Neither do I - but as we've seen matching on the name isn't and since we have a cast iron code why not use it and bypass any potential slip-ups?

[The agg isn't missing from Oligia strigalis - just not displayed in the processed information]

Hi Charlie,
My current impression (which remains pure conjecture based on the anomalies we are seeing) is that, when occurrence records were imported, the TVK associated with the record was used to try to make a match with the UK Species Inventory. However, it seems as if that copy of UKSI did not contain some of the TVK/names that the Gateway had employed meaning the the code could not be used.
Jim.

Hi Chris

I see what you are saying about the synonomy of Alophora barbifrons.

Likewise I see Tephrochlaena oraria is down as a synonym of Syntormon pseudospicatum not Tephrochlaena halterata.

I can't see any way of telling whether this a problem caused in migration to the Atlas or was a pre-existing problem. Presumably you think this is new, Chris, but I can't for the moment think how that might have arisen.

I guess we have to wait for David Martin to be back on the scene to look in to it all.

Jim Bacon.

Hi Charlie

I don't see any assertions about the taxon matching process but I can get some more information about the results. The query

https://records-ws.nbnatlas.org/occurrences/search?q=*:*&facets=name_match_metric

returns eight different results for the name matching method and the number of records to which it was applied. They are

+-------------------+-------------------+
| Name match metric | Number of records |
+-------------------+-------------------+
| taxonIdMatch      | 213427156         |
+-------------------+-------------------+
| exactMatch        | 2438552           |
+-------------------+-------------------+
| higherMatch       | 675574            |
+-------------------+-------------------+
| noMatch           | 514150            |
+-------------------+-------------------+
| canonicalMatch    | 59858             |
+-------------------+-------------------+
| fuzzyMatch        | 11790             |
+-------------------+-------------------+
| vernacularMatch   | 2820              |
+-------------------+-------------------+
| Unknown           | 2080              |
+-------------------+-------------------+

I'm reeling slightly at the size of those numbers... are there really 217 million records in there?
We can do a bit of digging in to any of these, for example with this query

https://records-ws.nbnatlas.org/occurrences/search?q=name_match_metric:noMatch&facets=raw_name

This reveals, top of the list, 32k records of Korscheltellus lupulina, NHMSYS0021142343, which have not found a match. Since neither this species nor this genus can be found in UKSI, and consequently not the TVK either, it fits my theory that there should be no match. Example record. There is no name you can search on to access these records. Meanwhile, according to the Atlas species list, this name is accepted in UKSI.

Jim Bacon.

Sounds about right - 152 million in the new BTO dataset, and there were over 100 million on the Gateway (included records with no public view which mostly haven't been published on the atlas, accounting for the "missing" records).

Thanks again for your detective work into the taxa matching issues everyone and particularly Jim, I hope it helps the team solve the issues quickly. There seem to be a lot of lepidoptera being affected - could it have anything to do with the new lepidoptera checklist (see threads in the species dictionary forum here and here)?

Hi Jim

Yeah, like you I could understand if the link was just broken complete but to establish a new, erroneous link seems weird. The UKSI database is fine and we've never had issues with the taxa that currently have problems so I doubt that anything was wrong with UKSI as a source of their data. The export was requested in an odd format (Text CSV) so it's possible that something happened during the export but I can easily reissue the files OR just give them an original Access copy which they can use to extract what they need.

I will see what David wants to do.

Chris R.