New URL for NEMO forge!   http://forge.nemo-ocean.eu

Since March 2022 along with NEMO 4.2 release, the code development moved to a self-hosted GitLab.
This present forge is now archived and remained online for history.
2020WP/ENHANCE-10_acc_fix_traqsr – NEMO
wiki:2020WP/ENHANCE-10_acc_fix_traqsr

Version 10 (modified by acc, 5 years ago) (diff)

--

Name and subject of the action

Last edition: Wikinfo(changed_ts)? by Wikinfo(changed_by)?

The PI is responsible to closely follow the progress of the action, and especially to contact NEMO project manager if the delay on preview (or review) are longer than the 2 weeks expected.

  1. Summary
  2. Option 2 revisited
  3. Preview
  4. Tests
  5. Review

Summary

Action ENHANCE-10_acc_fix_traqsr
PI(S) acc
Digest Reduce use of 3D allocatable arrays in RGB light penetration schemes
Dependencies If any
Branch source:/NEMO/branches/{YEAR}/dev_r{REV}_{ACTION_NAME}
Previewer(s) Names
Reviewer(s) Names
Ticket #XXXX

Description

The current implementation of RGB light penetration in traqsr (either varying or constant chlorophyll) uses 6, domain-sized 3D, temporary arrays which can be reduced to a few 2D arrays. The impact of the current implementation is most evident at lower processor counts where the impact of the extra 3D arrays can cause cache-misses and memory band-width issues. In an extreme case traqsr can switch from consuming 2% of run-time to 68% (comparing ORCA025 running on 200 cores vs 48 cores).

A simple redesign of the algorithm should remove this behaviour.

Implementation

The current code is structured thus:

      CASE( np_RGB , np_RGBc )         !==  R-G-B fluxes  ==!
         !
         ALLOCATE( zekb(jpi,jpj)     , zekg(jpi,jpj)     , zekr  (jpi,jpj)     , &
            &      ze0 (jpi,jpj,jpk) , ze1 (jpi,jpj,jpk) , ze2   (jpi,jpj,jpk) , &
            &      ze3 (jpi,jpj,jpk) , zea (jpi,jpj,jpk) , zchl3d(jpi,jpj,jpk)   )
         !
         ! code to set zchl3d(:,:,1:nskr+1)
         ! (the chlorophyll value projected from the surface values)
         !
         zcoef  = ( 1. - rn_abs ) / 3._wp    !* surface equi-partition in R-G-B
         DO_2D_00_00
            ze0(ji,jj,1) = rn_abs * qsr(ji,jj)
            ze1(ji,jj,1) = zcoef  * qsr(ji,jj)
            ze2(ji,jj,1) = zcoef  * qsr(ji,jj)
            ze3(ji,jj,1) = zcoef  * qsr(ji,jj)
            zea(ji,jj,1) =          qsr(ji,jj)
         END_2D
         !
         DO jk = 2, nksr+1                   !* interior equi-partition in R-G-B depending of vertical profile of Chl
            DO_2D_00_00
               zchl = MIN( 10. , MAX( 0.03, zchl3d(ji,jj,jk) ) )
               irgb = NINT( 41 + 20.*LOG10(zchl) + 1.e-15 )
               zekb(ji,jj) = rkrgb(1,irgb)
               zekg(ji,jj) = rkrgb(2,irgb)
               zekr(ji,jj) = rkrgb(3,irgb)
            END_2D

            DO_2D_00_00
               zc0 = ze0(ji,jj,jk-1) * EXP( - e3t(ji,jj,jk-1,Kmm) * xsi0r       )
               zc1 = ze1(ji,jj,jk-1) * EXP( - e3t(ji,jj,jk-1,Kmm) * zekb(ji,jj) )
               zc2 = ze2(ji,jj,jk-1) * EXP( - e3t(ji,jj,jk-1,Kmm) * zekg(ji,jj) )
               zc3 = ze3(ji,jj,jk-1) * EXP( - e3t(ji,jj,jk-1,Kmm) * zekr(ji,jj) )
               ze0(ji,jj,jk) = zc0
               ze1(ji,jj,jk) = zc1
               ze2(ji,jj,jk) = zc2
               ze3(ji,jj,jk) = zc3
               zea(ji,jj,jk) = ( zc0 + zc1 + zc2 + zc3 ) * wmask(ji,jj,jk)
            END_2D
         END DO
         !
         DO_3D_00_00( 1, nksr )
            qsr_hc(ji,jj,jk) = r1_rho0_rcp * ( zea(ji,jj,jk) - zea(ji,jj,jk+1) )
         END_3D
         !
         DEALLOCATE( zekb , zekg , zekr , ze0 , ze1 , ze2 , ze3 , zea , zchl3d )
         !

Where most of the temporary, full-depth arrays are not necessary because only two vertical levels are required at any one time. In fact even the zea array is unnecessary since the zchl3d array could be repurposed once its value has been used.

Option 1: Minmum memory usage

By rearranging the loop order and placing the vertical loop innermost then the code can be greatly simplified to an equivalent using minimal temporary storage. A couple of other changes seem sensible too, namely:

  • rename the zchl3d array to ztmp3d (since it is now used for two purposes)
  • only allocate ztmp3d to nksr+1; values below this are not used and nksr + 1 is likely << jpk
  • calculate and store the attenuation coefficient look-up table index as soon as the sub-surface chlorophyll value is known. This keeps all LOG operations in one loop and, in the case of constant chlorophyll, removes the LOG from the loop altogether.
      CASE( np_RGB , np_RGBc )         !==  R-G-B fluxes  ==!
         !
         ALLOCATE( ztmp3d(jpi,jpj,nksr + 1)   )
         !
         ! code to set ztmp3d(:,:,1:nskr+1)

         ! including the following changes after
         IF( nqsr == np_RGBc ) THEN          !*  Variable Chlorophyll
            .
            DO_3D_00_00( 1, nksr + 1 )
            .
            . no change until after
               zCze    = 1.12  * (zchl)**0.803
               !
               ! NB. make sure zchl value is such that: zchl = MIN( 10. , MAX( 0.03, zchl ) )
               zchl = MIN( 10. , MAX( 0.03, zCze * ( zCb + zCmax * EXP( -( (zpsi - zpsimax) / zdelpsi )**2 ) ) ) )
               ! Convert chlorophyll value to attenuation coefficient look-up table index
               ztmp3d(ji,jj,jk) = 41 + 20.*LOG10(zchl) + 1.e-15
            END_3D
         ELSE                                !* constant chrlorophyll
           zchl = 0.05
           ! NB. make sure constant value is such that:
           zchl = MIN( 10. , MAX( 0.03, zchl ) )
           ! Convert chlorophyll value to attenuation coefficient look-up table index
           zlui = 41 + 20.*LOG10(zchl) + 1.e-15
           DO jk = 1, nksr + 1
              ztmp3d(:,:,jk) = zlui
           END DO
         ENDIF
         ! 
         !
         zcoef  = ( 1. - rn_abs ) / 3._wp    !* surface equi-partition in R-G-B
         ! store the surface SW radiation; re-purpose the surface ztmp3d array
         ! since the surface attenuation coefficient is not used
         ztmp3d(:,:,1) = qsr(:,:)
         !
         !* interior equi-partition in R-G-B depending of vertical profile of Chl
         DO_2D_00_00
            zc0 = rn_abs * qsr(ji,jj)
            zc1 = zcoef  * qsr(ji,jj)
            zc2 = zc1
            zc3 = zc1
            zc4 = e3t(ji,jj,1,Kmm)
            DO jk = 2, nksr+1
               irgb = NINT( ztmp3d(ji,jj,jk) )
               zc0 = zc0 * EXP( - zc4 * xsi0r       )
               zc1 = zc1 * EXP( - zc4 * rkrgb(1,irgb) )
               zc2 = zc2 * EXP( - zc4 * rkrgb(2,irgb) )
               zc3 = zc3 * EXP( - zc4 * rkrgb(3,irgb) )
               zc4 = e3t(ji,jj,jk,Kmm)
               ! store the SW radiation penetrating to this location
               ! re-purpose the ztmp3d array since the attenuation coefficient
               ! at this point will not be needed again
               ztmp3d(ji,jj,jk) = ( zc0 + zc1 + zc2 + zc3 ) * wmask(ji,jj,jk)
            END DO
         END_2D
         !
         DO_3D_00_00( 1, nksr )
            qsr_hc(ji,jj,jk) = r1_rho0_rcp * ( ztmp3d(ji,jj,jk) - ztmp3d(ji,jj,jk+1) )
         END_3D
         !
         DEALLOCATE( ztmp3d )

This is code and memory efficient but may perform poorly due to non-contiguous access to the array elements (see performance section below).

Option 2: Low memory use (retain loop order).

A compromise solution, which reduces memory use and should perform better is to remove all unnecessary full-depth arrays but maintain loop order by keeping a few 2D arrays. The same additional changes listed above are also made.

       CASE( np_RGB , np_RGBc )         !==  R-G-B fluxes  ==!
         !
         ALLOCATE( ze0 (jpi,jpj)           , ze1 (jpi,jpj) ,   &
            &      ze2 (jpi,jpj)           , ze3 (jpi,jpj) ,   &
            &      ztmp3d(jpi,jpj,nksr + 1)                     )
         !
         ! code to set ztmp3d(:,:,1:nskr+1)
         ! including the following changes after
         IF( nqsr == np_RGBc ) THEN          !*  Variable Chlorophyll
            .
            DO_3D_00_00( 1, nksr + 1 )
            .
            . no change until after
               zCze    = 1.12  * (zchl)**0.803
               !
               ! NB. make sure zchl value is such that: zchl = MIN( 10. , MAX( 0.03, zchl ) )
               zchl = MIN( 10. , MAX( 0.03, zCze * ( zCb + zCmax * EXP( -( (zpsi - zpsimax) / zdelpsi )**2 ) ) ) )
               ! Convert chlorophyll value to attenuation coefficient look-up table index
               ztmp3d(ji,jj,jk) = 41 + 20.*LOG10(zchl) + 1.e-15
            END_3D
         ELSE                                !* constant chlorophyll
            zchl = 0.05
            ! NB. make sure constant value is such that:
            zchl = MIN( 10. , MAX( 0.03, zchl ) )
            ! Convert chlorophyll value to attenuation coefficient look-up table index
            zlui = 41 + 20.*LOG10(zchl) + 1.e-15
            DO jk = 1, nksr + 1
               ztmp3d(:,:,jk) = zlui
            END DO
         ENDIF
         !
         zcoef  = ( 1. - rn_abs ) / 3._wp    !* surface equi-partition in R-G-B
         DO_2D_00_00
            ze0(ji,jj) = rn_abs * qsr(ji,jj)
            ze1(ji,jj) = zcoef  * qsr(ji,jj)
            ze2(ji,jj) = zcoef  * qsr(ji,jj)
            ze3(ji,jj) = zcoef  * qsr(ji,jj)
            ! store the surface SW radiation; re-use the surface ztmp3d array
            ! since the surface attenuation coefficient is not used
            ztmp3d(ji,jj,1) =       qsr(ji,jj)
         END_2D
         !
          !* interior equi-partition in R-G-B depending of vertical profile of Chl
         DO_3D_00_00( 2, nksr+1 )
            irgb = NINT( ztmp3d(ji,jj,jk) )
            ze3t = e3t(ji,jj,jk-1,Kmm)
            ze0(ji,jj) = ze0(ji,jj) * EXP( - ze3t * xsi0r )
            ze1(ji,jj) = ze1(ji,jj) * EXP( - ze3t * rkrgb(1,irgb) )
            ze2(ji,jj) = ze2(ji,jj) * EXP( - ze3t * rkrgb(2,irgb) )
            ze3(ji,jj) = ze3(ji,jj) * EXP( - ze3t * rkrgb(3,irgb) )
            ! store the SW radiation penetrating to this location
            ! re-use the ztmp3d array since the attenuation coefficient
            ! at this point will not be needed again
            ztmp3d(ji,jj,jk) = ( ze0(ji,jj) + ze1(ji,jj) + ze2(ji,jj) + ze3(ji,jj) ) * wmask(ji,jj,jk)
         END_3D
         !
         DO_3D_00_00( 1, nksr )
            qsr_hc(ji,jj,jk) = r1_rho0_rcp * ( ztmp3d(ji,jj,jk) - ztmp3d(ji,jj,jk+1) )
         END_3D
         !
         DEALLOCATE( ze0 , ze1 , ze2 , ze3 , ztmp3d )

Performance and evaluation

Both these options produce identical results to the original code (based on an ORCA2_ICE_PISCES test using SETTE (which includes variable surface chlorophyll inputs). ln_timing was activated and the CPU time (averaged across all processors) spent in tra_qsr used as a simple measure of performance. Unfortunately, variations in runtime between successive tests (even with the same code) on the NOC cluster were almost as great as any difference arising from algorithmic differences. Each test was repeated 6 times with the following results:

code option CPU seconds spent in tra_qsr Average
original code 0.34 0.34 0.35 0.35 0.34 0.34 0.3433
minimum memory option 0.36 0.36 0.37 0.36 0.36 0.37 0.3633
low memory option 0.35 0.35 0.35 0.36 0.36 0.35 0.3533

from which the tentative conclusion is that the minimum memory option does perform consistently worst but the low memory option appears to be a suitable replacement to the original code. More stringent tests are require to confirm this.

These initial tests were performed using the standard 32 processor SETTE test for ORCA2_ICE_PISCES. To search for a better distinction between the options further tests were made by varying the number of processors. Tests with 2, 8, 32 and 60 processors were performed (3 for each option at each core count). The following table shows the percentage of CPU time spent in tra_qsr and the rank of the tra_qsr routine in the CPU time-sorted list of routines (a higher rank means tra_qsr is taking proportionally less of the overall CPU time). In each case the average of the 3 samples is given.:

% CPU spent in tra_qsr
#CPUs original min-mem low-mem
2 1.76 1.82 1.83
8 1.38 1.48 1.46
32 0.48 0.49 0.5
60 0.24 0.26 0.26


Rank in sorted list of routines by CPU usage
#CPUs original min-mem low-mem
2 14 12.67 12
8 16.33 15.67 15
32 22.33 21.33 23.33
60 26 25 25

Unfortunately the message is still mixed and, although the new codes reduce the size and number of temporary arrays; they have a slight detrimental effect on performance and do not alter the tendency for tra_qsr to consume higher percentages of the CPU time as domain size increases.

The low-memory option does, at least, lend itself to tiling.

Option 2 revisited

Following discussions with the previewer, it was decided that low-memory option should be the best approach but the slight deterioration in performance over the original code may be down to the over-zealous replacement of temporary scalars within the second 3D loop. On reflection there are also opportunities to reduce the number of floating point operations and load and store instructions within the first 3D loop.

Although significant variation between identical runs on the NOC cluster means the evidence is not conclusive; this second version of the low-memory option does appear to improve on the original code and is certainly no worse whilst using less storage. Here are the tables with the new results added. Graphs are shown below the code differences.

% CPU spent in tra_qsr
#CPUs original min-mem low-mem low-men v2
2 1.76 1.82 1.83 1.68
8 1.38 1.48 1.46 1.14
32 0.48 0.49 0.5 0.44
60 0.24 0.26 0.26 0.13


Rank in sorted list of routines by CPU usage
#CPUs original min-mem low-mem low-men v2
2 14 12.67 12 14
8 16.33 15.67 15 17.33
32 22.33 21.33 23.33 23
60 26 25 25 26

Here is the final set of differences between this improved low-memory solution and the original traqsr.F90:

  • traqsr.F90

    old new  
    109109      REAL(wp) ::   zchl, zcoef, z1_2        ! local scalars 
    110110      REAL(wp) ::   zc0 , zc1 , zc2 , zc3    !    -         - 
    111111      REAL(wp) ::   zzc0, zzc1, zzc2, zzc3   !    -         - 
    112       REAL(wp) ::   zz0 , zz1                !    -         - 
     112      REAL(wp) ::   zz0 , zz1 , ze3t, zlui   !    -         - 
    113113      REAL(wp) ::   zCb, zCmax, zze, zpsi, zpsimax, zdelpsi, zCtot, zCze 
    114       REAL(wp) ::   zlogc, zlogc2, zlogc3 
    115       REAL(wp), ALLOCATABLE, DIMENSION(:,:)   :: zekb, zekg, zekr 
    116       REAL(wp), ALLOCATABLE, DIMENSION(:,:,:) :: ze0, ze1, ze2, ze3, zea, ztrdt 
    117       REAL(wp), ALLOCATABLE, DIMENSION(:,:,:) :: zetot, zchl3d 
     114      REAL(wp) ::   zlogc 
     115      REAL(wp), ALLOCATABLE, DIMENSION(:,:)   :: ze0, ze1, ze2, ze3 
     116      REAL(wp), ALLOCATABLE, DIMENSION(:,:,:) :: ztrdt, zetot, ztmp3d 
    118117      !!---------------------------------------------------------------------- 
    119118      ! 
    120119      IF( ln_timing )   CALL timing_start('tra_qsr') 
     
    159158         ! 
    160159      CASE( np_RGB , np_RGBc )         !==  R-G-B fluxes  ==! 
    161160         ! 
    162          ALLOCATE( zekb(jpi,jpj)     , zekg(jpi,jpj)     , zekr  (jpi,jpj)     , & 
    163             &      ze0 (jpi,jpj,jpk) , ze1 (jpi,jpj,jpk) , ze2   (jpi,jpj,jpk) , & 
    164             &      ze3 (jpi,jpj,jpk) , zea (jpi,jpj,jpk) , zchl3d(jpi,jpj,jpk)   ) 
     161         ALLOCATE( ze0 (jpi,jpj)           , ze1 (jpi,jpj) ,  & 
     162            &      ze2 (jpi,jpj)           , ze3 (jpi,jpj) ,  & 
     163            &      ztmp3d(jpi,jpj,nksr + 1)                     ) 
    165164         ! 
    166165         IF( nqsr == np_RGBc ) THEN          !*  Variable Chlorophyll 
    167166            CALL fld_read( kt, 1, sf_chl )         ! Read Chl data and provides it at the current time step 
     167            ! Separation in R-G-B depending of the surface Chl 
     168            DO_3D_00_00 ( 1, nksr + 1 ) 
     169               zchl    = MIN( 10. , MAX( 0.03, sf_chl(1)%fnow(ji,jj,1) ) ) 
     170               zCze    = 1.12  * zchl**0.803 
     171               zCtot   = 40.6  * zchl**0.459 
     172               zlogc   = LOG( zchl ) 
     173               ! 
     174               zCb     = 0.768 + zlogc * ( 0.087 - zlogc * ( 0.179 + zlogc * 0.025 ) ) 
     175               zCmax   = 0.299 - zlogc * ( 0.289 - zlogc * 0.579 ) 
     176               zpsimax = 0.6   - zlogc * ( 0.640 - zlogc * ( 0.021 + zlogc * 0.115 ) ) 
     177               zdelpsi = 0.710 + zlogc * ( 0.159 + zlogc * 0.021 ) 
     178               ! 
     179               zze     = 568.2 * zCtot**(-0.746) 
     180               IF( zze > 102. ) zze = 200.0 * zCtot**(-0.293) 
     181               zpsi    = gdepw(ji,jj,jk,Kmm) / zze 
     182               ! 
     183               ! NB. make sure zchl value is such that: zchl = MIN( 10. , MAX( 0.03, zchl ) ) 
     184               zchl = MIN( 10. , MAX( 0.03, zCze * ( zCb + zCmax * EXP( -( (zpsi - zpsimax) / zdelpsi )**2 ) ) ) ) 
     185               ! Convert chlorophyll value to attenuation coefficient look-up table index 
     186               ztmp3d(ji,jj,jk) = 41 + 20.*LOG10(zchl) + 1.e-15 
     187            END_3D 
     188         ELSE                                !* constant chlorophyll 
     189            zchl = 0.05 
     190            ! NB. make sure constant value is such that: 
     191            zchl = MIN( 10. , MAX( 0.03, zchl ) ) 
     192            ! Convert chlorophyll value to attenuation coefficient look-up table index 
     193            zlui = 41 + 20.*LOG10(zchl) + 1.e-15 
    168194            DO jk = 1, nksr + 1 
    169                DO jj = 2, jpjm1                       ! Separation in R-G-B depending of the surface Chl 
    170                   DO ji = 2, jpim1 
    171                      zchl    = MIN( 10. , MAX( 0.03, sf_chl(1)%fnow(ji,jj,1) ) ) 
    172                      zCtot   = 40.6  * zchl**0.459 
    173                      zze     = 568.2 * zCtot**(-0.746) 
    174                      IF( zze > 102. ) zze = 200.0 * zCtot**(-0.293) 
    175                      zpsi    = gdepw(ji,jj,jk,Kmm) / zze 
    176                      ! 
    177                      zlogc   = LOG( zchl ) 
    178                      zlogc2  = zlogc * zlogc 
    179                      zlogc3  = zlogc * zlogc * zlogc 
    180                      zCb     = 0.768 + 0.087 * zlogc - 0.179 * zlogc2 - 0.025 * zlogc3 
    181                      zCmax   = 0.299 - 0.289 * zlogc + 0.579 * zlogc2 
    182                      zpsimax = 0.6   - 0.640 * zlogc + 0.021 * zlogc2 + 0.115 * zlogc3 
    183                      zdelpsi = 0.710 + 0.159 * zlogc + 0.021 * zlogc2 
    184                      zCze    = 1.12  * (zchl)**0.803 
    185                      ! 
    186                      zchl3d(ji,jj,jk) = zCze * ( zCb + zCmax * EXP( -( (zpsi - zpsimax) / zdelpsi )**2 ) ) 
    187                   END DO 
    188                   ! 
    189                END DO 
     195               ztmp3d(:,:,jk) = zlui 
    190196            END DO 
    191          ELSE                                !* constant chrlorophyll 
    192            DO jk = 1, nksr + 1 
    193               zchl3d(:,:,jk) = 0.05 
    194             ENDDO 
    195197         ENDIF 
    196198         ! 
    197199         zcoef  = ( 1. - rn_abs ) / 3._wp    !* surface equi-partition in R-G-B 
    198200         DO_2D_00_00 
    199             ze0(ji,jj,1) = rn_abs * qsr(ji,jj) 
    200             ze1(ji,jj,1) = zcoef  * qsr(ji,jj) 
    201             ze2(ji,jj,1) = zcoef  * qsr(ji,jj) 
    202             ze3(ji,jj,1) = zcoef  * qsr(ji,jj) 
    203             zea(ji,jj,1) =          qsr(ji,jj) 
     201            ze0(ji,jj) = rn_abs * qsr(ji,jj) 
     202            ze1(ji,jj) = zcoef  * qsr(ji,jj) 
     203            ze2(ji,jj) = zcoef  * qsr(ji,jj) 
     204            ze3(ji,jj) = zcoef  * qsr(ji,jj) 
     205            ! store the surface SW radiation; re-use the surface ztmp3d array 
     206            ! since the surface attenuation coefficient is not used 
     207            ztmp3d(ji,jj,1) =       qsr(ji,jj) 
    204208         END_2D 
    205209         ! 
    206          DO jk = 2, nksr+1                   !* interior equi-partition in R-G-B depending of vertical profile of Chl 
    207             DO_2D_00_00 
    208                zchl = MIN( 10. , MAX( 0.03, zchl3d(ji,jj,jk) ) ) 
    209                irgb = NINT( 41 + 20.*LOG10(zchl) + 1.e-15 ) 
    210                zekb(ji,jj) = rkrgb(1,irgb) 
    211                zekg(ji,jj) = rkrgb(2,irgb) 
    212                zekr(ji,jj) = rkrgb(3,irgb) 
    213             END_2D 
    214  
    215             DO_2D_00_00 
    216                zc0 = ze0(ji,jj,jk-1) * EXP( - e3t(ji,jj,jk-1,Kmm) * xsi0r       ) 
    217                zc1 = ze1(ji,jj,jk-1) * EXP( - e3t(ji,jj,jk-1,Kmm) * zekb(ji,jj) ) 
    218                zc2 = ze2(ji,jj,jk-1) * EXP( - e3t(ji,jj,jk-1,Kmm) * zekg(ji,jj) ) 
    219                zc3 = ze3(ji,jj,jk-1) * EXP( - e3t(ji,jj,jk-1,Kmm) * zekr(ji,jj) ) 
    220                ze0(ji,jj,jk) = zc0 
    221                ze1(ji,jj,jk) = zc1 
    222                ze2(ji,jj,jk) = zc2 
    223                ze3(ji,jj,jk) = zc3 
    224                zea(ji,jj,jk) = ( zc0 + zc1 + zc2 + zc3 ) * wmask(ji,jj,jk) 
    225             END_2D 
    226          END DO 
     210         !* interior equi-partition in R-G-B depending of vertical profile of Chl 
     211         DO_3D_00_00 ( 2, nksr + 1 ) 
     212            ze3t = e3t(ji,jj,jk-1,Kmm) 
     213            irgb = NINT( ztmp3d(ji,jj,jk) ) 
     214            zc0 = ze0(ji,jj) * EXP( - ze3t * xsi0r ) 
     215            zc1 = ze1(ji,jj) * EXP( - ze3t * rkrgb(1,irgb) ) 
     216            zc2 = ze2(ji,jj) * EXP( - ze3t * rkrgb(2,irgb) ) 
     217            zc3 = ze3(ji,jj) * EXP( - ze3t * rkrgb(3,irgb) ) 
     218            ze0(ji,jj) = zc0 
     219            ze1(ji,jj) = zc1 
     220            ze2(ji,jj) = zc2 
     221            ze3(ji,jj) = zc3 
     222            ztmp3d(ji,jj,jk) = ( zc0 + zc1 + zc2 + zc3 ) * wmask(ji,jj,jk) 
     223         END_3D 
    227224         ! 
    228225         DO_3D_00_00( 1, nksr ) 
    229             qsr_hc(ji,jj,jk) = r1_rho0_rcp * ( zea(ji,jj,jk) - zea(ji,jj,jk+1) ) 
     226            qsr_hc(ji,jj,jk) = r1_rho0_rcp * ( ztmp3d(ji,jj,jk) - ztmp3d(ji,jj,jk+1) ) 
    230227         END_3D 
    231228         ! 
    232          DEALLOCATE( zekb , zekg , zekr , ze0 , ze1 , ze2 , ze3 , zea , zchl3d ) 
     229         DEALLOCATE( ze0 , ze1 , ze2 , ze3 , ztmp3d ) 
    233230         ! 
    234231      CASE( np_2BD  )            !==  2-bands fluxes  ==! 
    235232         ! 

...

Documentation updates

Error: Failed to load processor box
No macro or processor named 'box' found

...

Preview

Error: Failed to load processor box
No macro or processor named 'box' found

...

Tests

Error: Failed to load processor box
No macro or processor named 'box' found

...

Review

Error: Failed to load processor box
No macro or processor named 'box' found

...

Attachments (13)

Download all attachments as: .zip