Interesting thanks. That is something that should be changed then, Null should not be the same as 0,0.
This seems like a good idea. The “Null” project, as someone mentioned, captures these, but it does not capture observations with location (0,0) which are auto-obscured. You can see this clearly here, for example this observation. Some of these observations achieved Research Grade - I just went through and marked these “location inaccurate”.
In most cases, observers seem not to be aware there is a problem. Could there be a way of alerting observers that some of their locations have been mapped to the middle of the ocean, perhaps with a notification?
I went and left a comment with instructions as well as voting no to the location is accurate on the records which were : research grade, birds, reptiles or mammals (I grew tired of the repetition, so others can try other families)
The reality is almost all of these come from users who
- use the mobile app so are unlikely to see the comment
- or have very low observation counts suggesting they used the app once or twice and have since stopped
So the reality is a very low number will likely ever get fixed by the user.
Thanks Chris. I have done the same with the rest of the Research Grade observations there for other taxa.
One other thing I noticed: observations stay in the “NULL” project even after their location has been fixed. It would be good to convert the NULL project to a collection project, an option which was not available when this project was created - @silversea_starsong as the creator of this project, is that something you agree with? It should be a square of the same size as the bounding box that encompasses obscured observations.
I still think an automated way to fix this problem would be good. Another idea is that once a day/week/month, a script could check for newly created (0,0) observations and add a vote for “location inaccurate” to the DQA. This could be easily over-ridden by the observer in the unlikely case that anyone actually observes something at precisely (0,0).
I will look into this. Thanks for the notification!
I updated the project – it now also includes observations that are auto-obscured and marked at Lat/Long (0,0).
FWIW I was recently alerted by someone commenting in half a dozen of my obs that they were in the null project. I had no idea they were there, and missing from their collection Projects. After that I searched for any more in the null Project, but it would have been good to know straight away. And I have not remembered to check that project since - must do that.
I don’t use the mobile app, and have over 11,000 obs. I don’t know how I failed to provide a location, but I have had other obs since lose their location data, while I was editing location I think. But with those ones I was aware of it and fixed it straight away.
Great! Does that mean that if due to a Conservation status something becomes auto-obscured, we will see it in the Null Project?
I checked in again in the region of 0,0 and 13 observations had reached Research Grade in the last week: about two per day. I’ve marked them as location inaccurate with a note to the observer, but if anyone else wants to check in periodically to do the same, you can do so here.
Not certain but I suspect the 16,000 kilometer accuracy circle might be the cause of it not being in the project.
I added this as one of the ways to help out on iNat, as well as drafted some example text to leave for users who mistakenly saved their obs at 0,0:
Nice! Thanks Cassi!
As mentioned here, there seems to be somewhat of an influx of observations at 0,0. To help fix these:
- Go to this link: https://www.inaturalist.org/observations/identify?reviewed=any&quality_grade=research%2Cneeds_id&order=asc&project_id=null
- Click the first image to open the Identify modal
- Click the “Data Quality” tab
- Give it a thumbs down next to “Location is accurate”
- Go back to info and inform the user their location is inaccurate. Here’s the text I use below, but feel free to modify it of course:
Hi, it looks like the location was accidentally saved at coordinates 0,0 — in the Atlantic Ocean! :) To add a location, click the “Edit” button, and use the map to drop a push-pin to more accurately depict where you were when you made this observation. Even just a general location is helpful. For now, I marked this observation as having an incorrect location in the Data Quality Assessment section, so please reply here letting me know once you’ve fixed it. Thanks!
Progress / Backlog for Observations at 0,0
(Excludes observations that were corrected in the meantime)
|Date||Observed in previous week||Total Needs ID/RG||Obs in NULL project|
|March 13th, 2019||12-ish? (via gorilla test server)||409||15,601|
|June 10th, 2019||164||591||unknown|
|June 21st, 2019||266||805||unknown|
|June 28th, 2019||151||307||unknown|
|July 6th, 2019||226||208||18,225|
|July 14th, 2019||134||36||18,289|
|July 21st, 2019||217||168||18,525|
|July 28th, 2019||213||352||18,700|
|September 5th, 2019||0||959||19,519|
The project has 18,000 observations right now. 805 sounds a bit optimistic!
Yep, though of those, around 17,000 are casual grade.
wow… are these all updated manually one by one? that seems super tedious and inefficient. seems like this is something that could be automated, especially if it’s an ongoing thing. are there other such tedious and ongoing things like this?
There are several listed here: https://forum.inaturalist.org/t/ways-to-help-out-on-inat-wiki/1983
The total number of 0, 0 observations at Needs ID/Research Grade are now down to zero, but new ones continue to reach RG every day.
I requested a way to automatically deal with observations at 0, 0 here: https://forum.inaturalist.org/t/automatically-deal-with-obs-whose-coordinates-are-0-0/4579
This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.