It is possible that some of the radar cases presented to the panel have a natural explanation. It seems likely that some possible natural explanations could be investigated without cooperation or assistance from the controlling military authorities except for a time record of unidentified traces that occur during designated test periods.
Some of the observations suggest that time-variable atmospheric ducting may on occasion result in echoes being obtained from distant ground locations as a result of refraction. Some of accounts described (a) groups or swarms of echoes that persist for some time in the same general location; (b) apparent trajectories of echo sources that exhibit sudden changes in the vertical and/or horizontal positions; and in particular (c) the tendency of apparent echo sources to concentrate over mountain tops. These are all characteristics to be expected of ducting conditions due to weather. These effects can come and go over long periods of time and they can also lead to scintillation or other changes over short time periods. (See, for instance, Hall & Barklay 1989.)
An atmosphere is said to be "superrefractive" when a horizontal light or radio ray curves downward with a radius of curvature that is less than the distance to the center of the planet. The atmosphere of the planet Venus is at all times globally superrefractive below an altitude of about 30 kilometers. In principle, echoes could be obtained from every area of the spherical surface of Venus from a radar system located at any position on the surface. If the air of Venus were perfectly clear, an observer would see all areas of the surface, all areas repeating in range to indefinite distances. In the four giant planets also, the large gradients of refractivity (or density) in their atmospheres produce superrefractive conditions.
The Earth's atmosphere is normally not superrefractive. However, common weather effects (in particular thermal inversions, where the air temperature increases with altitude, and/or the water-vapor content decreases with altitude) can and do produce regions of superrefraction that are localized geographically and in height. As a result, atmospheric ducts (channels that trap and conduct radar waves) can form that carry the signals far beyond the normal horizon. Such ducts can bend rays down to a distant surface area or, more easily, to a distant mountain top. Backscattering of the radar energy from the ground or from discrete objects on the ground then results in echoes that appear to the radar to be due to a target that is far away and (if the angle of elevation of the returning energy is measured) high in the atmosphere. A similar transient ducting of sound can produce the experience of hearing the whistle of only one particular train out of the many that originate at difference times from a busy track in the next valley.
As is well known, atmospheric ducting is the explanation for certain optical mirages, and in particular the arctic illusion called "fata morgana" where distant ocean or surface ice, which is essentially flat, appears to the viewer in the form of vertical columns and spires, or "castles in the air."
People often assume that mirages occur only rarely. This may be true of optical mirages, but conditions for radar mirages are more common, due to the role played by water vapor which strongly affects the atmospheric refractivity in relation to radio waves. Since clouds are closely associated with high levels of water vapor, optical mirages due to water vapor are often rendered undetectable by the accompanying opaque cloud. On the other hand, radar propagation is essentially unaffected by the water droplets of the cloud so that changes in water vapor content with altitude are very effective in producing atmospheric ducting and radar mirages.
With regard to "impossible" flight paths that may appear to be indicated by some of the echoes obtained by military radars, it is important to note that the records presented to the panel are based on measured time delays and measured elevation and azimuth angles-of-arrival of the reflected energy from the echoing object. As presented, certain target positions were plotted as height versus time. But height is computed from two parameters: (1) the measured time delay, which is a very good indication of range; and (2) the measured vertical angle of arrival, which may not be a valid representation of the vertical direction to the target. In particular, when ducting occurs, reflections from distant and distinct surface targets (buildings, bridges, trucks, etc.) may be received at elevation angles of several degrees, so that a ground target at a range of 100 kilometers, for example, would appear to represent an object at a height of several kilometers. Atmospheric turbulence would distort the duct and could cause sudden changes in angle of perhaps a few tenths of a degree, which would be interpreted as a sudden change in altitude of the order of half a kilometer. The horizontal angle of arrival would also be affected by turbulence, adding to the chaotic character of the apparent flight path.
Ducting to and from distant mountain tops requires less refractive bending than echoes to and from lower surface areas, and should therefore be more common. This may explain the concentration of apparent targets over mountains. A test of this hypothesis would be to place a radio receiver, tuned to the radar frequency, on or near the top of a mountain associated with unidentified targets. It should be connected to an antenna that has its unobstructed receiving lobe centered in the azimuthal direction of the radar and its vertical pattern extending from zero to at least several degrees in elevation. If ducting does in fact occur, the occurrence of unidentified radar echoes would be found to be correlated with major increases in the strengths of the radar signals measured by this receiver.