Analyze differences with updates for fractional grid for non-fractional case

[JessicaMeixner-NOAA]

Analyze effects of PR for Fractional landmask: Changing ice fraction definition for ufs-s2s-model when running with non-fractional masks. Either confirm there are no answer changes or if there are answer changes, run multiple 35 day runs with benchmark set-up.

[JessicaMeixner-NOAA]

There are answer changes. One source of difference is the line in atmos_model.F90:

if (IPD_Data(nb)%Coupling%ficein_cpl(ix) > one-epsln) IPD_Data(nb)%Coupling%ficein_cpl(ix)=one

still working to determine other ways the benchmarks will change from NOAA-EMC/fv3atm#81

[JessicaMeixner-NOAA]

The following is copied from an email from @shansun6 where she describes what has to be changed in NOAA-EMC/fv3atm#81 to obtain b4b results with the previous commit for non-fractional grid. All changes will be kept.

(1) atmos_model.F90: comment out this newly added line 1750 (we have discussed and would like to keep this line to avoid tiny open water fraction <epsln, where epsln=1.e-10), i.e.,

if (IPD_Data(nb)%Coupling%ficein_cpl(ix) > one-epsln) IPD_Data(nb)%Coupling%ficein_cpl(ix)=one

(2) GFS_physics_driver.F90: LHS is latest code, RHS is previous code. Pink is chosen in order to get b4b results
reverting are done in 3 sections:
Lines 1523, 1526, 1530 (still don't understand why this is necessary to get b4b results as frac-grid=F. But without it tiny difference is shown on every tile, e.g. diff of mediator_FBAtm_a_restart.tile5.nc on tile5 in "/scratch2/BMC/gsd-fv3-dev/Shan.Sun/rtgen.221502/tmp/cpld_fv3_384_mom6_cice_cold_bmark_rt/t5.nc".)

 Lines 1539, 1542, 1548, 2119 (local variables should be used as fice and Sfcprop%fice are not equal when fice<Model%min_seaice)

Line 2841 (epsln is introduced)

b4b results can be obtained if all pinks are chosen from above 3 sections.

[JessicaMeixner-NOAA]

As the source of differences have been determined (and documented above thanks @shansun6) and the code updates are desired, I'm closing this issue. @junwang-noaa @shansun6 @DeniseWorthen if you think this needs to be re-opened please do so.