#1072 closed Bug (fixed)
Calculation of griffies triads in ML gives NaNs where ML hits bottom
Reported by: | agn | Owned by: | George Nurser |
---|---|---|---|
Priority: | low | Milestone: | Unscheduled |
Component: | OCE | Version: | trunk |
Severity: | Keywords: | ||
Cc: |
Description
Rachel Furner and Enda O'Dea found that when calculating isoneutral diffusion with the griffies triads (ln_traldf_grif = .true. in the namelist) NaNs? were generated where the ML base reached the ocean floor.
The problem has been traced to the calculation of the isoneutral slopes just below the mixed-layer in lines 492-509 of ldfslp.F90. Where the ML went down to the bottom of the entire water column, nmln(ji+ip,jj)=jpkm1, unset values of zdzrho and zd{x,yr}ho were used, giving NaNs?.
Commit History (0)
(No commits)
Change History (2)
comment:1 Changed 11 years ago by agn
- Resolution set to fixed
- Status changed from new to closed
comment:2 Changed 11 years ago by agn
Corrected also in dev_v3_4_STABLE_2012
Note: See
TracTickets for help on using
tickets.
The bug was fixed by explicitly setting zt{i,j}_mlb = 0 whenever the ML hit the bottom at the U- or V-points, nmln(ji+ip,jj)=mbkt(ji+ip,jj) or nmln(ji,jj+jp)=mbkt(ji,jj+jp) respectively