search

cmip6dr / CMIP6_DataRequest_VariableDefinitions

Definitions of variables in the CMIP6 Data Request
7 stars 0 forks source link

correction of time axis specification for tasmax and tasmin in Amon #400

Closed taylor13 closed 4 years ago

taylor13 commented 5 years ago

@larsbarring noted in https://github.com/cf-convention/discuss/issues/371 that in the Amon table, the monthly mean daily max and daily min temperatures (tasmax and tasmin) should have a "climatology" axis rather than a regular axis. I'm not sure whether changing this for CMIP6 will mess up those in the middle of writing output (I don't think it would), but we should definitely correct this for CMIP7.

The difference in a "climatology" time axis and a regular one is that the name of the "bounds" variable is stored in the climatology attribute, not in the bounds attribute. I'm not sure if tasmax and tasmin in Amon should be functions of time3, time4, or some newly defined time axis.

Note that for these variables the actual values that should be stored in the variable pointed to by the climatology attribute would be identical to the values currently written to the variable pointed to by the bounds attribute.

martinjuckes commented 4 years ago

I agree that we should wait for CMIP7. This is a feature inherited from CMIP5, so users will not be supplied. I agree that it is likely that data producers would cope, but it would be confusing to data users if we end up with two different implementations for this variable.