3 |
Notes on tags used in MITgcmUV |
Notes on tags used in MITgcmUV |
4 |
============================== |
============================== |
5 |
|
|
6 |
untagged |
untagged: |
7 |
o Added net flux to downward flux conversion to pkg/exf/exf_getffields.F |
o minor changes to pkg/seaice pkg/cost/cost_ice.F |
8 |
o Added SEAICE_initialHEFF to pkg/seaice |
and verification/natl_box_adjoint/README |
9 |
o SEAICE_WRITE_PICKUP needs to be called before WRITE_CHECKPOINT |
o Modif's to enable tangent linear run in ECCO environment |
10 |
|
including time-dependent forcing |
11 |
|
* main problem is missing call mdsreadfield out of exf_set_gen |
12 |
|
(i.e. computation of trajectory in TLM mode) |
13 |
|
* incorrect g_active_write_... |
14 |
|
* (re-) init of bar files and g_bar files |
15 |
|
o Reverted back to 3-dim. tbar, sbar for SST, SSS bar files. |
16 |
|
o MERGES: |
17 |
|
* obcs/ ctrl/ G. Gebbie code |
18 |
|
(major changes in ctrl handling, volume flux balancing etc.) |
19 |
|
* cost/ ctrl/ A. Koehl code |
20 |
|
- new cost contributions (new data types) |
21 |
|
- efficient pack/unpack for time-dependent 2-d fields |
22 |
|
o replaced mutiple ctrl_get... by single generic ctrl_get_gen.F |
23 |
|
(except obcs) |
24 |
|
o somewhat cleaned package initialisation sequence for |
25 |
|
ctrl/ cost/ ecco/ |
26 |
|
o extended grdchk to enable obcs gradient checks |
27 |
|
o some generalizations for easier TLM <-> ADM switches |
28 |
|
|
29 |
|
ecco_c50_e32 |
30 |
|
o Patrick's adjoint-model storage directives for pkg/seaice |
31 |
|
o merged with checkpoint50g_post (pkg/seaice and verification/lab_sea) |
32 |
|
- added SEAICE_MULTILEVEL for 8-category sea-ice thermodynamics |
33 |
|
- LSR sea-ice dynamic solver moved to SouthWest B-grid location and |
34 |
|
made the default because of faster convergence than ADI |
35 |
|
o Modifications for adjoint of SEAICE_ALLOW_DYNAMICS |
36 |
|
- extended state in checkpoint_lev... |
37 |
|
- in lsr.F, replaced GO TO by fixed loop index (also avoid DO WHILE) |
38 |
|
- YEAH: in lsr.F, (first) successful use of CADJ LOOP = ITERATION |
39 |
|
- avoid stoing in S/R budget for now |
40 |
|
- in dynsolver.F, avoid NPSEUDO loop for adjoint |
41 |
|
- new comlev1_lsr in the_main_loop |
42 |
|
o Small modif's in lsopt/, optim/ to ease compilation |
43 |
|
New script to run optim (lsopt) |
44 |
|
|
45 |
|
ecco_c50_e31 |
46 |
|
o exf totally restructured |
47 |
|
- replaced all exf_set_ routines by one generic exf_set_gen.F |
48 |
|
and made sure it's still differentiable |
49 |
|
- split exf_getffields.F into new exf_bulkformulae.F |
50 |
|
- exf_inscal_/exf_outscal_ suffixes consistent with field names |
51 |
|
o New cost function: Tini-Tfin to do model spinup optimization. |
52 |
|
o Started new verification exp for Tini-Tfin optimization. |
53 |
|
|
54 |
|
ecco_c50_e30 |
55 |
|
o corrected CPP option mismatch in exf_Bulk... functions for case |
56 |
|
#define ALLOW_ATM_WIND && #undef ALLOW_ATM_TEMP |
57 |
|
o Merging from release1_p13 and from icebear5 |
58 |
|
- bug fix for pkg/seaice dynamic solver |
59 |
|
- Added SEAICE_initialHEFF to pkg/seaice |
60 |
|
- Additional options for pkg/cost/cost_ice.F |
61 |
|
- Updates to verification/natl_box_adjoint |
62 |
|
- pkg/ctrl/ctrl_init.F |
63 |
|
Fix for determination of startrec and endrec. |
64 |
|
- pkg/cal/cal_toseconds.F |
65 |
|
Fix for negative intervals shorter than 1 day. |
66 |
|
|
67 |
|
ecco_c50_e29 |
68 |
|
Bug fix for merging between c50 and KPP. |
69 |
|
ikey was passed from thermodynamics to kpp_calc via |
70 |
|
common block rather than being recomputed in kpp_calc, |
71 |
|
in contradiction with new key itdkey. |
72 |
|
New key ikppkey created, and tamc.h headers updated. |
73 |
|
|
74 |
|
ecco_c50_e28 |
75 |
|
o major merge: |
76 |
|
ecco-branch merged with checkpoint50 |
77 |
|
gmredi now available, grdchk, DIVA up-to-date for MPI |
78 |
|
However, two outstanding problems: |
79 |
|
- kpp adjoint is broken, needs fixing! |
80 |
|
- a few minor modif's in handling of CPP option |
81 |
|
ALLOW_TANGENTLINEAR_RUN in conjunction with |
82 |
|
ALLOW_ECCO_OPTIMIZATION are necessary to enable |
83 |
|
TLM runs with ECCO setup |
84 |
|
|
85 |
release1_p12 |
ecco_c44_e27 |
86 |
o Added missing /* */ to CPP comments in pkg/seaice, pkg/exf, |
o exf: |
87 |
kpp_transport_t.F, forward_step.F, and the_main_loop.F |
- Enable initialisation of forcing fields to constant |
88 |
o pkg/seaice: |
(runtime) values. |
89 |
- adjoint-friendly modifications |
- in exf_getffields.F |
90 |
- added a SEAICE_WRITE_PICKUP at end of the_model_main.F |
Reduce i-/j-loop to interior domain, discarding overlaps. |
91 |
|
That also fixes wrong TAF-key computations for key_1, key_2 |
92 |
release1_p12_pre |
with bulf formulae. |
93 |
o Updated useSingleCpuIO option in mdsio_writefield.F to |
- exf_init.F modify #ifdef for exf_init_evap |
94 |
work with multi-field files, e.g., for single-file pickup. |
- exf_getffieldrec.F, ctrl_getrec.F |
95 |
o pkg/seaice: |
The following INT-usages are not safe: |
96 |
- bug fix in growth.F: QNET for no shortwave case |
fldsecs = int(fldsecs/fldperiod)*fldperiod |
97 |
- added HeffFile for specifying initial sea-ice thickness |
fldcount = int(fldsecs/fldperiod) + 1 |
98 |
- changed SEAICE_EXTERNAL_FLUXES wind stress implementation |
and were modified. |
99 |
|
- added repeatPeriod (from MAIN branch) |
100 |
release1_p11 |
o grdchk: updated and merged |
101 |
o Modifications for using pkg/exf with pkg/seaice |
- has both ADM and TLM checks |
102 |
- improved description of the various forcing configurations |
- works for single- and multi-proc. |
103 |
- added basic radiation bulk formulae to pkg/exf |
- output cleaned |
104 |
- units/sign fix for evap computation in exf_getffields.F |
- worked successfully for parallel DIVA |
105 |
- updated verification/global_with_exf/results/output.txt |
Needed one update in EEPARAMS.h and updates in ctrl package |
106 |
o Added pkg/sbo for computing IERS Special Bureau for the Oceans |
More modif's may be necessary for ECCO code to support TLM. |
107 |
(SBO) core products, including oceanic mass, center-of-mass, |
o autodiff: checkpoint_lev... |
108 |
angular, and bottom pressure (see pkg/sbo/README.sbo). |
- updated storing for bulk formulae (some arrays were missing) |
109 |
o Lower bound for viscosity/diffusivity in pkg/kpp/kpp_routines.F |
- added state for seaice package |
110 |
to avoid negative values in shallow regions. |
- synchronized with main branch |
111 |
- updated verification/natl_box/results/output.txt |
o DIVA: |
112 |
- updated verification/lab_sea/results/output.txt |
Modif.s in |
113 |
o MPI gather, scatter: eesupp/src/gather_2d.F and scatter_2d.F |
- cost_init.F |
114 |
o Added useSingleCpuIO option (see PARAMS.h). |
- cost_averagesfields.F |
115 |
|
- ecco_the_main_loop.F |
116 |
release1_p10 |
- ecco_the_model_main.F |
117 |
o modifications for using pkg/exf with pkg/seaice |
and added cost_final_restore.F cost_final_store.F |
118 |
- pkg/seaice CPP options SEAICE_EXTERNAL_FORCING |
o With these changes, it is possible to fully separate |
119 |
and SEAICE_EXTERNAL_FLUXES |
the first forward run to compute the cost function (mdthe_main_loop) |
120 |
- pkg/exf CPP options EXF_READ_EVAP and |
and adjoint runs over individual DIVA intervals (adthe_main_loop) |
121 |
EXF_NO_BULK_COMPUTATIONS |
into separate model executions. |
122 |
- usage examples are Experiments 8 and 9 in |
This enables better tuning of DIVA to given queue sizes. |
123 |
verification/lab_sea/README |
o Loss of cost function value in consecutive DIVA runs is |
124 |
- verification/lab_sea default experiment now uses |
fixed by storing to 'costfinal' file. |
125 |
pkg/gmredi, pkg/kpp, pkg/seaice, and pkg/exf |
o Appropriate call of ctrl_unpack/_pack,grdchk_main at |
126 |
|
beginning/end of adjoint integration is accounted for. |
127 |
release1_p9 |
The current changes apply to TAF-1.4.26. |
128 |
|
TAF has one remaining bug, requiring one manual change |
129 |
|
(see MITgcm manual). |
130 |
|
o bug fix plus modif for ubar, vbar in cost_averagesfields |
131 |
|
o cal: synchronized between MAIN and ecco-branch |
132 |
|
|
133 |
|
icebear4 |
134 |
|
o Added cost_ice for sea-ice volume sensitivity experiment. |
135 |
|
o ecco_the_model_main.F: SEAICE_WRITE_PICKUP must precede WRITE_CHECKPOINT |
136 |
|
o Reduced derivative singularities in pkg/seaice. |
137 |
|
o Updated verification/natl_box_adjoint |
138 |
|
|
139 |
|
icebear3 |
140 |
|
o Removed spurious iikey and idkey |
141 |
|
o pkg/exf |
142 |
|
- bug fix in treatment of downward radiation |
143 |
o pkg/seaice |
o pkg/seaice |
144 |
- removed GOTO's and added taf directives |
- added initial sea-ice thickness parameter SEAICE_initialHEFF |
145 |
- double precision constants to reduce the g77 (Linux) |
o updated verification/natl_box_adjoint |
146 |
to F77 (SGI) differences reported in release1_p8 |
|
147 |
o tools/genmake |
icebear2 |
148 |
- added SGI options |
o Added missing /* */ to CPP comments in pkg/seaice, pkg/exf, |
149 |
o verification/testscript |
kpp_transport_t.F,and ecco_the_main_loop.F |
150 |
- updated to that of checkpoint47a_post |
o adjoint-friendly modifications to pkg/seaice |
151 |
o verification/global_ocean.90x40x15/input/eedata |
o added verification/lab_sea as per release1_p12_pre |
152 |
- modified for SGI f77 compatibility |
|
153 |
o verification/lab_sea |
ecco_c44_e26 |
154 |
- added description of sea-ice model |
o Merging pkg/exf and pkg/seaice from cvs tag ecco_ice2 |
155 |
- added missing matlab routines |
o exf: Modifications to correctly handle evap field for adjoint |
156 |
- added test of thermodynamics parallelization |
o grdchk: |
157 |
|
Added code to perform gradient checks for bulk formulae/atmos. state |
158 |
release1_p8 |
(atemp, aqh, uwind, vwind). |
159 |
o verification/natl_box: |
NOTE: |
160 |
updating new external_fields_load routine |
This package is out of synch with c48 package. |
161 |
o New package: pkg/seaice |
The latter also has the tangent linear gradient checks. |
162 |
Sea ice model by D. Menemenlis (JPL) and Jinlun Zhang (Seattle). |
Need to be merged. |
163 |
The sea-ice code is based on Hibler (1979-1980). |
o verification/natl_box_adjoint: |
164 |
Two sea-ice dynamic solvers, ADI and LSR, are included. |
* added new setup: code_bulk/, input_bulk, results_bulk |
165 |
In addition to computing prognostic sea-ice variables and diagnosing |
(Experiment 2, described in README) |
166 |
the forcing/external data fields that drive the ocean model, |
uses bulk formulae and atmos. state as control instead of |
167 |
SEAICE_MODEL also sets theta to the freezing point under sea-ice. |
air-sea fluxes. |
168 |
The implied surface heat flux is then stored in variable |
Performs gradient checks for first 4 elements of atemp control. |
169 |
surfaceTendencyTice, which is needed by KPP package (kpp_calc.F and |
* changed standard verification to perform |
170 |
kpp_transport_t.F) to diagnose surface buoyancy fluxes and for the |
gradient checks on first 4 elements on hflux control. |
171 |
non-local transport term. Because this call precedes model |
* gmredi switched off (will come after merge with c48). |
172 |
thermodynamics, temperature under sea-ice may not be "exactly" at |
|
173 |
the freezing point by the time theta is dumped or time-averaged. |
ecco_c44_e25 |
174 |
N.B.: fairly large differences in accuracy occur across |
o Merging some GM changes from main trunk (c47) to ecco-branch. |
175 |
different platforms/compilers; comparison between |
o Updating verif. natl_box_adjoint with GMRedi. |
176 |
g77 (Linux) and F77 (SGI Origin 2000/MIPSpro) |
o Adding setup with explicit differentiaton of pressure solver. |
177 |
generated output gives: |
o Bug fix: missing OBCS_OPTIONS.h for sponge layer code. |
178 |
T S U V |
o (Re-?) Commiting old bug fixes in kpp to ecco-branch |
179 |
C D M c m s m s m s m s |
(kpp_calc.F, kpp_routines.F) |
|
n p a R g m m e . m m e . m m e . m m e . |
|
|
f n k u 2 i a a d i a a d i a a d i a a d |
|
|
g d e n d n x n . n x n . n x n . n x n . |
|
|
|
|
|
Y Y Y Y 5 5 7 7 7 8 10 9 6 6 6 6 7 5 7 5 7 FAIL lab_sea |
|
|
|
|
|
release1_p7 |
|
|
o GAD: |
|
|
- generated new common blocks to account for call of |
|
|
same gad routines with differing traceridentities |
|
|
(needed to modify tracerIdentity indices in GAD.h) |
|
|
- generated separate common blocks for case useCubedSphereExchange |
|
|
(Department of Futurology) |
|
|
- parameter lists to gmredi_?transport: added tracerIdentity |
|
|
- added new key indices to tamc.h |
|
|
o external_fields_load: |
|
|
- added this routine to TAF list |
|
|
- needed to make some common blocks global and additional storing |
|
|
along the same lines as exf package (checkpoint_lev?_directives.h) |
|
180 |
o exf: |
o exf: |
181 |
updated external forcing package |
updated external forcing package |
182 |
- bug fixes carried over from ecco-branch |
- cleaned old exfa stuff |
183 |
|
- bug fixes |
184 |
(missing OBCS_OPTIONS.h in two routines) |
(missing OBCS_OPTIONS.h in two routines) |
185 |
- enable easy to use "no forcing". |
- enable easy to use "no forcing". |
186 |
- added exf I/O for atmospheric loading |
- added exf I/O for atmospheric loading |
187 |
- added exf I/O for runoff data |
- added exf I/O for runoff data |
188 |
- transfered scaling between exf <-> MITgcm to exf namelist |
- transfered scaling between exf <-> MITgcm to exf namelist |
189 |
o Adding new verification experiment global_with_exf. |
|
190 |
This verification experiment is almost identical to the |
ecco_c44_e24 |
|
global_ocean.90x40x15 experiment, except that it uses |
|
|
the exf (and cal) package instead of the |
|
|
external_fields_load routine (cf README file). |
|
|
|
|
|
release1_p6 |
|
|
o fixed the verification/global_ocean.90x40x15 experiment: |
|
|
- new bathymetry (the world according to A., JMC, and M.) |
|
|
- new initial fields and forcing fields (*.bin files) |
|
|
- new POLY3.COEFFS (for the next release one should switch to a full |
|
|
equation of state) |
|
|
- fixed several errors and redundancies in the data file |
|
|
- experiment uses looped cells |
|
|
- added matlab directory with diagnostic scripts for plotting of output |
|
|
|
|
|
release1_p5 |
|
|
o import recent fix from MAIN branch: |
|
|
- use recip_dx*,recip_dy* instead of /dx*,/dy* in orlanski_E,W,N,S |
|
|
- change phiMin in exp1 (to agree with documentation) |
|
|
from checkpoint45 to checkpoint45a_post: |
|
|
- specific exchange for X-slice domain (case Ny=1) |
|
|
- argument futureIter of S/R obcs_calc missing in |
|
|
exp4 & internal_wave /code |
|
|
- solve_for_pressure : set cg2d_x to zero where OBCS are applied |
|
|
from checkpoint44f_post to checkpoint44g_post: |
|
|
- fix surface correction term in multiDimAdvection |
|
|
(affects plume_on_slope results |
|
|
o bug fix in packages_readparms (missing ALLOW_CALENDAR) |
|
|
o bug fix in CG2D.h (correct array size for LETS_MAKE_JAM |
|
|
o Adjoint-related bug fixes in kpp: |
|
|
- kpp_calc: sore of kpphbl avoids recomputation/call to S/R kppmix |
|
|
- kpp_routines: store of Rib avoids partial recomputation bug of TAF. |
|
|
o added Eliassen-Palm flux hooks (dynamics,ini_fields) |
|
|
o removed unused TAF keys iikey, idkey (dynamics,thermodynamics) |
|
|
o flow directives: |
|
|
- new for S/R do_the_model_io |
|
|
- enhanced for S/R checkpoint |
|
|
o added Eliassen Palm flux controls to gradient check package |
|
|
o new ctrl package |
|
|
- adopted from ECCO environment to enable optimization |
|
|
- added Eliassen Palm fluxes to controls |
|
|
- adjusted global_ocean.90x40x15/ and carbon/ accordingly |
|
|
|
|
|
release1_p4 |
|
|
o Removed f90 comments ("!") |
|
|
o Modified initialisations to enable exf package for MITgcm |
|
|
(without invoking ECCO_PACKAGE). |
|
191 |
o modifications to gradient check package (Martin Losch) |
o modifications to gradient check package (Martin Losch) |
192 |
- enable centered differences |
- enable centered differences |
193 |
- modified format of standard output |
- modified format of standard output |
194 |
|
o bug fix for nwettile contents |
195 |
|
so far assumed global sum in first entry (1,1,k), |
196 |
|
needed for global control vector |
197 |
|
-> introduced new variable nwet[c/w/s]global(nr) |
198 |
|
o shifts of calls to cal/exf package to be accessible |
199 |
|
in MITgcm stand-alone |
200 |
|
o shifted packages_init_variables before convective_adjustment_ini |
201 |
|
(corresponding update of TAF tape initialisation). |
202 |
|
o removed auxiliary field in impldiff |
203 |
|
(caused numerical differences for GMRedi cases) |
204 |
|
o updated obcs control part (G. Gebbie) |
205 |
|
o removed some f90 comments ("!") |
206 |
|
o Bug fixes and updates in cost package |
207 |
|
* cost_final: correct multiplier mult_tauv, mult_sflux |
208 |
|
* cost_hyd: added call cost_ssh |
209 |
|
* cost_init_barfiles.F: replaced arrays by tmpfld's |
210 |
|
* cost_readtopexmean.F: ersmask -> tpmeanmask |
211 |
|
* cost_weights.F: corrected wtheta[2], wsalt[2] |
212 |
|
o cleaned exchanges in cg2d.F |
213 |
|
|
214 |
release1_p3 |
ecco_c44_e23 |
215 |
o added tangent linear setup to verification/carbon/ |
merged patches between release1_p1 and release1_p2: |
|
o added adjoint sensitivity setup to verification/global_ocean... |
|
|
o moved packages_init_variables before convective_adjustment |
|
|
o output of verif. which use GMRedi brought up-to-date with |
|
|
MAIN trunk (those were re-created after changes in GMRedi) |
|
|
o included CPP option SINGLE_LAYER_MODE |
|
|
to configure barotropic setup (Martin Losch). |
|
|
o Removed auxiliary variable in impldiff |
|
|
(caused numerical differences with MAIN trunk routine). |
|
|
o Introduced CPP option ALLOW_AUTODIFF_MONITOR to |
|
|
disable adcommon.h/g_common.h by default. |
|
|
o Cleaned exchange calls in cg2d. |
|
|
|
|
|
release1_p2 |
|
216 |
o fixed array boundary problem for adjoint |
o fixed array boundary problem for adjoint |
217 |
(related to partial recomputation within subroutine |
(related to partial recomputation within subroutine |
218 |
related to dynamics -> calc_viscosity -> mom_... |
related to dynamics -> calc_viscosity -> mom_... |
220 |
o prepare (but not complete) recomputation problem in impldiff |
o prepare (but not complete) recomputation problem in impldiff |
221 |
o fixed problem for adjoint of global_max |
o fixed problem for adjoint of global_max |
222 |
o switched off sponge layer code in external_forcing |
o switched off sponge layer code in external_forcing |
223 |
|
Bug fix in ctrl_mask_set_yz (spotted by G. Gebbie) |
224 |
|
|
225 |
|
ecco_c44_e22 |
226 |
|
o Finally, the merges with the Scripps cost function terms: |
227 |
|
Various additions, modifications, quality checks. |
228 |
|
o Plus: Included switch ALLOW_AUTODIFF_MONITOR to handle |
229 |
|
monitoring of adjoint state. |
230 |
|
o Plus: removed comm_stats, ini_communication_patterns from diff. list |
231 |
|
(no more need to "keep" MPI headers for adjoint generation) |
232 |
|
o Plus: exf changes: |
233 |
|
o added run-off routine |
234 |
|
o additional storing needed for ustress,vstess |
235 |
|
because of state-dependent quality check |
236 |
|
o added missing initialisations |
237 |
|
o Plus: genmake with MAKEDEPEND variable was still missing in ecco-branch |
238 |
|
NB: - verification/natl_box_adjoint/ 240 steps run |
239 |
|
* looks ok under Linux (twain.lcs.mit.edu) |
240 |
|
* looks ok on MIT SUN's (fjord.mit.edu) |
241 |
|
* NaN in reverse mode on Scripps SUN's at iter=214 |
242 |
|
* NaN in reverse mode on IBM SP3 (BlueHorizon) at iter=204 |
243 |
|
- 1 year forward int. global 2x2 deg. produces similar cost functions |
244 |
|
* all cost terms of new code are slightly smaller (0.5 to 3%). |
245 |
|
|
246 |
|
ecco_c44_e21 |
247 |
|
o reversed order of ctrl/cost initialisation |
248 |
|
o gmredi_slope_psi.F: Inconsistent types in MIN/MAX expressions |
249 |
|
o modified to enable ctrl field swapping |
250 |
|
o bug fix for global sum of ncvarlength in ctrl_init |
251 |
|
o no more call to ctrl_unpack from ctrl_init |
252 |
|
o no more write of ctrl vector in ctrl_pack |
253 |
|
however, init. of control vector from ctrl_unpack |
254 |
|
o various bug fixes & consistent nobcs handling in |
255 |
|
ctrl_pack/unpack routines (by G. Gebbie) |
256 |
|
o optim package: Added obcs control part for lsopt I/O |
257 |
|
(by G. Gebbie) |
258 |
|
o missing call to obcs_diags |
259 |
|
o moved calls in packages_init_fixed to packages_readparms |
260 |
|
because some quantities are needed earlier. |
261 |
|
|
262 |
|
ecco_c44_e20 |
263 |
|
o before major merges from the Scripps setup |
264 |
|
here are two small updates: |
265 |
|
- field swapping for surface flux controls enabled |
266 |
|
(remains t.b.d. for atmos. state plus obcs) |
267 |
|
- bug fix in active_file.F for xz-/yz- routines |
268 |
|
- added missing xz-/yz- routines to active_file_ad.F |
269 |
|
|
270 |
release1_p1 |
ecco_c44_e19 |
271 |
o merged necessary patches: |
o merged necessary patches: |
272 |
- AD-related changes for GMRedi |
- AD-related changes for GMRedi |
273 |
- fixes i KPP (delZ -> drF) |
- fixes i KPP (delZ -> drF) |
274 |
- hook to OBCS songe layer code in external_forcing |
- hook to OBCS songe layer code in external_forcing |
275 |
- errorMessageUnit non-zero in eeboot.F |
- errorMessageUnit non-zero in eeboot.F |
276 |
- modified test cost function and carbon verif. |
- included GMREDI, EXACT_CONSERV to natl_box_adjoint verif. |
|
|
|
|
release1_chkpt44d_post = chkpt44d_post |
|
277 |
|
|
278 |
chkpt44d_pre,post |
ecco_c44_e18 |
279 |
o added missing EXCLUDE_MONITOR flags |
o sponge layer code for obcs was missing |
280 |
o changed "e" to "_d" in gmredi_slope_limit, gmredi_slope_psi |
- added hook in external_forcing.F |
281 |
(incompatible typ in MIN/MAX expressions caused problems |
- added storing for TAF |
282 |
on IBM SP3) |
o new routine cost_init_barfiles to initialise |
283 |
o in genmake added variable MAKEDEPEND |
adjoint fields adtbar, adsbar, adpsbar |
284 |
plus resetting for case SunOS |
(needed in conjunction with corrected ladinit for active I/O) |
285 |
o added timer_stats.c routine for IBM SP3 |
o cleaned up and corrected order in packages_readparms, ecco_readparms |
286 |
o removed variables in dynamics |
o corrected and simplified weights for obcs controls |
287 |
|
- weight fields simplified |
288 |
o real fresh water flux implemented with non-linear free-surface. |
- bug fix in ctrl_pack |
289 |
|
- no more write of control vector in ctrl_pack |
290 |
chkpt44c_pre,post |
o i/j index bug fixes in cost_ssh_mean |
291 |
o few fix (mask in shap_s2, EmPmR in external_field_load, |
|
292 |
USE_NATURAL_BCS in solve_for_P); |
ecco_c44_e17 |
293 |
o add arguments myIter & myTime to S/R obcs_calc & solve_for_P |
o bug fixes in cost package |
294 |
|
* wrong identifier for active I/O of monthly mean psbar |
295 |
checkpoint44b_pre/_post |
(cost_averagesfields, cost_ssh) xx_psbar_dummy |
296 |
o merge of relevant stuff from the ecco-branch: |
* cleaned all identifier for active I/O of monthly fields |
297 |
|
* wrong ladinit in various cost routines |
298 |
|
o merge with interval (chkpt44c_post - chkpt44d_post) |
299 |
|
* added missing EXCLUDE_MONITOR flags |
300 |
|
* changed "e" to "_d" in gmredi_slope_limit, gmredi_slope_psi |
301 |
|
(incompatible typ in MIN/MAX expressions caused problems |
302 |
|
on IBM SP3) |
303 |
|
* in genmake added variable MAKEDEPEND |
304 |
|
plus resetting for case SunOS |
305 |
|
added timer_stats.c routine for IBM SP3 |
306 |
|
* removed variables in dynamics |
307 |
|
|
308 |
|
ecco_c44_e16 |
309 |
|
o new ECCO checkpoint, the first since ecco_c43_e15 |
310 |
|
which is branched from MITgcm. |
311 |
|
All changes were merged in between (see ecco-branch-mod? tags). |
312 |
|
o last changes: |
313 |
|
* assuming path for TAF and TAMC are known (e.g. in $path) |
314 |
|
* genmake: |
315 |
|
- added tools/xmakedepend to circumvene shortcoming of |
316 |
|
SUN's makedepend (size too small) |
317 |
|
- invoking ../tools/xmakedepend instead of makedepend |
318 |
|
- changed compiler flags for case SunOS |
319 |
|
|
320 |
|
ecco-branch-mod5 |
321 |
|
o mod's to reproduce open boundary run |
322 |
|
o enable field swapping for OB's and climatology |
323 |
|
o update store directives for OB control |
324 |
|
o added option ALLOW_OBCS_PRESCRIBE |
325 |
|
o call of getobcs now out of obcs_calc (param. list modified) |
326 |
|
|
327 |
|
ecco-branch-mod4 |
328 |
|
o this is the second part of changes necessary to reconcile |
329 |
|
the natl_box_adjoint verification (with and without grdchk). |
330 |
|
>>>!!! CHANGES BETWEEN THE TAGS ecco-branch-mod3 !!!<<< |
331 |
|
>>>!!! AND ecco-branch-mod4 !!!<<< |
332 |
|
>>>!!! SHOULD BE MERGED INTO THE MAIN TRUNK !!!<<< |
333 |
- genmake: removed $S64 overwrite for case SunOS |
- genmake: removed $S64 overwrite for case SunOS |
334 |
- pkg/exf: update and corrections for field swapping and obcs |
- pkg/exf: update and corrections for field swapping and obcs |
335 |
- pkg/ecco: parameter lists for the_model_main, the_main_loop |
- pkg/ecco: parameter lists for the_model_main, the_main_loop |
337 |
- pkg/autodiff: added flow directives for obcs, mdsio_gl_slice |
- pkg/autodiff: added flow directives for obcs, mdsio_gl_slice |
338 |
updated checkpointing_lev... lists for obcs |
updated checkpointing_lev... lists for obcs |
339 |
- model/src: minor changes in forward_step, plot_field |
- model/src: minor changes in forward_step, plot_field |
|
added directive for divided adjoint in the_main_loop |
|
340 |
- pkg/mdsio: added mdsio_gl_slice |
- pkg/mdsio: added mdsio_gl_slice |
341 |
|
|
342 |
o check parameters & config (chkpt44a_pre,post) |
ecco-branch-mod3 |
343 |
o OBC and NonLin_FrSurf. |
o this is the first part of changes necessary to reconcile |
344 |
|
natl_box_adjoint verification (with and without grdchk). |
345 |
|
This set of changes is relevant for ECCO only. |
346 |
|
The next set/tag of changes (ecco-branch-mod4) |
347 |
|
should be merged into the main trunk. |
348 |
|
- update some options |
349 |
|
- parameter lists of ecco_the_model_main and ecco_the_main_loop |
350 |
|
differed from those of the MITgcm counterparts. |
351 |
|
|
352 |
|
ecco-branch-mod2 |
353 |
|
o update ecco-branch-mod1 to checkpoint44 |
354 |
|
o exclude from updates: |
355 |
|
- verification/ |
356 |
|
- adjoint/ |
357 |
|
- pkg/cost |
358 |
|
- pkg/ctrl |
359 |
|
- pkg/ecco |
360 |
|
- pkg/obcs |
361 |
|
- model/src/the_model_main.F |
362 |
|
- model/src/the_main_loop.F |
363 |
|
|
364 |
checkpoint44 |
checkpoint44 |
365 |
|
o apply Shapiro filter on T&S before calc_PhiHyd (stagger time stepping) |
366 |
o fix bug in mom_vi_del2uv |
o fix bug in mom_vi_del2uv |
|
o select when filters are applied ; add options to zonal_filter (data.zonfilt) |
|
|
o gmredi: fix Pb in the adiabatic form ; add options (.e.g. Bolus advection) |
|
|
o update AIM experiments (NCEP input files) |
|
|
o improve and extend diagnostics (Monitor, TimeAve with NonLin-FrSurf) |
|
367 |
o added some stuff for AD |
o added some stuff for AD |
368 |
|
|
369 |
|
ecco-branch-mod1 |
370 |
|
o based on c43 |
371 |
|
o replaced packages cost, ctrl, ecco, obcs |
372 |
|
o removed MITgcm verif.'s, added ECCO verif.'s |
373 |
|
o added lsopt, optim for line search optimization. |
374 |
|
|
375 |
checkpoint43 |
checkpoint43 |
376 |
o added two new advection tests |
o added two new advection tests |
377 |
o added global 4x4 degree ocean test |
o added global 4x4 degree ocean test |