More of the shit I get up to

Back in 2015, I raised a question, or at least, pointed out that an answer could be determined, based on the details visible within a photo shown therein, reproduced in more detail below:

crescent moon by Union Station, Nashville, 1991
This was a photo that I took, with my cheesy little camera and crappy print film, while on a training seminar to Nashville, Tennessee, and in the post, I claimed that the date could be determined just from the details within the photo.

Which I did: November 9th, 1991. Maybe.

Okay, not all of that was from details within the photo, though I maintain that it could have been. Maybe. Given some more sources of info, anyway. But here’s how I arrived there.

First off, I knew it was 1991, because I bought a souvenir there that a psychotic girlfriend stole in 1992, and in 1990 I’d just moved to NC and wasn’t even working for the place that sponsored the trip. And I knew it was late in the year since one of the taller buildings was advertising ‘Visit xmas Village’ by leaving the lights on in certain offices at night. That’s about the extent of the information that I contributed through personal recollection, and I’m not sure these couldn’t have been determined from the photo. For instance, the clock tower is missing the statue of Mercury atop it, which pinned down at least a bracket of certain years.

Screenshot from Google Maps of Union Station, Nashville, TNThe shooting location was determined through Google Earth, confirmed through Google Maps Street View (normally, you could do this through Google Earth too, except that the latest versions don’t work in Linux, and Street View won’t work anymore on the older version that does work on Linux.) Most especially, the angles of the various points on the fa├žade pinned down the spot on the sidewalk, within maybe a handful meters or less, which led the next step.

In Google Earth, the shooting angle and the compass direction of the moon could be determined from this latitude and longitude (222°, give or take – notice how the moon lines up directly above that chimney-like peak.) The clock reads 5:17, obviously PM from the afterglow towards the west, and Nashville is on Central Time.

I could not determine for sure the height that the moon was above the horizon, because I could find no info on how high that clock tower is, which would have given me the elevation of the moon. Eventually, I found a full height photo of Union Station (from ground to above the clock tower) with someone in the photo for a reference, and in GIMP translated the pixel position in rough length/height references. Very rough estimate: the clock sits 34 meters up.

Back to Google Earth. Measurement of the clock from my shooting location was about 147 meters. Some simple calculations (done with an online right-angle calculator, though, because my math skills are untrustworthy) gave an elevation for the moon of 13°. That info carried over into Stellarium, where I could input the location (including ground altitude) and the time, then begin sliding days back and forth until the moon got into the right position.

No day was a tight match, but this was not unexpected. Using Google Earth and Google Maps, with the Streetview function, to pin down the shooting location has some slop, especially since the vehicle producing their photographs obviously wasn’t on the sidewalk where I was, and there’s a bit of wide-angle lens distortion from their photos. The same can be said for the aerial photos – change the dates on any given location with taller buildings and watch the positions shift depending on how far they were from the camera axis. Significant changes in distance produce only trivial changes in apparent angle, and of course, my tower height measurements were pretty off-the-cuff. But November 9th has the moon at 222° compass bearing, just over 15° elevation, and that’s the closest I can produce for that particular time of day (perhaps the most accurate part of all this) and location. I feel comfortable with it, at least until I can get anything more accurate to work with. November 9th of 1991 was a Saturday, which fits with the three-day seminar as well.

Anyway, that was two hours of sleuthing for no good reason, other than to see if I could do it with the tools at hand.