local modification of the Gib strait, Danish strait and other no more in DOMAINcfg ?

[mathiot]

BE CAREFUL !!! Due to dynamic behaviour of this ticket creation page, it is highly recommend to set first all other fields before writing the ticket description below.
If you have lost your draft after an unwanted reload, you can click on the link 'Restore Form' in the contextual menu upper right to recover it.
Remove these lines after reading.

Context

ORCA2 e1 and e2 used for SETTE is locally different to the one used in 3.6. At the Met Office we spot this also in our test with ORCA1 (large change in Med Sea and Black Sea).

Analysis

All the local modifications of e1 and e2 at some specific straits present in version 3.6 in domhgr.F90 are not present in the DOMAINcfg source code. It affects ORCA2, ORCA1 and ORCA05.

Here is the king of block I am talking about.

         !                                                ! =====================
         IF( cp_cfg == "orca" .AND. jp_cfg == 2 ) THEN    ! ORCA R2 configuration
            !                                             ! =====================
            IF( nn_cla == 0 ) THEN
               !
               ii0 = 139   ;   ii1 = 140        ! Gibraltar Strait (e2u = 20 km)
               ij0 = 102   ;   ij1 = 102   ;   e2u( mi0(ii0):mi1(ii1) , mj0(ij0):mj1(ij1) ) =  20.e3
               IF(lwp) WRITE(numout,*)
               IF(lwp) WRITE(numout,*) '             orca_r2: Gibraltar    : e2u reduced to 20 km'
               !
               ii0 = 160   ;   ii1 = 160        ! Bab el Mandeb (e2u = 18 km)
               ij0 =  88   ;   ij1 =  88   ;   e1v( mi0(ii0):mi1(ii1) , mj0(ij0):mj1(ij1) ) =  18.e3
                                               e2u( mi0(ii0):mi1(ii1) , mj0(ij0):mj1(ij1) ) =  30.e3
               IF(lwp) WRITE(numout,*)
               IF(lwp) WRITE(numout,*) '             orca_r2: Bab el Mandeb: e2u reduced to 30 km'
               IF(lwp) WRITE(numout,*) '                                     e1v reduced to 18 km'
            ENDIF

            ii0 = 145   ;   ii1 = 146        ! Danish Straits (e2u = 10 km)
            ij0 = 116   ;   ij1 = 116   ;   e2u( mi0(ii0):mi1(ii1) , mj0(ij0):mj1(ij1) ) =  10.e3
            IF(lwp) WRITE(numout,*)
            IF(lwp) WRITE(numout,*) '             orca_r2: Danish Straits : e2u reduced to 10 km'
            !
         ENDIF

Furthermore the README is a bit missleading on this:

2) copy in DOMAINcfg directory same input files (of related configuration) required in v3.6_stable.

DOMAINcfg package is EXACTLY what does exist in NEMO version 3.6 to define a model domain (both domain related namelist and initialization).

Recommendation

• If it is a deliberate action to remove this code from DOMAINcfg, I suggest to update the README and maybe update the ORCA2 input file distributed with the corrected e1 and e2 for Gibraltar, Bab el Mandeb and Danish straits.
• If it is NOT a deliberate action to remove this code from DOMAINcfg, I suggest to put it in DOMAINcdf.

[mathiot]

Additional comments:
It appears that in addition to the local horizontal scale factor modification, the local changes of the slip condition are also missing:

• Slip condition in some orca configuration is modify locally in dommsk.F90 in NEMO_3.6
• This part of the code (local modifications) has been move to DOMAINcfg, however fmask is still computed in NEMO and not read for the DOMAINcfg (in the latest version of DOAMINcfg, fmask is even not saved in the domcfg.nc file). So it means the local change of slip condition (fmask value along the coast) applied in 3.6 are no more used in 4.0

Suggestion:

• add a shlat2d input file to NEMO. It will allow user to play more easily with this.

If consensus on a solution, I am happy to implement it as part of ENHANCE-03_PierreMathiot_DomcfgTools.