| 4 |
Notes on tags used in MITgcmUV |
Notes on tags used in MITgcmUV |
| 5 |
============================== |
============================== |
| 6 |
|
|
| 7 |
|
checkpoint64e (2013/03/05) |
| 8 |
|
o pkg/ecco and several verification experiments: |
| 9 |
|
Remove special versions of the_model_main.F, the_main_loop.F, forward_step.F |
| 10 |
|
from ecco package, and revert to unified routine in model/src/ |
| 11 |
|
Adjust verification experiments accordingly, especially gloal_oce_llc90 |
| 12 |
|
Should conclude long-standing work by Gael to reconsile the_main_loop with |
| 13 |
|
ECCO and will enable OpenAD's revolve loop in ECCO setup. |
| 14 |
|
o pkg/generic_advdiff: |
| 15 |
|
- implement compressible flow method for multi-dim advection (similar to |
| 16 |
|
gad_som_advect.F); add new option "GAD_MULTIDIM_COMPRESSIBLE" (since |
| 17 |
|
TAF generates many recomputations) to use it; no yet coded with implicit |
| 18 |
|
vertical advection. |
| 19 |
|
- change experiment advect_cs (new output) and advect_xy (same output) |
| 20 |
|
to test compressible flow multi-dim advection. |
| 21 |
|
o pkg/generic_advdiff: |
| 22 |
|
- fix SOM advection for unusual cases (linear Free-Surf in r*, other |
| 23 |
|
cases where linear Free-Surf is not @ k=1, e.g., useShelfIce) |
| 24 |
|
o pkg/seaice: |
| 25 |
|
- fix bug in the no-slip implementation in seaice_calc_strainrates.F, this |
| 26 |
|
bug affects two verification experiments (lab_sea and lab_sea.hb87), |
| 27 |
|
recover old (wrong) results by setting noSlipFac = 2. in |
| 28 |
|
seaice_calc_strainrates.F; |
| 29 |
|
- add another averaging method for etaZ (etaZmethod=3), a mix of 2 and 0 |
| 30 |
|
- add runtime parameter SEAICE_JFNKepsilon for that finite difference |
| 31 |
|
gradient in the Jacobian times vector operation (s/r seaice_jacvec) |
| 32 |
|
- fix line search in seaice_jfnk.F (s/r seaice_jfnk_update) |
| 33 |
|
o pkg/seaice: reorganize seaice_lsr.F to be able to reuse more code: |
| 34 |
|
- new subroutines seaice_lsr_tridiagu/v |
| 35 |
|
- introduce loop bounds i/jmin/max to be able to use a restricted |
| 36 |
|
additive Schwarz method, but not used for now. |
| 37 |
|
- rename SEAICE_CALC_LSR_COEFFS into SEAICE_LSR_CALC_COEFFS |
| 38 |
|
- use new subroutines seaice_lsr_tridiagu/v in s/r seaice_preconditioner |
| 39 |
|
o OpenAD support |
| 40 |
|
Reorganize S/R call for overall time-stepping loop to facilitate |
| 41 |
|
OpenAD revolve loop replacement |
| 42 |
|
(new S/R main_do_loop.F) |
| 43 |
|
o model/src: |
| 44 |
|
- add 3-D array 3rd dimension to argument list of S/R ADAMS_BASHFORTH2/3; |
| 45 |
|
- pkg/cheapaml: remove local S/R ADAMS2d and use instead standard S/R |
| 46 |
|
ADAMS_BASHFORTH2. |
| 47 |
|
o pkg/cheapaml: |
| 48 |
|
- add new parameter (default=F, same as before) to disable full domain |
| 49 |
|
(including land) periodicity in X & Y dir. |
| 50 |
|
- new S/R CHEAPAML_COPY_EDGES to copy relaxation field into edges of |
| 51 |
|
prognostics field ; merge the 2 S/R cycle_2d_tracer and timestep_2d_tracer |
| 52 |
|
into single S/R CHEAPAML_TIMESTEP. |
| 53 |
|
|
| 54 |
checkpoint64d (2013/02/17) |
checkpoint64d (2013/02/17) |
| 55 |
o pkg/monitor (angular momentum): |
o pkg/monitor (angular momentum): |
| 56 |
- do not add Eta contribution to the total AM if using rigid-lid; |
- do not add Eta contribution to the total AM if using rigid-lid; |
Parent Directory
| 
Revision Log
| 
Revision Graph
| 
Patch