Figre out how to get tasmax and tasmin data for CESM2-WACCM for 2015-2100

[kyleredilla]

In finalizing our ensemble, we opted for CESM2-WACCM over CESM2 because data representation appears similar and -WACCM version had slightly better skill. According to the variant selection process (notebook), the r1i1p1f1 variant had the most representation, so we chose for this one. However, this variant does not show any holdings for 2015-2100 data for tasmax and tasmin, so we need to come up with a solution. I have contacted one of the data engineers listed as a contact for the dataset to see if they have plans to make that data available.

There are kind of 2 problems to consider here:

1. Main thing is that we should select a different variant for CESM2-WACCM if there is one that has solid data availability (pending a negative response / no response from NCAR data engineer). This could be done in a separate notebook, or we could update the notebook for all models (2)
2. This one is kinda a more generalized version of 1., in that we could maybe update the variant selection to include some analysis of what time periods are actually available, and do so for some set of higher priority variables (which would include tasmax and tasmin) instead of just simply looking at the number of variables available. I.e. this taught us that just because a variable is available doesn't mean it is for the time periods we need it for. The temporal coverage information is not part of the directory structure in any way, it can only be deduced from the file grouping / naming, and this varies between institutions, so this will take some hacking. But instead of doing it for only CESM2-WACCM, perhaps do it for all models.

[kyleredilla]

Yeah okay, data contact with NCAR said those data are not available due to model config errors, so we will need to look at other variants, for sure.

[kyleredilla]

We are bailing on -WACCM version in favor of regular CESM2 given the relative lack of data in ESGF.