3 |
Notes on tags used in MITgcmUV |
Notes on tags used in MITgcmUV |
4 |
============================== |
============================== |
5 |
|
|
6 |
|
checkpoint57a_post |
7 |
|
o this set of changes restores TAMC compatibility |
8 |
|
tested for global_ocean.90x40x15 adjoint on trough |
9 |
|
(what a pain) |
10 |
|
|
11 |
|
checkpoint57a_pre |
12 |
|
o some modifs in tools/build_options for sunos |
13 |
|
o call ctrl_pack for ctrl at end for optimcycle=0 |
14 |
|
|
15 |
|
checkpoint57 |
16 |
|
o add horizontal bi-harmonic viscosity for vertical velocity |
17 |
|
- new parameter viscA4W (defaults to viscA4) |
18 |
|
- affects only non-hydrostatic code |
19 |
|
- to be done: no-slip lateral boundary conditions |
20 |
|
o switch back to "sflux = sflux - runoff" in exf_getforcing.F |
21 |
|
o implement Implicit Vertical advection for pTracers |
22 |
|
o depth convergence accelerator: replace deltaTtracer by dTtracerLev(k) |
23 |
|
(implemented also for implicit vertical diffusion & advection) |
24 |
|
o OBCS: extend application of tracer open boundary conditions |
25 |
|
to a strip beyond the boundary of width Olx/Oly |
26 |
|
o add call monitor to TAF set of required calls (flow directives) |
27 |
|
and include in natl_box_adjoint |
28 |
|
o untangle a few ifdefs related to cost, ctrl, autodiff |
29 |
|
untangel a few ALLOW_ECCO_OPTIMIZATION |
30 |
|
o Apply Cooper & Haines representative estimates also at sub-surface (k=Nr) |
31 |
|
(theta2, salt2 in ecco_cost_weights) |
32 |
|
|
33 |
|
checkpoint56c_post |
34 |
|
o OBCS as control variables |
35 |
|
- update ctrl_ad_diff.list, obcs_ad_diff.list |
36 |
|
- remove balance of obcs controls from default |
37 |
|
- fix index bug nobcs in ctrl_init |
38 |
|
- fix dummy fields filen in ctrl_pack |
39 |
|
- add dummy weights for obcs |
40 |
|
o lsopt: comment out fort.94 output |
41 |
|
o optim: |
42 |
|
- remove unnecessary header files |
43 |
|
- adjusted namelists |
44 |
|
- add xerbla.F to Makefile |
45 |
|
o exf: |
46 |
|
- multiple modifications of exf_check_range |
47 |
|
- unit change in exf_init_runoff from [m/year] to [m/s] |
48 |
|
o mdsio: |
49 |
|
- i/o-vector change debugLevel from debLevA to debLevB |
50 |
|
|
51 |
|
checkpoint56b_post |
52 |
|
o collect all ptracers output in an new routine (PTRACERS_OUTPUT) that |
53 |
|
is called from DO_THE_MODEL_IO |
54 |
|
- rename PTRACERS_STATV_WRITE to PTRACERS_WRITE_TIMEAVE |
55 |
|
- introduce new parameter PTRACERS_dumpFreq (defaults to dumpFreq) |
56 |
|
- clean up PTRACERS_WRITE_TIMEAVE a little |
57 |
|
o add another verification experiment that tests both the cfc-pkg and |
58 |
|
the gchem-pkg with GCHEM_SEPARATE_FORCING undefined, that is with |
59 |
|
reactive tendencies treated inline with regular timestepping |
60 |
|
o testreport previously did a lousy job of comparing small numbers, so |
61 |
|
we (Ed and Martin) changed the way the cut-off of 1.e-12 is |
62 |
|
implemented in the little c-progam tmp_cmpnum.c which is embedded in |
63 |
|
testreport. |
64 |
|
o ptracers_monitor: bug fix for MPI |
65 |
|
|
66 |
|
checkpoint56a_post |
67 |
|
o GCHEM: finish reorganizating the package |
68 |
|
- forward_step calls GCHEM_CALC_TENDENDY, which computes gchemTendency |
69 |
|
(introduces another 3D-array for each passive tracer, but only if |
70 |
|
GCHEM_SEPARATE_FORCING is undefined. For GCHEM_SEPARATE_FORCING |
71 |
|
gchemTendency is not needed because the timestep is done separately) |
72 |
|
- gchemTendency is added to gPtr in GCHEM_ADD_TENDENCY |
73 |
|
- GCHEM is now---more or less cleanly---separate from PTRACERS |
74 |
|
- to bo done: handling of surfaceForcingPtr that is needed for KPP not |
75 |
|
so clear to me how that should be done in a general way. |
76 |
|
o CFC: improve formatting of output in CFC_ATMOS |
77 |
|
o PTRACERS: |
78 |
|
- rename GCHEM_MONITOR to PTRACERS_MONITOR and call it from MONITOR, so |
79 |
|
that ALL experiments with ptracers enable can be checked. This makes |
80 |
|
GCHEM_MONITOR obsolete. |
81 |
|
- include a runtime parameter PTRACERS_monitorFreq that defaults to |
82 |
|
monitorFreq |
83 |
|
- set default PTRACERS_write_mdsio to false if PTRACERS_write_mnc is true |
84 |
|
o add PTRACERS (PTR0[1-5]) to testreport output |
85 |
|
o exf_check_range: |
86 |
|
* default changed to .TRUE. |
87 |
|
* will check values for niter0 if debLevA, every timestep if debLevB |
88 |
|
* STOP in exf_check_range if 'out of range' |
89 |
|
* range chosen generous to allow for potential extema |
90 |
|
TBD: a version with stricter range |
91 |
|
permitted range would be runtime parameters |
92 |
|
o GMredi: |
93 |
|
- fix ldd97 slope limit ; extend valid domain of Psi-Bolus ; |
94 |
|
- change S/R GRAD_SIGMA to be valid in the overlap on CS-grid. |
95 |
|
- change the default GMREDI_OPTIONS.h: compile everything except Visbeck |
96 |
|
o GCHEM: |
97 |
|
- move call to GCHEM_FORCING_SEP, so that it is now called before |
98 |
|
the blocking exchanges to avoid computing halos in GCHEM_FORCING_SEP |
99 |
|
- introduced a useGCHEM runtime flag and applied it everywhere. |
100 |
|
- renamed PTRACERS_SEPARATE_FORCING to GCHEM_SEPARATE_FORCING; |
101 |
|
this flag is now only visible within GCHEM-pkg routines. |
102 |
|
- moved gchem_forcing_int temporarily into ptracers_forcing, this will |
103 |
|
be replaced by a 3D array that is computed before thermodynamics; for |
104 |
|
this, the tracer ID iTracer had to be included into the parameter list |
105 |
|
of ptracers_forcing |
106 |
|
- set up forward_step to call gchem_forcing, that will replace |
107 |
|
gchem_forcing_int in ptracers_forcing |
108 |
|
- cleaning up |
109 |
|
o small cleaning of exch2: remove files that are also in eesupp/src |
110 |
|
|
111 |
|
checkpoint56 |
112 |
|
o AIM: read land-Fraction earlier (from ini-fixed) and monitor land at nIter0 |
113 |
|
o mnc: cleanup "full" files--fixes out-of-memory problem |
114 |
|
o ECCO-GODAE |
115 |
|
* model/, pkg/, code/, input/ now complete w.r.t. ECCO-GODAE production setup |
116 |
|
* fully merged and updated from SIO code |
117 |
|
windstress sign change affects |
118 |
|
- exf_inscal_[u/v]stress |
119 |
|
- sign of xx_tau[u/v] adjustments |
120 |
|
- signs in cost_scat |
121 |
|
* ctrl_get_gen: avoid update of tau[u/v] for first few days |
122 |
|
* natl_box_adjoint updated and compressed |
123 |
|
* 2-level checkpoint switch added |
124 |
|
(tested in EXP. 1 of natl_box_adjoint) |
125 |
|
* new switch doInitXX to avoid init. of xx_... in optimcycle=0 |
126 |
|
o SVD |
127 |
|
* first complete, non-hacked version for SVD calculations |
128 |
|
* all interfaces clean w.r.t to handling of xx_..., adxx_..., g_xx_... |
129 |
|
performed via direct access files admtlm_vector.it???? |
130 |
|
(via ctrl_[un]pack) |
131 |
|
* optimcycle is used for ARPACK iteration counter |
132 |
|
(note: adxx_... is iterated, rather than xx_...; latter is constant!) |
133 |
|
* EV problem ok w.r.t. L2-norm in tangent space; need to check other norms |
134 |
|
and add linear transforms where necessary |
135 |
|
* controls are initial/final SST, SSS (can be extended; need to check |
136 |
|
ctrl_set_[un]pack_xyz.F according to ctrl_set_[un]pack_xy.F |
137 |
|
o mdsio: filename info to both stderr, stdout |
138 |
|
o exf: exf_inscal_runoff: separated default conversion [m/year] -> [m/sec] |
139 |
|
from sign change |
140 |
|
|
141 |
|
checkpoint55j_post |
142 |
|
o isolate dissipation tendency (allow to keep it out off AB) in V.I form. |
143 |
|
affects (truncation level) output.txt of global_ocean.cs32x15(&.thsice) |
144 |
|
o add consistent build directories and .cvsignore files to verification |
145 |
|
o added viscA4GridMin and viscA4GridMax for more control on biharmonic viscosity |
146 |
|
o make all subroutines 30 or less characters in length for certain older |
147 |
|
compilers to work (a thorough search only turns up two routines in |
148 |
|
pkg/ptracers) |
149 |
|
o allow (vecinv only) a different horiz. viscosity for Divergence and Vorticity; |
150 |
|
affects (truncation level) output.txt of global_ocean.cs32x15(&.thsice) |
151 |
|
|
152 |
|
checkpoint55i_post |
153 |
|
o change arrays for total vertical diffusivity (2.D if explicit, |
154 |
|
3.D if implicit) ; each ptracer uses its own background value. |
155 |
|
o added mnc/NetCDF output capability to ocn_compon_interf |
156 |
|
- added bits to verification/cpl_aim+ocn so that it can run on |
157 |
|
ITRDA and other clusters without the mpirun "-wd" option |
158 |
|
o fixed wayward endif in mom_fluxform.F that messed up cyclindrical |
159 |
|
configuration |
160 |
|
o give MNC the ability to create additional files rather than |
161 |
|
exceeding a specified file size limit "MNC_MAX_FILESIZE" |
162 |
|
- the default is just less than 2GB which is a practical limit |
163 |
|
on many systems |
164 |
|
o Compute the non-local term using either KPP vertical diffusivity (=default) |
165 |
|
or the total vertical diffusivity (when KPP_ghatUseTotalDiffus=T) |
166 |
|
|
167 |
|
checkpoint55h_post |
168 |
|
o add verification test "vermix" |
169 |
|
o remove all the *_ioinc flags and replace with the single global |
170 |
|
outputTypesInclusive flag |
171 |
|
o allow to set a vertical profile of vertical diffusivity for T & S |
172 |
|
o call PACKAGE_BOOT & PACKAGES_READPARMS just after INI_PARMS |
173 |
|
o useMNC moved from data.mnc to data.pkg |
174 |
|
|
175 |
|
checkpoint55g_post |
176 |
|
o Renaming carbon AD experiments to comply with testreport conventions |
177 |
|
o Updating adjoint of Held-Suarez hs94.1x64x5 |
178 |
|
o Fixing directives for non-ECCO forward_step, after ECCO directives |
179 |
|
have been fixed to avoid recomp. of do_oceanic_phys |
180 |
|
Works well for global_ocean.90x40x15, |
181 |
|
but recomp. issue remains for carbon |
182 |
|
o Updating adjoint output for natl_box_adjoint after |
183 |
|
pkg/ecco has been modified (see taga c55d_post) |
184 |
|
|
185 |
|
checkpoint55f_post |
186 |
|
o rename to timeave_init_fixed |
187 |
|
|
188 |
|
checkpoint55e_post |
189 |
|
o some delicate re-shuffle of store directives to avoid one |
190 |
|
extra call of do_oceanic_physics |
191 |
|
affects ecco_the_main_loop.F, do_oceanic_physics.F |
192 |
|
checkpoint_lev1_directives.h |
193 |
|
o NB: this may break global_ocean adjoint temporarily, |
194 |
|
but it is clear how to fix it. Will do later, need this now. |
195 |
|
|
196 |
|
checkpoint55d_post |
197 |
|
o ECCO specific cost function terms (up-to-date with 1x1 runs) |
198 |
|
o ecco_cost_weights is modified to 1x1 runs |
199 |
|
o modifs to allow observations to be read in as |
200 |
|
single file or yearly files |
201 |
|
o enable to read exf forcing fields as either |
202 |
|
single file or yearly files (flag useExfYearlyFields) |
203 |
|
checkpoint55d_pre |
204 |
|
|
205 |
|
o switch between 2-level and 3-level checkpointing |
206 |
|
o add ability of MNC to write local and "partial" (eg. 2D slices where |
207 |
|
the full 3D field is never actually stored) arrays to NetCDF files |
208 |
|
with the correct (that is, the complete multi-dimensional) set of |
209 |
|
array indicies |
210 |
|
- used in mom_vecinv() to write the diagFreq output |
211 |
|
- tested (demonstrated) in verification/aim.5l_cs |
212 |
o further mnc-ification |
o further mnc-ification |
213 |
- add verification/lab_sea to the list with mnc setups |
- add verification/lab_sea to the list with mnc setups |
214 |
- fix undefined mnc v-names in cd_code |
- fix undefined mnc v-names in cd_code |