A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
From: Frank Reed
Date: 2018 Jun 14, 11:37 -0700
David, you wrote: "Are the shadows on the satellite photographs on Google Maps genuine?"
Normally, yes. But some regions are rendered as full 3D models. If that's the case, the lighting can be generated for a selected date and time of day. In Google Maps, this feature is probably not easily accessible. You can find it in Google Earth views.
Also, even in sections that have only minimal 3D rendering (almost all images are draped on a topographic model of the area), the photos are not necessarily taken from directly above, and this can throw off the apparent geometry. It can be difficult to spot this. Try looking for features like tall chimneys and utility poles. You know they're vertical. In the attached image of a view near where I live, if you look at the shadow of the small garage (highlighted in the red box on the left), you might conclude the photo was taken at about 1:30 local sun time since the angle from the roof peak to the shadow of the peak is about 23°. But look at the utility poles along the street (one in the red box on the right). Their shadows are aligned quite close to true north implying that it's noon local sun time (Local Apparent Time). Looking at the utility pole itself, we can see the source of the discrepancy. The pole appears tilted because the photo was taken with a slant angle of 20-30°.
"Shadows to the east of St Peter’s Catholic Church indicate the Sun in the south; shadows to the west of the Church indicate the Sun in the west. I suppose it could have been a long sermon."
Ha ha ha. Yes, or it may be that the Church has some sort of Old Testament refractive power... sanctified ground and all that. :)
Note that these "satellite views" are not usually created by satellites, except in remote parts of the world. Most are aerial photography which Google has licensed or commissioned itself. The aircraft fly at relatively high altitudes, maybe 10 - 20,000 feet so that they reduce distortion and yet stay under some of the clouds and also maximize resolution. Survey flights like these occur on a regular basis, and you can see updates every year or two in Google Maps. Of course they have to be stitched together. An aircraft will fly along a band a mile or two wide and perhaps a hundred miles long, then turn around at the end and photograph the next parallel band (the sort of job that will be taken over by drones in the very near future). Then software removes angular distortion as much as possible and fits the parallel bands together. Usually this is not noticeable in the final product, but stitching errors do occur, and even with no errors you can sometimes spot a significant change in season or ground shadows or, more dramatically, a construction project at groundbreaking on one side of a seam and completed on the other side. That probably explains the shadow mismatch you found.