Changes between Version 20 and Version 21 of ticket/1613_NOC3_vvl_by_default


Ignore:
Timestamp:
2015-11-01T11:03:26+01:00 (5 years ago)
Author:
gm
Comment:

Legend:

Unmodified
Added
Removed
Modified
  • ticket/1613_NOC3_vvl_by_default

    v20 v21  
    2525 
    2626== I. Before/Now/After vertical coordinate system by default == 
    27   ===>>> runs ORCA2_LIM3 trunk and this branch  identical to the last digit  ===>>> '''         OK'''          see revision   [https://forge.ipsl.jussieu.fr/nemo/log/branches/2015/dev_r5836_NOC3_vvl_by_default?rev= 5845] 
     27  ===>>> runs ORCA2_LIM3 trunk and this branch  identical to the last digit  ===>>> '''          OK'''           see revision   [https://forge.ipsl.jussieu.fr/nemo/log/branches/2015/dev_r5836_NOC3_vvl_by_default?rev= 5845] 
    2828 
    29 '''   [[BR]]   Nevertheless'''          for both the trunk & this branch, ORCA2_LIM3 blow up atter 138 time step....[[BR]]        NB: to obtain the same results 1 things : revert domhpg.F90 optimization (use of r1_e...) [[BR]]        =========>>>>>>>>>>>>  try to find the origine of the model explosion:[[BR]]        (1) dynvor.F90 reverted to the v3.6_stable  version, i.e. with use of fmask at the proper place, the trunk still explode round the same time-step...  So the problem does not come from dynvor.[[BR]]        (2) iso neutral diffusion switch off (as well as div): it is not the issue the model still blow up (pdt 139)[[BR]]        (3) partial cells switch off (ln_zco=T): it is not the issue the model still blow up (pdt 142)[[BR]]      (4) tra_adv : use 2nd order centered scheme: blow up pdt 134[[BR]]      (5) just set hat as a constant= 2000  (instead of grid depend coeff  nn_aht_ijk = 20) : blow up pdt 134[[BR]]      (6) dynldf off (ln_dynldf_lap=F) & use of UBS (ln_dynadv_vec= F ; ln_dynad_ubs=T): '''      it is working !!!! '''(I stop the run after 3531)[[BR]]The error is thus probably in dynldf_lap_blp.F90. Note that if the code does not blow up it is very noisy at mid latitudes with such a setting... (probably logic with UBS on such a low resolution...).  
     29'''   [[BR]]   Nevertheless'''          for both the trunk & this branch, ORCA2_LIM3 blow up atter 138 time step....[[BR]]        NB: to obtain the same results 1 things : revert domhpg.F90 optimization (use of r1_e...) [[BR]]        =========>>>>>>>>>>>>  try to find the origine of the model explosion:[[BR]]        (1) dynvor.F90 reverted to the v3.6_stable  version, i.e. with use of fmask at the proper place, the trunk still explode round the same time-step...  So the problem does not come from dynvor.[[BR]]        (2) iso neutral diffusion switch off (as well as div): it is not the issue the model still blow up (pdt 139)[[BR]]        (3) partial cells switch off (ln_zco=T): it is not the issue the model still blow up (pdt 142)[[BR]]      (4) tra_adv : use 2nd order centered scheme: blow up pdt 134[[BR]]      (5) just set hat as a constant= 2000  (instead of grid depend coeff  nn_aht_ijk = 20) : blow up pdt 134[[BR]]      (6) dynldf off (ln_dynldf_lap=F) & use of UBS (ln_dynadv_vec= F ; ln_dynad_ubs=T): '''      it is working !!!! '''(I stop the run after 3531)[[BR]]The error is thus probably in dynldf_lap_blp.F90. Note that if the code does not blow up it is very noisy at mid latitudes with such a setting... (probably logic with UBS on such a low resolution...). 
    3030 
    3131 * dom_oce.F90 :[[BR]](1) systematic declaration of before/now/after vertical coordinate [[BR]](2) uniformisation of the name: hur (hvr) becomes r1_hu_n (r1_hv_n) ; idem for before and after[[BR]](3) gdep3w becomes gde3w === 
     
    3535 
    3636== [[BR]]II. Non-linear free surface by default == 
    37  * introduce an namelist parameter (ln_linssh) for linear free surface 
    38  * remove key_vvl and replace the "IF( lk_vvl ) THEN...ELSE...ENDIF" by "IF( ln_linssh ) THEN...ELSE...ENDIF" 
     37 * replace lk_vvl by .NOT.ln_linssh everywhere  
     38 * remove key_vvl and introduce ln_linssh in nam_zgr namelist 
    3939 
    4040----