I believe most GPS units will report a vertical dilution of position (VDOP) which is a function of the position of the GPS unit relative to the visible satellite constellation, and gives a rough indication of the vertical accuracy. I’m not sure how it relates to an absolute vertical accuracy in metres. The VDOP and several other dilutions of position are standard output in GPX files.
1 Like
Thanks, good to know! I’m guessing any vertical accuracy measures probably don’t get incorporated into the EXIF headers of smartphone photos, but I’d love to be wrong about that too.
Sure, but when I put the altitude into an observation field (which I do), I am getting it from the same place. No worse off then. Although I realise your main point was about renaming the horizontal accuracy, which makes perfect sense.
That’s a good question
I know when I down load photos from my Apple phone onto a storage file the elevations are on each photo. But that could be Apple and software working its magic in relation to the location to come up with a elevation, the numbers sometimes are way off
Just speaking for myself, (1) working mainly in parts of the world with good digital elevation models (DEMs) available, (2) working mostly in higher elevation areas, and (3) knowing the relatively poor accuracy of GPS elevations, I will always be getting my elevations from the DEMs (via various applications). Even if iNat starts importing GPS elevation and any associated vertical accuracy measures from EXIF headers, I would still want to enter mine manually.
But there are definitely parts of the world for which the GPS elevation will be the best available data.
4 Likes
I think I might do some investigating on my own observations, as I can get DEM elevations to compare to the exif values…
I’ve done this with my observations, and the difference between DEM and GPS elevations ranges mostly between 0 and about +/-30 meters (with a few outliers) for points with horizontal accuracy of 5 meters or better.
Not a huge difference in the great scheme of things (except when working near the sea coast), but I like to be as accurate as feasible.
1 Like
I had a project where I had to geo-locate to +/- 1m, but it was probably +/-2m realistically. I compared a couple of them. My DEM readings, which I took from LINZ National Imagery for New Zealand via a mapping app, are only +/- 5m to the photo exif values at circa 280m. I can live with that!
1 Like
I could definitely live with that too, if all my GPS waypoints were that close. But the variation over >1000 points has been much greater for me. Hopefully your sample wasn’t just a fluke…
More the latter. I don’t think it needs to be a feature, but some sort of prompt when it is appropriate. As I mentioned, I live in the flatlands, so elevation is not really an issue. In a sense, I was thinking out loud. I hadn’t really considered that it could be a factor in identification.
@mamestraconfigurata @tiwane Since this no longer seems to be intended as a feature request, should we close the topic?
1 Like
Yes, sorry about that. I’ll close it.