best-practice example workflow for loading ensemble ACCESS-ESM1.5 data

[Thomas-Moore-Creative]

Title:

write best-practice example workflow for loading ensemble ACCESS-ESM1.5 data

Description:

40 ensemble ACCESS-ESM1.5 data is available on NCI and is cataloged, see: https://access-nri-intake-catalog.readthedocs.io/en/latest/

Goals:

• Write a Jupyter notebook example to show and explore best-practice
• importantly, maintain the original names for the 40 ensembles in the resulting xarray object, for example r1i1p1f1

Environment:

• Pangeo-style: python, dask, xarray
• ARE @ NCI
• Module directories: /g/data/hh5/public/modules
• Modules: conda/analysis3

Additional context:

• adding implementation of a json file for providing settings like xarray_kwargs
• set this up as a start to full ARD workflow for ACCESS-ESM1.5 data

[Thomas-Moore-Creative]

The motivation here is (A) current project work that requires ACCESS-ESM1.5 and (B) efforts here by @jemmajeffree > https://github.com/COSIMA/cosima-recipes/issues/444

[Thomas-Moore-Creative]

able to extract the member names from the catalog keys and naively assign them after concatenation - but are they correct? CHECK THIS