Explaining phenomena such as this involves some knowledge, some guesswork, and some experimentation. One of the things I've always enjoyed about paranormal investigating is that there is often no definitively correct answer, only a probability one way or another.
Someone wishing to answer an "Is this a ghost?" type question has to examine the evidence presented and use their knowledge of the natural world and how it works, to eliminate the probable, natural causes.
Assuming that an anomoly has a natural cause should be the default starting point. That's not to say you would immediately exclude non-naturalistic causes altogether, just that they are statistically less likely.
With this picture, you know as well as I do that lenses have internal reflections and suffer from "ghosting" and lens flare, which, in all likelyhood, is what this is.
For further proof of the suggestion that this is the cause, Sam would need to replicate the event as near as is practically possible. That is, take some photos using the same equipment as used in this photo, and of a subject as close to the original as possible, and in similar conditions. So a rectangular outside light on a dark night might do the trick. Then compare the results of the experiment with the original anomolous image to see if there is any similarity.
Cheers,
Shaun