search

cmip6dr / Request

Request specification (which variables for which experiments, etc)
2 stars 0 forks source link

Handling variable choices #61

Open matthew-mizielinski opened 6 years ago

matthew-mizielinski commented 6 years ago

There are a number of instances within the data request for which two cmor variables have the same variable name and hence would produce the same output file. One example is ua7h and ua in 6hrPlevPt for several of the DECK experiments and HighResMIP (discussed in PCMDI/cmor#246).

I think this is down to the need for data providers to make a choice between two options. In several cases this is down to different variable shapes, or different pressure level sets as in the example mentioned above.

It would be really useful to have a clear view of the choices that need to be made when deciding on output to produce. A quick analysis of the data request (at v01.00.15) suggests that there are 69 entries with this issue. Within the data request there is the varChoiceLinkR class designed to hold this information, but I'd like to request a simple document (e.g. google doc or pdf) describing the choices that need to be made for each type of choice.

martinjuckes commented 6 years ago

See also https://github.com/cmip6dr/CMIP6_DataRequest_VariableDefinitions/issues/63

martinjuckes commented 6 years ago

See also https://github.com/cmip6dr/CMIP6_DataRequest_VariableDefinitions/issues/167 hfgeo