List of suggested species "Expected nearby" in drop down menu shows species ocurring hundreds of kilometers away

Bug Reports
1

List of suggested species “Expected nearby” in drop down menu shows species ocurring hundreds of kilometers away, some of them endemic species which never would occur in the area. For example in the High Tatra (Slovakia) Campanula scheuchzeri (only in the alps) is shown in the first position (responsible for 48 misidentifications), also Campanula pulla (endemic of austrian alps, 350 km distance minimum) is suggested, while Campanula tatrae does not occur.

315
3151000×1333 170 KB

This is already responsible for some definitely false records in GBIF …!
How does the algorithm work and why does it show species never recorded in an area as “nearby” but omitted species with undreds of obs. in an area???

Platform: windows 11

App version number, if a mobile app issue:

Browser, if a website issue:

URLs (aka web addresses) of any relevant observations or pages:

Screenshots of what you are seeing:

Description of problem:

Step 1:

Step 2:

Step 3:

2

what are the coordinates of that record?

3

May we have a link to the obs ?

The first suggestion used this Geomodel map. If the sp is included in CV already, the map both answers and raises questions, depending on how you evaluate the map. Hexagons are 60km so for local Endemics that gets complicated.

4

https://www.inaturalist.org/observations/374243230

5

grafik
grafik1817×912 558 KB

grafik
grafik1557×918 373 KB

here an export from GBIF for the 2 species

6

Thanks. Your sp tatrae has 54 obs.

We need 60 for CV update

https://inaturalist.freshdesk.com/en/support/solutions/articles/151000170368-which-taxa-are-included-in-the-computer-vision-suggestions-

So your sp does not have a Geomodel map for CV yet (no Nearby to offer). If you can find another 6 obs, then you may have to wait a month or 2. Previous CV update was https://www.inaturalist.org/posts/130939-updated-computer-vision-model-and-geomodel-with-over-1-300-new-taxa which uses data from April and was published in May, so we could get the next one soon, and your extra 6 would come in later.

7

Thanks for clearing the matter!!!

8

If it comes up again – it worth going to the taxon page – click About – is it Included (click to Geomodel map) – is it Pending (how many obs do you need to get to 60 ?) You can see comments from taxon specialists on the blog post each time – worked hard – glad to see mine included this time.

Random example where we need 3 more obs https://www.inaturalist.org/observations/357433825

9

But there’s also a higher number of ‘at least x photos’ or something, right?

10

Yes – a hundred photos was the original rule – but if you click to iNat’s Help article – we can lean on 60 obs which is clear on the taxon page. How many photos is not info which is immediately visible.

11

Fair - I just don’t want people being disappointed if they’re expecting to have it added to the CV on 60 and it isn’t…

12

It is a bit like taking a photo of a flower in a breeze, which picks up just as you click the shutter. Moving target.