Descriptions of MFX and CX appear to be incorrect

[yuanjianz]

MFX and CX (mass flux and courant number) IMPORT are listed as "eastward/northward" and implies on zonal/meridional direction.

FVdycoreCubed_GridComp/AdvCore_GridCompMod.F90

Line 150 in 329ab6d

LONG_NAME = 'pressure_weighted_eastward_mass_flux', &

However, they are directly passed to offline_tracer_advection, where fluxes on cubed-sphere X-Y direction are expected.

Reading dyn_core where MFX and CX are produced, they appear to be on original cubed-sphere X-Y direction without basis vector change. I am wondering if the descriptions of MFX/Y, CX/Y in AdvCore_GridComp and DynCore_GridComp are incorrect in terms of basis vector? Tagging @tclune @wmputman for confirmation.

P.S. I notice the incorrect-unit issue in #174 are not fixed for AdvCore_GridComp

FVdycoreCubed_GridComp/AdvCore_GridCompMod.F90

Line 151 in 329ab6d

UNITS = 'Pa m+2 s-1', &

[mathomp4]

Since you mention AdvCore, I'm also going to tag in @mmanyin as he knows about that as well...