Something (`splitnc`?) needs to handle instantaneous variables missing `cell_methods = time: point`

[joshuatorrance]

At the moment there's some instantaneous data missing time: point from the cell_methods which is recommended from CF-Conventions.

This will hopefully be fixed upstream soon in CICE5 and um2nc but there's existing data that ideally should have time: point added to cell_methods for instantaneous ice and atmos data.

These data will be processed with splitnc anyway so potentially this issue could be rectified here. Alternatively another post-processing script could correct this.

Instantaneous ice data can be identified by time_rep = "instantaneous".

For atmos data may need to assume that any data missing time cell_methods (i.e. does NOT have cell_methods = time: mean etc.) is instantaneous.

See also:
#6 (comment)