2 |
|
|
3 |
Notes on tags used in MITgcmUV |
Notes on tags used in MITgcmUV |
4 |
============================== |
============================== |
5 |
|
|
6 |
|
checkpoint46a_post |
7 |
|
o Added (atmospheric) pressure loading |
8 |
|
- new field in FFIELDS.h, etc... |
9 |
|
- new cpp flag ATMOSPHERIC_LOADING |
10 |
|
o Changed hFacC to _hFacC in calc_phi_hyd.F |
11 |
|
o Added SHORTWAVE_HEATING to some files for consistency |
12 |
|
checkpoint46a_pre |
13 |
|
|
14 |
|
checkpoint46 |
15 |
|
further merges from release1_p5 |
16 |
|
o new ctrl package |
17 |
|
- adopted from ECCO environment to enable optimization |
18 |
|
- added Eliassen Palm fluxes to controls |
19 |
|
o added Eliassen Palm flux controls to gradient check package |
20 |
|
o cost package |
21 |
|
- Compatible with new ctrl package |
22 |
|
- added Eliassen Palm cost hooks |
23 |
|
- modif's of existing cost functions |
24 |
|
o Adjoint-related bug fixes in kpp: |
25 |
|
- kpp_calc: sore of kpphbl avoids recomputation/call to S/R kppmix |
26 |
|
- kpp_routines: store of Rib avoids partial recomputation bug of TAF. |
27 |
|
o autodiff package |
28 |
|
- flow directives: |
29 |
|
+ new for S/R do_the_model_io |
30 |
|
+ enhanced for S/R checkpoint |
31 |
|
+ new gmredi.flow |
32 |
|
- Introduced CPP option ALLOW_AUTODIFF_MONITOR to |
33 |
|
disable adcommon.h/g_common.h by default. |
34 |
|
- fixed problem for adjoint of global_max |
35 |
|
o modified verification/carbon/ |
36 |
|
o New setup for adjoint of global_ocean |
37 |
|
o added hooks for Eliassen Palm fluxes (dynamics.F, ini_fields.F) |
38 |
|
o cleaned TAF keys iikey, idkey (dynamics.F, thermodynamics.F) |
39 |
|
o incorporated hooks for sponge layer code (external_forcing.F) |
40 |
|
|
41 |
|
checkpoint45d_post |
42 |
|
o import 2 fixes from release1_p5 (CG2D.h & packages_readparms.F) |
43 |
|
o use recip_dx*,recip_dy* instead of /dx*,/dy* in orlanski_E,W,N,S |
44 |
|
(affects plume_on_slope "cg2d" output) |
45 |
|
o add new parameter deltaTfreesurf for free-surface time stepping |
46 |
|
o exchange gT,gS when using staggerTimeStep & GM & Oly<4 |
47 |
|
o NONLIN_FRSURF: rescale gNm1 to get a better conservation with A-B |
48 |
|
o change phiMin in exp1 (to agree with documentation) |
49 |
|
o add new flags: |
50 |
|
- T,S forcing inside/outside Adams-Bashforth |
51 |
|
- turn off temp,salt advection and/or forcing |
52 |
|
- for each tracer, define internal flag for multiDimAdvection & A-B |
53 |
|
o monitor: change definition of KE to check conservation of Energy |
54 |
|
|
55 |
|
checkpoint45c_post |
56 |
|
o Removed f90 comments ("!") |
57 |
|
o Modified initialisations to enable exf package for MITgcm |
58 |
|
(without invoking ECCO_PACKAGE). |
59 |
|
o modifications to gradient check package (Martin Losch) |
60 |
|
- enable centered differences |
61 |
|
- modified format of standard output |
62 |
|
|
63 |
|
checkpoint45b_post |
64 |
|
o cleaned exchange calls in cg2d |
65 |
|
o Included CPP option ALLOW_AUTODIFF_MONITOR in forward_step |
66 |
|
o included CPP option SINGLE_LAYER_MODE in thermodynamics |
67 |
|
to configure barotropic setup (Martin Losch) |
68 |
|
o moved some initialisations within bi,bj loop in dynamics |
69 |
|
(TAF flow dependency) |
70 |
|
o in initialise_varia, moved call packages_init_variables |
71 |
|
before convective_adjustment_init |
72 |
|
(corresponding adjustment of TAF tape initialisation). |
73 |
|
o in calc_diffusivity/calc_viscosity extended domain to full overlap. |
74 |
|
|
75 |
|
checkpoint45a_post |
76 |
|
o compute Rho over the whole domain (extend iMin,jMin indices) |
77 |
|
o specific exchange for X-slice domain (case Ny=1) |
78 |
|
o argument futureIter of S/R obcs_calc missing in exp4 & internal_wave /code |
79 |
|
o solve_for_pressure : set cg2d_x to zero where OBCS are applied |
80 |
|
|
81 |
|
checkpoint44h_post = checkpoint45 |
82 |
|
o Fix to restore differentiability: |
83 |
|
place at which packages_init_variables is called in |
84 |
|
initialise_varia is crucial for reverse control flow. |
85 |
|
o modifications in GMRedi to enable stable adjoint |
86 |
|
(corresponding modif's in thermodynamics, dynamics) |
87 |
|
o added missing hook for sponge layer code in external_forcing |
88 |
|
o modified test cost functions |
89 |
|
o storing in gad_advection (plus removal of write(0,*)! |
90 |
|
o missing headers TR1.h added in convective_adjustment routines |
91 |
|
o errorMessageUnit set non-equal zero in eeboot |
92 |
|
(conflict on some platforms, e.g. IBM SP3) |
93 |
|
o modified carbon verif. |
94 |
|
|
95 |
|
checkpoint44h_pre |
96 |
|
o vertical grid option: allow to put the Interface at the middle between |
97 |
|
2 cell-centers ; replace delZ in pkg/kpp by drF. |
98 |
|
o GM Advective form: Tracers are advected using the residual transport (= |
99 |
|
Euler+GM-bolus); set param GM_AdvSeparate=T to return to previous form. |
100 |
|
|
101 |
|
checkpoint44g_post |
102 |
|
o fix surface correction term in multiDimAdvection (affects plume_on_slope |
103 |
|
results) ; ifdef missing in thermodynamics.F |
104 |
|
|
105 |
|
checkpoint44f_pre,checkpoint44f_post |
106 |
|
o added PTRACERS package |
107 |
|
This allows an arbitrary number of passive tracers to be integrated |
108 |
|
forward simultaneously with the dynamicaly model. |
109 |
|
+ Implemented so far: |
110 |
|
- basic forward algorithm (time-stepping, advection, diffusion, convection) |
111 |
|
- I/O and checkpointing |
112 |
|
- GM/Redi *but* using the GM/Redi coefficient of Salt |
113 |
|
+ Not implemented so far: |
114 |
|
- KPP |
115 |
|
- OBCS |
116 |
|
+ No specific example supplied (yet) but global_ocean.90x40x15 has the |
117 |
|
necessary data.ptracer file. Simply use -enable=ptracers and uncomment |
118 |
|
line in data.pkg. |
119 |
|
+ This package is disabled by default since it increases storage. |
120 |
|
|
121 |
|
|
122 |
|
o testscript: added -cleanup option |
123 |
|
o Shap_filt: a) new shap_filt S/R to use no-slip BC with S2 filter |
124 |
|
b) enable to filter 2D fields. |
125 |
|
|
126 |
|
checkpoint44e_pre,post |
127 |
|
o re-wrote convective adjustment to use pre-calculated weights |
128 |
|
Affects: model/src/convective_adjustment.F, convective_adjustment_ini.F |
129 |
|
Adds: model/src/convective_weights.F, convectively_mixtracer.F |
130 |
|
Original calls to convect() still in place. |
131 |
|
|
132 |
|
chkpt44d_pre,post |
133 |
|
o added missing EXCLUDE_MONITOR flags |
134 |
|
o changed "e" to "_d" in gmredi_slope_limit, gmredi_slope_psi |
135 |
|
(incompatible typ in MIN/MAX expressions caused problems |
136 |
|
on IBM SP3) |
137 |
|
o in genmake added variable MAKEDEPEND |
138 |
|
plus resetting for case SunOS |
139 |
|
o added timer_stats.c routine for IBM SP3 |
140 |
|
o removed variables in dynamics |
141 |
|
!!! I forgot to create tag chkpt44d_pre !!! |
142 |
|
!!! But chkpt44d_pre = chkpt44c_post !!! |
143 |
|
|
144 |
|
o real fresh water flux implemented with non-linear free-surface. |
145 |
|
|
146 |
|
chkpt44c_pre,post |
147 |
|
o few fix (mask in shap_s2, EmPmR in external_field_load, |
148 |
|
USE_NATURAL_BCS in solve_for_P); |
149 |
|
o add arguments myIter & myTime to S/R obcs_calc & solve_for_P |
150 |
|
|
151 |
|
checkpoint44b_pre/_post |
152 |
|
o merge of relevant stuff from the ecco-branch: |
153 |
|
- genmake: removed $S64 overwrite for case SunOS |
154 |
|
- pkg/exf: update and corrections for field swapping and obcs |
155 |
|
- pkg/ecco: parameter lists for the_model_main, the_main_loop |
156 |
|
harmonized between ECCO and MITgcm |
157 |
|
- pkg/autodiff: added flow directives for obcs, mdsio_gl_slice |
158 |
|
updated checkpointing_lev... lists for obcs |
159 |
|
- model/src: minor changes in forward_step, plot_field |
160 |
|
added directive for divided adjoint in the_main_loop |
161 |
|
- pkg/mdsio: added mdsio_gl_slice |
162 |
|
|
163 |
|
o check parameters & config (chkpt44a_pre,post) |
164 |
|
o OBC and NonLin_FrSurf. |
165 |
|
|
166 |
|
checkpoint44 |
167 |
|
o fix bug in mom_vi_del2uv |
168 |
|
o select when filters are applied ; add options to zonal_filter (data.zonfilt) |
169 |
|
o gmredi: fix Pb in the adiabatic form ; add options (.e.g. Bolus advection) |
170 |
|
o update AIM experiments (NCEP input files) |
171 |
|
o improve and extend diagnostics (Monitor, TimeAve with NonLin-FrSurf) |
172 |
|
o added some stuff for AD |
173 |
|
|
174 |
|
checkpoint43 |
175 |
|
o added two new advection tests |
176 |
|
o added global 4x4 degree ocean test |
177 |
|
o added test of H&S94 with free surface on cube |
178 |
|
o some final mods for gradient check, adjoint,... |
179 |
|
|
180 |
|
checkpoint42 |
181 |
|
Fixed AD-related problems: |
182 |
|
o Store directives up-to-date with re-arranged Adams-Bashforth |
183 |
|
(mainly thermodynamics.F) |
184 |
|
o New store directives for multi-dim. advection schemes |
185 |
|
* new CPP flag ALLOW_MULTI_DIM_ADVECTION |
186 |
|
* new common block and key passkey |
187 |
|
(mainly gad_advection.F) |
188 |
|
o Modified store directives for split of dynamics/thermodynamics |
189 |
|
for the case ALLOW_KPP |
190 |
|
o Cleaned argument list for timestep_tracer.F |
191 |
|
|
192 |
|
checkpoint41 |
193 |
|
o re-formatted/added comments in prologues of key subroutines |
194 |
|
to work with protex |
195 |
|
|
196 |
|
checkpoint40 |
197 |
|
o moved calc_gs, calc_gt and calc_tr1 back to model/src |
198 |
|
o added FLT (floats) package from Arne along with demo. experiment |
199 |
|
o re-arranged Adams-Bashforth scheme for tracers to allow easier |
200 |
|
implementation of forward-in-time forcing terms |
201 |
|
o more mods for non-linear free-surface including new variable |
202 |
|
o modified multi-dim method to work on cube |
203 |
|
o bug fix in KPP (from DM) |
204 |
|
|
205 |
|
checkpoint40pre9 |
206 |
|
o added DST advection schemes |
207 |
|
o fix MPI bug on SGI's |
208 |
|
o modifications for non-linear free-surface |
209 |
|
o fixed conditionals for fixed flow/passive tracer only mode |
210 |
|
o added tension/strain form of stress tensor |
211 |
|
o changes to commenting of arguments for TAF/TAMC |
212 |
|
o added/updated many matlab scripts |
213 |
|
o added multi-dimensional advection method for tracers |
214 |
|
|
215 |
|
checkpoint40pre8 |
216 |
|
o new package rw/ |
217 |
|
Added method for dumping intermediate local arrays: |
218 |
|
mdsio_writetile - same as mdsio_writefield except works from inside |
219 |
|
bi,bj loop |
220 |
|
mdsio_writelocal - same as mdsio_writetile except works for local |
221 |
|
arrays |
222 |
|
write_local_r? - higher-level wrapper for mdsio_writelocal |
223 |
|
Controlled by diagFreq. Defaults to zero (ie. no dumps) |
224 |
|
Example given at end of mom_vecinv.F that dumps some local arrays. |
225 |
|
o replaced calc_mom_rhs.F by mom_fluxform.F/mom_vecinc.F to |
226 |
|
control flux-form versus vector invariant code. |
227 |
|
Switch is runtime flag vectorInvariantMomentum (defaults to false) |
228 |
|
or genmake -DISABLE ... |
229 |
|
o TAMC-related update to cope with mom_fluxform.F/mom_vecinc.F |
230 |
|
(additional storing is now needed for case vecinv; |
231 |
|
currently nothing implemented to switch this storing off |
232 |
|
in case fluxform is used since no CPP flag available). |
233 |
|
|
234 |
|
checkpoint40pre7 |
235 |
|
o bug fix in dynamics.F |
236 |
|
calc_viscosity called before k=1,Nr loop. |
237 |
|
o added some #ifdef's ALLOW_PASSIVE_TRACER |
238 |
|
o cleaned initialisations in ctrl_map_ini |
239 |
|
related to TAF |
240 |
|
|
241 |
|
checkpoint40pre6 |
242 |
|
o split dynamics into thermodynamics & dynamics |
243 |
|
o corresponding split of calc_diffusivity into |
244 |
|
calc_diffusivity, calc_viscosity |
245 |
|
(kpp_calc_diff, kpp_calc_visc). |
246 |
|
o Added diffkr, kapgm to control variables |
247 |
|
o bug fix in gmredi_slope_limit for ldd97 |
248 |
|
|
249 |
|
checkpoint40pre5 |
250 |
|
o proper initialization of Adams-Bashforth time stepping |
251 |
|
+ uses forward step for very first step of model (iter=0) |
252 |
|
o re-generated *all* output since almost all output is modified by this |
253 |
|
change (sorry but we've been putting this one off for about 2 years) |
254 |
|
+ natl_box, aim.5l_Equatorial_Channel and aim.5l_zon-ave were |
255 |
|
*not* regenerated since they are already out of date and I don't |
256 |
|
want to create the impression that they are suddenly working. |
257 |
|
o removed KAP stuff for DEC's |
258 |
|
o modified/added many .cvsignore files in */code and */input |
259 |
|
o added new expt "solid-body.cs-32x32x1" |
260 |
|
+ this uses an alternative form of ini_curvilinear_grid.F which |
261 |
|
is not necessarily any better than the current one... |
262 |
|
|
263 |
|
checkpoint40pre4 |
264 |
|
o added re-initialisations of utrans,vtrans,rtrans in |
265 |
|
calc_common_factors which are crucial for TAF. |
266 |
|
o extended iMin,jMin range for calc_common_factors, |
267 |
|
calc_diffusivity |
268 |
|
o added option ALLOW_PASSIVE_TRACER |
269 |
|
o a few minor updates and fixes |
270 |
|
|
271 |
|
checkpoint40pre3 |
272 |
|
o created calc_gtr1.F for gad package |
273 |
|
o split the_main_loop and re-introduced forward_step |
274 |
|
o some recomputations of nIter0 for TAMC |
275 |
|
|
276 |
|
checkpoint40pre2 |
277 |
|
o Updated to latest branch of AJA, JMC |
278 |
|
o Added gradient check package |
279 |
|
o Added routines for passive tracer handling |
280 |
|
o Added carbon sequestration sensitivity experiment |
281 |
|
|
282 |
checkpoint39 |
checkpoint39 |
283 |
o Modifications and fixes to support TAMC |
o Modifications and fixes to support TAMC |
284 |
(mainly missing or wrong directives) |
(mainly missing or wrong directives) |