I have found the occasional observation where the map view shows someplace drastically different than the location name - in some cases 500 or 1,000 miles apart.
I just found one batch of observations where the location name is somewhere in Mexico but the map view shows one in the ocean off the coast of Greenland and three in the ocean between Iceland and Norway (highly improbable locations given the subjects of the observations, background vegetation, etc).
Curious what the best thing to do is - bug report? Alert the user? Is the location considered not accurate if the name is plausible but the map says something completely different from the name?
I found a couple of bug reports but they were from 2019 so I wasnāt sure if this might be a new/different issue than the ones back then.
Iāve seen a few observations like this. Actually, I somehow accidentally put one of my observations as taken in the middle of the ocean (yeah, a bit off there!), but luckily was notified by an identifier.
I find that this happens if I change the location after the observation is already submitted. I had a couple where I chose the wrong āGreat Oak Parkā. I corrected the location, but the map retained the old location. I had to delete and re-enter the obs to fix the error.
I found a few observations like this where the location on the map was obviously incorrect but the species probably occurred at the location associated with the name. I donāt know if it was an app glitch or from the user editing the location. I tried to contact the user but they had not been active in several years and did not respond. Since I know the map location well and knew that the species did not occur there I ended up marking āNoā for Location is accurate in the data quality checklist which moves the observation to casual.
Happens to me frequently if I have to edit the location. The actual location changes but the text stays what it was before the correction. Also I suspect newer users think that typing the name of the location as text is sufficient to indicate the location- so they might have the point dropped in their house and type āX State Parkā as the location text.
If you enter/submit an observation, then move the map pin, the location descriptor is not updated automatically.
If you enter/submit an observation, then change the location descriptor, the map pin is not moved automatically.
As far as the location descriptor generated by iNat not being āhelpfulā, it may be difficult to create an algorithm that consistently generates a āhelpfulā location descriptor. I would certainly do things differently (wrt what iNat currently does), but Iāve got very strong ideas of what constitutes a helpful location descriptor. I have considered writing some software to do this on my end (ie. simply replacing iNatās location descriptor), but itās down my to-do list. I would be happily to work with an iNat developer to create such an algorithm for iNat, but I have a feeling they would be unwilling to apply it retroactively to existing observations (there may be no easy way to know if an existing descriptor was generated automatically or was entered/edited by the observer).
One thing I would consider a ābugā is the fact that currently, it appears that the location descriptor is generated using a different algorithm depending on what interface you use to submit your observation. ie. same lat/long will get you a different location descriptor depending on whether you submit your observation via one of the apps, or via the web interface.
I have seen these and itās always because the user searched place name and picked something that popped out in search results instead of using direct map picker and placing the point
Itās because multiple places in world can have identical names or slight differences, so in your example you and user both know this species is say mexican but the search system doesnāt and it popped someother location in world with similar name; so simply reminding them to edit observation with map picker is the only correct way.
For my pinned location, with a careful text - then I have to move the pin on the map to the right place (I donāt have GPS). My careful text is replaced by Googleās default for there. Elsieās Peak - oh no - Google has Erf something Farm so and so - Sigh. If our pinned location could default to keep OUR text.
Is that before you submitted the observation? I donāt have a lot of experience with entering observations, and I havenāt tested out the different permutations to see how iNat behaves in every case. Iām pretty sure Iāve seen cases where folks go back and move a location pin and iNat keeps the old location descriptor (regardless of how it was originally created).
A user once alerted me to an inaccuracy in location on one of my observations. It was made in a cedar glade, but the map had it in a neighborhood some distance away. Investigation led to the discovery that over 500 of my observations had similarly incorrect location data attributed to them.
It turned out to be an issue with the settings I had for photo location data on my phone. It was obfuscating the geographic location of each image and causing iNat to have a field day with them. Great for privacy, not so much for scientific data. Thankfully, the descriptions I left and contextual clues in the photos jogged my memory enough to help correct the issue, but the process was pretty labor-intensive.
Unless the names were entered manually, they are irrelevant. When they are just pulled from Google Maps, they often pure nonsense. Just ignore. I ignore the names or mine. If I care, I enter something manually. If I donāt, only the coordinates are relevant.
The description is correct, it happens to me as well. It is very annoying. Whenever you move the pin, a new irrelevant name from Google is generated again.
I just tried it out on an old observation. First, I saved the lat/long of the observation outside iNat, then I tried moving the pin. I saved the result and the existing location descriptor (which had been automatically generated when I originally submitted the observation) didnāt change. So I ended up with a completely inappropriate location descriptor, given where the pin ended up. Then I edited the observation again to restore the original lat/long.
Perhaps if youāre in the middle of entering an (as yet) unsaved observation, and you move the pin, iNat will alter the location descriptor to match the location of the pin.
What would probably be required is for iNat to keep track of whether the current location descriptor was generated automatically or entered/edited by the observer. If the location descriptor is one that was automatically generated, then iNat should updated if the pin is moved (or the lat/long is edited). If the location descriptor is user entered/edited, then it should become static. That functionality would probably satisfy the maximum number of users. (keeping in mind that most users donāt even know they can enter/edit the location descriptor, and even if they did, they probably couldnāt be bothered to do so).