Instantaneous variables

[MargaridaS]

Dear CF Team,
I contact you because some instantaneous data from the ESGF contain time boundaries. I guess that instantaneous variables should not embed time boundaries, is this correct? Then, should the variable time_bnds be removed from file?

For example:
http://crd-esgf-drc.ec.gc.ca/thredds/fileServer/esgA_dataroot/AR6/CMIP6/CMIP/CCCma/CanESM5/historical/r1i1p2f1/3hr/uas/gn/v20190429/uas_3hr_CanESM5_historical_r1i1p2f1_gn_185001010300-186101010000.nc

Thank you very much in advance.
Best regards,
Marga

[JonathanGregory]

Dear @MargaridaS

It is fine to have cell boundaries in time for instantaneous data, or in space for data that applies at a point. I understand your concern that the boundaries are irrelevant to locating the data, but they may still be useful for other calculations that analysts would like to do.

Best wishes

Jonathan

[MargaridaS]

Dear @JonathanGregory
Thank you very much for your prompt reply. Then, I'll leave the time_bnds in case some people would like to do different analyses. Thank you very much!
Kind regards,
Marga