Anomalous Echoes Captured by a B-52 Airborne Radarscope Camera

7. Conclusions

The reconstruction in Section 5-3 means that the supposed fast track 771-772 occurs not at the start of the event but towards the end, and bracketed by two periods of extended stationing by the 40-degree Phase B echo. Interestingly, in no instance (including the possible similar echo athwart the 1.75 NM range ring at ~350 degs on frame 775; see Section 3) are Phase A and Phase B echoes shown on scope at the same time. So although the echo presentations and displacements are different in the two phases, the possibility remains that all echoes are due to movements of a single target through the radar cover.

7-1. The Phase A Echoes

As mentioned in Section 6-1 discussing meteors, an assumption that the same single target moved at high speed from scan to scan cannot be justified simply on the basis of two possibly unrelated echoes. But in the case of 771 to 772 we can to some extent test the hypothesis against other evidence contained in the detailed echo presentation. In fact there is an elongation of both echoes in the approximate direction of the inferred velocity which could be consistent with a smearing due to rapid passage through the beam in the implied direction.

Both these echoes are roughly elliptical. See Fig. 22 below. At first glance frame 771 appears to show a roundish echo adjacent to the range ring, but close examination and contrast-enhancement bring out the brightness due to spot-integration where the "tail" of the elliptical echo overlaps the range ring (painted of course on the phosphor by the same electron beam trace). The major axis of each ellipse appears to make a roughly similar angle with a line connecting both echoes, but rotated about 10 degrees clockwise in 771 and 10 degrees anticlockwise in 772. At the same time, each of these axes is rotated with respect to a radius drawn through it from the scope centre, this time anticlockwise in 771 and clockwise in 772.

These angular relationships are arbitrary in relation to the scope centre and the orientation of the echoes is not that of the ordinary "target arc" indicative of a stationary or slowmoving point target. The spray of pulses returned from a point target that is stationary or slow-moving relative to the angular rate at which the beam scans past it produces a sequence of spots on the tube phosphor all at the same range over an angular width approximately equal to or less than the nominal beam width, and these spots appear integrated into a short arc of brightness lying normal to the scope radius. Echoes 771 and 772 on the other hand are orientated obliquely to the radar line of sight.

Qualitatively speaking, this could suggest targets moving through the radar cover rapidly enough to show a small range rate during their dwell-time in the beam. Moreover the directions of motion implied by the range rate in both cases - having components of velocity approaching the radar in 771 and receding from the radar in 772 - would be consistent with consecutive scans of a single target, crossing the scope on the trajectory of a line connecting the two echoes.

Quantitatively the scenario is less clear: Rotating at 120 degrees/sec the edge of the beam would be travelling at around 8000 knots at the 1.05 NM range of 772. The target displacement of roughly 2.1 miles (assuming zero degrees relative elevation) from 771-772 indicates a maximum relative average speed of nearly 1900 knots in roughly the same clockwise direction, so the target dwell-time will be extended. A 1.6-degree beam rotating at 120 deg/sec would scan past a stationary point in about 0.013 sec., but because it is overtaking a co-moving target (~100 degs in 3.8 sec is approximately 26 deg/sec) the dwell-time will be ~22% longer, = 0.015 sec. In the limit case of a co-altitudinal target averaging ~1900 knots this allows a likely range displacement of no more than a few tens of feet, only a fraction of the 123 ft theoretical electromagnetic resolution in range of a single 0.25 microsec pulse.

Two objects statically separated by this small range differential would not be resolvable in principle; however, the pulse repetition rate of 1617 pps means that during the ~0.015 second dwell time some 24 pulses are sprayed across the moving target, so it is possible that it could be detectable. The signal amplitudes of successive pulses returned from a target with a significant range rate would not add directly across the beamwidth, and the resulting integration loss would tend to reduce echo brightness slightly on the PPI. This could be consistent with the weaker presentation of echoes 771 and 772 relative to echo 773. The latter may be bright because, in part, the target is relatively stationary and does not change slant range through the beamwidth, so that the returned signal amplitudes do add on the PPI.

Nevertheless for an effective point target, i.e. much smaller than the resolution cell on both range and azimuth dimensions, the fastest relative target speed consistent with the geometry would not alone cause smearing of echoes 771 and 772 over a range differential of about 400 ft or more, as photographed, more like 1/10 of this value, so we conclude that the radar echoing area of the target had an intrinsic length in the direction of motion which was already equal to or greater than the likely resolution. In this case subtracting the relatively small motion blur would indicate the underlying physical length as "seen" by 3cm radar. This model suggests that an overall range differential in the region of ~ 400 ft, as displayed, could have been caused by a reflector whose echoing area had a true major axis of perhaps 300 ft or more (400 ft minus a motion blur component of several tens of feet) passing through the coverage on a level trajectory between the two echo positions at around 11,500 ft MSL at a true average groundspeed of ~1900 knots.

(Alternatively, the same target could be detected at a steep depression angle in the bottom of the radar cover, passing below the right wing at an altitude very much lower than the B-52. In this case the target speed could be as low as around 1000 knots. The reduction in motion blur by about 50% due to the lower speed represents a difference of probably no more than about 10% of the total echo size, so bearing in mind the likely margin of error the difference in the implied true target length is not significant. It would be several hundred feet on either scenario.)

It might be thought defensible to separate these Phase A echoes from the better-defined Phase B sequence and disregard them as of doubtful significance. After all, blips that pop up for a scan or two could be stochastic noise and it can't be proven that they have anything to do with Phase B. On the other hand, there is a suggestion of pattern in the tendency (one can put it no more strongly in the case of such a small sample) for these echoes to appear on scope at times when the Phase B echo absents itself. They are both well-defined blips with interesting structure. And finally, the point should be made that the inference we have drawn here to complete the internal consistency of a "real target" model of Phase A - i.e., that such a target probably has an effective length of a few hundred feet aligned in the direction of motion - turns out to be consistent also with an elongation depicted by the primary echo in Phase B (greater, certainly in 773, but still probably in the order of hundreds of feet), with ground-visual witness reports of a "slender" or "wiener shaped" object, and with the air-visual report of an elongated egg-shaped object (estimated >200 ft in cue-reduced dark conditions) on the ground.

It's possible then to interpret the radar sequence as showing an object stationed off the left wing of the B-52 accelerating ahead, turning around the nose of the B-52 and giving a smaller, rapidly inbound echo off the right nose on frame 771, travelling to a position aft of the right wing on 772 then back to its station on 773, then again accelerating out of the altitude hole for another two scans (or one at least; the orientation of the possible echo touching the 1.75 NM range ring at ~350 degs on 775 implies an inbound vector which would be consistent with the sequence) before returning a second time to its station on 776 for a final seven frames then vanishing from the radar for good.

Turning then to possible interpretations of the Phase B echoes:

7-2. The Phase B Echoes

The likelihood that these echoes indicate a compact target, narrow in azimuth, with a significant radar cross-section, at first-trip range and thus aloft inside the altitude hole, has already been argued on several grounds. But although witness and photo evidence indicate a very strong echo, we note that the echo strength indicated in frame 773 is unique in the Phase B sequence, and generally there is a wide variation in the presentation of the primary echo during the photography. In fact this variation is for practical purposes 100%. Neglecting deception jamming effects of the sort mentioned in Section 6-6 (which ought not be our first resort) change in echo strength can be interpreted as:

a) variation in aspect (i.e. a relative rotation) of a single large anisotropic reflector;
b) varying augmentation by some external means of the return from a relatively small reflector of constant efficiency; or
c) varying attenuation by some external means of the return from a relatively large reflector of constant efficiency.

Or some combination of the above.

Option c) could include signal attenuation due to variation in range and/or variation in elevation (changing antenna gain). There is variation in displayed range over the sequence, though it is quite small, less than 20%; but it is in the wrong direction to account for the fact that the strongest echo occurs on frame 773 when the range is greatest. Change of target elevation is hard to rule out, but is questionable in view of the near-monotonic character of the reducing slant range and its strong correlation with the descent rate of the B-52 (see Fig.23 below), which suggest that these two variables are causally coupled.

The simplest explanation of this coupling would be that during the photo sequence (though not of course during the entire event beginning near FL200) the target is maintaining constant altitude whilst pacing the descending B-52, in which case the target is below the B-52 altitude and is increasing in relative elevation. Given the known radiation pattern it would thus be moving from a region of lower antenna gain into a region of higher antenna gain at the same time as closing slant range. This is at least a natural relation, but these range and elevation changes would both lead us to expect a systematic increase in echo strength from 773-783. The weak trend observed, as shown in the graph of photometric values in Fig. 24, is in the opposite direction.

So option a) is also attractive. The near-radial extent of the principal echo on 773 corresponds to perhaps ~ 800 ft on the PPI, and since this would be a projection in plan of an elongated object with an unknown orientation at an unknown negative elevation, changes in attitude could presumably result in large changes in echoing area.

On the other hand, the intermittent nature of the signal changes, and the occasional presence of a fugitive secondary echo generally connected to the nearer primary echo, suggest that we also explore option b).

On frame 773 there is a near discontinuity between the inner primary echo and the secondary echo, though they are faintly connected by the suggestion of a narrow bridge. But on frames 776 and 781, for example, the echoes become continuous. The nearer edge appears to remain constant in position (or to follow a roughly monotonic decrease in slant range from 1.05 to 0.87 NM) whilst the outer partner is fugitive. But the outer feature does display at roughly the same additional slant range each time it appears. This suggests that the possibility of a direct echo from a primary reflector whose presentation is being augmented by a ghost echo from a nearby secondary reflector, reaching the antenna by a slightly longer raypath.

The nature of such a secondary reflector is problematic because of the small echo separation (order of perhaps 1000 ft). As described in Section 6-12 ghosts normally arise due to an unusually efficient secondary ground reflector; and in a situation where the primary reflector and/or the radar are in motion, a persistent ghost requires a special kind of secondary reflector, a 'corner reflector' like an empty metal truck body or similar, which might be efficient over a range of changing incidence angles. But the slant range to the primary target in this case (mean approx. 5800 ft) does not allow it to be close enough to the ground. Even if we locate the primary reflector beneath the bottom edge of the main beam, close to the nadir, its altitude could not possibly be less than about 3200 ft, and realistically will be much greater. At -50 degrees its mean altitude during the photo sequence would be about 4600 ft. Moreover a remote corner reflector of any imaginable type will remain at a (more or less) fixed location relative to the aircraft, and as the range to the primary reflector from the 250 mph aircraft varies, so the separation of primary echo and ghost echo on the PPI changes in direct proportion. This would be very pronounced, yet no change in separation is detectable. These facts indicate a secondary reflector that is very close to the altitude of the primary reflector, and remains so, despite variations in efficiency, over a distance of at least a couple of miles. What could such a reflector be?

As discussed in Section 6-11 it seems inconceivable that an undetected layer of RI discontinuity, no matter how sharp, could have a power reflection coefficient high enough to account for the primary Phase B echo, and the marked bearing anisotropy cannot plausibly be explained by such direct backscatter either. But it seems possible that such a layer, which would be expected to be of wide horizontal extent, might be responsible for producing a secondary ghost echo from a very efficient airborne reflector pacing the B-52 at an altitude just above the layer.

For this it is further necessary, first that the target altitude be constant relative to the layer altitude, and second that the constant difference in height can be some fraction (to depend on angle of incidence and reflection) of the constant PPI range interval between primary and ghost echoes. The first condition has the consequence that there will be a systematic relationship between the changing slant range to the primary target on the PPI and the reducing vertical distance to the layer as the B-52 descends.

Such a correlation was shown in Fig. 23. The product moment correlation coefficient r = 0.83 is a very good positive correlation, though naturally short of a perfect functional relationship (the idealised relation would be a sine function) which could only occur if all measurements were precise and if the primary target remained directly beneath the aircraft (-90 degrees elevation) at all times. In the real case, this result indicates the likelihood of a target at a significant depression angle (else the B-52 descent would not contribute significantly to reduction of slant range) and so is not inconsistent with a target maintaining a constant altitude of around 4000 ft or more above the terrain, and therefore some significant fraction of 1000 ft (echo separation distance) above a reflecting layer at something over ~3200 ft.

The possible geometry of this situation is shown in Fig.25. Such a model would require that the primary target or object was at a fairly steep depression angle below the aircraft, consistent with the absence of any simultaneous aircrew visual reports, consistent with ground visual reports placing the UFO below the B-52, and consistent also with the subsequent "landing" scenario which would imply an object leaving the radar cover by dropping out of the beam.

Note that the conditions for producing an attenuated ghost of a large nearby target by secondary scattering from a layer below it are much less strict than would be required for direct backscatter. The geometry necessary to limit the ghost separation to about 15% of the primary echo range means that the incidence at the layer can't be grazing, as can be seen by studying Fig.25, but it does not require the unrealistic normal incidence condition for the "hot-spot" direct backscatter theory either. Because this is now a forward scatter situation, the primary reflector is permitted to be offset from the nadir sufficiently that it can be inside the Station Keep main beam coverage, much alleviating the problem of needing an extreme power reflection efficiency from a layer.

In forward scatter the inherent layer efficiency for partial reflection is much greater than for direct backscatter at or near normal incidence, improving rapidly as the 6th power of the cosecant of the reducing angle (e.g., ratio 4.2/1 at 52 degs; 8/1 at 45 degs; 625/1 at 20 degs). In addition the radar now receives signals by multiple routes, reflected from object to layer and back, and from layer to object and back, over the same raypath.

The literature of experimental radar meteorology certainly encourages the view that very sharp laminar RI discontinuities are not only much more common than balloon soundings might indicate but may have power reflection coefficients much greater than the highest values ever directly measured. In the present case no evidence of such a sharp scattering layer was detected by radiosonde; the intriguing implication is that such a layer could not have been observed at all without the fortuitous proximity of the large unidentified primary target which remains unexplained.

Finally there are some open questions. Col. Werlich indicates that there was no visual sighting of the object from the tower, even though a controller was following the flight of the B-52 with binoculars. Of course there is no direct evidence that the object detected on airborne radar was an optical emitter, and there was broken cloud above about 10,000 ft. But other ground observers in the area to the N and NE of the flight path did report bright lights which logic suggests may have been the same object, and the radar echo presentation seems to indicate a large body, so a tower sighting would not have been surprising. However from this vague, second-hand report it isn't even certain that the tower controller was looking at the B-52 at a time when the object was near it.

We should also consider that no object was seen in the air from the flight deck of the B-52 even when radar showed it approaching to within about a statute mile. This might be explainable if it was at a steep depression angle out of view from the flight deck windows, and it is also true that they were flying in layers of cloud and haze during parts of the event. But there was apparently no scattered illumination of clouds or haze observed at any time. One cannot exclude the possibility of directional light emission of course (i.e, principally downward), or there may be other explanations.

Other questions involve reports of echoes on independent weather and airborne gunnery radars (see Note 8). The weather radar report mentioned on the RAPCON radio tapes is interesting and potentially extremely significant. Unfortunately Blue Book's several attempts to obtain details were met with obstructive and uninformative responses from the SAC base. It seems very likely that a target of the size detected, with 1-mile to 3-mile separations from the B-52 and at altitudes generally between about 8-20,000 ft, should have been detectable and resolvable at various times by a variety of ATC and defence surveillance radars within a radius of 1-200 miles, including nearby SAGE air defence radars at Minot AFS; but if meaningful information was sought on other possible radar contacts there is no evidence of it in the file. Even information requested by FTD on the Minot AFB air traffic and weather radars, required under AFR 80-17, was not supplied. Col. Werlich only remarks that any echo on the RAPCON airfield control radar would have been obscured by the "pretty big blip" due to the B-52's IFF transponder signal. But this is very unsatisfactory.

There are other contextual factors which are largely beyond the limited scope of this report. Two of these - the prior and coincident ground-visual sightings of unidentified lights and structured objects, and the subsequent air-visual sighting of a large structured object on or near the ground by members of the crew of the B-52 - have been mentioned briefly. Obviously the present report is not final and is only a small part of a much larger investigation.

However, the following summary is possible:

A mobile, compact, airborne target or targets of unknown nature and of large radar cross-section (order of 100m², ), flying within about 1 - 3 NM of the B-52, seems the most likely explanation of the radar echoes photographed.

(Documentary evidence of independent ground radar contact with an unknown target near the B-52 does exist, but is too vague to be evaluable owing to what appear to have been constraints placed on the original investigation).

There is evidence that might be consistent with an intermittent radar ghost echo of this unknown primary target, and a hybrid model involving an efficient elevated scattering layer seems to be one possible interpretation of this, but there is no meteorological evidence for such a layer. (It is also true that the photographed structure of the echo could be related to the two-part structure of the object subsequently observed visually from the air by the pilot and copilot).

Endnotes ››

Bibliography ››