Custom Query (2349 matches)

Filters
 
Or
 
  
 
Columns

Show under each result:


Results (25 - 27 of 2349)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Ticket Resolution Summary Owner Reporter
#2499 fixed Major bug when activating bio-model light penetration systeam cetlod
Description

Context

When using the simulated time varying chlorophyll by TOP biogeochemical model ( ln_qsr_bio =.true. ) to compute the light absorption in the ocean, the heat content trend due to Qsr flux is always zero everytime

Analysis

In tra_qsr_init, the level of light extinction ( nksr ) is not computed in case of BIO light penetration (np_BIO) and then set to 0 by default. The heat content trend is not computed and stay equal to its initial values

  DO jk = 1, nksr
     qsr_hc(:,:,jk) = r1_rau0_rcp * ( etot3(:,:,jk) - etot3(:,:,jk+1) )
  END DO

Recommendation

Compute the tabulated attenuation coef and the level of light extinction exactly in the same way than the RGB light penetration

     CASE( np_BIO )                   !==  BIO light penetration  ==!
         !
         IF(lwp) WRITE(numout,*) '   ==>>>   bio-model light penetration'
         IF( .NOT.lk_top )   CALL ctl_stop( 'No bio model : ln_qsr_bio = true impossible ' )
         CALL trc_oce_rgb( rkrgb )                 ! tabulated attenuation coef.
         !
         nksr = trc_oce_ext_lev( r_si2, 33._wp )   ! level of light extinction
         !
         IF(lwp) WRITE(numout,*) '        level of light extinction = ', nksr, ' ref depth = ', gdepw_1d(nksr+1), ' m'

         !
      END SELECT

#2498 fixed SETTE test of the trunk do not pass on jean zay techene techene
Description

Context

Sette tests with NEMO's trunk do not pass on jean zay because of an error in the group name set up.

sbatch: error: Batch job submission failed: Invalid account or account/partition combination specified

Analysis

sette tests run on jean zay using BATCH_TEMPLATE/batch-X64_JEANZAY BATCH_TEMPLATE/batch-X64_JEANZAY uses the keyword GROUP_IDRIS GROUP_IDRIS is initialised by prepare_job.sh with the line

export GROUP_IDRIS=`echo ${USER} |cut -c 2-4`

where USER is the user account

Fix

GROUP_IDRIS should be initialised by prepare_job.sh with the line

export GROUP_IDRIS=`echo ${IDRPROJ}`

where IDRPROJ is the user group

#2497 fixed Incorrect time-series output in file tracer.stat when key_iomput is undefined smueller smueller
Description

Context

The file tracer.stat produced by a standard ORCA2_ICE_PISCES run with undefined key_iomput differs from that produced by a corresponding run with defined key_iomput. Further, SETTE reports the failure of the ORCA2_ICE_PISCES restartability test if key_iomput is undefined (e.g., in SETTE with environment variables SETTE_XIOS and SETTE_MPMD set to "no") due to differences in the output file tracer.stat.

Analysis

When time-series output into file tracer.stat is active, local variable ll_trcstat in source:/NEMO/releases/r4.0/r4.0-HEAD/src/TOP/trcstp.F90 is set to .TRUE. during timesteps which compute and output tracer.stat time-series values. In the ORCA2_ICE_PISCES configuration of the model, variable areatot, which is used in the time-series computation, varies in time. Without key_iomput, however, areatot is not updated during all timesteps in which ln_trcstat is set to .TRUE., which leads to the computation of incorrect time-series values. In contrast, if key_iomput is defined and the output of at least one of a set of specific PISCES diagnostic variables (pno3tot, ppo4tot, psiltot, palktot, and pfertot) is enabled, areatot is updated during all timesteps and the tracer.stat time series is correctly computed.

Fix

Line source:/NEMO/releases/r4.0/r4.0-HEAD/src/TOP/trcstp.F90:#L81 could be brought into line with the corresponding trunk version (source:/NEMO/trunk/src/TOP/trcstp.F90@13286:#L83) by substituting ll_trcstat for ln_ctl in order to ensure the correct computation of the tracer.stat time-series values.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Note: See TracQuery for help on using queries.