Go here to see some work I did on sorting out a sense of scale for this debate. I think the 'bug' theory is up in smoke, if you'll pardon the jab.
http://homepage.mac.com/b.a.freeman/Dar ... bum14.html
Far more likely -- but boring -- is a distant pole, perhaps a hundred yards beyond the foreground pipes/conveyor/whatever in a loading dock area.
What may be a very simple explanation is a utility pole shorting out, followed by a very bright flash of light and smoke which would of course wash out any part of an image shot at 1/20 and f4 to f5.6., resulting in a big white smudge we see here.
That same too-bright light would likely cause the algorithm for the digital sensor to make some heavy adjustments to the likely result: lens flare. This could easily be represented as a dark, purple-ish line we see here. It comes from the light out, not the other way around. Think of how often you've seen dark bands emanating downward from the sun over some storm clouds. I think we're seeing it here from a short-lived "artificial sun".
Also remember the bright area may already have been burning when the shutter opened! He (Mr Pryde) was looking up at the clouds and might easily not have noticed until he saw the pictures at home.
He's using a tripod for the camera (all 3 images are movementless, relative to the shore), so he didn't need to compose for each shot; he was just waiting 15 seconds and then would trip the shutter again.
For comparison sake among the non-photographers: Consider that a daytime shot of your pal might be 1/500 or 1/125 at f5.6 (ISO ~200 speed film/sensor setting) on a brightly overcast day.
Since our intrepid Mr Pryde was essentially shooting the clouds and darkening sky, he was exposing for them -- to capture the colors and gradations of clouds -- rather than exposing for the land, which is why the land area is in silhouette.
Cheers,
Brett.
