CMIP6 Volcello is its own cell measure.

[ledm]

In the CMIP6 tables, CMIP6_Oyr.json, CMIP6_Omon.json, CMIP6_Odec.json, and CMIP6_Ofx.json, the cell volume volcello is set to be it's own cell measure.

ie, in CMIP6_Oyr.json:

        "volcello": {
            "frequency": "yr", 
            "modeling_realm": "ocean", 
            "standard_name": "ocean_volume", 
            "units": "m3", 
            "cell_methods": "area: mean where sea time: mean", 
            "cell_measures": "area: areacello volume: volcello", 
            "long_name": "Ocean Grid-Cell Volume", 
            "comment": "grid-cell volume ca. 2000.", 
            "dimensions": "longitude latitude olevel time", 
            "out_name": "volcello", 
            "type": "real", 
            "positive": "", 
            "valid_min": "", 
            "valid_max": "", 
            "ok_min_mean_abs": "", 
            "ok_max_mean_abs": ""
        },

This seems like a strange choice and have caused some trouble for Iris loading these files. We have a discussion in the iris issues page, here:
SciTools/iris#3544

[durack1]

@taylor13 can you take a peek at this, we could have similar issues with other area/volume variables

[taylor13]

This issue has been raised before as issue 216. If something is to be done, it must be done in the data request. Then it will propagate to the CMOR tables. On the github data-request repository, I reported this as issue 394. I see someone has already reported the iris problem there, but I will reinforce that with a comment.

[ledm]

Thanks @taylor13, @durack1.