| 4 |
Notes on tags used in MITgcmUV |
Notes on tags used in MITgcmUV |
| 5 |
============================== |
============================== |
| 6 |
|
|
| 7 |
|
o pkg/thsice |
| 8 |
|
- corrected store directive fixes thsice adjoint gradient checks |
| 9 |
|
o genmake2 & testreport: |
| 10 |
|
- add genmake2 option "-oad" to generate a Makefile for OpenAD built and |
| 11 |
|
in this case, use tools/adjoint_options/adjoint_oad as default AD optfile. |
| 12 |
|
Update testreport accordingly. No need to set "export OPENAD=true" |
| 13 |
|
in adjoint_options/adjoint_oad anymore (will be removed later). |
| 14 |
|
o pkg/shelfice: |
| 15 |
|
- put T & S shelfice forcing into model surfaceForcing_T,_S (for now, |
| 16 |
|
only if SHELFICEboundaryLayer=F); Change model/src/external_forcing.F |
| 17 |
|
so that surface forcing applies to k=kSurf if using pkg/shelfice. |
| 18 |
|
o pkg/ctrl, pkg/grdchk, pkg/ecco: |
| 19 |
|
- Changes to make GENARR more modular (i.e. invisible if not used) |
| 20 |
|
o model/src/temp_integrate.F and salt_integrate.F |
| 21 |
|
- re-include store directives that were removed on Dec 27th. |
| 22 |
|
o model/src/external_forcing_surf.F |
| 23 |
|
- add storage directives for EmPmR and PmEpR to avoid recomputation of |
| 24 |
|
seaice_model in the adjoint that is problematic when seaice_model is |
| 25 |
|
turned off in the adjoint. The resolved issue dated back to July when |
| 26 |
|
EXTERNAL_FORCING_SURF call was moved. |
| 27 |
|
Update global_ocean.cs32x15/*/output_adm.seaice_dynmix.txt accordingly. |
| 28 |
|
o pkg/exf/exf_check.F |
| 29 |
|
- fix a small bug again (originally found by Wentao Liu, now found again) |
| 30 |
|
- streamline error output |
| 31 |
|
o pkg/streamice and aim_v23: |
| 32 |
|
- replace a few "_EXCH_XY_RL (" with "_EXCH_XY_RL(", so that |
| 33 |
|
Apple's /usr/bin/cpp knows what to do: |
| 34 |
|
Apple LLVM version 5.0 (clang-500.2.79) (based on LLVM 3.3svn) |
| 35 |
|
Target: x86_64-apple-darwin13.0.0 |
| 36 |
|
|
| 37 |
|
checkpoint64s (2014/01/05) |
| 38 |
|
o tools/adjoint_options: |
| 39 |
|
- switch to fastopt.net TAF server (instead of default fastopt.de) for |
| 40 |
|
AD optfiles adjoint_diva & adjoint_f95 (similar to adjoint_default). |
| 41 |
|
o model/src: |
| 42 |
|
- remove some unnecessary TAF storage directives in forward_step.F and |
| 43 |
|
in dynamics.F (note: not always removed by TAF, e.g., double storage |
| 44 |
|
of kappaRU,kappaRV in former version of dynamics.F). |
| 45 |
|
o model/src & pkg/ptracers: |
| 46 |
|
- call directly GAD_IMPLICIT_R & IMPLDIFF + DWNSLP_APPLY from temp_integrate.F |
| 47 |
|
salt_integrate.F and ptracers_integrate.F (instead of from thermodynamics.F) |
| 48 |
|
and remove ptracers_dwnslp_apply.F & ptracers_implicit.F |
| 49 |
|
- call MATRIX_STORE_TENDENCY_EXP for each tracer (in ptracers_integrate.F) |
| 50 |
|
- add fZon & fMer as output argument of S/R GAD_CALC_RHS. |
| 51 |
|
- update TAF AD-output of experiments global_ocean.90x40x15 (13 for AD-grad). |
| 52 |
|
o useSingleCpuInput separates single-CPU input from single-CPU output; |
| 53 |
|
for now only implemented for in exf_interp_read.F |
| 54 |
|
o verification/advect_xy: |
| 55 |
|
- fix missing EXCH call in local version of ini_salt.F and update output. |
| 56 |
|
o momentum diagnostics: |
| 57 |
|
- fix which deltaT is used in Total-Tendency diagnostics. |
| 58 |
|
- fix typo in diagnostics for momentum tendency from AB. |
| 59 |
|
o pkg/seaice/seaice_lsr.F: fix an initialisation bug for the |
| 60 |
|
SEAICE_VECTORIZE_LSR_ZEBRA option that lead to NaNs on a NEC-SX8 |
| 61 |
|
|
| 62 |
|
checkpoint64r (2013/12/08) |
| 63 |
|
o verification/*/results: |
| 64 |
|
- Update reference output of experiments that were passing with just 13 digits |
| 65 |
|
(make it easier to detect future changes in results) |
| 66 |
|
o model/src & pkg/ptracers: |
| 67 |
|
- move calls to GAD_ADVECTION (Multi-Dim advection) from thermodynamics.F to |
| 68 |
|
inside temp_integrate.F and salt_integrate.F and from ptracers_advection.F |
| 69 |
|
to inside ptracers_integrate.F ; remove ptracers_advection.F |
| 70 |
|
o coupling interface pkgs: |
| 71 |
|
[atm_compon_interf, ocn_compon_interf & atm_ocn_coupler] |
| 72 |
|
- change the config check to test the agreement of Ocean Depth |
| 73 |
|
with ATM land/sea mask (instead of ATM orography); |
| 74 |
|
- send to coupler the number of coupler time-steps to do for this run: |
| 75 |
|
this allows to check for consistency and don't need anymore to be |
| 76 |
|
specified in parameter file "data"; |
| 77 |
|
in coupler: |
| 78 |
|
- check consistency of RunOff map versus Ocean Depth and ATM land/sea mask. |
| 79 |
|
- improve report to log-file of config setting (pass log-file unit to few S/R) |
| 80 |
|
o pkg/compon_communic: |
| 81 |
|
- add coupling send/receive for non-tiled integer vector; |
| 82 |
|
- add S/R to check and stop if any proc has an error (similar to |
| 83 |
|
eesupp/src/stop_if_error.F but single thread and across Comm_World). |
| 84 |
|
o pkg/exch2: |
| 85 |
|
- in exch2*_get_rx[1,2]: use new list "W2_tileIndex" to replace search |
| 86 |
|
through list of all tiles my proc owns. |
| 87 |
|
o model/src: |
| 88 |
|
- AUTODIFF only: in calc_adv_flow.F, recompute rTransKp (instead of copying |
| 89 |
|
from rTrans) to simplify dependencies (no need to store rTrans anymore in |
| 90 |
|
calling S/R); this affects AD results at machine truncation level: |
| 91 |
|
update AD-output (TAF) of experiments: 1D_ocean_ice_column, |
| 92 |
|
bottom_ctrl_5x5 & obcs_ctrl (12 digits for AD grad, same FWD grad). |
| 93 |
|
o tools/build_options: |
| 94 |
|
- simplify Warnings options (-ieee & -devel cases) in linux_amd64_gfortran |
| 95 |
|
o coupling interface pkgs: |
| 96 |
|
[compon_communic, atm_compon_interf & ocn_compon_interf] |
| 97 |
|
- move declaration of local array "ibuf" out of header file "CPLR_SIG.h" |
| 98 |
|
to each S/R that needs it; remove un-used variables. |
| 99 |
|
- change component tile registration (cpl_register.F in both compon_interf |
| 100 |
|
and mitcomponent_tile_register.F) to pass the full list of tiles |
| 101 |
|
information (including bi,bj) in 1 vector (instead of in a bi,bj matrix). |
| 102 |
|
- with pkg/exch2: adopt the same convention used for I/O to define mapping |
| 103 |
|
towards coupler global array (unless W2_useE2ioLayOut=F): |
| 104 |
|
a) can now use different tiling between atm & ocn comp. |
| 105 |
|
b) no needs to modify the run-Off map when changing tiling (was needed |
| 106 |
|
even if using the same tiling between atm & ocn). |
| 107 |
|
c) will allow to run with blank-tiles in 1 component (e.g., for ocean |
| 108 |
|
full land tiles); not yet tested. |
| 109 |
|
- check for valid mapping from comp. tile to coupler global array |
| 110 |
|
- improve report of tile registration results to log file. |
| 111 |
|
o pkg/exch2: |
| 112 |
|
- in eesupp/src/eeboot.F: move call to W2_EEBOOT before Coupler |
| 113 |
|
tile-registration (CPL_REGISTER); |
| 114 |
|
- activate (un-comment) definition and setting of "W2_tileIndex" list to |
| 115 |
|
store the local bi,bj indices of each tile; |
| 116 |
|
- also print bi,bj in local report to log file; |
| 117 |
|
- fix pkg/debug/debug_cs_corner_uv.F for the case nSy > 1 (left from |
| 118 |
|
Jun 2009 modif of pkg/exch2). |
| 119 |
|
o testing on baudelaire: |
| 120 |
|
- run 2 tests at a time: no significant slow down (since enough cpu) and |
| 121 |
|
get the full set of test done earlier |
| 122 |
|
- switch the ifort test to "-devel" ; run the OpenAD test everyday. |
| 123 |
|
o model/src/thermodynamics.F and routines called from it: |
| 124 |
|
- pkg/gmredi: gmredi_residual_flow.F (replacing gmredi_calc_uvflow.F & |
| 125 |
|
gmredi_calc_wflow.F) to update eulerian velocity (input) into residual |
| 126 |
|
velocity (outp) by adding bolus velocity. |
| 127 |
|
- model/src: calc_adv_flow.F (replacing calc_common_factors.F) to compute |
| 128 |
|
mass transport from velocity + lateral grid cell area |
| 129 |
|
- rename calc_gt.F to temp_integrate.F (includes call to TIMESTEP_TRACER) |
| 130 |
|
- rename calc_gs.F to salt_integrate.F (includes call to TIMESTEP_TRACER) |
| 131 |
|
in thermodynamics.F |
| 132 |
|
- remove k from thermodynamics.F and move it to temp_integrate.F |
| 133 |
|
salt_integrate.F and ptracers_integrate.F (now done inside the tracer |
| 134 |
|
loop). |
| 135 |
|
- compute locally (in thermodynamics.F) 3-D velocity field that is used to |
| 136 |
|
advect tracers; pass it as argument to GAD_ADVECTION, GAD_SOM_ADVECT, |
| 137 |
|
PTRACERS_ADVECTION, TEMP_INTEGRATE, SALT_INTEGRATE, PTRACERS_INTEGRATE, |
| 138 |
|
GAD_IMPLICIT_R and PTRACERS_IMPLICIT |
| 139 |
|
- update AD-output (TAF & OAD) of experiments: 1D_ocean_ice_column, |
| 140 |
|
global_ocean.90x40x15, tutorial_global_oce_biogeo and |
| 141 |
|
tutorial_tracer_adjsens (+ TLM output). |
| 142 |
|
o tools/OAD_support and (verification/*/) code_oad: |
| 143 |
|
- OAD_support/cb2mGetModules.csh: add PACKAGES_CONFIG.h in header file to |
| 144 |
|
module conversion ; remove unneeded CPP_EEOPTIONS.h. |
| 145 |
|
- code_oad/CPP_OPTIONS.h: remove PACKAGES_CONFIG.h and use a more standard |
| 146 |
|
version (closer to model/inc/CPP_OPTIONS.h). |
| 147 |
|
o model/src: |
| 148 |
|
- remove parameter "useOldFreezing" from PARAMS.h and push it to the |
| 149 |
|
retired list of parameters (routine FREEZE has been replace by surface |
| 150 |
|
freezing (10.yrs ago) and by pkg/frazil (interior freezing). |
| 151 |
|
o pkg/gmredi: |
| 152 |
|
- add option (GM_CALC_TRANSP_FROM_VEL) to compute transport directly from |
| 153 |
|
velocity (same scheme but affects machine truncation) for GM_AdvForm. |
| 154 |
|
- update output of FWD experiments: front_relax.bvp, |
| 155 |
|
global_ocean.90x40x15.dwnslp, global_ocean.cs32x15 and ideal_2D_oce. |
| 156 |
|
|
| 157 |
|
checkpoint64q (2013/11/17) |
| 158 |
|
o verification: |
| 159 |
|
- generate new output on baudelaire (gfortran, -ieee) for OpenAD experiments |
| 160 |
|
tutorial_global_oce_biogeo and global_ocean.90x40x15. |
| 161 |
|
- generate new output (growth & melt) for seaice_itd experiment after |
| 162 |
|
updating the input pickup_seaice files (previously wrong). |
| 163 |
|
o pkg/cost: |
| 164 |
|
- define a "final cost function" for each tile, and use GLOBAL_SUM_TILE_RL |
| 165 |
|
(instead of _GLOBAL_SUM_RL) to get the final cost from all procs & threads: |
| 166 |
|
o this fixes global-sum value if multi-threads (since GLOBAL_SUM_RL |
| 167 |
|
cannot be applied to var in common block). |
| 168 |
|
o allows to get results independent of tile-per-proc setting (if using |
| 169 |
|
slower GLOBAL_SUM_SEND_RECV option). |
| 170 |
|
- update accordingly: seaice_cost_final.F, shelfice_cost_final.F, |
| 171 |
|
thsice_cost_final.F and ecco_cost_final.F (this one affects lab_sea |
| 172 |
|
forward gradients: update output). |
| 173 |
|
- update */code_oad/externalDummies.F by adding dummy GLOBAL_SUM_TILE_RL |
| 174 |
|
S/R and, for most exp., also comment out GLOBAL_SUM_R8. |
| 175 |
|
o pkg/mom_common: |
| 176 |
|
- move to pkg/mom_common and model/src (previously in tutorial_deep_convection |
| 177 |
|
code) 2nd version of isotropic 3-D Smagorinsky code interface: strain and |
| 178 |
|
viscosity are locally declared in dynmics.F and pass as argument to CALC_GW; |
| 179 |
|
ensure that all field value that are used are set. |
| 180 |
|
o tutorial_deep_convection (smag3d): |
| 181 |
|
- add missing scaling in mom_calc_smag_3d.F; |
| 182 |
|
- set smag3D_coeff in input.smag3d such that results do not change. |
| 183 |
|
o pkg/dic: |
| 184 |
|
- change dic_cost.F to compute cost on each tile first and then apply |
| 185 |
|
GLOBAL_SUM_TILE_RL: this affects forward gradient of AD experiments |
| 186 |
|
tutorial_dic_adjoffline & tutorial_global_oce_biogeo; update output. |
| 187 |
|
- remove kLev argument and use GLOBAL_SUM_TILE_RL in tracer_meanarea. |
| 188 |
|
o eesupp/src & pkg/mdsio: |
| 189 |
|
- remove gather/scatter_vector.F (from eesupp) and mdsio_read/writevector.F |
| 190 |
|
(from pkg/mdsio) since they are no longer used for tape IO. |
| 191 |
|
- move simple C subroutine "setdir.c" from pkg/compon_communic to eesupp/src |
| 192 |
|
(+ keep a copy in pkg/atm_ocn_coupler). |
| 193 |
|
o verification/isomip: |
| 194 |
|
- split domain in 8 squared tiles (instead of just 2) for FWD, AD & OAD tests. |
| 195 |
|
o tools/build_options: |
| 196 |
|
- optfile linux_amd64_ifort11, fast (no ieee) case: |
| 197 |
|
to get a working restart for all experiments, add "-fp-model source" |
| 198 |
|
to FOPTIM (also remove "-opt-streaming-stores auto" <- this is the default). |
| 199 |
|
|
| 200 |
|
checkpoint64p (2013/10/20) |
| 201 |
|
o pkg/autodiff: |
| 202 |
|
- adread_adwrite.F: when type and length do not match ALLOW_AUTODIFF_WHTAPEIO |
| 203 |
|
criteria, return to simple (default) tape IO format. |
| 204 |
|
o pkg/mdsio (+ eesupp/src): |
| 205 |
|
- rewrite/simplify tape-IO S/R (new S/R MDS_READ/WRITE_TAPE replace |
| 206 |
|
previous MDSREAD/WRITEVECTOR) with 2 array argument from each type (R4/R8); |
| 207 |
|
- fix globalFile and singleCpuIO options using simpler global mapping. |
| 208 |
|
- new pairs of scatter/gather vector S/R (one for each type, generated from |
| 209 |
|
template) corresponding to simpler global mapping. |
| 210 |
|
- adread_write.F: switch to new S/R (but leave old calls commented) and write |
| 211 |
|
to globalFile when parameter globalFiles is set to True. |
| 212 |
|
- switch on useSingleCpuIO in AD experiments: global_ocean.90x40x15 & isomip. |
| 213 |
|
o pkg/autodiff: |
| 214 |
|
- adread_adwrite.F: add a hack for RS arrays (when RS=real*4); |
| 215 |
|
with ALLOW_AUTODIFF_WHTAPEIO defined, since corresponding (real*4) is |
| 216 |
|
missing, just add a STOP if array type do not match. |
| 217 |
|
o pkg/exf: |
| 218 |
|
- to allow to use seaice-fraction from pkg/exf in other pkgs than pkg/seaice: |
| 219 |
|
rename CPP option EXF_ALLOW_SEAICE_RELAX to EXF_SEAICE_FRACTION |
| 220 |
|
and rename array obsSIce to exf_iceFraction. |
| 221 |
|
o tutorial_deep_convection: |
| 222 |
|
- add 1rst version of isotropic 3-D Smagorinsky code (from Louis-Philippe), |
| 223 |
|
for now all in this exp. code dir. |
| 224 |
|
- add secondary test using Smag-3D viscosity (keep very low backgroud visc for |
| 225 |
|
testing) and flux-limter advection scheme for Temp. (no explicit diffusion). |
| 226 |
|
o eesupp/src: |
| 227 |
|
- added SINGLE_DISK_IO option because reading of parameter files on |
| 228 |
|
very large number of processors did not work with existing set-up |
| 229 |
|
|
| 230 |
|
checkpoint64o (2013/09/25) |
| 231 |
|
o pkg/atm_phys & atm_compon_interf: |
| 232 |
|
- allows for coupling of gray-atmosphere (atm_phys) with ocean. |
| 233 |
|
o tools/build_options: |
| 234 |
|
- switch from "-fPIC" to "-mcmodel=medium" in optfiles: linux_amd64_gfortran |
| 235 |
|
& linux_amd64_ifort11 and in linux_amd64_g77 & linux_amd64_open64 also |
| 236 |
|
remove CPP setting (was identical to genmake2 default). |
| 237 |
|
o pkg/gmredi & model/src: Some updating and work done on the residual model. |
| 238 |
|
- created new file model/src/calc_eddy_stress.F to calculate the eddy stress. |
| 239 |
|
- fixed up the calculation of the eddy stress and how it is added to the |
| 240 |
|
external forcing via taueddy_external_forcing.F |
| 241 |
|
- at the moment, it is very experimental and only works with the variable eddy |
| 242 |
|
diffusivity calculated in gmredi_k3d.F |
| 243 |
|
o pkg/seaice/seaice_check.F: add a check to catch the case of specifying |
| 244 |
|
lw/swfluxfile instead of lw/swdownfile in data.exf together with pkg/seaice |
| 245 |
|
o testreport (& tools/do_tst_2+2): |
| 246 |
|
- for MPI test, set default COMMAND to 'mpirun -v -np TR_NPROC ./{executable}' |
| 247 |
|
o tools/build_options: linux_amd64_gfortran & linux_amd64_ifort11: |
| 248 |
|
- remove CPP='cpp -traditional -P' (this is genmake2 default) |
| 249 |
|
- add switch (for big setups) to use 'medium' memory-model (but leave it |
| 250 |
|
commented out for now). Note: needed with FC 19 to compile big setups. |
| 251 |
|
o pkg/atm_compon_interf: |
| 252 |
|
- regroup per field origin (model & pkgs) the coupling storage routines |
| 253 |
|
(which accumalate in time each coupling field) |
| 254 |
|
- move bi,bj loops inside atm_store_my_data.F and store also wind-stress; |
| 255 |
|
update accordingly + simplify aim_do_physics.F |
| 256 |
|
|
| 257 |
|
checkpoint64n (2013/08/26) |
| 258 |
|
o pkg/bbl: |
| 259 |
|
- update formulation to reduce dispersion of tracer properties in bbl |
| 260 |
|
- replaced horizontal and vertical transport parameters with velocities |
| 261 |
|
- changes results for verification/global_with_exf.yearly |
| 262 |
|
o verification/halfpipe_streamice: |
| 263 |
|
- change to matrix construction; update both Fwd & AD output. |
| 264 |
|
o verification/tutorial_tracer_adjsens: |
| 265 |
|
- OpenAD test: switch to vector-invariant momentum since fluxform with |
| 266 |
|
rStar causes problems with OpenAD; update output. |
| 267 |
|
o tools: |
| 268 |
|
- genmake2: define a common prefix for all temp-file names ; move them |
| 269 |
|
to local /tmp disk (but would be easy to move them back); |
| 270 |
|
- xmakedepend: apply the same switch for temp-files. |
| 271 |
|
o pkg/offline: |
| 272 |
|
- changes for use with rStar (NLFS): assume that we continue to provide |
| 273 |
|
UVELMASS/hFacW_ini & VVELMASS/hFacS_ini (hFacWi,S_ini as written in grid |
| 274 |
|
output files) and apply r* scaling at each time-step (function of the time |
| 275 |
|
varying Eta field); use normal pickup files (needed for restart with NLFS). |
| 276 |
|
o adjoint with OpenAD: |
| 277 |
|
- move OpenAD-specific files from verification/OpenAD/code_oad_all to a new |
| 278 |
|
package openad. This package has to be compiled for all runs with OpenAD |
| 279 |
|
instead of specifying code_oad_all as a modification directory. |
| 280 |
|
- link the fortran templates from tools/OAD_support to the build directory |
| 281 |
|
for preprocessing to avoid conflicts between different setups compiled |
| 282 |
|
simultaneously and simplify cleanup. |
| 283 |
|
o pkg/exf: |
| 284 |
|
- in exf_getforcing.F: comment out most of the EXCH calls (fields used by |
| 285 |
|
main model or other pkgs are exchanged in EXF_MAPFIELDS). |
| 286 |
|
o model/src: |
| 287 |
|
- do_stagger_fields_exchanges.F: skip EXCH calls when using pkg/offline |
| 288 |
|
(similar to do_fields_blocking_exchanges.F). |
| 289 |
|
o tools/genmake2: |
| 290 |
|
- remove TAF output src code just before calling staf (when not doing a fresh |
| 291 |
|
built, will stop if things go wrong with staf). |
| 292 |
|
- fix target "adobj" (left from TAF version-2 transition) |
| 293 |
|
o pkg/diagnostics: |
| 294 |
|
- add a parameter to track the status of pkg/diagnostics activation; |
| 295 |
|
this replace/extend the use of logical param "settingDiags" (to check |
| 296 |
|
when adding diag to the list) to also check any DIAGNOSTICS_[]FILL* call. |
| 297 |
|
New S/R to print error message and stop when pkgStatus is not right. |
| 298 |
|
- to try: add a warning when "data.diagnostics" exist but useDiagnostics=F. |
| 299 |
|
|
| 300 |
|
checkpoint64m (2013/08/13) |
| 301 |
|
o model/inc: |
| 302 |
|
- move FCORI_MACROS.h from PARAMS(_MACROS).h to GRID_MACROS.h (left from |
| 303 |
|
8 yrs ago modif when Coriolis arrays were moved from PARAMS.h to GRID.h) |
| 304 |
|
o pkg/kpp & aim_v23: |
| 305 |
|
- only call DIAGNOSTICS_FILL when useDiagnostics=TRUE. |
| 306 |
|
o pkg/seaice: |
| 307 |
|
- restore (last time on 2011/12/24) ability to compile pkg/seaice with |
| 308 |
|
undefined pkg/exf ALLOW_ATM_TEMP or ALLOW_RUNOFF CPP-options. |
| 309 |
|
- only call DIAGNOSTICS_FILL when useDiagnostics=TRUE. |
| 310 |
|
o eesupp/src: |
| 311 |
|
- fix check for overlap-size argument (myOLn.GT.MAX_OLY_EXCH) |
| 312 |
|
in exch1_rx[_ad].template (thanks to Francois Lefeuvre report). |
| 313 |
|
o switch to TAF version-2: |
| 314 |
|
- add flow directives to "freeze" naming convention for ad-routines that |
| 315 |
|
are used in I/O routines of adjoint state and associated common blocks |
| 316 |
|
that are used in adcommon.h (backward compatible with version-1) |
| 317 |
|
- update all ad-optfiles (in tools/adjoint_options + few local to some |
| 318 |
|
verification exp) to switch to latest TAF version (currently 2.5.5); |
| 319 |
|
merge "*_mpi" adoptfiles into the corresponding non-mpi (but commented |
| 320 |
|
out except for diva). |
| 321 |
|
- the major change with TAF version-2 is that both the differentiated |
| 322 |
|
routines and the forward routines (new) are returned as part of the TAF |
| 323 |
|
output source code. Update genmake2 accordingly. |
| 324 |
|
- new output for AD experiment tutorial_global_oce_biogeo (better AD gradient |
| 325 |
|
with version-2 when compared to TLM or FWD gradient); other experiments |
| 326 |
|
output show minor changes (machine truncation level); update the few AD & |
| 327 |
|
TLM output with 12 matching digits only (some prior to version-2 switch). |
| 328 |
|
|
| 329 |
|
checkpoint64l (2013/08/05) |
| 330 |
|
o model/src: |
| 331 |
|
- in dynamics.F, skip the call to CALC_VISCOSITY if momViscosity=F |
| 332 |
|
- main_do_loop.F: no longer includes additional (AUTODIFF) header files |
| 333 |
|
when using TAF but only if using OpenAD. |
| 334 |
|
o pkg/mom_common+fluxform+vecinv: |
| 335 |
|
- always set horiz. viscosity arrays to background value before calling |
| 336 |
|
MOM_CALC_VISC (in MOM_FLUXFORM & MOM_VECINV) and call S/R MOM_CALC_VISC |
| 337 |
|
only when using variable horiz. viscosity (useVariableVisc=T); |
| 338 |
|
- simplify mom_vecinv.F (only 1 block for momViscosity). |
| 339 |
|
o pkg/mom_common: variable horiz. viscosity in NH code |
| 340 |
|
- move horiz-viscosity 3-D arrays viscAh_W & viscA4_W (used in calc_gw.F) |
| 341 |
|
from model/inc/NH_VARS.h to pkg/mom_common/MOM_VISC.h |
| 342 |
|
- store in common block (in MOM_VISC.h) logical parameters: useHarmonicVisc, |
| 343 |
|
useBiharmonicVisc & useVariableVisc. |
| 344 |
|
- fix old bug (already in 1rst implementation of variable horizontal viscosity |
| 345 |
|
in calc_gw.F, version 1.27, 2006/06/20): |
| 346 |
|
horizontal bi-harmonic viscosity in gW was missing when wiscA4W=0 |
| 347 |
|
even if variable bi-harmonic viscosity (e.g., viscA4Grid) was non-zero. |
| 348 |
|
- change background horiz viscosity in vertical momentum (gW) to always be |
| 349 |
|
viscAhW,viscA4W (instead of switching to viscAhD, viscA4D when using |
| 350 |
|
any type of variable viscosity, e.g., viscAh/4Grid) |
| 351 |
|
- change default value of viscAhW,viscA4W to viscAhD,viscA4D (instead of |
| 352 |
|
viscAh,viscA4) which default to viscAh,viscA4 : this avoid missing |
| 353 |
|
horiz. viscosity in W eq when just setting viscAhD/Z,viscA4D/Z (since |
| 354 |
|
default viscAh,viscA4 is zero). |
| 355 |
|
o model/src: |
| 356 |
|
- extend length of drC & recip_drC from Nr to Nr+1. |
| 357 |
|
o Divided Adjoint (DIVA) with MPI: |
| 358 |
|
- genmake2: do not create local ./mpi_headers dir when no file to copy into. |
| 359 |
|
- remove option ALLOW_DIVIDED_ADJOINT_MPI (replaced by ALLOW_DIVIDED_ADJOINT |
| 360 |
|
and ALLOW_USE_MPI) from model/src code and from ECCO_CPPOPTIONS.h. |
| 361 |
|
|
| 362 |
|
checkpoint64k (2013/07/23) |
| 363 |
|
o tools/OAD_support: |
| 364 |
|
- avoid the need for file "oadtempflile" in insertTemplateDir.bash |
| 365 |
|
and remove it from genmake2. |
| 366 |
|
o tools/build_options: |
| 367 |
|
- remove MPI_HEADER_FILES_INC setting (no longer used) and comment out |
| 368 |
|
MPI_HEADER_FILES list (let genmake2 find the missing headers) |
| 369 |
|
o verification/halfpipe_streamice: |
| 370 |
|
- change to driving stress formulation; update both Fwd & AD output. |
| 371 |
|
o tools/genmake2 and Divided Adjoint with MPI: |
| 372 |
|
- remove default setting of MPIINCLUDEDIR from $MPIHOME (always been broken); |
| 373 |
|
- only create local ./mpi_headers dir if using MPI and MPIINCLUDEDIR is set; |
| 374 |
|
- add new variable (LOCAL_MPI_HEADERS, replace MPI_HEADER_FILES_INC) in |
| 375 |
|
Makefile, which is set by genmake2, and check which mpi-headers are needed. |
| 376 |
|
Note: need to update adjoint_options/adjoint_diva_mpi accordingly. |
| 377 |
|
- update target "CLEAN" (to also remove ./mpi_headers/ dir). |
| 378 |
|
o testreport: |
| 379 |
|
- changes for Divided Adjoint Run: when file "run_ADM_DIVA" is present, |
| 380 |
|
perform additional runs (number of additional runs is "add_DIVA_runs" taken |
| 381 |
|
from file run_ADM_DIVA ; should match "nchklev_3" in file "code_ad/tamc.h"). |
| 382 |
|
o verification/lab_sea (AD test): |
| 383 |
|
- change to test Divided Adjoint Run in this set-up (both standard AD test |
| 384 |
|
and 2 secondary AD test: noseaice + noseaicedyn); drop the evp AD test. |
| 385 |
|
o pkg atm_ & ocn_compon_interf: |
| 386 |
|
- remove cpl_earlyExpImpCall=F option (was mainly a hack added for the old |
| 387 |
|
staggerTimeStep implementation) and disable (retired) corresponding |
| 388 |
|
parameter. |
| 389 |
|
o pkg/offline: |
| 390 |
|
- comment out unused field-array: ICEM, ConvectCount, hflx[01], icem[01]; |
| 391 |
|
- add S/R OFFLINE_CHECK to print offline parameter summary |
| 392 |
|
and stop if loading a file that is not used; |
| 393 |
|
- enable to calculate GMRedi tensor, KPP mixing and convective index |
| 394 |
|
when the corresponding file(s) is/are not specified (but pkg is used). |
| 395 |
|
- comment out resetting of monitorFreq (to endTime) since it still useful |
| 396 |
|
for monitoring CFL numbers and checking offline fields. |
| 397 |
|
o OpenAD: |
| 398 |
|
- do not define OAD_DEBUG (in tools/adjoint_options/adjoint_oad) to get |
| 399 |
|
smaller standard output) and generate new (smaller) output (on baudelaire |
| 400 |
|
using gfortran with -ieee) for exp. tutorial_tracer_adjsens |
| 401 |
|
o model/src: |
| 402 |
|
- call EXTERNAL_FORCING_SURF outside bi,bj loops (in do_oceanic_phys.F), |
| 403 |
|
(where external_forcing_adjust was called) and change list of arguments; |
| 404 |
|
- merge external_forcing_adjust.F into external_forcing_surf.F |
| 405 |
|
- put surface relaxation of SST & SSS in new S/R (forcing_surf_relax.F) |
| 406 |
|
which also contains balancing surface relaxation (ALLOW_BALANCE_RELAX, |
| 407 |
|
previously in file "balance_relax.F", now removed); remove corresponding |
| 408 |
|
arrays in FFIELDS.h. |
| 409 |
|
- EXTERNAL_FORCING_SURF is now called when using pkg/offline; implement |
| 410 |
|
a hack for getting back surfaceForcingS (loaded from file) when using |
| 411 |
|
ALLOW_OLD_VIRTUALFLUX (pkg/dic, in DIC_OPTIONS.h) in offline set-up. |
| 412 |
|
- update output of AD test experiment global_ocean.cs32x15.seaice_dynmix |
| 413 |
|
(was getting only 11 digits for AD-grad, due to different truncation |
| 414 |
|
error in AD code). |
| 415 |
|
|
| 416 |
|
checkpoint64j (2013/07/04) |
| 417 |
|
o pkg/offline: |
| 418 |
|
- to enable the use of other features with pkg offline (e.g., obcs), |
| 419 |
|
also call S/R DO_OCEANIC_PHYS in offline run - more modifs to come. |
| 420 |
|
- comment out "allowFreezing" (now active) in 2 offline testing exp. |
| 421 |
|
o pkg/exf ALLOW_RUNOFTEMP |
| 422 |
|
- fix sign and make compatible with temp_EvPrRn .NE. UNSET_RL |
| 423 |
|
- ahdded RUNOFTEMP test to verification/global_ocean.cs32x15/input.seaice |
| 424 |
|
- changes verification/global_ocean.cs32x15/results/output.seaice.txt |
| 425 |
|
o pkg/atm_phys: |
| 426 |
|
- allows to add some damping on uVel,vVel (strength only function of k). |
| 427 |
|
o verification, ad-tests: |
| 428 |
|
- reduce length of the simulation for AD test: global_ocean.cs32x15.seaice, |
| 429 |
|
global_ocean_ebm and offline_exf_seaice.thsice. |
| 430 |
|
o verification/halfpipe_streamice: |
| 431 |
|
- found a bug in s/r pkg/streamice/streamice_advect_thickness which caused |
| 432 |
|
flux input boundary conditions to be ignored; a fix for this bug changed |
| 433 |
|
verification results for forward test. |
| 434 |
|
- add adjoint test. |
| 435 |
|
o pkg/ptracers: |
| 436 |
|
- allow to restart without SOM pickup (resetting SOM to zero). |
| 437 |
|
o pkg/gmredi: |
| 438 |
|
- New parameterisation for the calculation of K and (GM_PsiX,GM_PsiY) added. |
| 439 |
|
- It uses mixing length theory and suppression by steering level effects to |
| 440 |
|
calculate a diffusivity used in the Redi tensor (see Bates et al., submitted |
| 441 |
|
to JPO, for more details). |
| 442 |
|
- It also uses the diffusivity to calculate an eddy induced streamfunction |
| 443 |
|
based on a PV closure (rather than a buoyancy closure as in GM). |
| 444 |
|
- New files added to pkg/gmredi: |
| 445 |
|
gmredi_k3d.F calculates the diffusivity for the Redi tensor and the eddy |
| 446 |
|
induced overturning streamfunction (called from gmredi_calc_tensor) |
| 447 |
|
gmredi_calc_eigs.F calculates the eigenvalues and eigenvectors required |
| 448 |
|
for use in gmredi_k3d |
| 449 |
|
gmredi_calc_urms.F calculates the the rms eddy velocity used to calculate |
| 450 |
|
the eddy diffusivity |
| 451 |
|
|
| 452 |
|
checkpoint64i (2013/06/18) |
| 453 |
|
o pkg/cheapaml: |
| 454 |
|
- take src files from MITgcm_contrib/verification_other/offline_cheapaml/code |
| 455 |
|
that enable the use of seaice (pkg/thsice thermo & pkg/seaice dynamics) |
| 456 |
|
with pkg/cheapaml. |
| 457 |
|
o utils/python: add subpackage for cubed sphere function, first example is |
| 458 |
|
simple cs.pcol to plot 2D fields on a map |
| 459 |
|
o pkg/streamice: |
| 460 |
|
- This is a new package in MITgcm, which solves dynamic equations for ice |
| 461 |
|
sheets and ice streams |
| 462 |
|
- At the moment it is independent of atmosphere/ocean dynamics, or any |
| 463 |
|
other physical packages in MITgcm, even ShelfIce (although it is anticipat- |
| 464 |
|
ed that this will change) |
| 465 |
|
- There is a new verification exercise that tests it, halfpipe_streamice, |
| 466 |
|
evolves a 40x80 km ice stream/ice shelf (at coarse resolution) in a |
| 467 |
|
rectangular embayment |
| 468 |
|
- halfpipe_streamice is forward-only for now, will have TAF test soon |
| 469 |
|
- for questions, email dgoldber@mit.edu |
| 470 |
|
o pkg/thsice: |
| 471 |
|
- fix filling of snow-precip diagnostic (SIsnwPrc) (left from the removing |
| 472 |
|
of snowPrc from common bloc, on May 2nd, thsice_step_fwd.F revision 1.40); |
| 473 |
|
- remove diagnostics 'SI_FrcFx' ( now identical to 'SI_Fract' with new |
| 474 |
|
call sequence: ice-Thermo(thsice),ice-Dyn,ice-Advect ); |
| 475 |
|
- allow precip from pkg/cheapaml to make snow over seaice. |
| 476 |
|
o pkg/cfc: |
| 477 |
|
- move in separated header file (CFC_ATMOS.h) time-serie arrays of Atmos CFC |
| 478 |
|
- remove atmos CFC 2-D arrays from common bloc (now local variables) |
| 479 |
|
- change some hard-coded constant to run time parameter for reading and |
| 480 |
|
and interpolating atmospheric CFC ; keep same default as previously. |
| 481 |
|
- more flexible reading of Atmos CFC (number of records no longer hard-coded) |
| 482 |
|
- put arguments in standard order: myTime, myIter, myThid |
| 483 |
|
- use standard S/R GET_PERIODIC_INTERVAL for time interpolation of |
| 484 |
|
atmospheric CFC ; fix for the case myIter < PTRACERS_iter0. |
| 485 |
|
- fix calculation of surf tendencys for partial-cell and/or Non-Lin Free-Surf |
| 486 |
|
- new routine to print summary of CFC params and check for consistency. |
| 487 |
|
o pkg/seaice: add uice/vice to the parameter list of s/r seaice_advdiff, |
| 488 |
|
similar to thsice_advdiff. |
| 489 |
|
Note that for the B-grid case the routine parameters uc/vc are overwritten. |
| 490 |
|
o pkg/cfc: |
| 491 |
|
- change time interpolation for atmospheric CFC-11 (now same as for CFC-12); |
| 492 |
|
update output (CFC-11) of experiment cfc_example and tutorial_cfc_offline. |
| 493 |
|
- fix interpolation of atmphspheric CFC between Sorthern and Northern |
| 494 |
|
hemisphere value. |
| 495 |
|
o pkg/seaice: |
| 496 |
|
- allow to use seaice dynamics without pkg/exf (e.g., using pkg/cheapaml) |
| 497 |
|
o pkg/seaice: add Adams-Bashforth2 time discretization for ice dynamics; |
| 498 |
|
so far only for JFNK-solver, turn on with SEAICEuseAB2 = .TRUE. |
| 499 |
|
- requires an additional time level that is stored in the pickup |
| 500 |
|
- you can start with AB2 from a pickup without this time level with |
| 501 |
|
pickupStrictlyMatch = .false. |
| 502 |
|
|
| 503 |
|
checkpoint64h (2013/05/26) |
| 504 |
|
o pkg/cheapaml: |
| 505 |
|
- replace one ()**.5 with SQRT + use temp var "cdu = tau/Du" (coare3_flux); |
| 506 |
|
affect machine truncation (but exp cheapAML_box still pass with 13 digits). |
| 507 |
|
- consistent treatment of uWind,vWind location: assume everywhere in |
| 508 |
|
pkg/cheapaml that they are on C-grid (@ uVel,vVel location) |
| 509 |
|
(Note: was already the case for Tair,Qair advection); |
| 510 |
|
add option (INCONSISTENT_WIND_LOCATION) to recover old results. |
| 511 |
|
- compute diabatic forcing only in the interior (excluding overlap regions) |
| 512 |
|
- fill in array "Qsw" (for short-wave heating); |
| 513 |
|
- add diagnostic for precip ; fill-in diagnostics for cheapAML state vars |
| 514 |
|
before any update (consistent with MITgcm state var diagnostics); |
| 515 |
|
- cheapaml_fields_load.F & cheapaml_init_varia.F : only fill outer edges if |
| 516 |
|
non periodic ; + adjusted for uWind,vWind (C-grid). |
| 517 |
|
o pkg/seaice: |
| 518 |
|
- remove dependence of pkg/seaice versus pkg/exf (file: pkg/pkg_depend) and |
| 519 |
|
add explicitly pkg/exf in experiments that currently use pkg/seaice. |
| 520 |
|
- allows to compile pkg/seaice without pkg/exf. |
| 521 |
|
o tools/OAD_support: |
| 522 |
|
- use a local copy (ourRWCPinBd) to ensure index ourRWCP stays within bounds |
| 523 |
|
of array "ourStepOf" in function rvNextAction (revolve.F90). |
| 524 |
|
o pkg/aim_compon_interf renamed to pkg/atm_compon_interf |
| 525 |
|
- change pkg name (to be used with other atmospheric physics pkg) |
| 526 |
|
but by-pass CVS when making the new pkg so that we keep the CVS history. |
| 527 |
|
o pkg/offline: |
| 528 |
|
- call OFFLINE_FIELDS_LOAD from pkg/offline/offline_init_varia.F |
| 529 |
|
(previously called from pkg/dic_surfforcing_init.F) |
| 530 |
|
o model/src: |
| 531 |
|
- re-order sequence of calls within packages_init_variables.F, using |
| 532 |
|
standard order, except for: |
| 533 |
|
DIAGNOSTICS_INIT_VARIA and OFFLINE_INIT_VARIA (called @ the top), |
| 534 |
|
CTRL_MAP_INI and OBCS_INIT_VARIABLES (called @ the bottom) |
| 535 |
|
o new pkg: atm_phys |
| 536 |
|
- add entry point and source code (previously in contrib |
| 537 |
|
verification_other/atm_gray/code_gray/) for new pkg "atm_phys" |
| 538 |
|
(atmospheric physics pkg from P. O'Gorman and T. Schneider, JCl, 2008). |
| 539 |
|
- re-order sequence of calls within packages_readparms.F, |
| 540 |
|
packages_init_fixed.F and packages_write_pickup.F |
| 541 |
|
(but skip packages_init_variables.F because of TAF issues). |
| 542 |
|
- update contrib/*/atm_gray/ experiment. |
| 543 |
|
o pkg/seaice (ITD code): |
| 544 |
|
- simplify and fix the writing of pickup-files for ITD case |
| 545 |
|
o pkg/seaice (ITD code): |
| 546 |
|
- introduce floe size dependent lateral melt for ITD case; |
| 547 |
|
- cleaning (avoid division by zero, change to f77 syntax, remove tabs) |
| 548 |
|
- update seaice_itd experiment main output. |
| 549 |
|
o pkg/thsice: |
| 550 |
|
- remove snowPrc from common block THSICE_FLUX, in THSICE_VARS.h and |
| 551 |
|
declared it locally (without bi,bj) in thsice_main.F & aim_do_physics.F; |
| 552 |
|
changed accordingly pkg/atm2d. |
| 553 |
|
|
| 554 |
|
checkpoint64g (2013/05/01) |
| 555 |
|
o pkg/exf ALLOW_RUNOFTEMP |
| 556 |
|
- if defined, river and glacier runoff temperature can be read-in from |
| 557 |
|
runoftempfile; runoftempfile needs to have same format as runofffile |
| 558 |
|
o pkg/thsice (& pkg/bulk_force): |
| 559 |
|
- In case Energy-Reference-Level (ERL) is used (temp_EvPrRn=0), account |
| 560 |
|
for energy content of Precip + RunOff & Evap (like with pkg/exf). |
| 561 |
|
- change computation domain to always be the interior (without overlap). |
| 562 |
|
- fix bug (wrong overlap) when using EXF and stressReduction > 0. |
| 563 |
|
o verification/tutorial_global_oce_biogeo: |
| 564 |
|
- add preliminary OpenAD setup (still has problems) |
| 565 |
|
o pkg/seaice: |
| 566 |
|
- turn some parameters for choosing the convergence criterion of |
| 567 |
|
the inexact Newton method (JFNK) into runtime parameters for convenience |
| 568 |
|
- clean up seaice_calc_viscosities.F: remove/retire two options for |
| 569 |
|
SEAICetaZmethod (1,2), introduce area weighting for method 3; this changes |
| 570 |
|
the results of lab_sea/input because a division has been replaced by |
| 571 |
|
a multiplication with the reciprocal (sumNorm) |
| 572 |
|
- modify offline_exf_seaice/input.dyn_jfnk/data.seaice to improve |
| 573 |
|
convergence of solver (changes results, of course), and to test some |
| 574 |
|
new features |
| 575 |
|
o model/src: |
| 576 |
|
- remove unused (since Aug 2006) S/R SET_WRITE_GLOBAL_PICKUP and its call. |
| 577 |
|
- remove unused forcing field "dQdT" and associated parameter "dQdTFile". |
| 578 |
|
- move forcing adjustment (balancing surface forcing) in specific S/R |
| 579 |
|
where might also be applied exchanges (if needed) to forcing arrays. |
| 580 |
|
o pkg/shelfice: |
| 581 |
|
- allow to specify the Ice-Shelf mass per unit area (instead of |
| 582 |
|
shelficeLoadAnomaly) and do the conversion from one to the other |
| 583 |
|
- move loading from file of shelficeLoadAnomaly from SHELFICE_INIT_VARIA |
| 584 |
|
to SHELFICE_INIT_FIXED: (a) avoid some TAF recompution issues and |
| 585 |
|
(b) will be loaded in time (before calling INI_PRESSURE) in case it is |
| 586 |
|
needed in phi0surf for initial pressure calculation when useDynP_inEos_Zc |
| 587 |
|
- update isomip experiment forward output (get machine truncation error |
| 588 |
|
from conversion shelficeLoadAnomaly -> shelficeMass -> shelficeLoadAnomaly) |
| 589 |
|
o verification/offline_exf_seaice: |
| 590 |
|
revise data.seaice in input.thermo and input_ad and update results accordingly |
| 591 |
|
HO = 0.2, SEAICE_areaLossFormula = 2, |
| 592 |
|
SEAICE_growMeltByConv = .TRUE., |
| 593 |
|
SEAICE_doOpenWaterMelt = .TRUE., |
| 594 |
|
Improve consistency with input.thsice and fix the spurious 500m ice corners |
| 595 |
|
o verification/global_ocean.cs32x15: |
| 596 |
|
- set Energy Reference Level to liq. water @ 0 degC: temp_EvPrRn=0., |
| 597 |
|
in icedyn (thsice+seaice-dyn) test and update output. |
| 598 |
|
o verification/offline_exf_seaice: |
| 599 |
|
- adjust list of tested variables (with testreport) and update |
| 600 |
|
standard test output with better slection of monitored fields. |
| 601 |
|
o pkg/cd_code: |
| 602 |
|
- swtich off Adams-Bashforth on the 1rst iteration. Update output of |
| 603 |
|
experiment lab_sea.hb87 |
| 604 |
|
o tools/genmake2: |
| 605 |
|
- change default makedepend from the system default to tools/xmakedepend: |
| 606 |
|
When no makedepend is specified, select tools/xmakedepend it the test pass, |
| 607 |
|
otherwise select the system default (if exist and pass the test) or |
| 608 |
|
try to build cyrus-makedepend ; stop if none of the 3 works. |
| 609 |
|
o pkg/exf & pkg/thsice: |
| 610 |
|
- In case Energy-Reference-Level (ERL) is used (temp_EvPrRn=0), account |
| 611 |
|
for energy content of Precip + RunOff & Evap. Assumes: |
| 612 |
|
1) Rain has same temp as Air. |
| 613 |
|
2) Snow has no heat capacity (consistent with seaice & thsice pkgs) |
| 614 |
|
3) Run-Off comes at the temp of surface water (with same Cp) |
| 615 |
|
4) Evap over ocean is released to the Atmos @ surf-temp (=SST); should be |
| 616 |
|
put directly into Latent Heat flux & should use water-vapor heat capacity. |
| 617 |
|
5) Evap over snow/seaice (pkg/thsice) assumes constant Lf+Lv (consistent |
| 618 |
|
with snow having no Cp) => water vapor is released to Atmos @ 0^oC (=ERL). |
| 619 |
|
o pkg/seaice: simplify the use of CPP flags in JFNK-related files |
| 620 |
|
(compile when SEAICE_ALLOW_JFNK is defined) |
| 621 |
|
o pkg/thsice: |
| 622 |
|
- fix oceanic flux diagnostics & timeave (left from new sequence of calls |
| 623 |
|
implementation): move THSICE_AVE call after the advection (called directly |
| 624 |
|
from thsice_do_advect.F). |
| 625 |
|
o verification/global_ocean.90x40x15 and OpenAD: |
| 626 |
|
New experiment for OpenAD is identical to the one in input_ad except that |
| 627 |
|
multiDimAdvection=.FALSE., |
| 628 |
|
There is still an issue with this. |
| 629 |
|
o pkg/thsice: |
| 630 |
|
- account for ice velocity (ocean relative velocity) in turbulent heat flux |
| 631 |
|
at the base of the ice; update output of global_ocean.cs32x15 (icedyn) |
| 632 |
|
and offline_exf_seaice (main test). |
| 633 |
|
|
| 634 |
|
checkpoint64f (2013/03/31) |
| 635 |
|
o verification/OpenAD: |
| 636 |
|
- new (smaller, without OAD_DEBUG) output_oadm.txt generated using gfortran. |
| 637 |
|
o verification/isomip, pkg/shelfice and OpenAD: |
| 638 |
|
- add OpenAD setup to this exp and make shelfice package work with OpenAD |
| 639 |
|
o verification/tutorial_global_oce_biogeo: |
| 640 |
|
- update AD and TLM output after turning off ALLOW_FE (in DIC_OPTIONS.h) |
| 641 |
|
o verification/offline_exf_seaice: |
| 642 |
|
- replace adjoint experiment (now similar to fwd input.thermo/ setup). |
| 643 |
|
o verification/tutorial_tracer_adjsens, pkg/ptracers and OpenAD: |
| 644 |
|
- add OpenAD setup to this exp and make ptracers package work with OpenAD |
| 645 |
|
o pkg/seaice/seaice_evp.F: |
| 646 |
|
mask the asymmetric contribution of the ice-ocean stress (when |
| 647 |
|
the turningAngle is non-zero) over open water to avoid stripes |
| 648 |
|
in the EVP-solutions after many EVP-substeps. This does not change |
| 649 |
|
the verification experiment (because turningAngle=0 in that experiment), |
| 650 |
|
but will affect other EVP solutions with turningAngle.ne.0 |
| 651 |
|
o verification, AD & TLM output: |
| 652 |
|
- new TAF version 2.3.10 (previously 2.3.8) add more parenthesis in multiple |
| 653 |
|
additions and multiplications which affect results at machine truncation |
| 654 |
|
level. Update few AD output (1D_ocean_ice_column, bottom_ctrl_5x5, |
| 655 |
|
lab_sea.noseaicedyn & offline_exf_seaice) and TLM output |
| 656 |
|
(1D_ocean_ice_column & tutorial_tracer_adjsens.som81) which had only 12 |
| 657 |
|
digits of matching. |
| 658 |
|
o pkg/seaice: |
| 659 |
|
modify lsr-algorithm so that it vectorizes AND converges by using a |
| 660 |
|
"zebra" method: split vector loop into two with step size 2; this |
| 661 |
|
solves a general convergence problem for the JFNK solver on a vector |
| 662 |
|
computer. It has the potential to even improve convergence for |
| 663 |
|
non-vector machines, but it does change the results so it is turned |
| 664 |
|
off by default. To turn on define CPP-flags SEAICE_VECTORIZE_LSR and |
| 665 |
|
SEAICE_VECTORIZE_LSR_ZEBRA in SEAICE_OPTIONS.h |
| 666 |
|
|
| 667 |
|
checkpoint64e (2013/03/05) |
| 668 |
|
o pkg/ecco and several verification experiments: |
| 669 |
|
Remove special versions of the_model_main.F, the_main_loop.F, forward_step.F |
| 670 |
|
from ecco package, and revert to unified routine in model/src/ |
| 671 |
|
Adjust verification experiments accordingly, especially gloal_oce_llc90 |
| 672 |
|
Should conclude long-standing work by Gael to reconsile the_main_loop with |
| 673 |
|
ECCO and will enable OpenAD's revolve loop in ECCO setup. |
| 674 |
|
o pkg/generic_advdiff: |
| 675 |
|
- implement compressible flow method for multi-dim advection (similar to |
| 676 |
|
gad_som_advect.F); add new option "GAD_MULTIDIM_COMPRESSIBLE" (since |
| 677 |
|
TAF generates many recomputations) to use it; no yet coded with implicit |
| 678 |
|
vertical advection. |
| 679 |
|
- change experiment advect_cs (new output) and advect_xy (same output) |
| 680 |
|
to test compressible flow multi-dim advection. |
| 681 |
|
o pkg/generic_advdiff: |
| 682 |
|
- fix SOM advection for unusual cases (linear Free-Surf in r*, other |
| 683 |
|
cases where linear Free-Surf is not @ k=1, e.g., useShelfIce) |
| 684 |
|
o pkg/seaice: |
| 685 |
|
- fix bug in the no-slip implementation in seaice_calc_strainrates.F, this |
| 686 |
|
bug affects two verification experiments (lab_sea and lab_sea.hb87), |
| 687 |
|
recover old (wrong) results by setting noSlipFac = 2. in |
| 688 |
|
seaice_calc_strainrates.F; |
| 689 |
|
- add another averaging method for etaZ (etaZmethod=3), a mix of 2 and 0 |
| 690 |
|
- add runtime parameter SEAICE_JFNKepsilon for that finite difference |
| 691 |
|
gradient in the Jacobian times vector operation (s/r seaice_jacvec) |
| 692 |
|
- fix line search in seaice_jfnk.F (s/r seaice_jfnk_update) |
| 693 |
|
o pkg/seaice: reorganize seaice_lsr.F to be able to reuse more code: |
| 694 |
|
- new subroutines seaice_lsr_tridiagu/v |
| 695 |
|
- introduce loop bounds i/jmin/max to be able to use a restricted |
| 696 |
|
additive Schwarz method, but not used for now. |
| 697 |
|
- rename SEAICE_CALC_LSR_COEFFS into SEAICE_LSR_CALC_COEFFS |
| 698 |
|
- use new subroutines seaice_lsr_tridiagu/v in s/r seaice_preconditioner |
| 699 |
|
o OpenAD support |
| 700 |
|
Reorganize S/R call for overall time-stepping loop to facilitate |
| 701 |
|
OpenAD revolve loop replacement |
| 702 |
|
(new S/R main_do_loop.F) |
| 703 |
|
o model/src: |
| 704 |
|
- add 3-D array 3rd dimension to argument list of S/R ADAMS_BASHFORTH2/3; |
| 705 |
|
- pkg/cheapaml: remove local S/R ADAMS2d and use instead standard S/R |
| 706 |
|
ADAMS_BASHFORTH2. |
| 707 |
o pkg/cheapaml: |
o pkg/cheapaml: |
| 708 |
- add new parameter (default=F, same as before) to disable full domain |
- add new parameter (default=F, same as before) to disable full domain |
| 709 |
(including land) periodicity in X & Y dir. |
(including land) periodicity in X & Y dir. |
Parent Directory
| 
Revision Log
| 
Revision Graph
| 
Patch