| 2 |
|
|
| 3 |
Notes on tags used in MITgcmUV |
Notes on tags used in MITgcmUV |
| 4 |
============================== |
============================== |
| 5 |
|
|
| 6 |
|
checkpoint46f_post |
| 7 |
|
o Enable tangent linear (forward mode) gradient checks: |
| 8 |
|
- extended active file handling to g_... files |
| 9 |
|
- added TANGENT_SIMULATION to theSimulationMode |
| 10 |
|
- extended grdchk package accordingly |
| 11 |
|
o added tangent linear setup in verification/carbon/code_ftl |
| 12 |
|
o added adjoint setup in verification/front_relax/code_ad |
| 13 |
|
|
| 14 |
|
checkpoint46e_post |
| 15 |
|
o Added new equation of state -> MDJWF |
| 16 |
|
- EOS of McDougall et al., 2002, JAOT, submitted |
| 17 |
|
- caveat: the equation of state is only valid for a smaller (more |
| 18 |
|
realistic?) range of values than JMD95P/Z and UNESCO |
| 19 |
|
- added masks to the calculation of pressure in store_pressure |
| 20 |
|
- added more check values for density in check_eos (ini_eos.F), some of |
| 21 |
|
the old check values are out of the range of the MDJWF-eos, so don't |
| 22 |
|
expect perfect matches for those |
| 23 |
|
checkpoint46e_pre |
| 24 |
|
|
| 25 |
|
checkpoint46d_post |
| 26 |
|
o fixed store_pressure to work with both buoyancy relation = 'OCEANIC' and |
| 27 |
|
'OCEANICP', also initialised field pressure correctly in ini_eos in the |
| 28 |
|
case of pressure coordinates. eosType='JMD95Z' in |
| 29 |
|
combination with buoyancyRelation='OCEANICP' now causes an error. |
| 30 |
|
o Changed p = pressure(i,j,k,bi,bj) to p = pressure(i,j,kRef,bi,bj) |
| 31 |
|
in find_alpha/beta. |
| 32 |
|
checkpoint46d_pre |
| 33 |
|
|
| 34 |
|
o Changed p = pressure(i,j,k,bi,bj) to p = pressure(i,j,kRef,bi,bj) |
| 35 |
|
so that JMD95Z and JMD95P give approptiate static stability in find_rho. |
| 36 |
|
|
| 37 |
|
checkpoint46c_post |
| 38 |
|
o Changes necessary for ocean in p-coordinates |
| 39 |
|
- Added new buoyancy relation = 'OCEANICP' |
| 40 |
|
- Added new parameters = gravitySign (this used to be contained inside |
| 41 |
|
the factor dRdZ which I added when we first switched to R coordinates). |
| 42 |
|
X GM/Redi is not compatible (yet) |
| 43 |
|
X bottom drag and no-slip need to be debugged. |
| 44 |
|
checkpoint46c_pre |
| 45 |
|
|
| 46 |
|
checkpoint46b_post |
| 47 |
|
o Added new equation of state -> JMD95Z and JMD95P |
| 48 |
|
- EOS of Jackett and McDougall, 1995, JAOT |
| 49 |
|
- moved all EOS parameters into EOS.h |
| 50 |
|
- new routines ini_eos.F, store_pressure.F |
| 51 |
|
o Added UNESCO EOS, but not recommended because it requires |
| 52 |
|
in-situ temperature (see JMD95) |
| 53 |
|
o Modified formatting for knudsen2.f in utils/knudsen2 and added |
| 54 |
|
unesco.f to be used with POLY3 |
| 55 |
|
checkpoint46b_pre |
| 56 |
|
|
| 57 |
|
checkpoint46a_post |
| 58 |
|
o Added (atmospheric) pressure loading |
| 59 |
|
- new field in FFIELDS.h, etc... |
| 60 |
|
- new cpp flag ATMOSPHERIC_LOADING |
| 61 |
|
o Changed hFacC to _hFacC in calc_phi_hyd.F |
| 62 |
|
o Added SHORTWAVE_HEATING to some files for consistency |
| 63 |
|
checkpoint46a_pre |
| 64 |
|
|
| 65 |
|
checkpoint46 |
| 66 |
|
further merges from release1_p5 |
| 67 |
|
o new ctrl package |
| 68 |
|
- adopted from ECCO environment to enable optimization |
| 69 |
|
- added Eliassen Palm fluxes to controls |
| 70 |
|
o added Eliassen Palm flux controls to gradient check package |
| 71 |
|
o cost package |
| 72 |
|
- Compatible with new ctrl package |
| 73 |
|
- added Eliassen Palm cost hooks |
| 74 |
|
- modif's of existing cost functions |
| 75 |
|
o Adjoint-related bug fixes in kpp: |
| 76 |
|
- kpp_calc: sore of kpphbl avoids recomputation/call to S/R kppmix |
| 77 |
|
- kpp_routines: store of Rib avoids partial recomputation bug of TAF. |
| 78 |
|
o autodiff package |
| 79 |
|
- flow directives: |
| 80 |
|
+ new for S/R do_the_model_io |
| 81 |
|
+ enhanced for S/R checkpoint |
| 82 |
|
+ new gmredi.flow |
| 83 |
|
- Introduced CPP option ALLOW_AUTODIFF_MONITOR to |
| 84 |
|
disable adcommon.h/g_common.h by default. |
| 85 |
|
- fixed problem for adjoint of global_max |
| 86 |
|
o modified verification/carbon/ |
| 87 |
|
o New setup for adjoint of global_ocean |
| 88 |
|
o added hooks for Eliassen Palm fluxes (dynamics.F, ini_fields.F) |
| 89 |
|
o cleaned TAF keys iikey, idkey (dynamics.F, thermodynamics.F) |
| 90 |
|
o incorporated hooks for sponge layer code (external_forcing.F) |
| 91 |
|
|
| 92 |
|
checkpoint45d_post |
| 93 |
|
o import 2 fixes from release1_p5 (CG2D.h & packages_readparms.F) |
| 94 |
|
o use recip_dx*,recip_dy* instead of /dx*,/dy* in orlanski_E,W,N,S |
| 95 |
|
(affects plume_on_slope "cg2d" output) |
| 96 |
|
o add new parameter deltaTfreesurf for free-surface time stepping |
| 97 |
|
o exchange gT,gS when using staggerTimeStep & GM & Oly<4 |
| 98 |
|
o NONLIN_FRSURF: rescale gNm1 to get a better conservation with A-B |
| 99 |
|
o change phiMin in exp1 (to agree with documentation) |
| 100 |
|
o add new flags: |
| 101 |
|
- T,S forcing inside/outside Adams-Bashforth |
| 102 |
|
- turn off temp,salt advection and/or forcing |
| 103 |
|
- for each tracer, define internal flag for multiDimAdvection & A-B |
| 104 |
|
o monitor: change definition of KE to check conservation of Energy |
| 105 |
|
|
| 106 |
|
checkpoint45c_post |
| 107 |
|
o Removed f90 comments ("!") |
| 108 |
|
o Modified initialisations to enable exf package for MITgcm |
| 109 |
|
(without invoking ECCO_PACKAGE). |
| 110 |
|
o modifications to gradient check package (Martin Losch) |
| 111 |
|
- enable centered differences |
| 112 |
|
- modified format of standard output |
| 113 |
|
|
| 114 |
|
checkpoint45b_post |
| 115 |
|
o cleaned exchange calls in cg2d |
| 116 |
|
o Included CPP option ALLOW_AUTODIFF_MONITOR in forward_step |
| 117 |
|
o included CPP option SINGLE_LAYER_MODE in thermodynamics |
| 118 |
|
to configure barotropic setup (Martin Losch) |
| 119 |
|
o moved some initialisations within bi,bj loop in dynamics |
| 120 |
|
(TAF flow dependency) |
| 121 |
|
o in initialise_varia, moved call packages_init_variables |
| 122 |
|
before convective_adjustment_init |
| 123 |
|
(corresponding adjustment of TAF tape initialisation). |
| 124 |
|
o in calc_diffusivity/calc_viscosity extended domain to full overlap. |
| 125 |
|
|
| 126 |
|
checkpoint45a_post |
| 127 |
|
o compute Rho over the whole domain (extend iMin,jMin indices) |
| 128 |
|
o specific exchange for X-slice domain (case Ny=1) |
| 129 |
|
o argument futureIter of S/R obcs_calc missing in exp4 & internal_wave /code |
| 130 |
|
o solve_for_pressure : set cg2d_x to zero where OBCS are applied |
| 131 |
|
|
| 132 |
|
checkpoint44h_post = checkpoint45 |
| 133 |
|
o Fix to restore differentiability: |
| 134 |
|
place at which packages_init_variables is called in |
| 135 |
|
initialise_varia is crucial for reverse control flow. |
| 136 |
|
o modifications in GMRedi to enable stable adjoint |
| 137 |
|
(corresponding modif's in thermodynamics, dynamics) |
| 138 |
|
o added missing hook for sponge layer code in external_forcing |
| 139 |
|
o modified test cost functions |
| 140 |
|
o storing in gad_advection (plus removal of write(0,*)! |
| 141 |
|
o missing headers TR1.h added in convective_adjustment routines |
| 142 |
|
o errorMessageUnit set non-equal zero in eeboot |
| 143 |
|
(conflict on some platforms, e.g. IBM SP3) |
| 144 |
|
o modified carbon verif. |
| 145 |
|
|
| 146 |
|
checkpoint44h_pre |
| 147 |
|
o vertical grid option: allow to put the Interface at the middle between |
| 148 |
|
2 cell-centers ; replace delZ in pkg/kpp by drF. |
| 149 |
|
o GM Advective form: Tracers are advected using the residual transport (= |
| 150 |
|
Euler+GM-bolus); set param GM_AdvSeparate=T to return to previous form. |
| 151 |
|
|
| 152 |
|
checkpoint44g_post |
| 153 |
|
o fix surface correction term in multiDimAdvection (affects plume_on_slope |
| 154 |
|
results) ; ifdef missing in thermodynamics.F |
| 155 |
|
|
| 156 |
|
checkpoint44f_pre,checkpoint44f_post |
| 157 |
|
o added PTRACERS package |
| 158 |
|
This allows an arbitrary number of passive tracers to be integrated |
| 159 |
|
forward simultaneously with the dynamicaly model. |
| 160 |
|
+ Implemented so far: |
| 161 |
|
- basic forward algorithm (time-stepping, advection, diffusion, convection) |
| 162 |
|
- I/O and checkpointing |
| 163 |
|
- GM/Redi *but* using the GM/Redi coefficient of Salt |
| 164 |
|
+ Not implemented so far: |
| 165 |
|
- KPP |
| 166 |
|
- OBCS |
| 167 |
|
+ No specific example supplied (yet) but global_ocean.90x40x15 has the |
| 168 |
|
necessary data.ptracer file. Simply use -enable=ptracers and uncomment |
| 169 |
|
line in data.pkg. |
| 170 |
|
+ This package is disabled by default since it increases storage. |
| 171 |
|
|
| 172 |
|
|
| 173 |
|
o testscript: added -cleanup option |
| 174 |
|
o Shap_filt: a) new shap_filt S/R to use no-slip BC with S2 filter |
| 175 |
|
b) enable to filter 2D fields. |
| 176 |
|
|
| 177 |
|
checkpoint44e_pre,post |
| 178 |
|
o re-wrote convective adjustment to use pre-calculated weights |
| 179 |
|
Affects: model/src/convective_adjustment.F, convective_adjustment_ini.F |
| 180 |
|
Adds: model/src/convective_weights.F, convectively_mixtracer.F |
| 181 |
|
Original calls to convect() still in place. |
| 182 |
|
|
| 183 |
|
chkpt44d_pre,post |
| 184 |
|
o added missing EXCLUDE_MONITOR flags |
| 185 |
|
o changed "e" to "_d" in gmredi_slope_limit, gmredi_slope_psi |
| 186 |
|
(incompatible typ in MIN/MAX expressions caused problems |
| 187 |
|
on IBM SP3) |
| 188 |
|
o in genmake added variable MAKEDEPEND |
| 189 |
|
plus resetting for case SunOS |
| 190 |
|
o added timer_stats.c routine for IBM SP3 |
| 191 |
|
o removed variables in dynamics |
| 192 |
|
!!! I forgot to create tag chkpt44d_pre !!! |
| 193 |
|
!!! But chkpt44d_pre = chkpt44c_post !!! |
| 194 |
|
|
| 195 |
|
o real fresh water flux implemented with non-linear free-surface. |
| 196 |
|
|
| 197 |
|
chkpt44c_pre,post |
| 198 |
|
o few fix (mask in shap_s2, EmPmR in external_field_load, |
| 199 |
|
USE_NATURAL_BCS in solve_for_P); |
| 200 |
|
o add arguments myIter & myTime to S/R obcs_calc & solve_for_P |
| 201 |
|
|
| 202 |
|
checkpoint44b_pre/_post |
| 203 |
|
o merge of relevant stuff from the ecco-branch: |
| 204 |
|
- genmake: removed $S64 overwrite for case SunOS |
| 205 |
|
- pkg/exf: update and corrections for field swapping and obcs |
| 206 |
|
- pkg/ecco: parameter lists for the_model_main, the_main_loop |
| 207 |
|
harmonized between ECCO and MITgcm |
| 208 |
|
- pkg/autodiff: added flow directives for obcs, mdsio_gl_slice |
| 209 |
|
updated checkpointing_lev... lists for obcs |
| 210 |
|
- model/src: minor changes in forward_step, plot_field |
| 211 |
|
added directive for divided adjoint in the_main_loop |
| 212 |
|
- pkg/mdsio: added mdsio_gl_slice |
| 213 |
|
|
| 214 |
|
o check parameters & config (chkpt44a_pre,post) |
| 215 |
|
o OBC and NonLin_FrSurf. |
| 216 |
|
|
| 217 |
|
checkpoint44 |
| 218 |
|
o fix bug in mom_vi_del2uv |
| 219 |
|
o select when filters are applied ; add options to zonal_filter (data.zonfilt) |
| 220 |
|
o gmredi: fix Pb in the adiabatic form ; add options (.e.g. Bolus advection) |
| 221 |
|
o update AIM experiments (NCEP input files) |
| 222 |
|
o improve and extend diagnostics (Monitor, TimeAve with NonLin-FrSurf) |
| 223 |
|
o added some stuff for AD |
| 224 |
|
|
| 225 |
|
checkpoint43 |
| 226 |
|
o added two new advection tests |
| 227 |
|
o added global 4x4 degree ocean test |
| 228 |
|
o added test of H&S94 with free surface on cube |
| 229 |
|
o some final mods for gradient check, adjoint,... |
| 230 |
|
|
| 231 |
|
checkpoint42 |
| 232 |
|
Fixed AD-related problems: |
| 233 |
|
o Store directives up-to-date with re-arranged Adams-Bashforth |
| 234 |
|
(mainly thermodynamics.F) |
| 235 |
|
o New store directives for multi-dim. advection schemes |
| 236 |
|
* new CPP flag ALLOW_MULTI_DIM_ADVECTION |
| 237 |
|
* new common block and key passkey |
| 238 |
|
(mainly gad_advection.F) |
| 239 |
|
o Modified store directives for split of dynamics/thermodynamics |
| 240 |
|
for the case ALLOW_KPP |
| 241 |
|
o Cleaned argument list for timestep_tracer.F |
| 242 |
|
|
| 243 |
|
checkpoint41 |
| 244 |
|
o re-formatted/added comments in prologues of key subroutines |
| 245 |
|
to work with protex |
| 246 |
|
|
| 247 |
|
checkpoint40 |
| 248 |
|
o moved calc_gs, calc_gt and calc_tr1 back to model/src |
| 249 |
|
o added FLT (floats) package from Arne along with demo. experiment |
| 250 |
|
o re-arranged Adams-Bashforth scheme for tracers to allow easier |
| 251 |
|
implementation of forward-in-time forcing terms |
| 252 |
|
o more mods for non-linear free-surface including new variable |
| 253 |
|
o modified multi-dim method to work on cube |
| 254 |
|
o bug fix in KPP (from DM) |
| 255 |
|
|
| 256 |
|
checkpoint40pre9 |
| 257 |
|
o added DST advection schemes |
| 258 |
|
o fix MPI bug on SGI's |
| 259 |
|
o modifications for non-linear free-surface |
| 260 |
|
o fixed conditionals for fixed flow/passive tracer only mode |
| 261 |
|
o added tension/strain form of stress tensor |
| 262 |
|
o changes to commenting of arguments for TAF/TAMC |
| 263 |
|
o added/updated many matlab scripts |
| 264 |
|
o added multi-dimensional advection method for tracers |
| 265 |
|
|
| 266 |
|
checkpoint40pre8 |
| 267 |
|
o new package rw/ |
| 268 |
|
Added method for dumping intermediate local arrays: |
| 269 |
|
mdsio_writetile - same as mdsio_writefield except works from inside |
| 270 |
|
bi,bj loop |
| 271 |
|
mdsio_writelocal - same as mdsio_writetile except works for local |
| 272 |
|
arrays |
| 273 |
|
write_local_r? - higher-level wrapper for mdsio_writelocal |
| 274 |
|
Controlled by diagFreq. Defaults to zero (ie. no dumps) |
| 275 |
|
Example given at end of mom_vecinv.F that dumps some local arrays. |
| 276 |
|
o replaced calc_mom_rhs.F by mom_fluxform.F/mom_vecinc.F to |
| 277 |
|
control flux-form versus vector invariant code. |
| 278 |
|
Switch is runtime flag vectorInvariantMomentum (defaults to false) |
| 279 |
|
or genmake -DISABLE ... |
| 280 |
|
o TAMC-related update to cope with mom_fluxform.F/mom_vecinc.F |
| 281 |
|
(additional storing is now needed for case vecinv; |
| 282 |
|
currently nothing implemented to switch this storing off |
| 283 |
|
in case fluxform is used since no CPP flag available). |
| 284 |
|
|
| 285 |
|
checkpoint40pre7 |
| 286 |
|
o bug fix in dynamics.F |
| 287 |
|
calc_viscosity called before k=1,Nr loop. |
| 288 |
|
o added some #ifdef's ALLOW_PASSIVE_TRACER |
| 289 |
|
o cleaned initialisations in ctrl_map_ini |
| 290 |
|
related to TAF |
| 291 |
|
|
| 292 |
|
checkpoint40pre6 |
| 293 |
|
o split dynamics into thermodynamics & dynamics |
| 294 |
|
o corresponding split of calc_diffusivity into |
| 295 |
|
calc_diffusivity, calc_viscosity |
| 296 |
|
(kpp_calc_diff, kpp_calc_visc). |
| 297 |
|
o Added diffkr, kapgm to control variables |
| 298 |
|
o bug fix in gmredi_slope_limit for ldd97 |
| 299 |
|
|
| 300 |
|
checkpoint40pre5 |
| 301 |
|
o proper initialization of Adams-Bashforth time stepping |
| 302 |
|
+ uses forward step for very first step of model (iter=0) |
| 303 |
|
o re-generated *all* output since almost all output is modified by this |
| 304 |
|
change (sorry but we've been putting this one off for about 2 years) |
| 305 |
|
+ natl_box, aim.5l_Equatorial_Channel and aim.5l_zon-ave were |
| 306 |
|
*not* regenerated since they are already out of date and I don't |
| 307 |
|
want to create the impression that they are suddenly working. |
| 308 |
|
o removed KAP stuff for DEC's |
| 309 |
|
o modified/added many .cvsignore files in */code and */input |
| 310 |
|
o added new expt "solid-body.cs-32x32x1" |
| 311 |
|
+ this uses an alternative form of ini_curvilinear_grid.F which |
| 312 |
|
is not necessarily any better than the current one... |
| 313 |
|
|
| 314 |
|
checkpoint40pre4 |
| 315 |
|
o added re-initialisations of utrans,vtrans,rtrans in |
| 316 |
|
calc_common_factors which are crucial for TAF. |
| 317 |
|
o extended iMin,jMin range for calc_common_factors, |
| 318 |
|
calc_diffusivity |
| 319 |
|
o added option ALLOW_PASSIVE_TRACER |
| 320 |
|
o a few minor updates and fixes |
| 321 |
|
|
| 322 |
|
checkpoint40pre3 |
| 323 |
|
o created calc_gtr1.F for gad package |
| 324 |
|
o split the_main_loop and re-introduced forward_step |
| 325 |
|
o some recomputations of nIter0 for TAMC |
| 326 |
|
|
| 327 |
|
checkpoint40pre2 |
| 328 |
|
o Updated to latest branch of AJA, JMC |
| 329 |
|
o Added gradient check package |
| 330 |
|
o Added routines for passive tracer handling |
| 331 |
|
o Added carbon sequestration sensitivity experiment |
| 332 |
|
|
| 333 |
|
checkpoint39 |
| 334 |
|
o Modifications and fixes to support TAMC |
| 335 |
|
(mainly missing or wrong directives) |
| 336 |
|
o Corrected common blocks when fooling TAMC to produce adjoint I/O |
| 337 |
|
in addummy_in_stepping |
| 338 |
|
o Updated verification/vero/ (bug fixes in code/, correct pickups in input/) |
| 339 |
|
o Enabled "sliced" (xz-,yz-) exchanges, needed for obcs. |
| 340 |
|
o Included following packages: |
| 341 |
|
cal: calendar |
| 342 |
|
exf: external forcing |
| 343 |
|
ecco: essential routines to control use of cal, exf |
| 344 |
|
(NB: this is not the full ECCO environment) |
| 345 |
|
o disabled some packages by default in genmake |
| 346 |
|
|
| 347 |
checkpoint38 |
checkpoint38 |
| 348 |
o see notes_c37_adj.txt |
o see notes_c37_adj.txt |
| 349 |
o corrected wrong layer index in external_forcing_t |
o corrected wrong layer index in external_forcing_t |
Parent Directory
| 
Revision Log
| 
Revision Graph
| 
Patch