Xylar Asay-Davis

See https://github.com/E3SM-Ocean-Discussion/E3SM/pull/23 for additional discussion.

Follow-up work based on this PR will add `Time` and `Time_bnds` coordinate variables to MPAS-Ocean, MPAS-Seaice and MALI output for better CF compliance.

@akturner, @cbegeman, and @mark-petersen, you are free to test further if you want but you may also approve based on your review of https://github.com/E3SM-Ocean-Discussion/E3SM/pull/23 @matthewhoffman, I'm mostly wanting to make...

Thank you @matthewhoffman. I really appreciate your review and your helpful comments. > 1. Presumably `name_in_output` only functions as expected for pure output variables. If you tried to read in...

Thank you @mark-petersen, @cbegeman, @akturner and @matthewhoffman for taking the time to review this. @jonbob, I believe this is ready for you to integrate when it's a good time.

Thanks for posting this, @darincomeau. It remains a mystery why this is a problem in standalone and hasn't caused trouble in E3SM.

> One question for you @xylar -- for the velocity reconstruction arrays (MLEvelocityX,Y,Z) I wonder if a unique set of these is needed for normalVelocity, normalGMBolusVelocity, and normalMLEvelocity. Can I...

@mark-petersen, I agree with @vanroekel that the right fix is to make sure arrays related to `gm` are not being used when GM is off, rather than taking the arrays...

@jonbob, I will try to re-test this tomorrow. Do you want to rebase to get rid of the conflict in `config_pesall.xml`?

@milenaveneziani, is this something you have done already in a branch somewhere? I vaguely remember discussing this at some point.