Reroute missing fluxes in data ice-shelf melt fluxes (DISMF)

[xylar]

This merge adds the ability to reroute data melt fluxes remapped from the Adusumilli et al. (2020) dataset that were previously removed completely instead to the nearest sub-ice-shelf cell.

This requires first remapping to the base mesh, then rerouting, then culling land and masking land ice in the result.

In order to facilitate making DISMF datasets for existing meshes, the base mesh is also written out as part of files_for_e3sm and placed in share/meshes/mpas/ocean.

Checklist

• [x] User's Guide has been updated
• [x] Documentation has been built locally and changes look as expected
• [x] Document (in a comment titled Testing in this PR) any testing that was used to verify the changes

[xylar]

Testing on a new mesh

I successfully ran the IcoswISC mesh test and init/remap_ice_shelf_melt step. The resulting melt fluxes are now fully accounted for:

Rerouting fluxes from 839 cells
Captured flux (kg/s):         20170179.4
Rerouted flux (kg/s):          7507101.3
Total flux (kg/s):            27677280.8
Total after rerouting (kg/s): 27677280.8

[xylar]

Testing on an existing mesh

In the files_for_e3sm test case for an existing mesh, I tested the ocean_mesh and remap_ice_shelf_melt steps. I needed to recreate the IcoswISC30E3r5 base mesh, since I had deleted it. I have copied this file to share/meshes/mpas/ocean.

The remap_ice_shelf_melt results are nearly the same as above:

Rerouting fluxes from 840 cells
Captured flux (kg/s):         20170108.3
Rerouted flux (kg/s):          7507172.4
Total flux (kg/s):            27677280.8
Total after rerouting (kg/s): 27677280.8

I have copied the resulting DISMF dataset to:

/lcrc/group/e3sm/public_html/inputdata/ocn/mpas-o/IcoswISC30E3r5/prescribed_ismf_adusumilli2020.IcoswISC30E3r5.20240219.nc

[xylar]

@darincomeau and @cbegeman, we should probably start a design document for the next generation of DISMF/DIB forcing. This is my suggested first step in that direction -- making sure the total Adusumilli flux is the same as the total flux on the MPAS mesh.

[xylar]

For my test with a new IcoswISC30 mesh, here are the remapped Adusumilli fluxes on the base mesh: And here are the fluxes on the culled mesh, including rerouting:

[cbegeman]

@xylar From a quick read-through of the code changes and your testing, this looks great. Any other testing you would like me to do?

[xylar]

@cbegeman, thanks! No, I don't think additional testing is needed for now. I will try to figure out the area scaling that will keep the mass the same before and after remapping. I'll ping you again once I've got that.

[xylar]

. I will try to figure out the area scaling that will keep the mass the same before and after remapping. I'll ping you again once I've got that.

I have decided to fix that in #778, rather than here.

[xylar]

@cbegeman and @darincomeau, thank you both for all the reviews!