I often go on hikes and take photos to upload, but I don’t usually keep my location turned on on my phone, or I use an old digital camera, meaning the device doesn’t store the location of the photo
This means I only have a general sense of my observation’s location and I have to sort of choose a random spot along the trail.
I know it’s not ideal to be ballpark-ing the location, but without a selected location none of them can become research grade
Does anyone have insight into iNat’s policy/conventions on this topic?
I’m using a 13-year-old Canon EOS Rebel T3i, so what I do is photograph whatever critter with the old camera, and then take a pic of the display screen on the DSLR with my phone (with the location turned on). That provides an accurate time/location for the DSLR shot that can be added in the batch uploader, and I usually (unless I forget) delete that back-of-the-camera shot before uploading. The location and other info stays in the batch uploader even when the photo is deleted.
All that said, if you’re at a park, you could drop your location pin in the middle of the park and make the “circle of uncertainty” encompass the entire park. It’s not precise, but it’s within acceptable parameters.
I just ballpark 'em, yeah. Done this a few times for photos I uploaded that I took before I even knew what iNat was. I think it should still be fine as long as your accuracy isn’t set large enough to take up a whole province/state lol, which I imagine yours would not!
Is the circle of uncertainty a desktop feature? I haven’t encountered it yet, but I spend most of my time using the mobile app. I might start using my computer to upload observations if it means I can add a broader range
Mobile should have the option. For me, when I go to upload the observation on the app, there’s an option that mentions location right below the date/time option. The circle should automatically show up when you tap on it to pick a location
Easiest way is to have locstikn turned on in the camera app. Your GPS doesn’t have to run permanently but only when you take pictures.
But if you don’t have the exact location maybe try to also take a picture(s) of the surroundings or try to see when you are on a paper map and write down a comment. That might help you pinpoint the location again later.
even if you don’t have that try to remember best what part of the trail you took it. You can set the accuracy as wide as you need to be sure it’s somewhere in that area.
I was quite surprised to governments often don’t even care much about accuracy that’s better than 100 or 500m, which is ridiculous for modern times.
I can only speak for the Android app.
but if you edit an observation and change the location, you see a crosshair that indicates the precision of the point. Zooming in and out makes that crosshair cover more or less area and therefore increase or decreases the precision
The precision value can be seen in the top bar.
Ballparking the location is fine as long as you set the accuracy circle to be large enough that you are sure the true location falls within in.
I also don’t have a phone for location and do this with camera pics. For some spots easily seen from satellite, I can give quite a precise location. For others that are in the middle of woods with no landmarks visible via satellite, I sometimes have accuracy circles of 100s of meters or more. This is also true for older records. I try to be conservative and “round up” (make the circle larger than it needs to be) as this is what I would want someone to do if I were going to use their data.
As long as the accuracy circle is large enough you know the observation was somewhere in the circle, the observation is OK and can be RG, if you don’t know where in the park you took the photo, you can make the accuracy circle cover the whole park, I once had an observation on a road trip with multiple counties in the accuracy circle
One thing to note is that there might be a point where the accuracy circle is so large it doesn’t count as RG, but that would have to hundreds of kilometers, and I’ve even seen accuracy circles covering thousands of kilometers that I think were RG
I just increase the radius of accuracy. If I am say at a park and took multiple observations, then I can use the same circle to cover the approximate locations of all observations recorded there. This typically would be in the order of a few 100m to ~1km.
Sometimes I don’t leave my camera on long enough to get a GPS fix. However, I often have an observation just a few minutes before or after with one, so I drag a picture from that observation to help narrow down the location. I just got to remember to remove that image before uploading observation. Sometimes I forget and people are wondering why an unrelated image is in the image set for an observation.
For the uptight / high volume club, it’s also possible to inject GPS tags back into images if you have something else that creates a GPS track. I use “exiftool” to do the GPS tag manipulation, and download the track from my Garmin smart watch.
At one point I was having trouble where if my phone was in battery saver mode, and between the mountains, its GPS tags are absent or wrong.
I turn on the location for my phone when I know I will be making observations in a place I won’t be able to figure out the location of (such as when hiking or traveling). But sometimes I’ve found that yields incorrect location data anyways, if the phone is pinging off a cell tower nowhere near my location (or if there is no cell service)
My real camera doesn’t have GPS, so it always requires manually adding the location.
If am posting from my laptop, I make a “pinned location” (for a park, for example, or my yard) and set the circle large enough to encompass the area I was in. This is useful if you take a lot of pix in one location, or return to it often.
I don’t think you can do that from the mobile apps, at least with IOS. In which case I use the “ballpark” method.
this is exactly what i do. sometimes i find old photos of animals and i have no idea where i was that day, but i know i was in a certain general area. i just make a gigantic circle encompassing all the places that organism might’ve been.
Just get it as close as you can be reasonably sure about. Exact pinpoint accuracy is not needed in most cases. Increase the margin of error circle a little to make it clear that the exact location is not pinpoint, but somewhere within that circle.
In the past when submitting obs where I did not have GPS (older observations that I knew fairly accurately where I was) I always added “Location approximate” in the notes.
I just experimented with the circle of uncertainty (I did not realize that it was appropriate to change the default accuracy setting) and will implement this in the future (and go back to incorporate this into some of previous observations where needed)
Not to dispute that phone locations can be inaccurate, but your phone accurately locating you by GPS isn’t normally dependent on cell service / access to cell towers. It is possible to have a phone with a damaged GPS chip (in some models, it’s what a swelling battery damages first). The GPS limitation is often that it’ll need a view of multiple satellites and with bad conditions (in a mountain valley, with a lot of moisture in the air, in a region that hasn’t been prioritized for satellite coverage) it’ll be slow and power hungry to get a fix on where you actually are. There’s also a penalty (deliberately downgraded resolution) for being outside of the continental USA.
You will often do slightly better if you run a map app ahead of taking pictures, since that prompts it to start the process, but that comes at a battery life cost - battery saver mode often degrades GPS performance.
There’s absolutely regions without cell tower service that’ll still allow nearly perfect GPS usage. Not that any of this is really helpful - if I’m somewhere remote, I’m probably guarding my battery life fairly carefully and doing without geotagged images is a price to pay.
I do usually photograph where I parked and labelled waypoints, figuring that it’ll let me reproduce my itinerary.