Open emiliom opened 3 years ago
@emiliom - for version 2 of the convention, lat and lon have moved - there are now lat/lon in the /Sonar/Beam_groupX that at the ping times and lat/lon in the /Platform/Positions group(s) at the position sensor times. There are coordinate attributes added too, so I think things are a few steps closer to meeting CF H.4.2.
I don't quite understand how to do the cf_role="trajectory_id" - do you have any insight on this? Is it necessary for each trajectory_id in a file to be unique? Or unique across a dataset, which can be many files?
Point out explicitly in the convention document, under the
Platform
group, that the lat/lon data in SONAR-netCDF4 ver.1 is consistent with the CF H.4.2. Single trajectory specification (a CF Trajectory Discrete Sampling Geometry). Possibly suggest small additions that would move closer to actual CF adherence, particularly adding a trajectory variable with an attributecf_role = "trajectory_id"
and possibly acoordinate
attribute on trajectory "data" variables likespeed_ground
.From @gavinmacaulay: The consistency with the CF single trajectory specification was deliberate :) We tried to use CF conventions whenever it was sensible (and did attempt to meet CF-1.7, although it wasn't feasible for all of sonar-netcdf4 v1).