[Feature]: CDAT Migration: Add support for opening multiple time series datasets

[tomvothecoder]

Is your feature request related to a problem?

The CDAT codebase supports XML files to open up multiple time series datasets. The cdat-migration-fy24 branch uses xc.open_dataset() for single variables that don't need to be derived.

• Zppy specifies input paths as directories of .nc files (related comment)
• There might be some cases where variables are split up across multiple files and need to be concatenated as a single dataset object.

Describe the solution you'd like

Update _get_time_series_dataset_obj() to use xc.open_mfdataset()

https://github.com/E3SM-Project/e3sm_diags/blob/22cd0f569e7df4be147a8fe6fae3e00d5aeb9d8c/e3sm_diags/driver/utils/dataset_xr.py#L1035-L1037

Describe alternatives you've considered

No response

Additional context

No response

[forsyth2]

Zppy specifies input paths as directories of .nc files

Just to elaborate a bit:

• The core of https://github.com/E3SM-Project/zppy/pull/598 is the bash file change
• That takes zppy from A) concatenating variables into a txt file and then using cdscan to create an xml file to B) taking an entire directory of nc files. I.e., "process all the variables in this directory" rather than "process all the variables I listed in this file".
• This is incompatible with how Diags (on main) currently functions, which is why https://github.com/E3SM-Project/zppy/pull/598 can't be merged into zppy's main until Diags has completed its CDAT migration.