How to flag new species added to "expected nearby" CV/ML models?

General
, , ,
1

Can we generate lists of the top unexpected (and/or misidentified/reversed ID) species by region on iNat?

Is anyone aware of a way to get lists of organisms that are not expected in a particular area by species?

The closest I can get for the Monterey Bay area is through the “Geospatial” filter ‘not expected nearby’ in the Identify tool: https://www.inaturalist.org/observations/identify?place_id=6b5c39d2-21bc-4957-a332-06f54283d3ab&expected_nearby=false

This works OK, but I’d like to be able to generate a list of species by # that are being proposed/identified in a particular region that are not expected to be in the area by the geospatial model. Is there another way to display or look this up that’s not in the “Identify” tool?

Maybe this is more of a feature request for “geospatial” and “disagreement” options be integrated into the “explore” tool? If a feature request more appropriate or has already been asked before, can the moderators please let me know before deleting so I can rephrase the question/request?

I’m wondering because an inordinate amount of effort must be spent to correct deeply-entrenched incorrect IDs for a particular taxa in a region–by being able to lookup a list of the most commonly reversed/incorrect ID species, we might be able to head off this issue, or at least “stop the bleeding” by reducing the CV suggestions for a particular area in the “found nearby” suggestions.

Even better would be a list of recent species additions to the “found nearby” in the ML/CV models so that any new species (possibly erroneously) added to a region (by people choosing an incorrect ID and then getting RG by someone auto-agreeing) would be highlighted for a second look.

1 Like
2

You can apply the same query parameters in Explore and use the Species tab: https://www.inaturalist.org/observations?expected_nearby=false&place_id=118063&subview=table&view=species

image
image1312×856 412 KB

Rather than accepting new feature requests for the Explore page, suggestions are going in this topic: https://forum.inaturalist.org/t/ideas-for-a-revamped-explore-observations-search-page/8439

3 Likes
3

Thank you so much! I figured there might be some html-based solution, I really need to brush up on how to modify it for my needs. Maybe there’s a forum post out there on this? ;)

I did manage to modify it further based on the filters to get exactly what I wanted (plants at RG that aren’t expected in the area), which works great when scrolling down to some of the less commonly identified species (incorrectly RG observations [incorrect ID/not wild] highlighted):

image
image1212×596 272 KB

It also is useful for finding unusual species that are not necessarily with any data quality issues, but are notable for being waifs, disjuncts, etc potentially like the Kopsiopsis strobilacea, Viburnum ellipticum and Sanicula arguta above.

Thanks again!

3 Likes