4 |
Notes on tags used in MITgcmUV |
Notes on tags used in MITgcmUV |
5 |
============================== |
============================== |
6 |
|
|
7 |
|
o Implement 4-level checkpointing: AUTODIFF_4_LEVEL_CHECKPOINT |
8 |
|
Needed for high-res. runs with large global files |
9 |
|
to be able to limit tapelev size |
10 |
|
o Changed some tests for adjoint: |
11 |
|
* carbon/ now tests AUTODIFF_2_LEVEL_CHECKPOINT |
12 |
|
* natl_box_adjoint/code_bulk/ tests AUTODIFF_4_LEVEL_CHECKPOINT |
13 |
|
* global_ocean.90x40x15/ back to default 3-level checkpointing |
14 |
|
|
15 |
|
checkpoint57p_post |
16 |
|
o reading of delXfile & delYfile moved from ini_parms.F to ini_grid.F |
17 |
|
o move ${PKG}_MNC_INIT from ${PKG}_READ_PARAMS to ${PKG}_INIT_FIXED |
18 |
|
(already the case for some pkgs, including recent MNC_init, e.g. thsice) |
19 |
|
o call CALC_GW from DYNAMICS (instead of from FORWARD_STEP) |
20 |
|
o Dissipation & phiHyd gradient are always added to gU,gV in timestep.F |
21 |
|
(was already the case for dissipation with mom_vecinv, |
22 |
|
and also the case for grad.PhiHyd when staggered-timeStep) |
23 |
|
This will allow to put dissipation out-off the AB time-stepping. |
24 |
|
Unfortunately, affects truncation error. Update output of exp. that "fails" |
25 |
|
(cg2d matching number) and leave the others for later: front_relax(10), |
26 |
|
global_ocean_pressure(10),ideal_2D_oce(10),internal_wave(9),lab_sea(12), |
27 |
|
matrix_example(7!),natl_box(12),plume_on_slope(11),vermix(12x3,11x2). |
28 |
|
|
29 |
|
checkpoint57o_post |
30 |
|
o fix MNC/MDSIO flags in PTRACERS so that (1) it works, and |
31 |
|
(2) its ready to be documented during the 20050801 DocFest |
32 |
|
o MNC-ify the gmredi package -- tested and works w/ and w/o mnc |
33 |
|
for ideal_2D_oce and lab_sea on linux_ia32_g77 |
34 |
|
o set to #undef CTRL_SET_OLD_MAXCVARS_30 |
35 |
|
in global1x1_tot/code_seaice_adj to access indices 32, 33 |
36 |
|
(worth another tag) |
37 |
|
|
38 |
|
checkpoint57n_post |
39 |
|
o adding precip control xx_precip 32 |
40 |
|
o adding swflux control xx_swflux 33 |
41 |
|
o Replacing 10 forcing penalty cost routines by 1 generic. |
42 |
|
o Implementing partitioning in time-mean and variable adjustment |
43 |
|
for atmos. state. |
44 |
|
o Time-mean constant error is now read in data.ecco (instead of hard-coded) |
45 |
|
o in natl_box_adjoint: |
46 |
|
* updated exp 1 to 10 |
47 |
|
(but does not have time-mean vs. variable split results yet) |
48 |
|
o updated global1x1/*_seaice_adj/ |
49 |
|
to incorporate latest changes |
50 |
|
o enable useSingleCpuIO for mdsio vector routines |
51 |
|
(applies in particular to tapelev I/O) |
52 |
|
o removed inefficient K-loop in all active_file*control*.F |
53 |
|
|
54 |
|
checkpoint57m_post |
55 |
|
o re-arrange exp. adjustment.cs-32x32x1 (adjust_nlfs.cs-32x32x1 merge into) |
56 |
|
and advect_cs (use EXCH-2 and standard cs-32 grid files). |
57 |
|
o pkg/exch2: filling of face-corner halo regions is optional |
58 |
|
(ifdef W2_FILL_NULL_REGIONS) and is turned off by default. |
59 |
|
|
60 |
checkpoint57l_post |
checkpoint57l_post |
61 |
tag before updating tutorials (pre-release-2) |
tag before updating tutorials (pre-release-2) |
62 |
o add cos & sin of grid-direction orientation angle (curvilinear grid only, |
o add cos & sin of grid-direction orientation angle (curvilinear grid only, |