Sync with UFS WM version on 3/3

[uturuncoglu]

This PR aims to sync UFS Coastal with the authorative repository. Testing and fixing issues are still ongoing.

[mansurjisan]

Everything looks good to me. Test passed both on WCOSS2 and Hercules.

[SmithJos13]

@uturuncoglu Okay I rechecked out the model, and it seems to be working now. I'll let you know when I have checked the output!

[uturuncoglu]

@SmithJos13 Okay. Perfect!

[uturuncoglu]

@mansurjisan Thanks for the update. If you are fine with the results, please go ahead and review the PR and approve. So, I could merge it once I got confirmation from @SmithJos13 too. Thanks all of your help.

[mansurjisan]

Thanks, @uturuncoglu . I already reveiwed and approved the PR this morning. Let me know if you need any help.

[SmithJos13]

@uturuncoglu my first run doesn't quite look right I'm going to do some experimenting to see if I can find the issue
Left is the run with the sync branch right is my confirmed working run

[uturuncoglu]

@SmithJos13 Okay. Let me know if you find something. I'll hold this PR until Monday and merge it. If you find something, we could fix it before the merge. Otherwise, we could still put fixes after the merge. Let me know what you think?

[SmithJos13]

Okay the results are looking better, but I needed to comment out the preprocessor flags USE_CICE. I think there might be something wrong with the way the USE_CICE flag is being passed to the SCHISM and SCHISM-ESMF component. Left being sync branch with the preprocessor flags being commented out and the right being my working configuration.

[SmithJos13]

https://github.com/oceanmodeling/schism/blob/99627e7a6b32df979acf0a2f23ed58d0f628fbb9/src/Hydro/schism_init.F90#L1701

      allocate(aice(npa),tau_oi(2,npa),fresh_wa_flux(npa),salinity_flux(npa),net_heat_flux(npa), &
               srad_o(npa),srad_th_ice(npa),stat=istat)
      if(istat/=0) call parallel_abort('INIT: ice frac allocation failure')
      aice(:)          = real(0)
      tau_oi(:,:)      = real(0)
      fresh_wa_flux(:) = real(0)
      salinity_flux(:) = real(0)
      net_heat_flux(:) = real(0)
      srad_o(:)        = real(0)
      srad_th_ice(:)   = real(0)


       allocate(sflux_o(npa),stat=istat)
       sflux_o(:)     = real(0)

This is repeated code that apears within the macros just needs to be removed.

[uturuncoglu]

@SmithJos13 Okay. I also updated CDEPS DATM ERA5 mode for shortwave bands. So, we have SCHISM CPP option and CDEPS fix in the sync branch. Let me know if you ghee any issue. In the build you need to set USE_CICE=ON to activate -DUSE_ICE CPP option. By default, it is OFF. Anyway, just test the new version and let me know if you see anything wrong.

[uturuncoglu]

@SmithJos13 do you also want me to remove

       allocate(sflux_o(npa),stat=istat)
       sflux_o(:)     = real(0)

part?

[josephzhang8]

https://github.com/oceanmodeling/schism/blob/99627e7a6b32df979acf0a2f23ed58d0f628fbb9/src/Hydro/schism_init.F90#L1701

      allocate(aice(npa),tau_oi(2,npa),fresh_wa_flux(npa),salinity_flux(npa),net_heat_flux(npa), &
               srad_o(npa),srad_th_ice(npa),stat=istat)
      if(istat/=0) call parallel_abort('INIT: ice frac allocation failure')
      aice(:)          = real(0)
      tau_oi(:,:)      = real(0)
      fresh_wa_flux(:) = real(0)
      salinity_flux(:) = real(0)
      net_heat_flux(:) = real(0)
      srad_o(:)        = real(0)
      srad_th_ice(:)   = real(0)


       allocate(sflux_o(npa),stat=istat)
       sflux_o(:)     = real(0)

This is repeated code that apears within the macros just needs to be removed.

I need to do on SCHISM repo but prefer Ufuk to submit a PR once all is done. Thx!

[uturuncoglu]

@josephzhang8 let us push to our fork first and then I'll open an PR in the automative repository to bring all remaining changes.

[SmithJos13]

@uturuncoglu Yes you can remove that part

[uturuncoglu]

@SmithJos13 Okay. That is also in the sync branch. So, you have everything that you need in the sync branch. Please update your code and build with USE_CICE=ON. I hope you will get the correct results now. Thanks again all of your hard work to bring this capability.

[SmithJos13]

Of course!

looks like there might be some issues with new CDEPS functionality during the build. Probably just need to declare a variable.

/work2/noaa/vdatum/jsmith/Coastal_sync_check/ufs-weather-model/CDEPS-interface/CDEPS/datm/datm_datamode_era5_mod.F90(386): error #6404: This name does not have a type, and
 must have an explicit type.   [STRM_FAXA_SWDN]
    if (associated(strm_Faxa_swdn)) then

[uturuncoglu]

@SmithJos13 Okay. Fixed. Now it compiles. If you could check the mediator history file and make sure about fields produced by the CDEPS, that would be great.

[SmithJos13]

Will do! I'll report back!

[SmithJos13]

Good news!!! It worked! The results look consistent!

also looks like all the fields are in the mediator history files:

 double atmImp_Faxa_swdn(time, atmImp_ny, atmImp_nx) ;
                atmImp_Faxa_swdn:_FillValue = 1.e+30 ;
                atmImp_Faxa_swdn:units = "W m-2" ;
                atmImp_Faxa_swdn:standard_name = "atmImp_Faxa_swdn" ;
        double atmImp_Faxa_swndf(time, atmImp_ny, atmImp_nx) ;
                atmImp_Faxa_swndf:_FillValue = 1.e+30 ;
                atmImp_Faxa_swndf:units = "W m-2" ;
                atmImp_Faxa_swndf:standard_name = "atmImp_Faxa_swndf" ;
        double atmImp_Faxa_swndr(time, atmImp_ny, atmImp_nx) ;
                atmImp_Faxa_swndr:_FillValue = 1.e+30 ;
                atmImp_Faxa_swndr:units = "W m-2" ;
                atmImp_Faxa_swndr:standard_name = "atmImp_Faxa_swndr" ;
        double atmImp_Faxa_swnet(time, atmImp_ny, atmImp_nx) ;
                atmImp_Faxa_swnet:_FillValue = 1.e+30 ;
                atmImp_Faxa_swnet:units = "W m-2" ;
                atmImp_Faxa_swnet:standard_name = "atmImp_Faxa_swnet" ;
        double atmImp_Faxa_swvdf(time, atmImp_ny, atmImp_nx) ;
                atmImp_Faxa_swvdf:_FillValue = 1.e+30 ;
                atmImp_Faxa_swvdf:units = "W m-2" ;
                atmImp_Faxa_swvdf:standard_name = "atmImp_Faxa_swvdf" ;
        double atmImp_Faxa_swvdr(time, atmImp_ny, atmImp_nx) ;
                atmImp_Faxa_swvdr:_FillValue = 1.e+30 ;
                atmImp_Faxa_swvdr:units = "W m-2" ;
                atmImp_Faxa_swvdr:standard_name = "atmImp_Faxa_swvdr" ;

Update checked the values of the shortwave bands against the downwelling component from the mediator history files in MATLAB and looks like the agreement it is good!!

max(swvdf+swvdr+swndr+swndf-swnet,[],'all')

ans =

     1.136868377216160e-13

I think we are ready to merge! if there is anything else you want me to check first let me know.

[uturuncoglu]

@SmithJos13 That is great news. I'll create baselines with this version and then merge it. Thanks for all of your help. Please also share the latest version of your run directories that I could use for RTs.

[SmithJos13]

Here is the run directory for CICE-SCHISM /work2/noaa/vdatum/jsmith/Demos/Run_Dir_DATM_SCH_CICE and for standalone run /work2/noaa/vdatum/jsmith/Coastal_PR/Run_Dir/CICE_Standalone Let me know if you have any issues.