Resurrecting the mixed-layer trend diagnostics
Context
The mixed-layer diagnostics potentially available via ln_tra_mxl
are long since dead (but not entirely buried). The current code is incomplete and seems to be a partial port of the code last working in the CMIP5 branch: CMIP5 branch from 12 years ago.
-
Branches impacted: branch_4.2 initially -
Reference configuration/test case ORCA2_ICE_PISCES -
Any solution will be a starting point for fixing branch_5.0 and main (with additional considerations for RK3)
Analysis
It is possible to get this code to the point of producing plausible looking output but it needs deeper analysis to assess the accuracy and completeness of the solution. This need for expert input is shared with #487 and #395 (closed). The merge request and branch associated with this issue will detail the changes required to activate the option in an ORCA2_ICE_PISCES
SETTE type test. Further work will be needed to assess the solution. Areas of continued concern will be highlighted in the merge request comments.
Fix
Currently trd_mxl_zint
is called from trd_tra_mng
but does nothing useful.
It looks as if trd_tra_mxl
was being prepared as a replacement but is not currently used (it also lacks the one thing that trd_mxl_zint
still does)
My solution is to add the trd_mxl_zint
code to trd_tra_mxl
, fix a few bugs and change the call in trd_tra_mng
There are a few other changes related to new components (e.g. _zdfp
) that are needed and additions to namelist_cfg
, field_def_nemo-oce.xml
and file_def_nemo-oce.xml
are required to activate testing in ORCA2_ICE_PISCES
.