Analysis of the Intercepted Targets over Lakenheath - 2
An Opinion by Martin Shough
Track F. The 23 Squadron Venom AI radar targets
c) moon echoes
d) anomalous propagation/CAT
e) birds and insect angels
g) meteors, auroral echoes, lightning echoes and sferics
Evidence relating to the radar observations here designated Track F comes from the reports of the two 23 Squadron Venom crews scrambled from RAF Waterbeach at 0200Z and 0240Z.
It has been widely assumed hitherto that Track F is in fact cognate with Track E; however there are differences both in context and in content which go deep enough to breed serious doubt of this assumption in my opinion. These arguments can be examined in The Case of the 'Missing Interception', An Opinion on the Timing Problem and The 23 Squadron Scrambles - No GCI Control. Contemporary documentary justification for this position now exists in the form of evidence that another RAF fighter piloted by the 23 Squadron CO, Wing Commander Davis, was involved in a different radar-UFO interception that same night. It cannot presently be proved that this previously unknown incident does correspond to the Track E incident as described in what can be called the 'classical scenario'. But I believe that the independent arguments for treating Track F as a separate incident - or rather, two separate but related incidents - are in any case strong.
Both John Brady (rad. op. in Venom #1) and Ivan Logan (in #2) were inclined to interpret the AI radar echoes as probably from an airborne object of some sort. In an interview with Jenny Randles in 1996 Logan said:
We didn't attach too much importance to it. It certainly was unusual. At the time we thought - as I said - that it was a barrage balloon or a met balloon. I certainly did not put it down to atmospherics or interference. I didn't think it was anything like that.
Q: So you think it was a solid object?
A: I think it was an object on the radar, but whatever it was I don't know. It might have been a helicopter perhaps.
John Brady, interviewed by Dave Clarke and Andy Roberts in 2001, recalled:
. . . I suppose it would have to be something fairly solid. You get various meteorological things, but that would be unlikely at that position and at that altitude, on that radar.
DC: So it was not likely to have been the result of a temperature inversion at that altitude?
JB: Most certainly not.
DC: So we are down to Met balloons or something suspended in the airstream?
JB: Yes, something like that.
In the absence of an explicit ground radar description of this target or targets, the opinions of both aircrews that there was no appreciable movement cannot be definitely confirmed. Nevertheless, in each case the Venom radar operator observed his target 'come down the tube' of the APS-57 radar very rapidly and formed the impression that this rate of closure was entirely due to the airspeed of the Venom. Ground radar could in principle have had a different view of things. However each of the Venoms made several runs of this type at the target, over a combined period of perhaps 80 minutes (out of 100 minutes total flying time), with the same effect on every occasion, so the likelihood that between times the target was moving around and only happened to stop each time they reacquired it on the AI radar is presumably small. Moreover one would imagine that the controller on the ground at Lakenheath would at some stage have informed one of the pilots if this were the case. The fact that both Venoms were repeatedly directed back to the target by GCA radar, with reasonable apparent success despite the fact that GCA had no heightfinder radar, also suggests that the target was at least stationary in altitude.
This conclusion is inconsistent with a fixed-wing aircraft of any type. It should also be borne in mind that the total duration of the unsuccessful attempts to intercept the object exhausted most of the fuel of two Venoms in succession. This would imply an extraordinary endurance on the part of a conventional aircraft which, in addition to hanging around Lakenheath for something like 2 hours (allowing for a period of GCA radar tracking before the scramble and then the time for the first Venom to arrive from Waterbeach), had to be capable of flying in and out of the area.
Logan mentions half-seriously that the target could have been a helicopter. However apart from flight duration, the pilots reported complete absence of any visual contact with the target. A helicopter, even if for some strange and dangerous reason it was unlit, should have been visible at some stage against the starlight or ground lights - arguably one of the Venoms would even have collided with it - yet nothing was ever seen. Moreover a crew who were motivated to maintain darkness and silent running over a military airfield in spite of repeatedly being buzzed by attacking jets would presumbly have been 'hostile' intruders; but where would a Soviet helicopter have come from? It is also true of course that the radar return from a helicopter should have been very substantial, given the typically lumpish geometry and the large swept area of tail and lift rotors. But both radar operators stated that although the return from this object was fairly definite it was also fairly small. Finally, a helicopter hovering at a few thousand feet over the airfield in the early hours of the morning should have been easy enough to identify from the ground by the clatter of rotors.
There appears to be no way of explaining these radar contacts as due to aircraft.
A typical weather balloon would be climbing at 1000 fpm or more [Lally 1967]. During the course of the two interceptions it ought to have risen from just over the airfield to about 60,000' - which is about 150% of the normal maximum operating height of a Venom. Could one balloon have risen out of range after the first scramble, whilst the second Venom was vectored to a second balloon? No, because during the extended presence of each of the Venoms at Lakenheath a balloon would have risen some 30,000'. So there must have been a single balloon which was not positively buoyant and so failed to ascend. A balloon with a slow leak drifting at neutral buoyancy with 20-30 knot winds would be for all practical purposes stationary in relation to a jet closing on it at several hundred knots; but it would not of course be stationary in relation to Lakenheath. If it started out near the airfield at 0200 then by the time Venom #2 broke off it should have been most of the way to Norwich, yet the action in both cases was in the area of the Lakenheath airfield.
A truly stationary balloon would imply a tethered balloon - either a defensive barrage balloon, an experimental balloon of some kind, or a deliberate spoof.
According to Dave Chambers, pilot of Venom #1, they were directed to a position over or close to the airfield itself. All the airmen agreed that what it most resembled was something like a balloon, and Ivan Logan suggested more than once that it might have been a barrage balloon. But a barrage balloon over a busy SAC bomber base in peacetime seems very unlikely. Much more probable is a runaway tethered balloon from elsewhere that became trapped.
In Analysis of Track E, Section f. it is pointed out that Paul Fuller found documentation showing that both the RAF Balloons Unit and the Met. Office Research and Development Establishment flew tethered balloons and that from time to time balloons did escape. Both of these units operated at RAF Cardington. It turns out that the Balloons Unit had not flown a tethered parachute-training balloon for almost a month, probably due to winds; but the Met. Office experiments were conducted at the rate of two or more a day in August and they did record occasional losses (though no dates are specified). RAF Cardington is approximately 40 miles SW of Lakenheath, and winds on the afternoon of 13th August 1956 were from the SW.
The Met Office experimental balloons were tethered at heights up to 5000' at Cardington and one would expect them to have been lit for safety, at least during night-time measurements. If the balloon had escaped many hours earlier during daylight, when it was not lit, then one would expect it to have been further afield than Lakenheath by 0200 given the afternoon and evening winds. On the other hand, if power was supplied to the balloon by cable then perhaps it lost its hazard beacon when it broke free. Such a balloon could conceivably have stayed at moderate height, due either to a designed near-neutral buoyancy or to a slow leak, and have been blown towards Lakenheath trailing its tether, becoming snagged by chance near the airfield. But even a darkened balloon ought to have been picked up as a silhouette against ground lights or starlight at some point during a number of interception attempts. Also one has to wonder about the chances of both the balloon and the Venoms surviving perhaps eight separate close shaves at several hundred knots.
Finally it should be noted that the CPN-4 GCA surveillance radar was fitted with a moving target indicator. A stationary echo on the PPI would naturally be suspected ground clutter, so if MTI was initially switched off it would be switched on in order to remove it. This is what MTI was for. MTI had been similarly used to try to filter the Track E stationary target earlier in the night. MTI ought to have removed any target - such as a tethered balloon - which was showing a negligible ground speed. Evidently it did not. It seems unlikely given the stable weather conditions that a balloon would have been jerking around under the influence of wind gusts. One possibility is that a snagged balloon might develop small-amplitude oscillations due to clear air turbulence associated with an undetected refractive-discontinuity layer or pocket which happened to coincide with its altitude of a few thousand feet. Perhaps such oscillations could occur at a frequency sufficient to defeat the MTI filter. (For illustration a range rate of about 130 knots would introduce a doppler shift equal to half of phase, or 5 cm., in the CPN-4 interpulse period of 1/1200 sec.; a rate of 65 knots, a shift of one quarter phase, and so on pro rata. Practical sensitivity depends on the gating of the difference-taking circuit that compares the inputs from the phase detector.) But in the absence of aerological evidence for such a layer this adds another degree of speculation to what is arguably already a somewhat unrealistic hypothesis.
c) moon echoes
Echoes from the moon are theoretically possible (see Analysis of Track E, Section e, for a detailed discussion) but there is no point in pursuing the hypothesis in this case since the moon had long set and was already 34° below the SW horizon by 14: 0200Z.
d) anomalous propagation/CAT
If the GCA surveillance radar was picking up a ground reflector in AP conditions then it is possible that at low level the Venoms' APS-57 airborne radars might pick up the same ground target, which would explain how the controllers were able to guide the pilots successfully to the target. However the available radiosonde data do not give evidence of low-level superrefractive conditions. There is evidence however of a subrefractive surface layer, which would lift the bottom edge of the beam, reduce the radar horizon and so tend to reduce the likelihood of local AP clutter. Moreover the CPN-4 ground radar was fitted with MTI which would remove first-trip stationary targets. Phase-shifted AP returns by multiple trip from elevated superrefractive layers (>30,000') for which there is partial evidence, and which might not be filtered out by the MTI, would have no causal relation at all to the targets detected by the airborne radars, reducing the repeated correlations of the ground- and air-radar responses to pure coincidence.
Both Logan and Brady were of the definite opinion that the AI targets they detected could not have been due to anomalous propagation effects. Nevertheless the limits of such effects were not understood even by experts in radar meteorological theory in 1956 and it is entirely likely that rare effects might not be recognisable by operators. Scattering of radar energy to ground from an hypothetical narrow layer of refractive-index discontinuity, for example, could in principle have occurred both for a ground radar and for an AI radar in an aircraft flying just below the layer. In order to explain why AI radar echoes were only detected each time the the interceptors were directed towards the ground radar target, one could imagine a very discrete, stable domain of clear air turbulence at a layer boundary. The different wavelengths of the CPN-4 and APS-57 radars, 10 and 3 cm respectively, do bracket the optimum scattering efficiency from CAT which for sensitive experimental radars has been found to occur at about 5 cm [Atlas, Hardy & Naito, 1966], and even though the peak powers are low in this case the APS-57 in particular would be illuminating the CAT domain at an efficient grazing incidence which reduced to effectively zero degrees as the aircraft closed to intercept.
However, a little thought about the geometry will show that airborne radar echoes whose position appeared to correlate at least in relative azimuth with the position on ground radar could only be produced by such forward scattering in the case that the aircraft is flying on a radius directly towards or away from the ground radar antenna. The jet will be vectored towards the displayed location, but because the displayed range on the ground radar is always twice the slant range to the reflecting feature the jet will be at twice the range of the feature and, for all vectors except radial headings, will never find itself on a heading towards the true location of the feature. Thus they would be highly unlikely to ever pick up echoes from the same feature and the repeatedly-confirmed correlation of ground and airborne radar targets becomes once again a question of very awkward coincidence.
Where correlating targets are detected by different ray paths at changing azimuths and incidences on two radars to triangulate a position in local air space, this implies that both ground and airborne radars are displaying echoes directly backscattered from the same stable local feature. In other words the atmospheric feature behaves in some degree like a solid "object" capable of reradiating or reflecting significant energy at normal incidence. But direct backscatter efficiencies are very small indeed except for very sensitive research radars of the right characteristics in the right conditions; with these radars in these conditions direct backscatter seems out of the question in terms of the known limits of such effects. And could such a feature remain stable in approximately the same place over Lakenheath for well over an hour in the face of 20-30 knot winds? As well as avoid being disrupted by mixing during repeated passes by two jets?
e) birds, insect angels
Almost all insects are found in the first few thousand feet of the atmosphere and so insects might be found at the sort of altitude reported. The air temperature was also well above the approximately 6°C limit below which insects normally do not fly. Insect cross-sections tabulated in [Blackmer et al., 1969] at 3.2 cm - close to the wavelength of the Venom's APS-57 - are typically on the order of 0.1cm2, which suggests that a considerable density of insects would be required in a localised area to generate a definite echo on a radar set of quite low peak power. Cross-sections at 10 cm on the CPN-4 would be smaller, the echo fainter per-kilowatt at a given range; but there is no information on the presentation of the ground radar target in this case. In general however, insect motions will be dominated by the movement of the air stream in which they are flying, so in this respect an insect swarm is not consistent with a target that apparently remained fairly stationary and compactly-defined over the airfield for well over an hour in the face not only of 20-30 knot winds but also of two jet interceptors.
Flocking birds might conceivably maintain station despite the wind, circling over a restricted area for some time, and the echo might resemble a fairly discrete point target. But this would be unusual nocturnal behaviour for common flocking birds, such as gulls - which are anyway a little slow and would have to keep headed into winds of this speed in order to maintain station. Due to the inverse-4th-power signal attenuation, even single large birds at short range could present surprisingly respectable target cross-sections to both ground and airborne radars, and one thinks of a hunting owl moving in a small-radius orbit or a large hovering hawk. But aside from the extraordinary stamina of such a performance over many tens of minutes there is again the problem that one would not expect a bird or birds to keep their concentration without flinching whilst two jets thundered towards them repeatedly at several hundred knots.
Weather reports show that there was no chance of precipitation.
g) meteors, auroral echoes, lightning echoes and sferics
Weather reports also show that there were no thunderstorms anywhere in the area. Records of aurorae for the period show that the auroral geomagnetic index was globally "quiet" and no visual displays of north-hemisphere aurorae were observed on the night of August 13-14 1956. There is also no resemblance between the persistent localised radar indications reported and the transient echoes due to auroral streamers. Moreover it is highly unlikely that the APS-57 airborne radar would have the power to detect auroral echoes under any circumstances. There is also no resemblance whatsoever to any possible echoes from meteor-wake ionisation or turbulence.
Neither Logan's nor Brady's description of the radar target in any way resembles the lines, speckles and swirls typical of the display products of radio frequency interference. Both stated in interview that they discounted interference. It was a discrete echo like that from a compact reflective object. Such a discrete echo might in certain special circumstances be caused by remote interference from an emitter with similar pulse characteristics whose output is 'sampled' in a highly selective way, say by a highly anisotropic radio duct in the atmosphere. However this is not only highly unlikely a priori it is highly unlikely that such freak RFI would be picked up repeatedly in the same area by two different planes over the space of an hour or more, on a number of different headings, each time coincident with being directed towards a ground radar target observed in the same location. It is impossible for the same source of RFI to generate a phantom 'point target' simultaneously on the CPN-4 and the APS-57 (at any scope position, never mind coincident geographical locations) owing to the very different signal bandwidths and electromechanical display characteristics of the two equipments. Similar arguments apply to the hypothesis of internal noise sources or component failures. Deliberate deception jamming or "spoofing" is possible in principle (see Analysis of the Intercepted Targets over Lakenheath - 1, Section h.); but simultaneous spoofing of two such different fixed and mobile radar sets, on the ground and in the air, by any means likely to have been available in 1956 is implausible.
At least one member of the aircrews did toy with the idea that the incident was a set-up by the Americans as a test of RAF defence readiness, simply because it was unheard of to be sent in at low level over an airfield under USAF control, and because no 'natural' explanation seemed wholly convincing. Perhaps the target was a deliberate decoy - a drone or balloon of some kind flown for the purpose. But there is no evidence of this, and it seems unlikely that a workaday airfield landing radar would have been chosen to carry out such a politically delicate and potentially hazardous exercise. It also seems unlikely that this would be done over a SAC bomber base where only two weeks before a B-47 on landing and take-off exercises had crashed into a nuclear weapons storage igloo with fatal and near-catastrophic consequences.
© Martin Shough