Flux conserving, bin average of 2-D field about gcmfaces

[gaelforget]

I do not know of anyone who might have implemented this type of regridding but would definitely be a very useful addition. Are you thinking about schemes that deal with c-grid vector fields (U,V,W) or more common "conservative interpolation" schemes?

[menemenlis]

It is for surface fluxes. Dustin and I are writing code for llc270 to 1-deg-grid bin-averaging. Happy to share when done.

[gaelforget]

Would be great. Here is what I would suggest:

• 1. take a look at sample_processing/example_interp.m which is mentioned in the gcmfaces manual under tutorial examples. interp=gcmfaces_interp_coeffs(lon(:),lat(:)); should work with any of our llc, cs, etc grids, and give you the four nearest neighbors. No need to redo this part -- build on it as much as possible.
• 2. Fork, branch, and put your proposed implementation in the gcmfaces_devel/ folder.
• 3. PR back to this repo once you have an implementation that you want me to review.
• 4. Discuss and iterate on the PR if needed

Just like what we routinely do with MITgcm (see https://mitgcm.readthedocs.io/en/latest/contributing/contributing.html for detailed instructions if needed).

ok?

ps: could you please add, in this thread, the paper reference for the scheme you are looking at with Dustin.