Small numerical error in mass flux regridding

[sdeastham]

Name and Institution (Required)

Name: Seb Eastham
Institution: Imperial College London

Confirm you have reviewed the following documentation

• Support guidelines
• User manuals
• Debugging GEOS-Chem and HEMCO errors
• Current and past Github issues

Description of your issue or question

There is a small bug in the algorithm used to handle regridding of mass fluxes between resolutions, where an edge length multiplication is missing (see GEOS-ESM/MAPL#1199). This error is resolved by some work done by Tom Clune in GEOS-ESM/MAPL#2056 (partially described in GEOS-ESM/MAPL#1202).

[lizziel]

I implemented the changes in GEOS-ESM/MAPL#2056 and compared before and after using GEOS-FP mass fluxes. The result is no differences in diagnostics and restart files.

[lizziel]

I revisited this after discussion with Tom Clune at GMAO. I pared down the updates in GEOS-ESM/MAPL#2056 to only what we need and tried again. I now see differences, although I suspect the issue before what error on my part in putting in the update.

Here is an example of differences in daily mean after a 1-day run.

This update can go into 14.4.1 and I will create a PR.

[lizziel]

See MAPL PR geoschem/MAPL#36 for this update.

[lizziel]

Now merged.