Review and updates to NCI dask examples

[hot007]

I have reviewed the dask training materials.

• Many edits are minor language corrections, but some issues are noted in commit comments, mostly around data availability.
• I'd recommend including as many input datasets as possible in the repo so that it can be run independently of being a member of specific NCI projects. If this is done, it'll be important to include licensing information for each dataset stored in demo_data.
• It would also be good to rerun some of the notebooks to clear spammy output where clients have been restarted, e.g., and in some cases python warnings should be addressed.
• Where plots are made, inclusion of coastlines would help readability, and it's important to ensure maps are correctly oriented.
• Added a license file, the licence must be CC-BY-SA as it is inherited from the dask materials.
• I'm hoping many of the outputs creating extraneous .png, .html etc files are dynamically produced and removing/ignoring them should just keep the repo clean without affecting outputs for anyone running the code.

I was not able to test all notebooks, only those which did not rely on data other than CMIP in fs38/oi10.

Please consider and accept/reject these suggestions via git so that my contribution is recorded if possible.

Thank you, useful materials!

Claire

[hot007]

One more update to dask_02, forgot to change dataset name.