Hi all - so I downloaded a large amount of GPX data, and noticed after downloading that some of the species (these are all fungi) have multiple duplicate coordinate entries but for different dates.
It seems slightly unlikely to me that any given person’s GPS device would get the EXACT same coordinates as someone else who came by at a different time.
Is this an incorrect assumption? That is, if there were 5 different observations at a given spot, give or take a few meters, by 5 different people and devices, would we expect to see the exact same coordinates recorded?
How precise do the coordinates appear to be? If they’re in degrees with five digits after the point, which is how they’re displayed here, that’s 10/9 meter (1.111) in latitude and somewhat less (depending on latitude) in longitude.
What sort of GPS device? Did they have a receiver on the top of a pole, which is typical of survey-grade GPS, or were they handheld devices?
It’s unclear what the source of this data is - is it from iNat or from a specific device? If it’s from iNat, how do you know that the data are GPX waypoints?
If this is iNat data, same coordinates can occur if the observation locations were entered as text and the GPS coordinates generated by the system. For instance, if the observer put “San Francisco” for the location, the system generally will generate the same coordinates each time. Many commonly used locations have a pile of observations at the same coordinates that Google gives that location.
Could it be weird computer number storage? I’ve used systems where I enter 2.75 and it gets saved as 2.7499999. Maybe people are entering rough numbers and iNat makes them look more precise than they really are. Or it could be a processing step before it gets to iNat.
Depends on if the observation locations are recorded from the device itself, or added in later. If added later and someone just searches for a location, then everyone who does that would have the same location for the observation.
@phma I don’t know what device type - I’m just downloading them from iNat with ONLY lat/lon/date/name; I didn’t download the positioning device and method fields. I can do that and report back, though.
I downloaded them as CSV files (the only way I can get the data), and wrote a small Java program to convert them to GPX files, preserving the precision found in the CSV files.
The coords are varying in precision. Eg: (these aren’t dups, just examples of precision):
@cthawley I’ve responded elsewhere with some info about this data (I didn’t download device types/methods; I got them as CSV and converted them to GPX, so it’s the raw data from iNat I’m referring to, I don’t know what their original form was; precision on lat/lon for the dups I’m referring to is an EXACT match between observations, etc).
Thanks for the info about iNat itself generating the coords - I wasn’t aware that this happened. I’ll check out the specifics of the points to see if that’s what happened - do you know if the coord derivation is available in the data download, then? I know you can get device/method, but what would it show if iNat itself derived them?
@tallastro Interesting! I didn’t think of that - I wrote a program to find and display all the dups, so I can post some of them here as examples later. I’ll re-download the data with more info (device and method, etc) so I can compare what might have happened.
@earthknight Yeah, others have been saying that (or similar things) in this thread too - I’ll look into that by downloading more data to see how the coords were derived!
@deboas That would be annoying (the same observer marking an observation at the pinned location more than once); though I guess there’s plenty of instances of “I found the same thing as this other person, in the same place”. :-D
regular users don’t have a great way to see exactly how locations of observations were recorded. we can only infer based on other fields recorded on the observation, based on other observations, etc.
Why annoying? I use a pinned location with a large accuracy circle for observations made at my residence. It is not unusual for me to observe the same species multiple times, because I have repeat visitors so I might record them on different days over the course of the summer.
@spiphany “Annoying” is relative to the person being annoyed; it’s annoying to me and for the purposes I’m downloading the data; I’m converting large volumes of data into GPX files and uploading into Gaia so I can view the points offline, since iNaturalist requires you be online to do searches, and I’m often out away from signals. Having a lot of points that show up in the EXACT same location clutters up Gaia’s UI. I wasn’t aware I had to scrub the data of extraneous points showing the exact same thing at the exact same place, since I (wrongly) presumed that that wouldn’t be the case in the downloaded data. At worst, I thought maybe there might be clusters of the same species located near each other, if GPS accuracy varied between observations.
I’m allowed to be annoyed, even if you’re not. ;-)
I have several observations at home with the same coordinates and a 100 m circle, which are obscured, so the locations are scattered over a coordinate rectangle for anyone else. I also have an observation of an Argiope spider and fan clubmoss at the same coordinates (a traverse point in the woods) but different precisions; the spider was right on top of the nail, but the clubmoss is several meters away, though clubmoss is also in the background of the spider picture. I usually do something similar at other surveying sites.
Seems to me that there are at least three legitimate ways to get identical locations on multiple dates. I’ve done these, so of course they’re legimate, by definition.
Use iNaturalist’s / Google’s location for a site, e.g. all those records located at a park headquarters, wherever they may have been in the park. (OK if the accuracy circle is big enough.)
Copy/paste a location from one of my own observations, especially if it’s a photo of the same plant taken on multiple dates.
Use coordinates from my own list of locations for places I revisit, my personal equivalent of the list of sites from iNaturalist & Google.
In my opinion, the identical coordinates are probably not anything nefarious, though it’s worth checking.
