There is a post elsewhere about records reported from the middle of the North Sea, or North America. These are presumably incorrectly entered grid references. Since one of the steps in loading a record into the database is to assign it to a VC every OSGR entry must be checked. Those records with values outside the area for which it is defined can be identified as flagged as erroneous. Those which are not assignable to a VC may still be valid, e.g. marine sea-birds, and are should not be rejected automatically. However, a message could be sent to the recorder asking them to check the location.
Comments
I was intrigued by these…
I was intrigued by these postings so went looking for such erroneous records. Certainly there are some Vulpes vulpes attempting to swim the Atlantic!
I did note an interesting record 3665231 of a Small Tortoiseshell from the North Sea with a photograph clearly taken on board a ship. This record may well therefore be correct but it might not get verified because I doubt any County Recorder is enabled to verify records in the North Sea? I will raise this with Butterfly Conservation but there may be a similar issue with other recording groups.
also
Hi Keith Thanks for the speckled wood record you flagged. Interesting that the recorder used their phone to record it, and their gps put them 2km further north in the sea from where they actually were. I think it's a general problem with phone gps - I've put my phone down with the map showing at home and the dot showing my location has walked out of the house, across the road and into the next street without the phone moving.
'Onshoring' efforts
It's easy to spot dodgy map references when the dot lands in the sea. I've been busy over the winter onshoring dragonfly records that have a named Location. The errors seem to be caused in several ways:
- App users who submit their records when they get to the car park or back home, failing to adjust the map ref from the the system-supplied GPS coordinates.
- Phone GPS that hasn't 'settled' because the phone hasn't been stationary for long enough. (Anecdotal; unproven)
- The recorder can't read a map. No names here but one frequent misplacer has FRES after his name.
- Getting the myriad (initial two letters of an OSGB ref) wrong, so that the ref is out by 100 km.
- Transposing digits when typing in a grid ref. e.g. TQ234789 for TQ247489.
- Using a named location that has been used before, assuming that the map ref is correct. This results in clusters of records in the wrong place.
- Imported records where something has gone awry with the spreadsheet processing. (Anecdotal; unproven)
There are surely more ways in which these errors can happen and I suggest encouraging recorders to include a detailed Location name as a cross-check. It might be best to require a Location name before the record can be saved.
Since most records in the sea do not belong to a Vice County, most verifiers in recording schemes such as ours can only see records within their own group of VCs and the offshore records do not show up. To see them they have to go to Explore > All records, where verifiers rarely venture.
I'm now working on queries (in an offline database) looking for egregious map references attached to records with location names. This is far from foolproof and throws up many false positives where placenames such as 'Farnham' and 'Abbots Field' occur at several places across the country. But we'll give it a go.
Far too many records are getting entered to iRecord with impossible location, such as North Atlantic for terrestrial mammals, Greenland etc. The record usually states a UK location, eg a town, and then uses Lat-Long incorrectly.
Please can iRecord admin check data outside of the UK, and get it removed?
Also, some verifiers have ticked these records, despite not checking the location.