Originally posted by Chris Rozoff
While I'm not going to offer an explanation for this phenomenon, I will offer a hypothesis: Could it be there was some kind of horizontal area of concentrated horizontal vorticity that was tilted and stretched by two areas of enhanced updraft to create the odd funnel? One can imagine the boundary layer, especially in the presence of thunderstorm activity and shear, there are often invisible vortex tubes and all it might take is enough of a ping from a thunderstorm to make them visible and strong enough to be coherent. So anyway, imagine a horizontal vortex, and two updrafts tilting it and defining it into a ring-like structure.
Another idea might be just what Shane said: Vortex merger. It's easy to visualize 2-d vortices merging, but when you add a third dimension and high aspect ratio to the problem, it is much more complex (beyond my level of understanding) and the two funnels might merge in the fashion described above. Actually, this hypothesis seems just as likely, if not more likely than my first dumb hypothesis.
Any other ideas or more verifiable theory?