Create Filter by Location Accuracy

A observations filter for location accuracy would be really useful. Specifically I would like to be able to set:

  • A maximum and/or minimum uncertainty
  • A filter to show/hide observations that don’t have an uncertainty value
  • A filter to show/hide obscured and/or unobscured observations (and maybe even a way to distinguish auto-obscured vs. manually obscured observations.

Use cases:

  • If I’m browsing the map in search of interesting locations to visit, or for locations that have no observations, obscured and very uncertain observations make this very challenging
  • It helps in finding observations that are mapped wrong (e.g. most observations in the middle of a lake are probably mapped wrong. But it’s hard to find them when obscured observations are also shown on the same map)
  • It will provide a way to easily see what species are being auto-obscured
  • If you’re looking into microhabitats, observations above a certain accuracy value are useless
  • It would let me find my own observations that I haven’t provided a location accuracy for
  • Etc.

Seconded! I would definitely use this, much of what i do requires high spatial precision and it would be nice to be able to hide or filter things that don’t qualify.

1 Like

I agree this would be nice to have. Keep in mind that the Explore page/Observations search (https://www.inaturalist.org/observations) likely won’t have any filters added to it until it is entirely rewritten. It was written in a language we’ve decided to move on from and it doesn’t make sense to add more filters to something we’ll be scrapping (hopefully sooner rather than later).

3 Likes

That being said, perhaps something like the taxon page maps can include this in the meantime. I’ll see if it’s possible.

2 Likes

I trust that the filters will be added to rather than deprecated on the Explore/Observations page?

In addition to Locality resolution added - which I wholeheartedly endorse, the taxa rank filters need to be fixed.
It would also be nice to have the Observation fields and Annotations included in the filters. (and when it eventually arrives, a cutoff on the summed reputation of the identifiers).

1 Like

Ok, just added a few more API parameters that will work in obs search and Identify. No UI yet, maybe never since these features are very niche, IMO. Here they are, with examples:

Not quite the same, but the geoprivacy API parameter has existed for a while. It’s not in the UI because, well, like everything you’ve requested it’s really only of interest to a tiny fraction of power users, but here’s how you edit a URL to use it:

Searching by whether or not the coordinates were obscured because of a taxon’s conservation status (the taxon_geoprivacy param) was something we released like last week or something, but it works pretty much the same:

I should note that there’s still some work to be done here, specifically on the maps. I don’t think they’re using these new params just yet.

10 Likes

wow, awesome, thank you!

1 Like

Excellent stuff!!! Thanks!

1 Like

Wonderful, thank you! Should these be added to the URL wiki? (https://forum.inaturalist.org/t/how-to-use-inaturalists-search-urls-wiki)

2 Likes

If/when the maps are fixed to display properly that would be a good idea - right now I think it will just cause issues from people being confused.

2 Likes

One thing you may want to look into (if it is expected behaviour, then please ignore), but if I click on this link https://www.inaturalist.org/observations?geoprivacy=private
I get a very small subset of the private records, because I have a default search location set to Canada.

I’m not sure how small a default search location I could set and still have it actually apply the filter.

1 Like

For me, that is only showing my own records, not anybody else’s. Which seems fine.

1 Like

not for me, I see everyone’s.

1 Like

Theoretically all place-based searches should exclude observations with obscuration cells or positional accuracy circles that aren’t contained by the place’s bounding box, so there’s a limit to how small a place you could use for this kind of attempt to guess the true coordinates of an obscured observation.

1 Like

What is this “taxon geoprivacy”? I see it applies to 22 species. How does this differ from taxon_obscured? When does it apply? Who decides?
And why does it not apply across the taxon, but to only certain cases? (e.g. 72 out of 525 observations for Clemmys guttata)

1 Like

There is no decision beyond someone with curator powers going in and setting it. It applies only to a subset of C. guttata records as most jurisdictions are set to obscured, but some curator(s) have set it to private for 3. Those being the US states of Ohio, Vermont and the Canadian province of Ontario.

1 Like

Are there guidelines for when to use private vs obscured? Is this entirely at the discretion of some concerned curator? How is conflict between curators mediated?
This is the first I have heard from it, and I am on the South African Sensitive Species committee which regulates locality data access, which is in the southern African community iNat contract.

1 Like

There is no written info anywhere on the use of private I am aware of. The only guidelines are the ones mentioned here : https://forum.inaturalist.org/t/change-process-to-update-obscuring-level-of-taxa/558.

There is no tracking or display of who makes a change either adding, or removing a default geoprivacy.

Basically last curator who makes a change wins.

1 Like

Thanks - but there is nothing there about the “privacy” option (vs obscured) for taxa. That was entirely new to me from this thread. Certainly worthy of some guidelines.

1 Like

‘Thanks - but there is nothing there about the “privacy” option (vs obscured) for taxa.’

That’s exactly what my 1st sentence says.

1 Like