Archive Index


Analysis of Visual Sightings at Lakenheath

An Opinion by Martin Shough

1. Military Air-Visual sightings

Lakenheath teletype BOI-485, if taken as written, seems to indicate a separate air-ground radar/air-visual contact by interceptor #1 immediately prior to the main interception episode. This is quite absent from the accounts of Perkins and Wimbledon, and the brevity and ambiguity of BOI-485 render interpretation difficult. BOI-485 opens its description of "Interception or identification action taken" as follows:

Interception was undertaken by one British jet fighter on alert by 60th AAA Sector Control. Aircraft is believed to have been a Venom.
The aircraft flew over RAF station Lakenheath and was vectored toward a target on radar 6 miles east of the field. Pilot advised he had a bright white light in sight and would investigate.
At thirteen miles west he reported loss of target and white light.

Immediately after this there follows BOI-485's version of the interception episode analysed elsewhere.

The phrase "loss of target and white light" implies an AI radar contact with this target which was lost at the same time as the light disappeared. But who was vectoring this aircraft? It was evidently not the RAF Interception Control team (or at least, Wimbledon as Chief Controller was not aware of any such secondary action by his team), and Perkins recalls no such separate incident. It has been widely and tacitly assumed that BOI-485's "bright white light" and the "something" seen by a Venom #1 crewman in Perkins' account are the same thing, and that this visual sighting occurred during the main interception episode. However this sits awkwardly with the immediately consecutive paragraph of BOI-485, which appears to state that after the "loss of target and white light" the pilot was then vectored to another target not 6, but 10 miles east of the field. It does appear that Captain Stimson is here attempting to describe a separate, prior interception attempt by Venom #1.

So this reported visual sighting is not easy to integrate. According to Wimbledon it is possible that something was seen but he did not recall any real-time pilot report of visual contact. Perkins originally stated that reference to visual contact was only made during a subsequent air-to-air transmission between interceptor #1 and the inbound interceptor #2: "I saw something but I'll be damned if I know what it was." BOI-485 confirms independently that a pilot visual was reported over the radio, but does not contain this transmission verbatim, adding the detail (not recalled by Perkins) that it was of "a bright white light". Perkins suggested that the pilot statement he did recall was probably not quoted in BOI-485 "because it contained what was then considered profanity", but made no claim to total recall. He pointed out that in addition to his team's written reports "all telephones, microphones and even controller chit chat was recorded on an excellent recorder . . . . " It is possible that the description of a "bright white light" was contained somewhere amongst this material, but was either not heard or was forgotten by Perkins himself. It is also possible that this information was collected by Stimson from GCA personnel.

On the other hand one can argue that the air-to-air talk assumed by Perkins to have been pilot talk also contained interjections by the nav/rads of one or both interceptors, each of whom (assuming these were indeed both RAF Venoms) had their own access to the VHF transmitter by means of a footswitch which would in fact override the pilot's transmission. Whether the lead plane's radar had "seen" anything or not would be a very pertinent question. Then the statement "I saw something but I'll be damned if I know what it was" could be the interceptor #1 radar operator's response to the question "Did you see anything?" directed to him by the radar operator of interceptor # 2. So there may be no connection between the exchange reported by Perkins and the visual sighting reported in BOI-485. (It is even possible, though perhaps less plausible, that the "bright light" mentioned in BOI-485 as a visual contact is itself a misinterpretation of a reference to the brightness of a radar blip.)

There is a third interpretation of the statement "I saw something but I'll be damned if I know what it was", in which the pilot is speaking but is not referring to a sighting of a UFO. This depends on the fact (again, assuming that Capt. Stimson was correct in his "belief" that the jet was an RAF Venom) that the pilot will have one eye constantly on the radar-controlled GGS.5 gunsight during closure on the AI radar target. When the GGS.5 mechanism automatically acquires range lock under control of the radar at a distance of 800 yards the electromechanical response of the gunsight's illuminated aiming graticule will be visible to the pilot even if a target is not. This visual indication could explain a statement much like the one reported, consistent with Perkins' initial statement that the pilot had indeed, at the moment of closure, reported that his "radar gunsight locked on".

So there are different plausible interpretations and this confusion is difficult to unravel. Until recently the BOI-485/Perkins references to a pilot visual stood alone. The 1996 evidence of the two 23 Squadron crews who were scrambled from Waterbeach in the early hours did not explain it; neither of these pilots reported seeing anything at all. However, the recent discovery of the adventitious involvement of Air Commodore (then Wing Commander) Anthony N. Davis, CO of 23 Squadron, reopens this question in the context of the "missing interception".

Internal evidence had already suggested that the known 23 Squadron scrambles probably were not the whole story and Wing Commander Davis' information bears this out. He recorded that he was vectored by ground radar towards what appeared to be "a bright star". So it appears that there was indeed an air-visual sighting of a "bright white light" at some point during a "missing interception", although it is unproven that the interception involving Davis corresponds to the events reported in the classical scenario, and the question of exactly when this sighting occurred and its relationship to the surrounding events remains unanswered. (According to 23 Squadron navigator Ivan Logan, Davis was "commuting" from Coltishall to Stradishall and back for the duration of the crash inquiry, which might suggest a time earlier rather than later in the night.)

RAF Stradishall is some 20 miles south of Lakenheath, and Davis' route NE from Stradishall to Coltishall (near Neatishead) would probably have passed a few miles to the east of Lakenheath. But where the aircraft was when it was diverted, and its instant heading during the final vector towards the "bright star", are of course unknown. The aircraft altitude is also unknown. No substantial analysis is possible in these circumstances, but a few notes on possible astronomical sources are offered.

For an observer on the ground in the Lakenheath area at midnight GMT 13-14 August 1956 Capella (mag +0.21) stands at 23 elevation in the NE, Vega at 56 elevation in the W, and Arcturus (+0.24) low on the horizon at about 3 in the WNW. The plane of the ecliptic ran around the southern sky from the E to the SW, and by midnight the moon and Saturn had long set some 13 and 15 respectively below the western horizon at negative elevations many times the critical grazing angle (about 0.5) for superior mirage. The only visible planet was Mars at 24 elevation in the SE, then at opposition and quite bright (though at magnitude -1.1 far short of its mean opposition magnitude of -2.25, which is about the same as Jupiter). Mars, however, appears distinctly amber, particularly at opposition. Perhaps a bright star close to the point of setting, like Arcturus in the WNW, would be the most likely cause of a "bright white light" which quickly disappeared.

Of course these values are for ground observers and would be significantly changed for a pilot at high altitude. It is worth noting that from an aircraft at 20,000' over Lakenheath the effective celestial horizon would be roughly that of a ground observer some 250 miles away in any direction, with a gain of some 2 elevation. For example, the sky due east would resemble that seen by an observer at Arnhem in the Netherlands. Thus if Davis' Venom were at this altitude the planet Venus would rise at 55 azimuth in the NE at about 0043Z with an impressive magnitude of 3.8, not far from maximum brilliance, almost half an hour before becoming visible for observers on the ground. Optical mirage could also be a factor here for a source viewed through what was probably a significantly refractive layer at 54,000'. A superior mirage of Venus might be produced temporarily, even while still beyond the geometrical horizon (see Section c. below).

Other explanations are conceivable. In the context of BOI-485, for example, the pilot's initial confirmation of a light in the position of the radar target "east of the field" could indicate the 5-second sweep of the Orford Ness lighthouse 50 miles away on the Suffolk coast, quite possibly visible from altitude in the ESE in CAVU conditions. But the on-going operational training of QRA Squadron aircrews involved - as a matter of routine - frequent flying and navigation exercises, practice interceptions and Bomex drills. One would expect pilots and navigators - particularly a Commanding Officer and the Squadron's Leading Radar Navigator - to be thoroughly familiar with so singular a local landmark as the lighthouse.

So a "bright white light" could in principle have been any one of several first magnitude stars, Mars, Venus, the lighthouse, or other ground/shipboard lights of various kinds, not to mention lighted met. balloons and air traffic. It might have been something interesting, but this depends on how this aspect of the case is integrated with others - in particular, on whether the light seen bore any relation to the reported AI radar target. Since Davis and McIlwrath apparent did not acquire AI radar contact their Venom would not appear to have been the intercepting jet described by Perkins, BOI-485 or Wimbledon (they may have been in the second interceptor, which was "unable to make contact"). Therefore the "bright white light" reported in BOI-485 and by Perkins might, or might not, have had something to do with the "bright star" reported by Davis, and either or neither might have been related to the radar indication.

It is possible that Perkins and/or BOI-485 confuses the sequence of events. It is noteworthy that neither of the interception locations given in the latter corresponds with the location recalled by Perkins in 1968, which was about 16 miles SW of the field. Questioned about this by Stanton Friedman in 1975, Perkins was as confused as anybody and as a result replied that he was inclined to defer to BOI-485:

I cannot be sure about the first vector, where the pilot reported a bright white light. It rings a bell but I simply can't remember for sure. I was fairly confident of the position of intercept [but] I believe the Blue Book report is probably the more accurate.

But BOI-485 does contain errors, and is sometimes less than coherent as to sequences. It is possible, therefore, that the source of this confusion is not Perkins but Capt. Stimson or someone on his intelligence staff, encouraged by the fractured report format dictated by AFR 200-2. On the other hand if BOI-485 is accurate and Venom #1 was directed to another target prior to the primary interception, how could this be integrated with Perkins' and Wimbledon's accounts?

The implication in this case would be that Perkins forgot or missed entirely an initial air-intercept episode which was presumably over quite quickly. Could this have been the passing involvement of another aircraft, prior to the arrival of "Venom #1"? The implication is of an aircraft which was not fully fuelled or was otherwise not prepared for an extended commitment but happened to be nearby and was called on to attempt an identification, much as happened earlier with the T-33 at Bentwaters. RAF or even USAF planes might have been called on in these circumstances. Wing Commander R.G.Grocott, an Eastern Sector senior officer at the time, told Dave Clarke:

USAF fighters . . . may sometimes have been used ad hoc for visual track identification. We had great flexibility in filling our task and the Sector Controller, even at times the Chief Controller at a GCI station, might call on any available aircraft (whether already airborne or ready on the ground) to assist in identifying a track, but control would normally, though not invariably, remain with the RAF Air Defence Control system.

Since it is now a matter of record that Wing Commander A. N. Davis, en route to RAF Coltishall on a 'bus' trip, was diverted in just this manner, it is possible to suggest that his Venom was first on the scene, but was unable to hang about, perhaps owing to limited fuel. This Venom would also be unlikely to be carrying ammunition or a gun-camera, or might be otherwise not battle-worthy, and after a first inconclusive vector Davis might be released to go on his way, attention then turning to the first of two scrambled aircraft.

An alternative scenario would have Davis arriving to assist with a situation already underway, filling the role of Venom #2 just as "Venom #1" was retiring from the fray. As Perkins recalled:

We gave #2 the location of the UFO and advised him we still didn't have him on radar but probably would have shortly. He delayed answering for some seconds then finally said - "Lakenheath . . . . Returning to home Station. My engine is malfunctioning." He then left our frequency.

The equivalent episode is given in BOI-485 thus:

Target disappeared and second aircraft did not establish contact. First aircraft returned to home station due to being low on fuel. Second Venom was vectored to other radar targets but was unable to make contact. Shortly second fighter returned to home station due to malfunctions.

Both of these statement could be consistent with Davis' recollection that he was "vectored on to a suspected U.F.O. but made no radar contact and found myself chasing a star". If he then informed Lakenheath that his aircraft was not properly fit for active duty this might get interpreted as declaring a 'malfunction'.

In either case the origin of the other plane or planes is uncertain. The possibility that the first interception attempt might have been by a USAF jet has been considered, and this suggestion is credited in principle by Wing Commander Grocott (above), and also by Grahame Scofield who, according to the 23 Squadron Diary, had been scrambled after a UFO earlier that night. Said Scofield:

A possible scenario would have been that a USAF jet was scrambled first but could not close. We had airborne radar and were a night fighter squadron so were used to this type of situation. My guess (and I emphasise it is only a guess) is that the USAF tried their own intercept first, about which we would have known nothing. Then, having failed to locate and still reporting an unidentified object hovering over their main base, they would have requested help from the RAF. There would be a flurry of high level discussion about who pays etc and eventually 23 squadron were instructed to go.

Scofield's later response to the discovery of the Davis documents is interesting:

It is entirely consistent with the record that he would have been airborne at the time of the incident and that we would have been unaware of his presence. He would have most likely flown with his own navigator, a Flight Lieutenant Paddy McIlwrath, from Northern Ireland. He was another "old timer" who had flown Mosquitos during the war. He was squadron "Nav Rad Leader", which interprets to mean he was the lead radar navigator. . . . If they were airborne, they would have been the first choice of any duty controller who was responding to a call for assistance from Lakenheath. I am intrigued by his comment that he was "chasing a star." As most of our sorties were done at night we were pretty familiar with the night sky. I suppose if he made no radar contact and there was a planet or bright star low on the eastern horizon he might just have assumed that there had been a false reported sighting. It would not be difficult with an air almanac to check on the stellar sky at midnight on that night to see where the stars are positioned. I recall that it was a clear night and the stars were bright . . . . [I]t is entirely probable that the Davis intercept was the one noted at midnight, which would coincide with the reports from the ground radar stations [Neatishead & Lakenheath] . . . . So you now have some fairly clear facts. There undoubtedly was an incident. The US jet was probably up first, then came Davis, then came the two crews from 23 Squadron.

The Davis "bright star" may or may not be the origin of the "bright white light" episode recorded in BOI-485, and the light may or may not have been Venus or something else. There is little point in pursuing this exercise. Absence of crucial times, locations and headings - and uncertainties in the relations between times, locations and headings that are reported - mean that it is ultimately impossible to remove the ambiguity from the source materials in this instance.

2. Military Ground-Visual

The only source for these sightings is BOI-485. Reassembling the fragmentary information therein leads to the following account:

At some time prior to the first radar contacts at Lakenheath, USAF ground personnel at the airfield observed at least two fast-moving aerial objects. The initial number of visual observers is given as three, but the obviously-incomplete witness listing in BOI-485 contains, in addition to three radar personnel, four 307th Bomb Wing Intelligence Specialists - Erickson, Lynch, Godfrey and Fowler - who may well have been visual observers. All four airmen were considered "reliable" or "very reliable". (Richard T. Lynch was one of those contacted by the author with the assistance of the Military Officers' Association of America. However he did not respond.)

Whereas the radar observers were "alerted" by Bentwaters GCA "to be on lookout for unusual targets", the visual observers were not alerted according to BOI-485. Their attention was attracted when they simply "looked at sky and saw the objects". This is obviously consistent with the fact that radar contact is said to have occurred "later" than visual contact. In other words the visual observations were made independently of the radar observations. Whether or not the visual observers were independent of one another is not known.

Two objects appear to have been observed at the same time travelling at high speed on SW headings. They were "round white lights", each about the apparent size of a golf ball at arm's length (approximately 2 or about 4 times the diameter of the full moon). Their estimated altitude was 2000-2500'. (Note: an object of the angular size indicated at >2500' slant range would be on the order of 100' in diameter.) The objects had no discernable surface features of any kind, no visible tails, trails or exhausts and made no audible sound. The objects were observed to "stop" and then "immediately assume an easterly heading" (135 implied change of course) without apparent change in altitude. The two objects "joined up" with one another and moved off "in formation together", dwindling in size to a "pin point" as they "disappeared on easterly heading." The manner of flight was characterised as "rapid acceleration and abrupt stops" and "travelling at terrific speeds and then stopping and changing course immediately."

No binoculars or other optical aids were used. Visibility was 10 nautical miles, sky clear, ceiling unlimited. "All ground observers and reports from observers at Bentwaters agree on colour, maneuvers and shape of object" and the observed visual behaviour was "substantially [the] same" as the radar-tracked behaviour. The Lakenheath observers also reported an "unusual amount of shooting stars", adding that the objects seen were "definitely not shooting stars".

BOI-485 does not state how long this episode lasted, nor is it certain that this is the only visual episode referred to. There is therefore some doubt as to whether radar and ground-visual observations at Lakenheath were at any time actually concurrent. It appears that at the time of first visual contact no unknown radar targets were being reported, and the statement that "two radar sets [evidently CPS-5 & CPN-4 in this context] and three ground observers report substantially same" can be taken to indicate similarity rather than concurrency. But no exact times are given and Stimson's summary could be interpreted to mean that there was some degree of concurrency at some time: "The maneuvers of the object were extraordinary; however . . . radar and ground visual observations were made on its rapid acceleration and abrupt stops . . ." Indeed, the statement that "radars reported these facts to occur at later hours than the ground observers", given the meaning of "hours" in military-speak [i.e., "2300 hours", "2305 hours" and so on], does not necessarily imply that radar and visual observations did not overlap. This could be interpreted to mean that the objects were first sighted visually (presumably close in to the station at low estimated altitude), moved away to the E, and were then detected on radar when they left the shadow cone and entered (or re-entered) the surveillance coverage. On this interpretation the visual objects might relate to two of the "three to four additional [radar] objects" which GCA detected "doing the same maneuvers in the vicinity of this station" whilst the primary radar target was being tracked by GCA and RATCC at greater range.

That these targets were ex hypothesi not initially detected on radar but were seen visually inbound from the NE is not necessarily a difficulty, since there is no reason to assume that the GCA trailer was manned until after the Bentwaters alert. If the GCA operators were relaxing in the standby room with no expectation of inbound traffic, this would be consistent with Perkins' statements that there was "little or no traffic" on the RATCC radar, and that it was only after his unit was alerted that he had his controllers "start scanning the scopes [emphasis added]". Indeed he said in 1975: ". . . unless they [GCA operators] are expecting traffic or actually handling an arrival or departure they very rarely sit in the trailer at ten or eleven at night. Ours were in the trailer at the time but that was probably because we asked them to look for the target we had been alerted to look for." In fact the request came direct from Bentwaters according to BOI-485, but the point stands. Thus visual objects may well have been seen which were at some later stage also detected by radar, and there may well have been radar-visual concurrency at this time. But as usual BOI-485 is simply unclear.

a.) meteors

The "unusual amount of meteors" of course refers to the Perseid shower under way at the time. (Please refer to the more detailed discussion of the Perseid hypothesis, vis a vis the Bentwaters observations, given in Analysis of the Bentwaters Fast Radar Tracks and Analysis of the Bentwaters Visual Observations.) In this case the initial SW heading of the visual objects is not inconsistent with meteors from the Perseid radiant, and "terrific speed" could be taken to indicate meteoric angular rates.

Most people have never knowingly seen a meteor, although there is a shower radiant of some sort above the horizon of every observer on every night of the year. Routine shower meteors are obvious when one knows what one is supposed to be looking at, but these are swift streaks, not majestic fireballs, and quite easy to miss. The observers "looked at sky and saw the objects", and nothing else in particular "called their attention" to them. As mentioned, this is consistent with the statement that the visual observations began before the radar observations. But it doesn't rule out the possibility that they thought nothing much about them until they heard that UFOs were being tracked on the radar and only then decided they were looking at UFOs. However, this doesn't add up. The Lakenheath observers did know what they were looking at because according to BOI-485 they themselves drew attention to the meteors and took care to distinguish them from the 'UFOs':

8. Any other unusual activity or condition, meteorological, astronomical, or otherwise, which might account for the sighting:

Ground observers report unusual amount shooting stars in sky. Further state the objects seen were definitely not shooting stars as there were no trails behind as are usual with such a sighting.

Apart from the initial SW heading and high speed, no other features of the observation are diagnostic of Perseids, in particular the sudden stationarity, the joining up "in formation together", the specifically noted absence of any trails, and not least the easterly departure heading.

To explain the reported kinematics - arriving on SW heading, stopping, then departing to the E - we would have to assume that from the position of the Persid radiant at 30 elevation in the NNE one meteor flew 'upwards' towards the zenith, where it disappeared and attention was transferred to a bright star somewhere overhead, probably Vega which was the only noteable bright star not near the horizon at the time (mag 0.0, 5th brightest N hemisphere star, prominent at high elevation in the SW). After a while the observers ceased looking at Vega and transferred their attention again to another meteor which was on a trajectory slanting down towards the E horizon. But this can't realistically have been another Perseid because the Perseid radiant is of course still in the NNE, half the sky away from Vega, and an 'east heading' Perseid would start life a very long way away from the point where a previous 'southwest heading' Perseid had finished up. So one needs to invoke another meteor, a sporadic one unconnected with the Perseids, that happened to cross the sky near Vega (or whatever star had been identified as the hovering UFO) from a different direction, and drew the observers' eyes away with it.

Something of this sort seems to have been imagined by Blue Book in 1956. After a discussion with Capt. Gregory (recorded - presumably in error - as being on August 14, which interestingly is the day after the event and several days before ATIC officially received the report) Dr. L.V.Robinson of the Air Science Division wrote a Memorandum for the Record suggesting a coincidence of meteors from different parts of the sky:

a. The Perseids are prominent and apparently could be quite spectacular over England.

b. The Perseids are at their visual peak between the 11th and 20th of August

c. These meteors, in their individual flights, could appear to cross over at large angles to other meteors.

d. It is possible that individual meteor trails (ionized gases) may trace on radar scopes.

These points were incorporated in Capt. Gregory's evaluation for HQ Washington - along with some rather curious remarks about seasonal disturbances of the geomagnetic field - to support Blue Book's conclusion that the sightings were caused by meteors. The explicit precautionary advice by the witnesses that the objects were "definitely not" meteors, re-echoed in an evaluation by Blue Book's own scientific consultant, was ignored.

Of the other remarks (a) is rather misleading. The Perseids will look the same in anyone's sky if the radiant is on the right side of the earth during the local night. The elevation of the radiant will vary with the local time, being higher or lower in the sky. There is nothing special about the Perseids on this particular August 13-14 as viewed from England, except that the radiant is a few degrees higher in the sky at a given local time than for viewing locations south of the 49th parallel. From Boston the difference is one hour. That night the Perseids appeared in near enough the same positions in the skies of Washington DC, or St. Louis, Missouri, at 0200 local time, as they did in the sky of London, England, around midnight. They would have been visible at all these places during any of the hours of darkness, and the difference in latitude will be insignificant compared to how much of the sky is obstructed by buildings, trees, clouds, city lights etc.

It is also not the case that the date or the local time were particularly favourable for observing the Perseids. Further discussion of the Perseids is in Visual Observations at Bentwaters, 2255. And the remark about radar tracking should be considered sceptically in the light of Analysis of the Bentwaters Fast Radar Tracks, Section f., and Analysis of the Intercepted Targets at Lakenheath-1, Section g.

The estimated apparent size of the objects will be unreliable since over-estimation of visual angles is almost the rule even amongst experienced observers; nevertheless the impression given is not of the bright streaks of small particle meteors. A round white light the size of a golfball at arm's length suggest larger fireballs, which are spectacular visual objects, atypical of the Perseids, almost never occur in pairs, and appear not to have been reported at all (let alone reported widely, as they typically are) elsewhere in SE England despite the clear evening.

To summarise, it is noteworthy that: 1) the observers were aware of the meteor shower in progress and explicitly stated that the objects were "definitely not" meteors; 2) an easterly departure heading of the objects is opposite to the direction of meteors crossing the sky from the Perseid radiant; 3) Perseid shower meteors, and other prominent meteors, typically have plainly visible ionization trails, but the observers emphatically stated that there were "no trails"; and 4) the visual observations appear to have been made independently and spontaneously before radar observations began at Lakenheath, yet the behaviour reported was "substantially" the same as that of the subsequently-acquired radar echoes (which analysis shows cannot reasonably be attributed to meteors).

b.) aircraft

The statement that "one white light joined up with another and both disappeared in formation together" might suggest controlled flying. But the absence of sound on a calm summer's night tends to rule out aircraft as the cause of these sightings. "Terrific" angular rates would indicate proximity and very low altitude, consistent with the observers' estimate of 2000-2500' or even lower, and the roar of two jets would be audible even at some miles. The flight path is of course not consistent with a high-performance jet. The implied proximity of helicopters, which could stop and change course in something analogous to the fashion reported, would be closer still, and the engine/rotor noise obtrusive. The description of large "round white lights" is also inconsistent with conventional aircraft configurations or lighting. Overall there is no resemblance whatever to aircraft, and the "reliable" USAF observers would presumably be familiar with nocturnal overflights by USAF jet fighters, bombers and transports, on training flights, supply runs and in landing patterns of all kinds.

c.) optical mirage

As a general rule the critical grazing angle for optical mirage is taken to be about 0.5, based on maximum observed refractive-index changes of a few times 10-5 through several hundred meters [Viezee 1969] or a temperature gradient on the order of 10.0C/100 m of free atmosphere. In extreme conditions such a gradient may presumably be exceeded and the critical angle would widen, but probably not by more than a factor of two. Partial scattering is also possible for very bright sources at shallow angles, but the optical power reflection coefficients of inversion layers are very small. In general all optical mirage effects are of very small extent and have strict tolerances compared to those possible for microwave and longer wavelength radio waves.

The intelligence report contains no information on angles of elevation. It merely repeats that the objects were first seen approaching on a SW heading at an estimated 2000-2500' altitude, stopped, then immediately headed E where they disappeared with "no change in altitude".

Mirage images of small, bright sources close to the horizon can undergo changes in brightness which suggest motion towards and away from the observer and can even appear to wander laterally in extreme conditions. Given such conditions a bright celestial body on the ENE horizon might loom and recede in this way. However there are no bright stars or planets on or near the NE or E horizon during the couple of hours following the Bentwaters alert at 2255Z, and before that only Mars was present, rising in the ESE at about 2100 and moving further S during the night. The only prominent body in the NE, Capella (6th brightest N.hemisphere star at magnitude +0.1), doesn't set and this night is never lower than about 9, at which time (around 2000, still civil twilight) it is due N and thus nowhere near the required azimuth. By 2300Z Capella is at 34 azimith in the N-by-NE, but at 18 elevation and climbing - far too high to be subject to mirage effects. Venus is at this time still 14 below the horizon (already allowing for refraction in a standard atmosphere), which is many times the critical grazing angle for superior mirage.

Mundane explanations such as cars on hilltop roads are rules out by the extremely flat terrain. Extremely bright ground lights normally invisible far to the NE of Lakenheath might be glimpsed by superior mirage, and possible slow "pulsation" and image wander might suggest approach and recession. Lights associated with local structures would be expected to be familiar. There appear to be military ranges 10-15 miles NE in the Breckland area which might generate flares or other activities. And then there will be landing lights of military aircraft, both local traffic and planes in landing patterns for other airfields such as Thetford (about 10 miles away). How likely such light sources might be as explanations is really impossible to gauge, except to say that there is no evidence in the Hemsby radiosonde data of temperature gradients sufficient to cause mirage in the local atmosphere at low levels through which the ray paths from such sources would be viewed.

Refractivity at optical wavelengths (unlike radio wavelengths) is virtually insensitive to changes in humidity, meaning that over narrow layers of the atmosphere where the pressure can also be considered nearly a constant the indicators for mirage are temperature inversions or superautoconvective temperature lapses. Vertical temperature profiles for the nearest Met Office radiosonde are graphed here. There are no such gradients suggested in the troposphere.

There is a temperature inversion of 6.0 through 75 mb of pressure gradient at an altitude a little above 30,000', equivalent by rule-of-thumb to a gradient of about 1.0/100m. Table 1 suggests that this gradient is probably mildly superrefractive with a curvature of about 10 arc sec./km, but only slightly more than the standard 7.5 arc sec./km of air at homogenous temperature.

diff. (C/100m)

curvature ("/km)

- 3.4


- 1.0


- 0.5




+ 6.9


+ 11.6


Table 1. Light-ray curvature against vertical temperature gradient for standard pressure and temperature (1013.3 mb. 273K). Adapted from Viezee 1969

However, at 54,000' there is a 10 inversion through only 11 mb or about 300'. This corresponds to a curvature approximately equal to the 33"/km of the earth's surface. Thus a light ray entering this layer at very great slant range, and at a grazing angle, might be ducted for a considerable distance around the curvature of the earth, producing a superior mirage of a source below the geometric horizon. If the efficiency of this optical duct were to vary it is conceivable that a diffused mirage image of Venus, still many degrees below the NE horizon, might briefly loom and dwindle, giving the impression of a "round white light" which approached and then receded to a "pinpoint" before vanishing.

In this connection it is very interesting to note the following description of a phenomenon reported by the astronomer Donald Menzel:

Flying in the Arctic zone near Bering Strait on March 3, 1955, I observed a bright UFO shoot in toward the aircraft from the southwestern horizon. Flashing green and red lights, it came to a skidding stop about 300 feet, as nearly as I could judge, from the aircraft. Its apparent diameter was about one third that of the full moon. It executed evasive action, disappearing over the horizon and then ruturning until I suddenly recognized it as an out-of-focus image of the bright star Sirius. The sudden disappearance was due to the presence of a distant mountain that momentarily cut off the light from the star [Menzel 1972; see also Menzel & Boyd, 1963].

Evidently in this case the observer could have been "in" the elevated optical duct, and it is interesting that Menzel also invokes a "random walk" mechanism through a discontinuous series of layers to account for the axial symmetry of the magnified image and answer an objection raised by McDonald. McDonald regarded Menzel's explanation of his own sighting as effectively impossible in terms of atmospheric physics [McDonald 1968] and argued that no other astronomer had ever reported a similar observation. Menzel responded reasonably that few astronomers do their observing from 20,000' over the arctic. (For the record, from 20,000' over the Bering Strait area on March 3 1955 Sirius would have been setting due SW on the sea-level horizon at about midnight local time [from sea level itself Sirius would have set some 30 minutes earlier].)

I am not aware that a similar phenomenon would be considered commonplace, especially for observers at ground level in East Anglia, and given the likely large negative elevation of Venus in the present instance (-14 at 2300Z allowing for standard refraction; -14.5 geometric) it might be beyond the normally understood limits of the atmosphere. This is certainly the case for an image which loomed and receded with a 45 displacement in azimuth. Nevertheless the civilian visual report in Section 3 below should also be compared with Menzel's description.

Note that not until about 0100Z would Venus be just below the horizon, which is much too late to square with any interpretation of the evidence sources. That is, the visual sightings are said to have precursed the radar sightings, and the latest start time we have for the radar observations - given in BOI-485 - is 0010Z, meaning that visual observations cannot possibly have begun much after about midnight, when Venus was still at about -8. Moreover this start time is believed to be in error. In my opinion the true start time is approximately (or even exactly) one hour earlier.

Of course it is not at all easy to account for a pair of such mirage images which "joined up . . . and disappeared in formation together", even assuming that the report's description of objects on SW and E "headings" can be comfortably squared with the requirement for negligible change in elevation. In short it seems likely that some fairly large extensions to accepted optical mirage theory would be required.

d.) searchlights

Searchlight spots on a low cloudbase are probably the most attractive prima facie explanation of fast-moving, erratic circular lights over a military airfield. In clear air it is possible that insufficient light would be scattered by the beam to render it visible, so that only the patch of illumination appeared as a moving "round white light". There are very good reasons (including a witness report; see Section 3 below) for supposing that searchlights were operative once the 60th AAA was up and running in active defence of the airfield. Before any radar contact with unidentified targets had yet been made by USAF surveillance radars, the 60th AAA was "alerted" by Lakenheath GCA to "watch for unusual targets" when the initial call came from RAF Bentwaters. Whenever it was that the US Army's TPS-1D defence acquisition radar first acquired UFO targets, as BOI-485 says it did (see also here), it is entirely possible given the military sensitivity of Lakenheath that an airfield defence alert condition existed from the moment of the warning.

Unfortunately for the spotlight-on-cloud hypothesis, however, there was no low cloudbase at Lakenheath at this time. In fact there was no cloud at all. There had been remnant altocirrus or high cirrus at some coastal locations earlier in the evening (patchy cirrus at 25,000 had made observation of Mars "intermittent" at Bentwaters between 2100 and 2200) but this was dissolving, and inland skies particularly were completely clear by 2300. The midnight GMT weather reports for all Met Office surface weather stations, including nearby Mildenhall, are nil cloud at any height and "conditions unchanged in last hour". This corroborates the Lakenheath USAF weather report for the period midnight-0300: "clear sky", "ceiling unlimited".

It is still possible that a thin aerosol layer which otherwise escaped visual detection might scatter sufficient light from a powerful searchlight beam [see Thayer 1969b]. Most observers that night are generally agreed on the unusual clarity of the sky; but it is true that between 2130 and 2215Z Lts. Metz and Rowe, airborne at a few thousand feet near Bentwaters, did observe that the Orford Ness lighthouse was "flashing through a low haze along the east coast". However it is also true that coastal conditions at this time do not apply directly to inland conditions an hour or two later, and the Met Office coastal weather reports of 'moist' ground at 1800Z are uniformly 'dry' by midnight. No mist is reported anywhere. Nevertheless, in context with the synoptic development (see also here) the weather reports indicate coastal evaporation lagging behind inland locations to the west, where development of stratified haze at the upper boundary of a narrow surface humidity layer might possibly have been more advanced.

At the same time the contrast with the sky background of light patches scattered from such a haze layer would not be expected to be great, and the stratification would have to be very well defined for the searchlight beam not to be visible. Also, the "round white lights" appeared large when closest ("golfball" at arm's length) and shrank to a "pinpoint" as they moved away. This is rather the opposite of what would be expected from a spotlight on haze or cloud, where the area illuminated would become larger and more diffuse as the slant range to the layer increased.

e.) balloons

It is conceivable that a lighted balloon rising to the NE of an observer on the ground and encountering a faster westerly air flow might give the illusion of an object approaching on a SW heading, pausing, then veering back towards the east. Winds of around 15-20 knots are on the face of it inconsistent with the report of "terrific speed", and a typical 1.5 candle meteorological balloon lamp does not sound much like "a round white light the size of a golf ball at arm's length".

But perhaps the observers were seeing the balloons themselves in unusual conditions, rather than just their tracking lights - for example, suppose that they were picked out by 60th AAA searchlight batteries following the alert to "watch for unusual targets" (see Section 2.d above). However a balloon a couple of feet in diameter at several thousand feet slant range would still be a very small visual object no more than one or two arc minutes across and thus close to the limit of resolution of the naked eye (no optical aids were used according to BOI-485). On the other hand, by the inverse square law a light 10 times further away would be 100 times as faint, so if the distance to a nearby balloon lamp at just a few hundred feet range were overestimated as a few thousand feet, then the balloon could have a respectable angular diameter on the order of 10 arc minutes (about 1/3 the diameter of the full moon), the perceived absolute magnitude of the luminosity (of tracking light or of spot-lit balloon) will be 100 times as great, and for a given angular rate a tenfold increase in perceived distance equates to a tenfold increase in perceived velocity.

The period covered by the upper wind measurements in the Lakenheath weather report is midnight to 0600, and it is entirely probable that these would be routine balloon release times. Therefore if the start time of 14:0010 for the first radar sightings given in BOI-485 is accurate, and given that ground visual sightings began before the radar sightings, then a midnight balloon launch could fit this sighting quite well. However there is considerable doubt that the above date-time group is accurate, as a range of interlocking evidence points to the conclusion that radar observations at Lakenheath began on the order of one hour earlier, probably some minutes after 13:2300Z. In this case the visual reports would not coincide with a routine balloon release.

The fact that this theory would require (at least) two weather balloons is another awkward complication. And it may be thought unlikely that Air Force observers - and investigators - would fail to recognise their own rising weather balloon(s) in this way, from (ex hypothesi) rather close range, when they would be familiar with such releases several times a day.

3. Civilian Ground-Visual

The origin of this retrospective report dates to February 1978. It is contemporary with Wimbledon's letter to the Sunday Times and was in fact prompted by the same flurry of media interest. The principal witness, a resident of Ely, Cambridgeshire (about 10 miles west of Lakenheath), wrote to a "UFO Bureau" set up by the Daily Express after seeing a brief summary of the Lakenheath story (insofar as it was then known) in that newspaper. The witness remembered well the "Lakenheath flap of 1956":

. . . I have never seen such panic. From the Prickwillow Rd, Ely, we could see searchlights sweeping the sky in every direction. We saw a bright white star-like light speeding low across the fen straight towards us. When the light reached us it stopped dead. It did not slow down. It just stopped. The light went out, then [there was] a bright flash of light and it shot off on the exact course from which it came, as if on a slack string.
About ten minutes later we sat on a wall in New Barns Avenue opposite the phone-box. Looked up and [saw] almost above us three amber glowing objects - quite low - but enveloped in cloud, although the night was clear. After a few minutes the objects began to revolve in a clockwise direction, slowly at first, then much faster, and the objects were swapping places with each other at great speed - and it [the cluster of objects] went straight up and was gone. . . .

The witness also recalled that a jet fighter had overflown the town at low altitude, apparently in pursuit of the initial light. The letter was passed by the Express to Jenny Randles who published a note about this new information in 1981

In subsequent correspondence and a telephone interview with the present author, the witness, Mr. John H. Killock, recalled the incident in similar terms without obvious embroidery. He later gave written answers to specific questions, with the caveat that it was "a long time to remember every detail." Nevertheless it appeared that the event remained quite vivid to him.

He remembered the date of August 13 because he married shortly afterwards, but he could not remember the exact time of the initial sighting except that it was late (definitely after 9:30 pm, possibly as late as 11:30 pm - remember this would be BST, so 2030-2230 GMT). Four others were with him, all young men and women from working class backgrounds; he has since lost contact with two of them, one is dead, and the fourth is living in Germany. Three of the five were "totally disinterested" in what they had seen - "one was a young lady, the other two young men, all in their late teens and, as you said, no contest. My own opinion, and also that of my friend, was that we had seen 'UFOs'. We did not discuss the incident after that night." Mr. Killock explained that he has always been very interested in aircraft and at the time took particular note of the "distinctive" engine tone of the jet which overflew the city outskirts at low altitude after the first light had disappeared. The following account emerges from his narrative, sketch map and answers to supplementary questions.

They were at a position near the junction of Lisle Lane with the Prickwillow Rd., on the E edge of the town (Ordnance Survey map ref: TL547 807), looking out across the low fens to the E. The first light approached them at low apparent altitude from the east, looking initially like a bright star of the approximate magnitude of Sirius (-1.58), descending from an estimated 1500' "down to almost cutting the grass" at a speed that seemed to be several hundred miles an hour, growing rapidly in brightness. "My thought at the time was that it must be a jet aircraft and that the pilot must be a little crazy to fly so low and fast, and that it would roar over our heads. But this did not happen. There was no sound . . . . It lost height very quickly, and flew so low over the fen I looked down on it." (Note: OS Landranger sheet 143 indicates observer position approximately 20 metres above MSL, with most of the South Level to the east either at or below MSL.)

The light then gained a little in relative elevation, climbing to an estimated 200', and "stopped dead without slowing at about 70 [elevation] to my position", seeming to be very close by. Its line of flight from the E had not been directly towards them, so that it now appeared to the S. Its brilliance at this point was estimated as about twice that of the planet Venus (which at that date was magnitude 3.8 and very bright, though not above the horizon until about 0115Z/0215 BST) but it was still only a featureless light. Immediately upon stopping, the light went out for perhaps 2-3 seconds. Then it reappeared with a bright flash and "flew back towards Lakenheath", climbing along the same flightpath, dwindling in brilliance until it once again looked like a bright star before finally disappearing. The whole event took place in about 20 seconds.

"Soon after" this they heard a jet aircraft approaching from the SW. According to Mr. Killock he is "certain" that he recognised "the very distinctive sound of the de Havilland Ghost turbojet" and turned to look up. He could make out the "dim outline" of the aircraft and a red navigation light on one wing as it passed over Ely somewhat to the S of his position, at an estimated 400-500' altitude, then flew on to the E in the direction of Lakenheath.

Perhaps "20-30 minutes" later (note: this is inconsistent with the "about ten minutes" recalled by the witness in 1978), from a position opposite the cemetery in New Barns Avenue, Ely, a few hundred yards NW of the previous location, they saw "searchlights sweeping the sky from the direction of Lakenheath, and also Mildenhall, I think." There were three searchlight beams visible in the E - two from Lakenheath, the other a little to the right (Mildenhall). At this time Mr. Killock looked away to his right, due S, and was "amazed to see the glow of three amber objects hovering over the road at a height of 50-70 ft", apparently only some 30 yards away. His impression was that these were parts of one object. He said:

. . . the object seemed to generate its own cloud infolding on itself, completely silent and rock-steady. After some minutes [with] the sound of a metallic click and a whine the UFO began to rotate in a clockwise direction, but after only a few feet it stopped and returned to its original position. Again [we heard] the metallic click and whine but this time the rotation was very fast, and for a few seconds the objects seemed to swap places in what appeared to be a figure-of-eight before returning to spinning in the shape of a ring. The object began ascending until it was out of sight. I think it was one object and not three, but I cannot be sure as I could not see beyond the cloud that the object formed.

A resemblance of this description to searchlight spots on cloud is noteable, neglecting the apparent position in the sky and the sounds. Howsoever, after this object had disappeared the searchlights continued to be visible over the E skyline for a time but were soon extinguished, and nothing more was seen.

This report is interesting inasmuch as it indicates a level of general alert not implied by any media accounts widely available in 1978. Indeed not even the more detailed accounts in the specialist literature at that time (i.e, those by Thayer, McDonald, Hynek & Klass) contained any mention of an airfield-defence alert situation, or sufficient information to easily deduce the likelihood. Even the lengthy extracts from BOI-485 quoted in the Condon Report were so peppered with deletions (by Condon's hand) that of only two passages from which the active involvement of the 60th AAA might have been inferred, both were mutilated. In one place the acronym "AAA" remains but with the unit number and location both deleted, surrounded by so many other deletions and coy substitutions that the context is virtually unintelligible. In the second place the phrase "60th AAA" is deleted entire and replaced with the code-letter "A", elsewhere used to indicate simply "Lakenheath", with no possibility whatever of recovering the meaning from the context. Nor does reference to the AAA or any mention of airfield defence appear in any other pre-1978, or later, source known to the author. Only access to a full copy of BOI-485 made from the original Blue Book file could reasonably be expected to allow the inference that Army radars, anti-aircraft artillery and associated searchlights were deployed in defence of the airfield.

Two points to note are a) there is a small inconsistency in the witness' 1978 and 1988 recollections of elapsed time between the two sightings; and b) the gyrating amber lights of the second Ely sighting have no close counterpart among other visual observations reported that night.

On the other hand, the motion and appearance of the earlier bright white light are very characteristic of the pattern observed, according to BOI-485, by personnel at Lakenheath, who similarly reported lights which headed SW, stopped, then suddenly moved E.

It is also interesting to compare the description of the Ely object with the behaviour of the rare mirage image of Sirius described by Menzel (see Section 2.c above). Qualitatively the similarity is perhaps even more marked in this case, because the apparent movement of the light was confined to the same azimuth. But at the same time the witness describes very considerable changes in elevation during the sighting: From 70 down to zero, or even less than zero, and returning, is so extreme as to make the notion of "image wander" rather absurd. Could any conceivable displacement due to mirage cause something similar to this? A remembered azimuth a little north of E could perhaps be squared with the 57 azimuth of Venus rising, but since this would not happen until after 0100 there is a discrepancy of several hours also. At the mean of the witness's two bracketed time estimates - 2130 - Venus was more than 18 below the horizon, or fully 88 of elevation from the reported 'rest' position of the light in the sky. And it happens that there is circumstantial evidence suggesting that the reported time may be dependable.

This evidence is the observation of the jet. Killock was definite that the "very distinctive" engine note and the "dim outline" of the aircraft identified it as a Venom night fighter. Interestingly, we do have evidence that a Venom was scrambled from RAF Waterbeach, about 10 miles south of Ely, at 2120Z. This was Grahame Scofield and Les Arthur, who according to the Squadron Diary were scrambled to intercept a UFO. A southwesterly (windward) then northerly climb and turn out of RAF Waterbeach onto a NE heading might very well have taken a Venom in a circuit over the S of Ely as described. In fact this is exactly what Scofield stated:

I seem to recall that we were on standby for about an hour before we received the scramble instruction. The time recorded for take off was 2120 p.m [21.20 GMT on the 13th August] when it was fully dark.

We took off in a westerly direction and were vectored onto a heading of 045 degrees, which involved a climbing turn to the right. I do not recall whether we had more precise instructions regarding an intercept but the usual drill was to be sent to intercept incoming bombers, either Russian on probing missions or allied doing training exercises.

It would have taken about 20 minutes to reach maximum operating height and it was in the course of this climb that we discovered that our two wing tip tanks were missing.

The tip tanks were eventually recovered from a field outside Cambridge, certifying a direction of take-off which must have consumed a few miles in the turn back onto an almost reciprocal 45 (NE) heading, bringing the Venom in over the SE of Ely at, say, about 2125 or 2130, which is right in the middle of the time-window recalled by the witness on the ground.

Scofield recalled that they were on standby at Waterbeach for about an hour before 2120 and took off "when it was fully dark" - or it appeared so from the inside of the cockpit canopy. By convention, "civil twilight" had begun at 2106 with the sun 6 under the horizon (astronomical dusk would be at 2305) so it would probably not yet be pitch black at 2130 when the sun was 15 under the horizon. Also, the half-moon was still a few degrees from setting in the SW sky and would remain visible for a further 50 minutes or so. According to the ground witness it was dark at Ely, but there was sufficient sky brightness to allow the 'dim outline' of the aircraft to be seen. This is consistent with conditions at about 2130Z.

A possible difficulty here is that this witness thought the Venom was low, no more than about 500', and carried a red navigation light. Even allowing for inevitable inaccuracy in such judgements it was evidently low enough to be heard very clearly and distinguished in silhouette. But a Venom on a typical sortie would be both very much higher and unlikely to be showing lights. Scofield recalled that after take off from Waterbeach they climbed out to the northeast:

. . . and it was in the course of this climb that we discovered that our two wing tip tanks were missing. This was a severe matter as without them our airborne endurance was reduced from about an hour and a half to 45 minutes. We took some time verifying that the tanks were in fact missing as we always flew without lights and in the dark it was difficult to see the silhouette of the wings.

Scofield's pilot, Flying Officer Les Arthur, confirmed that it was after the turn onto their NE heading that they realised the tip tanks were missing. If therefore they were heading towards Ely at height, and were having difficulty seeing, it would be consistent if they dropped low in order to try to pick up the silhouette of the wing tips against the lights of the town, where, being under the radar close to an urban area, they were obliged to show a hazard light.

In other circumstances the Ely ground-visual report would not be assigned much weight. A rapid white light in the east and an amber object in the south would bring to mind the Perseid meteor shower and the planet Mars, and due allowance would be made for the effect of 22 years of imperfect memory on the experiences of excited youth. Moreover, although there were said to be several witnesses we only have the word of one of them. The report would be discarded as unevaluable and not probative. And so it is, considered alone; but the details of the Venom reported at a time and place consistent with a known sortie, and the report of searchlights, indicative of "panic" to the witness, are in my opinion intriguing pieces of evidence.

On the other hand, if the witness' confident recollection of the date were in fact wrong, then there may have been a cloudbase, and the three dancing amber lights might in this case be plausibly related to the three searchlight beams. But if there was low cloud in the east dense enough to reflect the dancing searchlight spots visibly at a range of 10 miles, then the bright white light in the east probably was neither a meteor (Perseid or not) nor a mirage of any astronomical source. And, if the date was not August 13, why were searchlights "sweeping the sky" in the first place? There may have been an airfield-defence exercise on a nearby date, and the presence of a Venom on a coincidental, or perhaps integrated, PI exercise is always possible. But on balance it is reasonable to conclude that these events, recalled by the witness as a very singular episode in his experience, probably did occur on the night of August 13 1956 and that they should be treated together with the similar visual reports made by personnel at Lakenheath.

No other explanation for these visual sightings suggests itself on the limited information available. Luminescent birds, insects, aurorae, lightning, missiles, or windborne debris - among other hypotheses - can all be ruled improbable for obvious reasons. These sightings should therefore be carried as tentative "unknowns". There are certain features of the civilian and military reports that hint at extraordinary optical mirage effects, but in view of other features this can hardly be said to be a promising speculation.

Archive Index