Closed MortenHofft closed 5 years ago
It could, but we don't have the taxonomy as context when interpreting the elevation. The limit is 17km for everything: https://github.com/gbif/parsers/blob/master/src/main/java/org/gbif/common/parsers/geospatial/MeterRangeParser.java#L80
All but 54 of them have elevation=9999, presumably some placeholder for "unknown". The 54 will mostly be where the measurement is in feet, we detect and convert the value if "ft" is present, but otherwise don't.
https://www.gbif.org/occurrence/datasets?elevation=9900,9998&taxon_key=6&taxon_key=5&advanced=1
@fmendezh In that case a suggestion for the new pipeline: Consider taxonomy when flagging elevation issues. I'm removing the API label as I consider it a won't fix for now
For georeferenced occurrences, we can even consider location.
It looks like Species Links is using 9999 as default value for some of the datasets that they host (see list below). I contacted them by email.
the Species Links issue is fixed and we are working with Burke university to republish a new version of their dataset.
content 250K plants and fungi above 9900m That sounds unlikely.
api We have a flag for unlikely - shouldn't it be set? Currently
elevation unlikely == 0
https://www.gbif.org/occurrence/datasets?elevation=9900,9999&taxon_key=6&taxon_key=5&advanced=1