We currently only support ingestion into HARP of the ASCII ACE data, and although this suffices, it might be more convenient (and future proof) to support the netcdf format.
There are two flavours, the complete time series and the 'time separated', but the format of both is the same.