KAN_B position

[jasonebox]

re: dataverse data, via ECMWF Simon Noone Simon.Noone@mu.ie asks

"Here is an example that I found of a station with no gps information at all. "

whereas the other stations include position data. I think it's worth including that position... what I have: 67.1252, -50.1832, 350 m

Jason

[BaptisteVandecrux]

Similar to #152, there is no GPS observations and there is a need to decide whether we fill a field called "gps_lon" with a value that is not measured continuously by a GPS.
On the dataverse, there is a metadata file that should have a set of "standard coordinates" for each station. The same standard, static coordinates should also be available as an attribute "lat" "lon" and "alt" in the netcdf files.

@sjnoone I suggest that you use the metadata file above and

1. if location_type is local glacier or ice sheet then you use the observed coordinates,
2. if location_type is tundra or bedrock and there is no observed coordinates, then the standard static coordinates can be used instead

[BaptisteVandecrux]

Closing here because it is not a data-related issue.
I don't think the variable gps_lat gps_lon gps_alt should contain other thing that continuous GPS measurements.
We could decide to replace them with more general lat lon alt which would contain either interpolated or static values.
Coordinates interpolation was requested here: GEUS-Glaciology-and-Climate/pypromice#236
and I am adding cases of static AWS there as well