinely active and well-established field: Lumières dans la nuit (LDLN), founded by amateur astronomer Raymond Veillith in February 1958 and by the late 1970s widely regarded as the principal specialized French UFO journal, had seen its subscriber base swell dramatically in the wake of France's major 1974 sighting wave, climbing toward roughly 4,750 subscribers by 1980, with an extensive network of regional delegates and volunteer investigators feeding it case material throughout this exact period. A genuinely thorough research pass on this case therefore has to ask whether LDLN, or comparable French ufological publications and archives active in this era, ever covered the Fauville-Évreux sighting independently of GEPAN's own file.

Honestly stated: no such independent secondary coverage of this specific case turned up in this research. Extensive searching located substantial general material on GEPAN's founding, mandate, and 1970s-era methodology, on LDLN's own institutional history, and on French ufology's broader activity in exactly this period — but nothing identifiably discussing the April 23, 1978 Évreux radar-visual sighting by name, whether in LDLN's archived back issues, in later French skeptical or academic retrospectives on GEPAN's early case history, or in any other identifiable published source outside GEIPAN's own official case file. This is, on its own, a meaningful and unsurprising finding rather than a gap to paper over: a single military-sourced sighting that generated no press coverage, no landing, no physical trace, and no public witness statement (the officer's report went directly into GEPAN's internal case file through military and CNES channels, not through the gendarme-and-press pipeline that put many of the era's more famous French cases into wider circulation) is exactly the profile of case that would plausibly escape independent ufological attention entirely, however genuinely and rigorously it was investigated internally. The absence of secondary literature here says something real about how this case reached the public record at all — quietly, institutionally, and only decades later through GEIPAN's own digitization of its historical archive — rather than reflecting any failure of this research pass to look. Readers should treat this case's evidentiary record as resting entirely on GEPAN/GEIPAN's own official file, with no independent journalistic, ufological, or academic cross-check available anywhere in the public record to corroborate or challenge it from outside the investigating agency itself.

Key Witnesses

The Five Observables Assessment

The Five Observables framework, developed by the Pentagon's Advanced Aerospace Threat Identification Program to describe performance characteristics beyond known conventional technology, is applied here for comparative consistency across the site's case files — though this case's evidentiary base (a fading point-source light and a single weak, low-resolution radar echo) offers very little of the kind of high-fidelity performance data the framework was originally built to describe.

Assessment: None of the Five Observables are affirmatively present in this case. The radar-derived speed estimate (30–40 knots) is well within the range of ordinary aircraft, and no maneuver, acceleration, or directional reversal of any kind was reported by either the visual witness or the radar log. The only Observable with even speculative relevance is Low Observability, on the strength of the radar echo's very weak (1/5) intensity — but this is at least as consistent with a marginal, borderline-detectable conventional target (or a spurious echo) as with a genuinely low-reflectivity object, and the case file offers no way to distinguish between those readings from the data available.

Evidence Assessment

Why "Open"

This case is stamped Open rather than Credible, Disputed, or Debunked because its own investigating agency's final classification is, itself, explicitly and deliberately unresolved — GEIPAN's narrative conclusion describes the phenomenon as "moderately strange, distant, but [with] testimony of good consistency" (D1), a classification that is not a synonym for "probably genuine" or "probably explained," but a formal acknowledgment that neither reading can be confidently sustained on the available evidence. The case earns real credit on several fronts: a professionally measured witness reporting through a military channel with no evident motive to embellish; a second group of observers whose existence the investigating agency itself documents; a radar contact whose bearing and timing match the visual account; the elimination of the two most locally obvious mundane explanations (helicopter, balloon); and a specific astronomical calculation that, in GEIPAN's judgment, makes Venus "extremely unlikely." This file's own check (see The Venus Check) finds Venus setting at almost exactly the reported bearing as the light faded, which keeps that question more open than GEIPAN's wording suggests.

Set against that, however, are weaknesses substantial enough that "Credible" would overstate the case: the radar detection is single-station, very weak, and was never cross-checked by the one air-defense-grade radar network node that could plausibly have settled the question either way; no recording of that radar return survives, leaving every technical detail dependent on one person's written account; the case's real multi-witness character never converted into more than one piece of formal testimony; and GEIPAN's own file explicitly leaves open the possibility that the radar contact itself was a false echo, a caution the investigating agency chose to publish rather than omit. A case where the agency that holds the primary evidence still declines to resolve it, in a classification last updated in 2015, is exactly what "Open" is meant to describe: not dismissed, not confirmed, and not likely to move in either direction without evidence that, at this remove, is unlikely to surface.

Sensor & Instrumentation Detection Profile

This case's only instrumentation is the single ground radar operated at Évreux Air Base 105, which logged the 20:44–20:51 UTC contact at bearing 300°, range 18 nautical miles, heading toward bearing 330°, speed 30–40 knots, intensity 1/5. No airborne radar, no gun-camera or photographic equipment, no radiation-detection instrument, and no independently instrumented meteorological sensor data beyond the general "partly cloudy" conditions recorded in GEIPAN's structured metadata played any role in this case. Critically, the one additional sensor system that plausibly could have contributed — the STRIDA-network air-defense radar at Cinq-Mars-la-Pile, roughly 190 kilometers to the south-southwest — is recorded by GEIPAN as not having confirmed the Évreux return, without the file specifying whether it was actively checked and returned a null result or simply never produced usable corroborating data. No physical trace, no photographic record, and no recording of the radar signal itself exist in any form; the entire technical record of the sighting survives only as descriptive text in the duty officer's written report and GEPAN's subsequent case analysis, with GEIPAN's own file explicitly flagging that the single available radar reading cannot be fully distinguished from a false echo on the evidence retained.

Environmental & Geospatial Context

The sighting took place over Normandy's Eure department in late April, under partly cloudy skies with some clearing ("quelques éclaircies ou nuageux" per GEIPAN's structured metadata) — conditions compatible with clear sightlines during the clearer intervals but not the fully unobstructed sky an unambiguous astronomical identification would ideally want. NASA JPL Horizons puts the Sun 14.5° to 16.4° below the horizon during the sighting (22:35–22:51 CEST), so the sky was in astronomical twilight, still faintly bright toward the northwest; a full Moon was rising in the east-southeast, behind an observer facing the light; and Venus, at magnitude −3.9, was setting at azimuth 296°–299°, 3.4° to 1.1° up. ERA5 hourly reanalysis gives a clear sky at 20:00–21:00 UTC, which differs from the witness record of partial cloud.

Geographically, the sighting's reported bearing (300°, roughly northwest of the base) and range (18 nautical miles for the radar contact) place the phenomenon over a mixed rural-and-small-settlement landscape typical of this part of Normandy — consistent with GEIPAN's own environmental classification of the setting as "territoires artificialisés, zones urbanisées" (urbanized/artificial land), reflecting the base's own developed character and its proximity to Évreux itself (a city of several tens of thousands) rather than deep open countryside. Évreux Air Base 105 sits within regular transport-aviation traffic patterns given its role as a Noratlas/Transall airlift hub, meaning the surrounding airspace was not unused or unfamiliar territory for the base's own radar and flight-following personnel — a relevant baseline for assessing how unusual an unidentified slow-moving contact at 18 nautical miles would have appeared against the base's normal traffic picture. No solar or geomagnetic anomaly, no documented meteor shower active on this date, and no other space-weather trigger is noted anywhere in the case's available record; late April carries no major annual meteor shower peak comparable to, for instance, August's Perseids, removing one of the more common mundane explanations that complicates many other radar-visual cases from consideration here.

Observer Credibility & Occupational Profile

The case's sole formally documented witness was a military duty officer on active shift at a French Air Force base — a professional, operationally responsible role rather than a private individual reporting an unsolicited observation. Duty officers in this kind of posting are trained to maintain situational awareness, respond to anomalies through defined procedures, and communicate clearly and precisely up an established chain of command; the officer's own decision to request a radar check partway through an ongoing visual sighting, rather than either dismissing it or waiting until it concluded, reflects exactly that kind of trained operational instinct. GEIPAN's structured metadata records his emotional reaction as "curiosity, interest — active" rather than fear, distress, or panic, a comparatively calm, engaged response profile consistent with a professional observer methodically tracking something unusual rather than reacting to a perceived threat.

Weighed against this credential-based strength is the case's central structural limitation: this is the only formally recorded account for what the investigating agency's own file describes as a multi-witness sighting. The lookout personnel who reportedly saw the same phenomenon independently were never individually identified, interviewed, or credited with their own separate statement — meaning the case's occupational-credibility profile, however strong for the one witness it does have, cannot be broadened to draw on whatever additional corroborating detail a second or third professional observer's account might have contributed. No civilian witnesses, and no witnesses outside the base's own military personnel, are recorded anywhere in the file; this is, in that specific sense, a purely military-internal sighting from intake through investigation, evaluated entirely within an institutional chain rather than surfacing through public reporting, press coverage, or independent civilian corroboration at any stage.

Historical Precedents & Archive Matches

Radar-visual corroboration — an independent instrument reading that lines up with a human observer's account of the same object, at the same time, on a compatible bearing — is widely treated within UAP research as one of the strongest categories of evidence available, precisely because it removes the sighting from the realm of a single person's subjective perception and ties it to a physical instrument's own independently generated data. The category is also, however, one where the specific quality of the corroboration varies enormously from case to case: a single weak return from one uncross-checked station is a fundamentally different evidentiary object than four independent radar systems tracking the same non-ballistic maneuvers, or two independent detection channels aboard the same aircraft cross-confirmed by a wholly separate ground station. This site has published several of the most significant radar-visual cases in the historical record; setting the Fauville-Évreux sighting against them is instructive precisely because of how differently their evidentiary weight breaks down.

The Lakenheath–Bentwaters Radar-Visual Incident (1956)

The Lakenheath–Bentwaters case (stamped Disputed, 65/100 on this site) represents the high end of instrumented redundancy: four physically independent radar systems across two separate airfields, ground-level visual sightings, a passing aircrew's own corroborating account, and an RAF Venom night-fighter that achieved an actual weapons-radar lock on the target. Yet even that far larger evidentiary base did not settle the case cleanly — a 2000s re-investigation tracked down the actual surviving Venom aircrews, who directly disputed the dramatic decades-later "chase" narrative that had entered the case's popular retelling through a single 1968 letter. The lesson for Fauville-Évreux is a cautionary one in the opposite direction: even a case with vastly more instrumented redundancy than this one can still collapse into genuine dispute once the underlying testimony is checked closely, which argues for real caution in how much weight Fauville-Évreux's single, much weaker radar return should be given on its own.

The RB-47 Bomber Chase (1957)

The RB-47 Bomber Chase (Open, 60/100) offers perhaps the closest structural parallel worth drawing, and the contrast is telling. There, three physically distinct detection systems — the bomber's own visual crew, two independent airborne ELINT receivers, and an entirely separate Air Defense Command ground radar station near Duncanville, Texas — reported the same target with repeated, simultaneous appearance-and-disappearance correlation over more than an hour. That is exactly the kind of independent, cross-system confirmation that Fauville-Évreux never obtained: where the RB-47 case's ground radar (Site Utah) independently validated what the aircraft's own instruments were already reporting, Cinq-Mars-la-Pile's silence left Évreux's single weak return exactly where it started, uncorroborated by the wider national radar network. Both cases share an important intellectual honesty, though: RB-47's own most serious skeptical rebuttal (the Kessler Air Force Base reception-artifact hypothesis) remains genuinely unresolved for lack of surviving primary duty logs, in the same way GEIPAN's own file leaves the "faux écho" possibility explicitly and honestly open for Fauville-Évreux rather than resolving it either way.

The Tehran UFO Incident (1976)

The Tehran UFO Incident (Credible, 62/100), which took place barely eighteen months before Fauville-Évreux and was itself reviewed at the time by the U.S. Defense Intelligence Agency, involved two scrambled fighter interceptors experiencing documented weapons and communications systems failures on approach to the target — instrumented anomalies affecting multiple separate aircraft systems, not merely a passive radar echo. The comparison underscores just how much further along the evidentiary spectrum a case needs to sit before "instrumented corroboration" becomes difficult to explain conventionally: Tehran's radar-visual pairing came with active systems interference across two aircraft; Fauville-Évreux's came with a single passive 1/5-intensity ground return that never triggered any comparable instrument anomaly anywhere else in the encounter.

The USS Nimitz / Tic Tac Encounter (2004)

The USS Nimitz encounter (Verified, 96/100), the highest-scored radar-visual case on this site, benefits from a scale of technical documentation entirely unavailable to any 1978 French case: the most advanced radar system fielded by the U.S. Navy at the time, infrared video recording, and multiple independent aircrew witnesses, later officially acknowledged as authentic by the Pentagon. Fauville-Évreux, investigated by a ten-person unit little more than a year old with none of that instrumentation, was never going to reach a comparable evidentiary tier — the comparison exists here specifically to illustrate the enormous range the "radar-visual case" category actually spans, from a passive, single-station 1/5-intensity ground echo at one end to multi-sensor, video-recorded, officially authenticated military intercepts at the other.

A Domestic French Parallel: Air France Flight 3532 (1994)

Perhaps the most directly relevant comparison, because it involves the exact same French air-defense radar station, is the Air France Flight 3532 / Capt. Duboc sighting of January 28, 1994, in which the airliner crew's own visual sighting was compared against a near-simultaneous unidentified track logged by Cinq-Mars-la-Pile's radar, relayed through the CODA air-defense command center at Taverny — sixteen years after Cinq-Mars-la-Pile's silence on the Évreux case. In the 1994 case, the same station that produced no confirming data for Évreux in 1978 did generate its own independent radar track for a different sighting entirely, though even there the aircraft's reported position and the radar's logged track disagreed on which side of the aircraft the object actually was — a reminder that even when Cinq-Mars-la-Pile's radar does produce a contact for a French UAP case, reconciling it precisely with an independent visual account is not automatically straightforward. Read together, the two cases suggest a station that is a genuinely active, capable node of France's air-defense radar network across at least two different decades, but one whose data, when it exists at all for a given UAP report, still requires careful independent reconciliation rather than treatment as an automatic, clean confirmation.

Taken as a set, these five comparisons place Fauville-Évreux clearly and honestly toward the weaker end of the radar-visual case spectrum: real, professionally sourced, and internally consistent, but resting on a single weak instrumented reading that never received the kind of independent second-system confirmation that elevates the strongest cases in this category into the "verified" or "credible" tier.

Material Analysis

This case involves no physical or trace evidence of any kind — no landing marks, no recovered material, no soil or vegetation anomaly, no metallurgical or isotopic sample, and no biological effect on any witness. That absence is itself informative context for how this case differs in kind from GEPAN's later, more elaborate physical-evidence investigations: the same organization that would, three years later, commission multi-laboratory soil and plant analysis for the Trans-en-Provence landing case simply had no comparable physical substrate to work with here. The Fauville-Évreux sighting was, from beginning to end, a distant aerial observation — nothing about the reported encounter (a light seen at 10–15° elevation, 18 nautical miles away at its radar-confirmed range) placed the phenomenon anywhere near the witnesses or the ground, so no material-evidence investigation was ever a realistic possibility for this specific case type, and none was attempted.

The Conventional Explanation Candidate: Venus Misidentification

Because the phenomenon was a red, low-elevation point of light at dusk, Venus is the natural first candidate for any skeptical review — and it is the candidate GEPAN's own investigators most rigorously tested using a specific, reproducible astronomical calculation rather than a general assumption. As detailed above, Venus's calculated position that evening (azimuth 296–299°, elevation 1.5–4°) sat close in bearing but dramatically lower in elevation than the reported object (azimuth ~300°, elevation 10–15°), a gap of roughly 8 to 13.5 degrees that is not a marginal or easily-confused discrepancy for a trained observer looking directly at the object in question.

Helicopter Misidentification

Given that Évreux's own Groupement aérien 1/56 Vaucluse operated Puma helicopters from the base in this era, a helicopter misidentification is a locally grounded, non-arbitrary candidate — and one GEIPAN's file states was explicitly tested and rejected, though without preserving the specific technical reasoning in the portion of the record publicly summarized.

Weather or Research Balloon

A drifting balloon, potentially with an attached light or instrument package, is another mundane candidate consistent with a slow, low-intensity radar return and a fading visual light source — and, like the helicopter hypothesis, GEIPAN's file records this as explicitly tested and rejected.

Radar False Echo / Anomalous Propagation ("Faux Écho")

Distinct from a misidentified physical object, this candidate holds that the radar return itself — independent of whatever the duty officer actually saw in the sky — may not represent a real airborne target at all. Ground radar systems are subject to a range of well-documented artifact sources: secondary/multipath reflections off terrain or structures, anomalous atmospheric propagation (temperature-inversion ducting) that can deliver spurious or displaced returns, and simple receiver noise presenting as a marginal target at very low signal intensity. GEIPAN's own case file explicitly raises this possibility and does not resolve it, which is precisely why it remains a live, unresolved candidate rather than a rejected one.

Related Cases

Further Reading — Recommended Literature

DISCLOSURE: Some links above are affiliate links. If you purchase through them, OverClassified receives a small commission at no additional cost to you. This helps support independent research.

Essential Viewing

Sources

• GEIPAN (CNES), "FAUVILLE (27) 23.04.1978", Case 1978-04-01879 — official case summary
• GEIPAN (CNES), Testimony Detail, Témoignage 2096, Case 1978-04-01879 — structured witness metadata
• GEIPAN (CNES), official case archive & public portal — geipan.fr
• GEIPAN (CNES), "Comprendre un Phénomène" — investigative methodology, astronomical position-checking guidance
• GEIPAN (CNES), Questionnaire astronomique — standardized astronomical observation form (PDF)
• Wikipédia (français), "Groupe d'études des phénomènes aérospatiaux non identifiés" — GEPAN founding and early institutional history
• Wikipedia (English), "GEIPAN" — organizational evolution, GEPAN → SEPRA → GEIPAN
• Wikipédia (français), "Claude Poher" — biography and role in founding GEPAN
• CarteOVNI, "Petite histoire de l'ufologie en France: GEPAN, SEPRA, GEIPAN" — institutional methodology evolution, 1977–present
• Institut national de l'audiovisuel (INA), "1978: le GEIPAN «en quête» d'ovnis" — archived 1978 French television coverage of GEPAN's contemporaneous methodology
• Wikipédia (français), "Base aérienne 105 Évreux-Fauville" — base history, 1978-era squadrons
• Wikipedia (English), "Évreux-Fauville Air Base" — Cold War operational history, NATO tenancy, return to French control
• avionslegendaires.net, "L'Armée de l'Air à l'automne 1978: organigramme et unités" — confirmed 1978 squadron assignments at Évreux BA 105
• EPR 118, "Une petite histoire de la Défense Aérienne française" — STRIDA network history and Cinq-Mars-la-Pile's 1970s-era CDC configuration
• Wikipedia (English), "French air defence radar systems" — STRIDA/CDC network overview, Cinq-Mars-la-Pile's designation
• ici.fr (formerly France Bleu), "Au centre de Cinq-Mars-la-Pile, les militaires inspectent le ciel français" — CDC role, staffing, and ongoing operations
• Wikipedia (English), "Cinq-Mars-la-Pile" — commune geography and CDC 07.927 designation
• Wikipédia (français), "Lumières dans la nuit" — LDLN institutional history, 1970s–1980s readership and role
• Wikipedia (English), "Fauville, Eure" — commune geography, coordinates, population
• Wikimedia Commons, Category: Base aérienne 105 Évreux-Fauville — image archive, including this file's hero photograph
• Wikimedia Commons, File: Aéroport militaire d'Evreux.jpg, photo by Ascobole (2011), CC BY-SA 3.0 — licensing and attribution
• This site, key-figures.html, entry on Claude Poher — GEPAN founding, 1977–1978 activity including the concurrent Cussac, Cantal reconstruction
• This site, case-files/trans-en-provence-1981.html — comparative reference for GEPAN's later, physical-evidence-based methodology under Jean-Jacques Velasco
• This site, case-files/lakenheath-bentwaters.html — comparative radar-visual case, multi-radar corroboration and its later re-investigation
• This site, case-files/rb-47-bomber-chase.html — comparative radar-visual case, multi-sensor independent corroboration model
• This site, case-files/tehran-1976.html — comparative radar-visual case, instrumented systems-failure corroboration
• This site, case-files/uss-nimitz.html — comparative radar-visual case, modern multi-sensor documentation standard
• This site, case-files/air-france-flight-3532-duboc-sighting.html — comparative case sharing the same Cinq-Mars-la-Pile radar station, 1994
• This site, locations/index.html, entry on Toulouse — CNES/GEPAN/SEPRA/GEIPAN institutional headquarters
• YouTube (AFP), "Ovnis: le Geipan, au CNES, enquête sur ces phénomènes" — oEmbed-verified video, general GEIPAN institutional coverage
• metar-taf.com, Évreux-Fauville (BA 105) Air Base, LFOE — aviation reference, base coordinates and identifiers
• Wikipédia (français), "Groupe d'études des phénomènes aérospatiaux non identifiés", section on Alain Esterle succession — Poher's December 1978 resignation and handover
• NASA JPL Horizons, observer ephemerides for Venus, Mars, the Sun and the Moon from Évreux (49.03°N, 1.22°E), 1978-04-23 20:35–20:51 UTC. ssd.jpl.nasa.gov
• Open-Meteo Historical Weather API (ERA5), hourly data for 49.03°N, 1.22°E, April 23, 1978. open-meteo.com
• Wikipedia (English), "Centre de détection et de contrôle militaire" — CDC 07.927 "Raki" at Cinq-Mars-la-Pile. en.wikipedia.org

Corrections (October 2026)

A full check of this page against Internet Archive copies of GEIPAN's case page (1978-04-01879) and testimony record (2096), NASA JPL Horizons, EPR 118's history of French air defence, the ici report on Cinq-Mars-la-Pile, French and English Wikipedia on GEPAN and BA 105, and this site's earlier corrections found several misreadings of the GEIPAN summary. The page quoted GEIPAN as saying the echo was "non confirmé par la station radar secondaire de Cinq-Mars-la-Pile"; the summary actually says "non confirmée par l'autre radar de Cinq-Mars-La-Pile" ("the other radar"). The page said the officer's report carried the wrong date; GEIPAN says it was the radar observation report. The page treated the radar's end at 20:51 as an independent match with the visual sighting; GEIPAN records 20:51 as the end of the observations "and the stopping of the radar." The page presented the officer's silence about Venus as corroboration; GEIPAN calls it surprising. A NASA JPL Horizons check confirms GEIPAN's Venus figures but shows Venus setting at almost exactly the reported bearing while the light faded, so this file no longer describes Venus as decisively ruled out. The sighting took place in astronomical twilight, not full darkness.

Following this review, the rating was changed from Open 48 to Open 38 (October 27, 2026), because the higher rating had rested partly on claims corrected or removed above.

Other corrections: GEPAN was created in 1977 (CNES dates it to 1 May 1977), not "formally authorized on March 23, 1977," and it was not the world's first state UFO study body; BA 105 lies a few kilometers east of Évreux, not two kilometers northeast; Cinq-Mars-la-Pile is about 190 km away (corrected in an earlier batch) and, by 1978, already had an AN 226 radar (1968) and the CESAR system (1973) in addition to its 1965 equipment; unsourced details of the modern underground site (depth, a "17-kilometer gallery," one of "only two" national centres) were removed; the lookout personnel "also" saw the light, which GEIPAN does not describe as independent; the comparison scores were updated to the current ratings of the Lakenheath–Bentwaters (65), RB-47 (60) and Tehran (Credible, 62) files; the computed weather entry ("mainly clear, 7%") was replaced with hourly reanalysis, and the meteor-shower note was corrected (the Lyrids peak around April 22–23). GEIPAN links now point to Internet Archive copies because geipan.fr has been returning errors.