“is this some sort of actual regulation or law?”
It is a fact - if information has not been gathered, it can’t be used.
I don’t believe I have said any records are of absolutely no value. I have said previously in this discussion that a location name that corroborates the position of the place marker can make up for the lack of an accuracy circle. I don’t believe in magic.
The problem why this and other fields don’t get set is how records are entered. I upload my records as a batch update, and there aren’t any fields to record accuracy of location, gender, life-stage, etc. Even though I have this information available and could include it in the batch update.
After uploading the records, I then go through and manually add the photographs. After a 2 week holiday, this is a lot of records. So please can we have some more fields in the batch update, so these field are more likley to get set and add value to the records.
This stresses me out. My phone adds the precise location of my photos, its one of the things I check when uploading them. With an exact, pinpoint location I never questioned whether they needed an accuracy radius or not. I scrolled back through my most recent observations and found the accuracy lacking so I spent an hour or so adding it my observations. Then I realized that it was miserable to manually add it to past observations so I would just focus on all newly uploaded ones. But as I check, and add, all my new observations it seems like an unnecessary step since my locations are exact for the most part. I obsessively add details and annotations to my observations and adding the extra one that (to me) seems arbitrary has broken my brain. Am I understanding this correctly that even though my location is very precise, it needs an accuracy radius just so other people know that my location is precise and not randomly selected?
the people who are asking others to add an accuracy value make 4 assumptions:
however, the truth is that those assumptions don’t reflect reality. it cannot be determined whether a location in iNaturalist is accurate or not, regardless of whether an accuracy value is input or not, unless someone independently verifies that location, or unless you have some special knowledge that a particular location was determined using a method that will give you an accurate location for a particular use case.
there are likely just as many observations recorded with accuracy values that have bad locations as observations recorded without accuracy values.
if you’ve checked a few of your observations and your coordinates look fine to you, i personally would not stress yourself out reviewing or changing any of the rest of your old observations. it’s good practice to always make sure your coordinates look reasonable, and it doesn’t hurt to add accuracy values if you want to, but, it’s really your call whether you should make the extra effort to record an accuracy value, if you’re not already capturing that information as part of your existing workflow.
the only special case where i would always make the effort to include an accuracy value is in a case where i am purposely recording coordinates that i know are not near the true location of an observation or that i know have a high level of uncertainty. for example, for some observations where i want to keep locations very rough, sometimes i’ll just use coordinates at the center of downtown, with a 50000m accuracy circle that tells people not to use this observation for geospatial analysis, unless they are comfortable with that high level of error.
…
thanks @bouteloua for changing the title of the topic. “can be important to some data users” is better than “is important”.
I appreciate your reply, it helps! And that’s what I was thinking too, the presence of an accuracy value does not mean the location is correct.
Yep, part of uploading my observations is to double-check the location to make sure its correct. Adding the accuracy value would be three extra mouse clicks, opening the map, adding the value, saving my change. When you do several dozen observations every day it adds up.
Thank you again.
You just did.
:-D I assume you are pulling our legs :-D I am very happy with my Garmin Oregon 600 if the point on the (inaccurate) photographic image of the map displayed is within 10m of where I know I took the picture.
If I am trying to find that spot again another day I am equally as happy if I land within 10m.
I’ve only ever put acc data in my observations if I was unable to pinpoint the location. According to this thread I guess that 90% of my observations are:
If #2 is correct, then not notifying people of this when uploading observations is kind of a big error for a website whose information is intended to be used in scientific research. Surely this should be highlighted as an error when uploading, just like missing dates and locations?
So … how do I go back to add acc data to my 3000 observations that don’t include it? Adding acc=false works for searching observations, but not for batch editing.
it’s not
Most GPS recorders are not accurate to more than about 3m, and usually much worse, especially if in a forest, close to a cliff face or building, etc. 5-10m is much more accurate. The “accuracy” that a GPS recorder gives you is actually the “precision” of the instrument and not the true accuracy wrt to a position on the ground.
Also - if you are taking a photo of an animal (say a bird) - you are generally not right where the bird actually is. If you are giving an accuracy of 1m - that is more likely the position of the photographer than of the object being photographed. There is no shame in giving an accuracy of 10m or so - or 100-200 m of a bird in the distance.
This may be your interpretation of this thread. But this thread includes many diverse viewpoints, many if not most of which don’t support that interpretation.
Yes this is unfortunate, and I’m not seeing a good workaround at the moment.
Is adding observations to iNaturalist just “extra mouse clicks”?
The reason for the “accuracy” field, is to allow for more information, if one chooses to, it is not mandatory.
A researcher dropping points solely based on Accuracy may be dropping records unnecessarily. When this topic came up previously, I edited one of my GPS points to add an Accuracy value. The system also changed the value of Src (Coordinate Source) to “Manual” and I don’t think the user can select the Src value. I decided not to edit more records. If I was filtering records out that had dubious accuracy, I’m keeping the GPS unless something else seems wrong. The device provided the location when the photo was taken or iNat observation created. It’s probably not that far off. Src set to “Manual” would get more scrutiny. I’d probably toss really large values and look carefully at the really small values. How could you manually drop a pin within 3m of your true location when out in a sea of trees? I’d need to see a clear landmark nearby for that kind of accuracy.
even if there’s little support for it, you can see how a casual reader of this thread might come to that conclusion though, right?
@invertebratist – since you started the thead, i’m curious how you first came to the notion that foks should make extra effort to input an accuracy value on their observations in order to make them useful for science. is this something that some other user told you? is there actually some research somewhere where someone compared the reliability of iNaturalist data with and without accuracy values? are you filtering out records without accuracy values in your own research? (and if so, how did you decide that that was the right methodology for your use case?)
i think it would also make sense to edit you original thread, if you can, to include your later clarification:
I’m trying to find a way to interpret your comment that isn’t a rude dig. It sounds like you’re implying I don’t want to put effort into my observations to insure their accuracy?
There is always a way to allow for more information, the question is whether or not it adds anything substantial to the observation. We could ask users to additionally check a box that they’re “really really sure that the location is accurate” but where do you stop? My concern was directed at the researchers who said they disregard observations that are lacking an accuracy value. I never thought it was mandatory.
The point of the accuracy field is to show an area within which an organism was definitely seen. A pin based on GPS is going to show approximately where the observation was made, but not how precise that approximation is. If your GPS device doesn’t calibrate perfectly, the pin could be quite a distance from the actual location.
If adding an accuracy circle to your observations is such a problem for your workflow that it would stop you from adding observations at all, don’t do it. Some data is much better than no data. However, if you can take the time to make sure the location shown includes the true location, please do.
To certain people on this thread, please don’t encourage others not to include this information. I’ve seen several comments saying things like, “It’s not really important. You can add it if you feel like it, but otherwise don’t bother.” This idea isn’t helpful. We want to encourage people to provide data that is as correct and precise as possible. Most people will not make observations the ideal way, which is to be expected on a citizen science site, but our job is to do the best we can to provide accurate data and encourage other users to do the same.
Whenever I find myself going on a long iNatting trip that will require lot of phone battery, I would take the pictures with my phone’s camera and then upload them when I get home. This method has me just adding the location later than a live gps system. I often am not accurate enough with the location such as in a trail in a forstest. I would just add the location on the trail in the middle of the trail or where ever I choose to do it. I don’t think it’s truly necessary because I have the general location where the habitat is pretty much the same throughout. However, on my shorter and walks that I use Seek or iNaturalist Next app, the location is accurate.
My question is for the later location addition that are not accurate, should I have the geoprivacy setting set to obscured instead?
Just add an accuracy circle that shows where the observation was made. E.g. if you know you were somewhere on a certain trail, put the center point at the middle of the trail and expand the circle so it includes the whole thing. That’s pretty much the whole point of the accuracy circle.
So yes, setting is obscured and resize the accuracy circle to better fit the possible location. Tested it on one of my observations.
I’m not sure why you’re obscuring the location. That shouldn’t be related to the precision; It’s meant to be used if you don’t want other people to know the exact location.