How exactly do you do that? I would like to do the same for my area but can’t figure it out.
You may find that you can have a similar experience in an easier way by IDing in your county, then moving on to a neighbouring one, and so on.
3 Likes
Yes, I was doing that, but it means I have to take large jumps in amount of observations. This way I could move in smaller increments.
2 Likes
I looked for a simple latitude close to where your last observation was and put a radius of 100 on it. In this case I used lat=48.0 and long=-122.50 which coincides to a place on Whidbey Island. Try this if you wish: https://www.inaturalist.org/observations/identify?iconic_taxa=unknown&lat=48.0&lng=-122.50&radius=100
There are 2,361 unknowns.
This is just across the pond from where I live.
If you find yourself going too far east as you expand your radius, I tend to use a corridor for identifications from northern California along a wide strip to almost the west of Alaska - the Pacific Northwest Coast zone. There is a lot of similarities in species throughout this: https://www.inaturalist.org/observations/identify?iconic_taxa=unknown&place_id=67725
2 Likes
Thanks!
1 Like
Yeah, the trick with the iNat search tool is that a lot of advanced commands are not available in the search interface, so that you have to manually “code” them in the url.
Looks like bobmcd answered your question, so have fun!
1 Like
Happy new year!
I am still working on the tool I presented above. I made several rounds of rework, that’s why it takes so much time. I hope to release it soon, but I would like to test it further, to make it as robust as possible against unexpected events.
I would like to show you here a view of the data collected with this tool.
So far, the tool downloaded the data of 178383 “Unknown” observations from these places:
- Central America, South America,
- Mexico,
- Alabama, Arizona, Arkansas, California, Florida, Georgia, Louisiana, Mississippi, New Mexico, North Carolina, Oklahoma, South Carolina, Tennessee, Texas,
- Hawaii.
The tool generates a “best ID” (similar to the “We are pretty sure this is in…”) for every observation and I was curious to generate a report to show the taxonomic distribution of these many “Unknown”.
This report shows the (guessed) distribution of these 178383 “Unknown” observations down to the rank Infraclass:
Every raw shows 3 numbers. For instance, the raw Animalia means that, among these 178383 observations, 2187 are guessed to be animals, and 25748 other ones are guessed to be animals with a best ID at a rank between Infraclass (included) and Kingdom (excluded). (Note that these 2187 ones may have a best ID at a lower rank, for instance a Family that would be grafted directly to the Kingdom, without a Class, if it exists). 27935 is simply the total of all of them.
This is part of a more detailed report showing the distribution of these 178383 observations down to the rank Family:
This is part of the most detailed report, showing all the 145 best IDs in the Family Urticaceae:
This means that if you are interested in identifying the Family Urticaceae, you may define this Family as a filter in this tool and then review and identify these 145 observations.
7 Likes
The tool also generates a report of all identifications made by the user.
For instance, this shows all my identifications of a few “Unknown” observations, done during the last 2 weeks while testing the tool:
The report includes all the details, with links to the observations identified and to the taxa pages. For instance:
Cassia leptophylla
https://www.inaturalist.org/taxa/335464
https://www.inaturalist.org/observations/7511649
https://www.inaturalist.org/observations/6911100
1 Like
For submitting IDs that are not in the IDs suggested by the AI, it is convenient to manage a “Bookmarks” taxonomy (there is one for every taxonomic filter defined in the tool settings). I use these bookmarks for instance to identify Cassia leptophylla, a taxon that never appears in the AI suggestions:
The combobox above shows all my bookmarks attached to the filter “Subfamily Caesalpinioideae”.
Using the checkbox to show all “Ancestors”, the button to show all “Children” of the taxon selected (can be applied recursively), it is possible to browse the whole taxonomy. Using the button “Add” to bookmarks, it is then possible to add any taxon to the bookmarks.
In addition to the AI suggestions and to the Bookmarks, it is also convenient to browse an “Overview” taxonomy attached to the filter selected and containing the filter taxon itself, all its brother taxa and all its children taxa:
The colored raw corresponds to the best match between the contents of this taxonomy and the best ID for the observation being reviewed.
The 3rd taxonomy available is the one generated from the AI suggestion for the observation being reviewed:
So, the user has 3 taxonomies available (1 entirely managed by the user) for selecting an ID to submit, plus the tools for browsing the whole taxonomy available in iNaturalist when this is not enough. The display of these taxonomies is automated, the user never has to order the taxa manually.
3 Likes
That seems quite amazing!
1 Like
I agree! I very recently got the idea (it just popped into my head) to look at the taxonomies for the most likely-looking suggestions and find a common ancestor. Before trying this, I usually identified to genus, species, or kingdom - rarely knowing anything in between. At this point it is not well tested (I’ve only been doing it a few days) but seems like a good plan. I wish there was a quicker way to do it, though.
4 Likes
I m trying to kingdom as much as i can to reduce it, while learning my way around here. Also on flowers, i try to add annotations like flowering. I see they are being noticable faster. At list i can do this for experts i guess.
If it is not good thing to remove them from there, feel free to speech me about ways to being helpful better :)
4 Likes
Does anyone have suggestions for making IDing more interesting? I’ve been working on Unknowns, but it’s gotten so I can’t handle more than one page at a time. I’ve tried filtering by area, but it didn’t help that much. I need to do this, but it’s just so boring now, it’s hard to manage. Also, I would prefer to filter out the non-unknown observations in the pile (life is OK, but bacteria etc I cannot identify). What URL additions do I need for that?
&without_taxon_id=131236%2C67333%2C151817
3 Likes
https://www.inaturalist.org/observations/identify?taxon_id=184884&lrank=superorder
That’s what I’m looking at for past month or so, - 90 pages (+2-3 each day), pretty easy to go through if you know insects at least a little.
My first reaction to seeing your comment here is, take a little break! Ask yourself what your “need” is. No worries, there are still other folks whittling down those Unknowns while you do something different and there will still be some to look at if/when you come back to it.
If it’s satisfying to see a pile get smaller as you id, then pick a different, more interesting pile. (Going to a still-broad category where you can help narrow down some things, like @fffffffff suggests, is a good strategy when I want a change of pace.)
If it’s satisfying to just be helping in a positive yet easy way, there’s a whole sphere of Annotations- you can pick something interesting yet relaxing to go through. Related to that (same pane in Identify), adding Observation Fields can be helpful too, and can be as complex or simple an operation as you want depending what you do with it.
My overarching message: Keep it fun, and if it does not feel fun or feels like just a job, switch up to something more fun.
11 Likes
I made a comment on the State of matter Life cleanup regarding homonym errors that may be helpful in decreasing “Unknown” records.
2 Likes
Check the number now (as of this minute, 219,003) 
It’s been pretty much halved, despite so many observations being added!
How many of those have you all reviewed? Me, I have a few over 12,000 that I personally can’t improve on and I’ve marked as reviewed.
3 Likes
How do you count the reviewed ones?