#542 closed Enhancement (fixed)
Computation of redistribution among categories in LIM3
Reported by: | nemo_user | Owned by: | vancop |
---|---|---|---|
Priority: | low | Milestone: | |
Component: | LIM3 | Version: | v3.1 |
Severity: | Keywords: | LIM* v3.1 | |
Cc: |
Description
Sometimes, due to ridging routine, ice thickness within one category may exceed limits, which is currently takcled by transferring ALL the volume and area contained in this category. Therefore, at subsequent time step, the category is empty, which the model does not seem to like very much. We found staircase-type patterns in mean ice thickness due to that in ORCA05.
The following patch is a simple way to smooth this type of category transfer. Instead of transferring all the ice area/volume, we transfer only half of it. 2 lines are commented and replaced by their corrected version:
The module involved is limitd_th.F90 and the routine is lim_itd_th_reb.
1137 !---------------------------------------
1138 ! identify thicknesses that are too big
1139 !---------------------------------------
1140 zshiftflag = 0
1142 DO jj = 1, jpj
1143 DO ji = 1, jpi
1144 IF (a_i(ji,jj,jl) .GT. zeps .AND. t_i(ji,jj,jl) .GT. hi_max(jl) ) THEN
1145 zshiftflag = 1
1146 zdonor(ji,jj,jl) = jl
1147 !!! zdaice(ji,jj,jl) = a_i(ji,jj,jl)
1148 !!! zdvice(ji,jj,jl) = v_i(ji,jj,jl)
! Those are the new lines
1149 ! fix new ice thickness to hmean and a_i to
a_i/2 (PM)
1150 zdaice(ji,jj,jl) = a_i(ji,jj,jl)/2 ! only half of category in transfert in the other category (PM)
1151 zdvice(ji,jj,jl) = v_i(ji,jj,jl)-zdaice(ji,jj,jl)*(hi_max(jl)+hi_max(jl-1))/2
1152 ENDIF
1153 END DO ! ji
1154 END DO ! jj
All the best
Pierre Mathiot (UCL)
Commit History (0)
(No commits)
Change History (6)
comment:1 Changed 12 years ago by clevy
- Owner changed from NEMO team to vancop
comment:2 Changed 11 years ago by barthele
comment:3 Changed 11 years ago by clem
- Resolution set to fixed
- Status changed from new to closed
fixed in branch dev_r4028_CNRS_LIM3 to be included in the next merge party (Nov. 2013)
comment:4 Changed 7 years ago by nemo
- Keywords LIM* added
comment:5 Changed 3 years ago by nemo
- Keywords r3.1 added
comment:6 Changed 3 years ago by nemo
- Keywords v3.1 added; r3.1 removed
Since I have never run NEMO in ORCA05, I have not seen the staircase-type patterns solved by this regularization. However :