3 |
Notes on tags used in MITgcmUV |
Notes on tags used in MITgcmUV |
4 |
============================== |
============================== |
5 |
|
|
6 |
|
checkpoint55g_post |
7 |
|
o change arrays for total vertical diffusivity (2.D if explicit, |
8 |
|
3.D if implicit) ; each ptracer uses its own background value. |
9 |
|
o added mnc/NetCDF output capability to ocn_compon_interf |
10 |
|
- added bits to verification/cpl_aim+ocn so that it can run on |
11 |
|
ITRDA and other clusters without the mpirun "-wd" option |
12 |
|
o fixed wayward endif in mom_fluxform.F that messed up cyclindrical |
13 |
|
configuration |
14 |
|
o give MNC the ability to create additional files rather than |
15 |
|
exceeding a specified file size limit "MNC_MAX_FILESIZE" |
16 |
|
- the default is just less than 2GB which is a practical limit |
17 |
|
on many systems |
18 |
|
o Compute the non-local term using either KPP vertical diffusivity (=default) |
19 |
|
or the total vertical diffusivity (when KPP_ghatUseTotalDiffus=T) |
20 |
|
|
21 |
|
checkpoint55h_post |
22 |
|
o add verification test "vermix" |
23 |
|
o remove all the *_ioinc flags and replace with the single global |
24 |
|
outputTypesInclusive flag |
25 |
|
o allow to set a vertical profile of vertical diffusivity for T & S |
26 |
|
o call PACKAGE_BOOT & PACKAGES_READPARMS just after INI_PARMS |
27 |
|
o useMNC moved from data.mnc to data.pkg |
28 |
|
|
29 |
|
checkpoint55g_post |
30 |
|
o Renaming carbon AD experiments to comply with testreport conventions |
31 |
|
o Updating adjoint of Held-Suarez hs94.1x64x5 |
32 |
|
o Fixing directives for non-ECCO forward_step, after ECCO directives |
33 |
|
have been fixed to avoid recomp. of do_oceanic_phys |
34 |
|
Works well for global_ocean.90x40x15, |
35 |
|
but recomp. issue remains for carbon |
36 |
|
o Updating adjoint output for natl_box_adjoint after |
37 |
|
pkg/ecco has been modified (see taga c55d_post) |
38 |
|
|
39 |
|
checkpoint55f_post |
40 |
|
o rename to timeave_init_fixed |
41 |
|
|
42 |
|
checkpoint55e_post |
43 |
|
o some delicate re-shuffle of store directives to avoid one |
44 |
|
extra call of do_oceanic_physics |
45 |
|
affects ecco_the_main_loop.F, do_oceanic_physics.F |
46 |
|
checkpoint_lev1_directives.h |
47 |
|
o NB: this may break global_ocean adjoint temporarily, |
48 |
|
but it is clear how to fix it. Will do later, need this now. |
49 |
|
|
50 |
|
checkpoint55d_post |
51 |
|
o ECCO specific cost function terms (up-to-date with 1x1 runs) |
52 |
|
o ecco_cost_weights is modified to 1x1 runs |
53 |
|
o modifs to allow observations to be read in as |
54 |
|
single file or yearly files |
55 |
|
o enable to read exf forcing fields as either |
56 |
|
single file or yearly files (flag useExfYearlyFields) |
57 |
|
checkpoint55d_pre |
58 |
|
|
59 |
|
o switch between 2-level and 3-level checkpointing |
60 |
|
o add ability of MNC to write local and "partial" (eg. 2D slices where |
61 |
|
the full 3D field is never actually stored) arrays to NetCDF files |
62 |
|
with the correct (that is, the complete multi-dimensional) set of |
63 |
|
array indicies |
64 |
|
- used in mom_vecinv() to write the diagFreq output |
65 |
|
- tested (demonstrated) in verification/aim.5l_cs |
66 |
|
o further mnc-ification |
67 |
|
- add verification/lab_sea to the list with mnc setups |
68 |
|
- fix undefined mnc v-names in cd_code |
69 |
|
- add kpp to mnc output (both snapshot and timeave) |
70 |
|
o non-hydrostatic code |
71 |
|
- added horizontal viscosity parameter viscAhW for w, that replaces viscAh |
72 |
|
in calc_gw.F |
73 |
|
- defaults to viscAh |
74 |
|
- this is useful, when I want to use viscAh=0 (e.g., only biharmonic |
75 |
|
friction or Leith) in non-hydrostatic mode, as long as there is |
76 |
|
no biharmonic friction and/or Leigth or equivalent for vertical velocity. |
77 |
|
o pkg/cost |
78 |
|
extend simple cost functions to treat time mean quantities |
79 |
|
(T,U,V) via cost_accumulate_state from cost_tile |
80 |
|
o multidimAdvection on CS-grid: implement the full 3 pass ; overlap needs to |
81 |
|
be 2 times larger than the minimum size the advection scheme requires. |
82 |
|
now get a perfect conservation of salt in advect_cs test-exp (updated) |
83 |
|
o GGL90 (Gaspar's TKE mixed layer) |
84 |
|
- add horizontal diffusion of TKE, requires exchanges in |
85 |
|
do_fields_blocking_exchanges, horizontal diffusivity is zero by |
86 |
|
default. In OPA there is no horizontal diffusion of TKE but the |
87 |
|
mixing coefficients are computed from a horizontal average of TKE of |
88 |
|
6 points or so. I think that diffusion has a little more physical |
89 |
|
justification. |
90 |
|
- clean up ggl90_calc in the hope of reducing memory usage (this hope |
91 |
|
was in vain) |
92 |
|
- mask tke-variable in ggl90_init |
93 |
|
o fix mask Pb in multidimAdvection on CS-grid (still a small Pb with ipass) |
94 |
|
o fix biharmonic diffusion on CS-grid. |
95 |
|
o more obcs stuff |
96 |
|
- added obcs_external_fields_load as an alternative to the EXF-pkg option |
97 |
|
for reading boundary values. This routine is called from |
98 |
|
obcs_prescribe_read |
99 |
|
- updated obcs_check |
100 |
|
- changed sponge layer relaxation parameter units for days to seconds |
101 |
|
- turned off sponge layer by default |
102 |
|
- changed argument list of OBCS_CALC to myTime+deltaTclock |
103 |
|
- fixed about 35 bugs in obcs_external_fields_load |
104 |
|
|
105 |
|
checkpoint55c_post |
106 |
|
o obcs stuff |
107 |
|
- new runtime flags useOBCSprescribe, useOBCSbalance |
108 |
|
- re-arranged obcs_calc |
109 |
|
- moved exf_getobcs to obcs_prescibe_read |
110 |
|
obcs_prescibe_read is called after ORLANSKI stuff |
111 |
|
and independent of it (executable can contain all). |
112 |
|
- updated obcs_check |
113 |
|
- added exchanges for U,V,gU,gV for useOBCSprescribe=.TRUE. |
114 |
|
These may be necessary only if tangential velocities are |
115 |
|
prescribed at OB, not sure, needs testing. |
116 |
|
o make the MNC_CW_ADD_VATTR_* subroutines all take scalar arguments |
117 |
|
which simplifies the calls and passes the ifc check-bounds test |
118 |
|
|
119 |
|
checkpoint55b_post |
120 |
|
o add a 2D scalar, Z location, exchange_2 routine. |
121 |
|
o EXCH_UV for CS-grid: add one u,v in the corner-halo region, allows |
122 |
|
to compute vorticity on a wider stencil [e.g., vort3(0,1)&(1,0)] |
123 |
|
affects advect_cs results (for the wrong reason): update output.txt |
124 |
|
o use a more standard discretization for biharmonic viscosity ; |
125 |
|
(original version still available with #define MOM_VI_ORIGINAL_VISCA4 ) |
126 |
|
affects results of dome set-up: update output.txt |
127 |
|
o fix gad_advection.F (CubedSphere part), back to version 1.23 |
128 |
|
o fix bug in mom_vi_del2uv.F (CubedSphere part). |
129 |
|
|
130 |
|
checkpoint55a_post |
131 |
|
o merged code to |
132 |
|
* prescribe/read time-dependent open boundaries |
133 |
|
(works in conjunction with exf, cal) |
134 |
|
* sponge layer code for open boundaries |
135 |
|
* each boundary N/S/E/W now has its own CPP option |
136 |
|
(healthy for the adjoint) |
137 |
|
|
138 |
|
checkpoint55 |
139 |
|
o up-to-date adjoint of previous breaks |
140 |
|
- natl_box_adjoint: problem for exp 12, 13: cost = 0. needs fix |
141 |
|
- global_ocean_ebm: adjoint still needs fix (TLM ok) |
142 |
|
- global_ocean.90x40x15: extended GMREDI_OPTIONS.h |
143 |
|
- carbon: * remove tr1-related adjoint verification |
144 |
|
* add ptracer as control variable |
145 |
|
o remove all tr1-related code (ALLOW_PASSIVE_TRACER) |
146 |
|
(adjoint stuff still has some tr1 'names', but all use ptracer arrays) |
147 |
|
|
148 |
|
checkpoint54f_post |
149 |
|
o in verification/carbin/ remove experiment related to |
150 |
|
ALLOW_PASSIVE_TRACER (array tr1 etc) |
151 |
|
code_ad/ input_ad/ results_ad |
152 |
|
|
153 |
|
o add two new packages |
154 |
|
- ggl90, TKE-model of Gaspar et al. (1990), Blanke+Delecluse (1993) |
155 |
|
+ subject to optimization (memory) |
156 |
|
- opps, OPPS (ocean penetrative plume scheme) of Paluszkiewicz+Romea (1997) |
157 |
|
+ subject to reordering (many GOTOs in the original code) |
158 |
|
+ so far opps is called for every (i,j) point, and then operates in |
159 |
|
the vertical; one may want to reorder the scheme in order operate in |
160 |
|
horizontal slabs at the cost of (a lot of) memory, at least ten new 3D |
161 |
|
fields) |
162 |
|
o overhaul of IO so that we now have flags for MDSIO and/or MNC |
163 |
|
- all verification tests compile and run with linux_ia32_g77 |
164 |
|
- defaults are compatible with current input files--nothing |
165 |
|
should change if you were not previously using MNC |
166 |
|
- MNC output has been added in numerous places (eg. timeave) |
167 |
|
but there are still a few writes not yet do-able with MNC |
168 |
|
(this is in progress) |
169 |
|
- flags now allow for either/or/both use of MDSIO and MNC and |
170 |
|
documentation will soon follow |
171 |
|
- numerous small formatting cleanups for ProTeX |
172 |
|
o add version info (via BUILD_INFO.h) to MNC and stdout |
173 |
|
|
174 |
checkpoint54e_post |
checkpoint54e_post |
175 |
o add two new packages |
o add two new packages |
176 |
- pp81 (Packanowski and Philander, 1981), Richardson number and |
- pp81 (Packanowski and Philander, 1981), Richardson number and |