Remove ecmwf_macc functions

[AdamRJensen]

• Closes ECMWF MACC will be retired #1691
• I am familiar with the contributing guidelines
• [ ] Tests added
• Updates entries in docs/sphinx/source/reference for API changes.
• Adds description and name entries in the appropriate "what's new" file in docs/sphinx/source/whatsnew for all changes. Includes link to the GitHub Issue with :issue:`num` or this Pull Request with :pull:`num`. Includes contributor name and/or GitHub username (link with :ghuser:`user`).
• [ ] New code is fully documented. Includes numpydoc compliant docstrings, examples, and comments where necessary.
• Pull request is nearly complete and ready for detailed review.
• Maintainer: Appropriate GitHub Labels (including remote-data) and Milestone are assigned to the Pull Request and linked Issue.

This PR removes the get_ecmwf_macc and read_ecmwf_macc functions. The reasoning is that the MACC dataset is now outdated (covers the period 2003-2012) and the functions are hardly ever used thus it is not worth maintaining these functions any longer.

For more recent reanalysis data, users are encouraged to check out ECMWF ERA5, CAMS, or NASA MERRA-2 as a starting point.

See this link for a comparison between MACC and CAMS.

[AdamRJensen]

@kanderso-nrel I guess you will insist we deprecate these functions instead of bluntly removing them? 😆

[cwhanse]

@kanderso-nrel I guess you will insist we deprecate these functions instead of bluntly removing them? 😆

if @kanderso-nrel doesn't, I will.

[kandersolar]

insist we deprecate these functions instead of bluntly removing them?

Perhaps not insist, but strongly endorse :) It costs little to maintain so I see little reason to not do a typical deprecation cycle.

[mikofski]

I should stress that:

the MACC dataset is now outdated (covers the period 2003-2012)

is not IMHO a valid reason for deprecating this function. EG: one could say that TMY2 is outdated, but I predict we will never deprecate read_TMY2.

However, this reason:

and the functions are hardly ever used thus it is not worth maintaining these functions any longer.

is IMHO a good reason to do away with something that costs us to maintain, but does not pay any dividends if no one is using it

[adriesse]

does not pay any dividends

Let me check my bank statement whether this is really true...

[mikofski]

Should we consider archiving the MACC data set somehow somewhere?

[AdamRJensen]

I think it's fair to merge this PR soon as the dataset is removed on June 1st and our next release will surely be after that.

[kandersolar]

LGTM, pending the actual dataset removal on the ECMWF side.

[kandersolar]

@AdamRJensen points out that the MACC dataset decommissioning is complete, so I think the time has come to merge this PR.