Is iNat App GPS getting less accurate?

[charlie]

I’ve observed the location accuracy has gotten worse with iNat over the last several years. I’m basing that based on comparisons with the map and points imported into GIS rather than the stated accuracy. It’s a lot more likely to put points in the wrong place than it used to be. I’m not sure what changed and it may have to do with GPS technology or security measures or something. I think the new app does a better job of waiting until there’s better accuracy, but still isnt’ perfect. I know a lot of people like the workflow of taking photos and importing them into iNat via the app but i don’t do it in large part bceause of this issue. I find that the camera app takes much worse GPS locations with its photos than the app especially if i keep the app open between observations.

In terms of external GPS i was using an external Garmin GPS with an iPad (no cell internet connectivity) and i found the GPS accuracy to be worse than a phone without an external GPS. It may b ethat you can get better ones or it may be they work better when the phone itself also has GPS, and there may be workarounds to do things lile also record a GPS track at the same time (but good luck with the phone battery then).

@jdmore

@earthknight

for iNat the relevant accuracy range is in the tens of meters to around a hundred meters even for sessile species.

This may be what is relevant for your uses of iNat, but I’ve lost count of how many times another observer’s ~5m accuracy (which is what I usually achieve) has helped me re-find an unusual observation in the field for verification and further documentation

Yeah it may vary by location and device type but in Vermont in the USA i always get about 10-20 m accuracy when it’s working as intended. Sometimes better. I definitely disagree we should write off as a precise mapping tool. Just need caveats.

[danly]

@charlie

It’s a lot more likely to put points in the wrong place than it used to be.

I’ve noticed that too, and just assumed that the fault was with my phone. It used to reliably place observations in my neighborhood, but now regularly places them miles away. (I haven’t changed my accuracy settings, they are just whatever the default is)
Now I always manually change the locations for my observations, especially for species where it’s important to show them in the correct habitat.

[upupa-epops]

I regularly have this as well, even more dramatically, where the first couple photos will have accuracy in the range of 2-3000 m before it settles down and goes to 5-20 m.

[tiwane]

I moved the above posts from here to a new topic. I’ll also note that the original post in the thread where this conversation originated is about the device’s camera app’s accuracy, not the accuracy recorded in iNat.

I personally haven’t noticed any issues with GPS accuracy in the app over the past few years, aside from some issues we had with iNat Next months ago during pre-alpha testing.

Keep in mind that the iNat app doesn’t have anything to do directly with GPS accuracy, that comes from your device and is usually due to local conditions. If there are issues getting accurate GPS coordinates within the iNat app, they’re likely due to the app not continuing to ask the device for more and more accurate locations. Currently iNat Next will stop asking for a location once the accuracy radius is below 10m.

[charlie]

Yeah, i am not sure the cause of the issue is from iNat’s end. I had one series of issues when I was in Quebec in Canada where i do not have a cell phone plan, and that seemed to confuse it more than no cell service at all. A bunch of points were very wrong though they had a huge uncertainty buffer which did include the correct location. I was not using iNat Next because as mentioned in the other thread it’s not currently very usable without cell reception. But when next is usuable it does prompt when GPS precision is low, which is helpful.

I finally got a new phone as my old one was dying, so we will see if that helps

[upupa-epops]

My response was for my camera app as well. This initial significant inaccuracy is why I care so much about adding or double-checking the accuracy when I upload images off my phone, since I’m assuming that if the accuracy is extremely low then the actual location is probably off as well.

[tiwane]

Last winter I was in a pretty deep redwood forest canyon and testing out an early version of iNat Next. No matter what, during a stretch of trail the app would never get below about 5km of accuracy. I then opened Apple Maps to see if that got a more accurate reading of my location but it also stubbornly stuck to one spot and with that 5km radius. So it seemed like a device problem, not one specific to iNat.

[earthknight]

So, smartphones generally use what’s called A-GPS, not pure GPS. They use a few GPS bands, and fine-tune using distances to cell towers.

A handheld GPS unit relies purely on the GPS satellite signals and uses up to 12 bands for most common handheld ones, with each band corresponding to one satellite. The more satellites connected the greater the accuracy as errors get corrected by the averaging of the signal. Some GPS units will also rapidly take several hundred readings and average those to give you one location point.

Since a phone is using fewer bands to connect to the satellites that means connecting to fewer satellites and a greater uncertainty as there is more variability in the signals. Hence the use of cell towers as a verifying and fine-tuning system for smartphones and such. I’m not sure, but I think that if you’re near a change in the coverage area that can throw off the accuracy of the ‘A’ part of the ‘A-GPS’.

I’m don’t know if they still do, but phones used to be able to get their location purely from the cell towers as well, if they couldn’t get GPS signal (or in some cases didn’t actually have GPS receivers in them).

Each of these systems has potential inaccuracies. Usually those will cancel out, but sometimes they reinforce each other.

And even a handheld GPS if left in one location for a long time will leave a track that jumps all over the place. Periodically the rangers we work with will forget to turn off a GPS and even though the GPS unit is sitting on the same table for hours without being moved the location will jump around by tens to hundreds of meters (we see this because the tracking feature is always left on in order to monitor patrols and routes taken).

This is, in part, because the satellites are moving and the GPS unit is constantly acquiring new satellites and losing other ones. Each time this happens it throws the location off a bit.

Then you have other factors at play. Different substrates or topography can throw off signals, causing them to bounce and the reflected signal getting picked up, sometimes wet vegetation can mess up a signal, and the further north you are the less accurate the location becomes as well as you’re contacting the satellites from an increasingly oblique angle resulting in more potential interference from the atmosphere and fewer satellites available for connection.

All this means that there are a lot of ways for a GPS signal to be compromised in some manner just from natural causes.

For what it’s worth, in the tropics I find that the location data from by GPS units and phone are generally pretty accurate, but being in the topics is pretty much the ideal location for good GPS signal acquisition and to have the most satellites available to connect to as well.

[Ec0pete]

This a common problem when taking GPS readings in tall forest or valleys, even in a city street surrounded by skyscrapers. It is all to do with the number of satellites your device can detect which is usually a function of the amount of open sky above you. Readings taken from a boat out at sea or when you are in the desert will give you the best accuracy.

[mikkohei13]

I maintain an app that records GPS tracks, and over the past year, we’ve also had a rising number of reports about declining accuracy—sometimes with locations off by several kilometers, even though the phone indicates that the location is accurate.

Our testing shows this issue affects all apps on the phone, including iNaturalist. The most plausible explanation we’ve found is intentional GPS jamming, which has become more common in certain areas globally. You can read more about it e.g. here: Flightradar24 GPS Jamming.

[streepvaren]

This is indeed a really annoying ‘bug’ in the INaturalist-app (at least in Android, which I use). It seems as if the current position has to be found again ‘from zero’ each time you open the app while it is running in the background. In contrast, the Obsmap-app (observation.org) doesn’t loose the GPX-fix during the time you have your smartphone in your trouser pocket (between two separate observations).

This problem can be bypassed by having your location tracked by another aplication like Geo Tracker while using the INat-app. But I think it would be much better if the INat-app was ‘fixed’ to prevent this location loss. Currently an unnecessary decrease in accuracy of mobile INat-observations is happening, often without the user realizing it. You really have to look carefully to check of the location is ‘fixed’ enough before saving the observation. I would post a lot more mobile observations of this wasn’t the case. But waiting each time for a proper GPS-fix is to time consuming for me, time which I prefer to use for observing nature.

Best,
Thomas

[mo0nsgreenthumb]

Do you have an iPhone or android? My GPS accuracy is pretty spot on. I have an android phone though.

[zygy]

Continually tracking your location in the background uses up a lot of battery, plus it’s a privacy red flag for some people. That said, I would love it if there was an option in the settings to turn on continuous tracking for when I’m doing a bioblitz or something. As it is, I do the same thing you do and start running GPX Tracker in the background.