Discussion

We have undeniably found some evidence suggestive of an atmospheric-optical explanation. In general "atmospheric-optical" means some effects on the propagation of light either by airborne particles (haze, mist or ice crystals) or by refractive index anomalies (unusual temperature gradients, causing mirage).

Unusual ice-halo effects are ruled out by the absence of ice in the line of sight. But there was a haze layer below the aircraft, probably associated with a weak temperature inversion in the CI area. That inversion would be the remnant of a much stronger advection inversion near the Breton coast, beyond the normal horizon, which was probably strong enough to form a localised optical duct.

Given the finding of a possible mirage-producing duct near the French coast one might feel that this cannot reasonably be a coincidence, and that mirage of sun-glitter on the sea near Brittany really ought to be a clear favourite. But we have placed this theory in Band 2 (barely plausible). Why?

• The gravity wave direction appears to have been near transverse to the LOS from the Trislander, and the wave slope negligible. However, the capillary wave orientation and capillary wave slope distribution are crucial factors in formation of a sun glitter pattern. Meteorological evidence suggests some likelihood of a sea breeze development, that may have generated near-shore surface winds parallel to the LOS from the Trislander (favourable orientation for transverse capillary crests despite adverse gravity wave orientation) by about 1400UTC; but low wind speeds of ~2 m/sec would suggest only a small tail of favourable ~20° slopes in the capillary slope distribution.
  
  This evidence is inconclusive but we cannot rule out a bright sun glitter pattern - perhaps aided by atmospheric-optical focusing or compression. However we find it unsatisfactory that the theory offers no interpretation of Capt Patterson's sighting;
  
  
• it is not easy to understand how the theory accounts for the sharp-edged outlines and "dark bands" of two identical reflection patterns several km apart (1° at ~150km = 2.6km).
  
  But these objections are perhaps not fatal, and it might seem justifiable to set them on one side for the sake of promoting the theory at least to Band 3 (somewhat plausible). The really serious problem is that during the course of 6 minutes Capt Bowyer observed the two UAPs steadily cross each other from left to right, horizontally, over an arc of a few degrees.
  
  We are satisfied that there is no refractive index mechanism in the literature - even of a very speculative nature - that would begin to explain this, and that it is a significant feature of the report which we have no good reason to dismiss n1 Only Capt Bowyer had an extended, uninterrupted view from the pilot's seat. Kate Russell, seated several rows back, had two shorter sightings, the first of perhaps 1 minute duration, the second of around 25sec. She also saw a relative motion of the two UAPs. Between the two sightings "the one above water [UAP#1}seemed to have moved" (Appendix B.v). Kate did not notice any relative motion during either sighting however. This may be explained by their brevity and by the witness's focus of attention during an unexpected event. The motion of ~3° in 6 minutes indicated by Capt Bowyer would correspond to an average rate of approximately 0.5°/min, or 0.5arcmin/sec. This would permit a relative motion of about 1/5 the width of UAP#1 during the second sighting and perhaps ½ the width of UAP#1 during the first sighting (assuming a constant rate). It therefore wouldn't be surprising if Kate had detected relative movement at least during her first sighting, if not the second. She did not. Nevertheless a relative displacement was noted between the sightings. .
  
  Which presents us with the classic dilemma of eyewitness evidence: What is its weight, balanced against conventional scientific models of the world? In this case we can get rid of a major problem, and have an interesting but unchallenging mirage, if only we disregard the description of the two identical images crossing laterally. This is one of those "significant features". Do we have a good reason to ignore it? Our position is that ad hoc trimming for the express purpose of "saving the phenomena" is not a good enough reason unless alternative explanations that do not require trimming can be ruled out as unacceptable on other grounds.
  
  

In judging whether it is good method to scrap significant features of the observation other factors come into play, such as the internal consistency of the prima facie sighting geometry in Section 3, where by respecting the reported lateral motions we find

• a consistent set of sightlines from the Trislander to a pair of locations in the Alderney-Guernsey area including the correct parallax due to the aircraft motion,
• relative angular sizes of the two UAPs consistent with the distances to these locations,
• the correct ratio of changing angular sizes
• and a UAP#1 location consistent with an independent sightline from an observer on a nearreciprocal bearing (obviously neither the localised Breton inversion to the S nor sun-glitter reflection are relevant to a sightline looking N from near Sark).

A mirage - even a scientifically unknown "lateral mirage in the free atmosphere" - doesn't explain these things in a natural way, whereas something like reflections on local haze, or lenticular clouds, or EQL in that area, could do so. And although Capt Patterson didn't see a high-definition object, he did see something of the right sort of shape and size and colour in the right place at the right time at the right apparent altitude (independently estimated), the like of which (he said) he had never seen before. It would certainly be preferable to take account of this sighting, too, if at all possible.

And preferring the haze-scattering theory does not mean that the potentially mirage-causing inversion on the Breton coast is a mere coincidence. The coastal advection duct is connected with the same warm NNE airflow producing the weak CI area inversion and the associated haze layer, so the coastal inversion is an indirect but necessary component of the haze-scattering theory, even if the light rays reaching the observers have not passed through that part of the atmosphere.

We think it would be exciting to be able to claim evidence of a completely new type of refractive index phenomenon, but we wish to emphasise that a mobile lateral mirage of the type implied would require horizontal temperature gradients of a severity and stability that seem inconceivable in the free atmosphere. Before adopting such a lateral mirage as a favourite one would wish to have ruled out the haze-scattering theory, the theory that observers were mistaken, and all other possible theories - including those that we have not yet thought of.

We are not convinced that the observations were mistaken, although we accept that this can never be ruled out by any objective test short of conclusively proving the presence of some phenomenon that explains them. During our investigation the overall cohesion and reliability of Capt Bowyer's account (in particular) has been tested in various small ways and it appears to us to have been careful and reliable. We think it possible that the UAPs did behave as described. This being so, we believe that the haze-scattering theory and the EQL theory are interesting alternative possibilities which could repay further study by experts