Open tcarval opened 1 year ago
Argo NetCDF uncertainty on observations is reported in the _ADJUSTED_ERROR variable. \_ADJUSTED_ERROR is an ancillary data for \_ADJUSTED. The CF convention "ancillary data" chapter recommends to use the variable attribute "ancillary_variables".
We may consider to manage "ancillary_variables":
float PSAL(N_PROF, N_LEVELS) ; PSAL:ancillary_variables = "PROFILE_PSAL_QC PSAL_QC" ; float PSAL_ADJUSTED(N_PROF, N_LEVELS) ; PSAL_ADJUSTED:ancillary_variables = "PROFILE_PSAL_ADJUSTED_QC PSAL_ADJUSTED_QC PSAL_ADJUSTED_ERROR" ;
float PSAL(N_PROF, N_LEVELS) ;
PSAL:ancillary_variables = "PROFILE_PSAL_QC PSAL_QC" ;
float PSAL_ADJUSTED(N_PROF, N_LEVELS) ;
PSAL_ADJUSTED:ancillary_variables = "PROFILE_PSAL_ADJUSTED_QC PSAL_ADJUSTED_QC PSAL_ADJUSTED_ERROR" ;
Argo NetCDF uncertainty on observations is reported in the _ADJUSTED_ERROR variable. \_ADJUSTED_ERROR is an ancillary data for \_ADJUSTED. The CF convention "ancillary data" chapter recommends to use the variable attribute "ancillary_variables".
We may consider to manage "ancillary_variables":