Indexing: misinterpretation/-representation of longitudes in the eastern hemisphere of Antarctica

18029
Reporter: ahahn
Assignee: cgendreau
Type: Bug
Summary: Indexing: misinterpretation/-representation of longitudes in the eastern hemisphere of Antarctica
Priority: Critical
Status: Reopened
Created: 2015-11-11 10:33:49.19
Updated: 2015-12-10 13:29:44.913
        
Description: reported by Dave Watts (10.11.15 via email to helpdesk). summary:

In http://www.gbif.org/dataset/86788d2c-f762-11e1-a439-00145eb45e9a, multiple longitudes have been wrongly negated to be in the western hemisphere and not in the correct eastern hemisphere as published by the source. In addition, they carry an error flag and a country mismatch indicator.

All the data are in the eastern part of the world. The same is the case for a number of other datasets:
- RMT Trawl catch from the 1982/83 V2 ADBEX I voyage
- RMT Trawl catch from the 1984/85 V5 SIBEX2 voyage
- http://www.gbif.org/dataset/8679c142-f762-11e1-a439-00145eb45e9a
- http://www.gbif.org/dataset/7b7269b6-f762-11e1-a439-00145eb45e9a
etc]]>
    

Author: jlegind@gbif.org
Created: 2015-11-11 16:07:40.383
Updated: 2015-11-12 10:10:02.617
        
[~cgendreau]
Let me just add some details on this:
If you access this record http://www.gbif.org/occurrence/201208312  you will see that it is placed under Chile and Argentina, while the actual trawl took place in the eastern hemisphere. Go to the verbatim record and you will see the true coordinates. It is safe to say that the longitude flip took place in the GBIF interpretation.
This relates to the issue http://dev.gbif.org/issues/browse/POR-2181 where it was decided that geographical look-up that return [ country==NULL and latitude <= -60 ] are assigned country code "AQ" (Antarctica). HOWEVER this was not implemented - please talk to [~omeyn@gbif.org] about it.

I suspect that if a record has the stated country as AQ AND the coordinates fall outside the Antarctica shape file, then the process starts flipping coordinates UNTIL it lands within the AQ coordinate polygon. In this case it was in the middle of Drakes Passage.

   
    

Author: cgendreau
Created: 2015-12-10 10:38:58.236
Updated: 2015-12-10 10:38:58.236
        
Fixed as mentioned in http://dev.gbif.org/issues/browse/POR-2181.
All records below -60 deg have been reinterpreted.
    

Author: cgendreau
Comment: Some records with latitude > 60 are still problematic.
Created: 2015-12-10 13:29:44.909
Updated: 2015-12-10 13:29:44.909