Satellite images of the world tend to have some interesting artifacts from time to time. Airplanes in particular look quite interesting—there used to be one of these “rainbow planes” over Jackson Park in Google Maps. It appears to be gone now. But there’s no mystery here—I know what causes these.
On the other hand, this Oak Park post office also looks mightily interesting. But for the life of me, I can’t figure out what’s going on here. It almost looks like the building is aflame, but I couldn’t find any references to such a fire in Google. What’s the cause of this building’s appearance in this particular satellite image?
I don’t know the optics involved, but I can see color separation there, and the “rays” go along perpendicular directions. The roof also appears abnormally biright.
I suspect that you’re getting almost specular reflection from the white rooftop (almost mirror-like, that is. The angle of incidence between the sun and the roof is equal to the angle of reflection betwen the rooftop and the camera – only it isn’t a mirror, so it’s not giving you an image of the sun, just a really bright reflection. Your rooftop isn’t one o’ them Lambertian scatterers), and this extreme brightness is giving you weird effects. The focal plane array, with its finely and uniformly spaced tiny detector elements (I call them “pixels”, even though that’s not really correct) act like a diffraction grating in two dimensions. Diffraction grating act kind of like prisms, spreading out the light in different colors along the directions. I suspect you’re getting a reflection off a lens surface or something that is giving you that crucifom diffractive spreading. You don’t see it from the other roofs because they’rre either not at the right angle, or they are better at scattering light evenly, rather than preferentially along one direction.
Yeah, just looks like lens flare to me, and CalMeacham explained it better than I could. I’ve been to that post office, and it’s definitely not aflame.
Good spot. Looking at where the “rays” intersect, and also the shadow angles, it looks like the sun is reflecting on the southern face of that long skylight.
That would be just the kind of thing to give you that specular reflection, instead of the diffuse reflection you’d expect from a painted roof or pebble-covered roof. I was wondering if threy had tile or something, but window glass is much better for getting high, directional reflection.
But I don’t think the example from the OP is a simple “lens flare”, which is the result of internal reflections in the lens itself. Such effects either have an oval profile (to each reflection) or show the symmetry of the iris (a lot of old NASA lens flares had the pentagonal shape of the five-leafed aperture stop). In this case you have a curciform shape with priosmatic color separation – which tells me that the CCD or FPA sensor itself is probably involved, probably due to back reflections from some lens surfaces. So it’s not purely a “lens” flare, but a “lens plus camera” flare.
