| 4 |
Notes on tags used in MITgcmUV |
Notes on tags used in MITgcmUV |
| 5 |
============================== |
============================== |
| 6 |
|
|
| 7 |
|
o pkg/obcs + seaice-dynamics: |
| 8 |
|
- new implementation calls OBCS_APPLY_UVICE (in seaice_dynsolver.F) before |
| 9 |
|
solving for uIce,vIce and prevent SEAICE_LSR or SEAICE_EVP to modify OB |
| 10 |
|
values (consistent with an implicit method). |
| 11 |
|
- add option OBCS_UVICE_OLD (in OBCS_OPTIONS.h, defined by default) to |
| 12 |
|
return to older implementation (same results as before); note that none of |
| 13 |
|
the "experimental" OBCS_SEAICE options are implemented in the new version. |
| 14 |
|
- switch test exp. seaice_obcs to use new version (update output). |
| 15 |
|
o pkg/seaice: |
| 16 |
|
- add few stop (B-grid & CS-exch, B-grid & obcs, AUTODIFF & SEAICE_maskRHS) |
| 17 |
|
- add parenthesis in Water-Ice drag calculations (evp & lsr); |
| 18 |
|
- in seaice_lsr.F, simplify and fix AA3 setting in case sNx or sNy=1; |
| 19 |
|
and fix case SEAICE_waterDrag=0 (but unstable if no or little seaice) |
| 20 |
|
- update output of test exp lab_sea that uses evp or lsr (fwd+ad) |
| 21 |
|
and global_ocean.cs32x15 (icedyn & seaice fwd+ad). |
| 22 |
|
o pkg/layers: |
| 23 |
|
- Include potential density as new coordinate (Thanks to David Munday) |
| 24 |
|
o pkg/dic: |
| 25 |
|
- Remove subroutine CALC_PCO2_APPROX_CO3 from carbon_chem.F and add |
| 26 |
|
carbonate computation/output to CALC_PCO2_APPROX |
| 27 |
|
o model/src: |
| 28 |
|
- improve vectorization also for adjoint code of |
| 29 |
|
solve_tridiagonal and solve_pentadiagonal. As this requires up to |
| 30 |
|
5 new 3D-fields, this code is only turned on for the combination of |
| 31 |
|
ALLOW_AUTODIFF_TAMC and TARGET_NEC_SX. |
| 32 |
|
- remove hard-wiring of CPP-flag ALLOW_SOLVERS_KLOOPINSIDE to |
| 33 |
|
ALLOW_AUTODIFF. This flag now needs to be set outside of |
| 34 |
|
solve_tri/pentadiagonal.F |
| 35 |
|
o pkg/exf: |
| 36 |
|
- fix previous modif in exf_readparms.F (revision 1.66: treatment of |
| 37 |
|
optional namelist EXF_NML_OBCS). |
| 38 |
|
- add many debug-print (to trace down where pkg/cal error msg comes from) |
| 39 |
|
o pkg/mom_vecinv/mom_vi_hdissip.F |
| 40 |
|
- rearrange loops so that vectorization is improved and in particular |
| 41 |
|
the adjoint routine vectorizes. The code is a little longer now, but I |
| 42 |
|
think easier to read. |
| 43 |
|
Note: CPP-flag MOM_VI_ORIGINAL_VISCA4 is not tested. |
| 44 |
|
o pkg/seaice: |
| 45 |
|
- allow to specify initial seaice velocity from files (uIceFile,vIceFile) |
| 46 |
|
o pkg/exch2: |
| 47 |
|
- alternative code (without tile x tile matrix) for cumulated (global) sum; |
| 48 |
|
used as default; keep using matrix version in adjustment.cs-32x32x1. |
| 49 |
|
o tools/build_options: |
| 50 |
|
- after merging some MPI and/or multi-threads specific optfiles so that |
| 51 |
|
the modified default optfile support both built (using genmake2 -mpi |
| 52 |
|
for the MPI built, using genmake2 -omp for OpenMP built), |
| 53 |
|
move now obsolete optfiles to tools/build_options/unsupported/ |
| 54 |
|
(add copy of email sent to support list in doc/optfiles_changes.txt) |
| 55 |
|
|
| 56 |
|
checkpoint63c (2011/09/28) |
| 57 |
|
o pkg/ebm: |
| 58 |
|
- fix ebm_zonalmean.F for several procs in Ydir (nPy > 1); update output of |
| 59 |
|
test exp. global_ocean_ebm (affects machine truncation, 12 digits for cg2d) |
| 60 |
|
o pkg/seaice (SItracer pickup) |
| 61 |
|
* Simple code for pickup of SItracer fields (#define ALLOW_SITRACER) |
| 62 |
|
To be done: |
| 63 |
|
- thorough testing of missing fields in S/R seaice_read_pickup |
| 64 |
|
- thorough matching between generic tracers and specific fields used |
| 65 |
|
(e.g. volume-weighted age, area-weighted age, salinity, ...) |
| 66 |
|
o pkg/grdchk: |
| 67 |
|
- all procs calculate and print finite-difference gradient (for testreport) |
| 68 |
|
- change description output used by testreport; |
| 69 |
|
- AD testing: also compare and report Finite-difference gradient; |
| 70 |
|
=> update all output_adm*.txt |
| 71 |
|
o test exp. global_ocean_ebm: |
| 72 |
|
- switch back useRealFreshWaterFlux (no problem anymore with run-off); |
| 73 |
|
and turn on exactConserve; update (fwd) output. |
| 74 |
|
o pkg/seaice (LSR): |
| 75 |
|
- print to STDOUT number of solver iterations at SEAICE_monFreq interval. |
| 76 |
|
- modify solver to deal cleanly with u/v switch at face edges (CS-grid): |
| 77 |
|
* identical results with non-CS topology (but save up to half of EXCH calls) |
| 78 |
|
* no much changes in cs-32 verification tests since no ice at face edges |
| 79 |
|
(differences due to convergence criteria, since nb of iterations changed) |
| 80 |
|
* expect some differences on lat-lon-cap grid. |
| 81 |
|
update FWD results of global_ocean.cs32x15 test exp. (icedyn & seaice) |
| 82 |
|
o pkg/exch2: |
| 83 |
|
- add new CPP option to disable Cumul-Sum matrix (to save storage when using |
| 84 |
|
many tiles); until alternative coding is added, just add a stop. |
| 85 |
|
o pkg/kpp: |
| 86 |
|
- move ptracer/salt correction to KPP diffusivity from calc_3d_diffusivity.F |
| 87 |
|
to S/R KPP_CALC_DIFF_PTR (+ add tracer-index to argument list). |
| 88 |
|
o optfiles: |
| 89 |
|
- merged few MPI specific (*+mpi*) and multi-threads specific (*+mth) optfiles |
| 90 |
|
into corresponding standard one, which now support different build: |
| 91 |
|
(when using "genmake2 -mpi" for MPI built, "genmake2 -omp" for OpenMP built) |
| 92 |
|
- add "darwin_amd64_gfortran" optfile (initially from Constantinos, tested by Gael) |
| 93 |
|
o pkg/obcs (seaice): |
| 94 |
|
- add missing EXCH for uIce,vIce after applying OBCS; |
| 95 |
|
- remove S/R EXF_SWAPFFIELDS_X/Y , disable S/R EXF_SET_OBCS_X/Y by adding |
| 96 |
|
argument = number of levels to S/R EXF_SWAPFFIELDS_XZ/YZ and S/R |
| 97 |
|
EXF_SET_OBCS_XZ/YZ ; update obcs_exf_load.F; |
| 98 |
|
- fix EXF_SET_OBCS_XZ/YZ for multi-threaded (assuming loaded arrays are shared) |
| 99 |
|
- verification exp. seaice_obcs: split domain in 2 tiles ; add SIZE.h_mpi |
| 100 |
|
and eedata.mth for MPI and mutlti-threaded testing. |
| 101 |
|
o pkg/flt: |
| 102 |
|
- make FLT EXCH working with multi-threads (and MPI + multi-threads); |
| 103 |
|
- use buffer for floats IO (trajectories & profiles); size set in FLT_SIZE.h; |
| 104 |
|
- fix pickup and trajectories + profiles output with multi-threads; |
| 105 |
|
- add parameter to select group of fields to write along trajectory and |
| 106 |
|
profile (default unchanged); |
| 107 |
|
- only call PORT_RAND* function when flt_noise <> 0; |
| 108 |
|
- add multi-threaded test for verification exp. flt_example. |
| 109 |
|
o eesupp/src: |
| 110 |
|
- fix S/R ALL_PROC_DIE for MPI+multi-threads. |
| 111 |
|
- USE_OMP_THREADING: check that number of threads (from eedata) matches |
| 112 |
|
actual number (from env.var. setting). |
| 113 |
|
o pkg/ebm: |
| 114 |
|
- fix multi-threading in ebm_area_t.F ; clean-up EBM.h header file. |
| 115 |
|
- add multi-threaded test for verification exp. global_ocean_ebm |
| 116 |
|
|
| 117 |
|
checkpoint63b (2011/08/29) |
| 118 |
|
o pkg/ebm: |
| 119 |
|
- fix sign of runoff; |
| 120 |
|
- update directly forcing fields (Qnet,EmPmR) to use standard |
| 121 |
|
external_forcing_surf.F (remove ebm_forcing_surf.F); |
| 122 |
|
- avoid loading forcing fields a 2nd time (1rst time from ini_forcing.F); |
| 123 |
|
- set either tauThetaClimRelax (in data) or tauThetaZonRelax (in data.ebm) |
| 124 |
|
and apply relaxation towards corresponding SST field (2-D map or ZonAver). |
| 125 |
|
- update global_ocean_ebm parameters file data and update output (fwd+ad). |
| 126 |
|
o verification/matrix_example: |
| 127 |
|
- split domain in 8 tiles (2x4); add SIZE.h_mpi and eedata.mth for |
| 128 |
|
MPI and mutlti-threaded testing. |
| 129 |
|
o pkg/matrix: |
| 130 |
|
- change name of S/R MATRIX_INIT to S/R MATRIX_INIT_VARIA ; |
| 131 |
|
- fix for multi-tiles (add bi,bj indices to exp/impMatrixCounter); |
| 132 |
|
- fix for multi-threaded (remove DATA statement); |
| 133 |
|
- use standard DIFFERENT_MULTIPLE function; |
| 134 |
|
- add MATRIX_OPTIONS.h |
| 135 |
|
o verification/matrix_example: |
| 136 |
|
To allow to split the domain in several tiles, change the domain size |
| 137 |
|
from 31x31 to 32x32 by adding 1 row (on Southern side) + 1 column (on |
| 138 |
|
Western side) of land points. |
| 139 |
|
o (*_)write_pickup: |
| 140 |
|
- fix few bugs in mypackage_read/write_pickup.F |
| 141 |
|
- in MDS_WR_METAFILES calls, replace "myTime" by a local copy into |
| 142 |
|
array of length 1 (was caught by compiler syntax-check). |
| 143 |
|
o pkg/diagnostics (velocity potential solver): |
| 144 |
|
- fix initialisation of main diagonal matrix (for overlap values) |
| 145 |
|
o tools/build_options: |
| 146 |
|
- towards single optfile for MPI and non-MPI built: merge few +mpi optfile |
| 147 |
|
into previously non MPI one (linux_am64 pgf77,open64,pathf90,ifort11) |
| 148 |
|
- linux_amd64_gfortran: merge +mpi_generic version into standard optfile: |
| 149 |
|
* move IEEE flags to DEVEL section; take IEEE flags from +mpi_generic ; |
| 150 |
|
* FAST case: for now, keep -O3 with ini_masks_etc.F in NOOPTFILES list; |
| 151 |
|
and use "-devel" testreport option for standard testing on baudelaire. |
| 152 |
|
o pkg/grdchk: |
| 153 |
|
- Assume that all gradient-check position are within the same tile |
| 154 |
|
corresponding to global indices (iGloTile,jGloTile) and that |
| 155 |
|
positions (i,j indices or nbeg) are local to this tile; |
| 156 |
|
- set tile local-indices and proc. number according to tile global-indices; |
| 157 |
|
- remove grdchkwhichproc from namelist (retired param); |
| 158 |
|
- send adjoint gradient to all procs so that all procs print it to STDOUT; |
| 159 |
|
- switch on MPI test for global_ocean.90x40x15 AD test experiment. |
| 160 |
|
o adding package bbl (Bottom Boundary Layer) |
| 161 |
|
description in MITgcm/pkg/bbl/bbl_description.tex |
| 162 |
|
example/test experiment in MITgcm_contrib/bbl |
| 163 |
|
o adjoint verifications |
| 164 |
|
- changed 5 experiments to multi-tiles |
| 165 |
|
* global_ocean.90x40x15 |
| 166 |
|
* obcs_ctrl |
| 167 |
|
* offline_exf_seaice |
| 168 |
|
* tutorial_dic_adjoffline |
| 169 |
|
* tutorial_tracer_adjsens |
| 170 |
|
changed reference output for those that crossed tile boundaries |
| 171 |
|
(changed choices of gradient check positions) |
| 172 |
|
- with same tiling, switch on MPI test for the last 4 (add SIZE.h_mpi) |
| 173 |
|
- temporary bug fix in grdchk_loc.F for obcs |
| 174 |
|
o genmake2 & testreport |
| 175 |
|
- add target "obj" (just for OBJFILES files) to Makefile; |
| 176 |
|
- add option to specify FEXTRAFLAGS on command line. |
| 177 |
|
o pkg/mypackage: |
| 178 |
|
- remove bi,bj arguments from argument list of S/R MYPACKAGE_CALC_RHS |
| 179 |
|
|
| 180 |
|
checkpoint63a (2011/08/02) |
| 181 |
|
o pkg/diagnostics (velocity potential): |
| 182 |
|
- fix velocity-potential calculation when using OBCS: balance net inflow |
| 183 |
|
at each level (new S/R OBCS_DIAG_BALANCE). |
| 184 |
|
- fix some of the problems with OBCS in stream-function calculation. |
| 185 |
|
o pkg/exf/exf_readparms.F |
| 186 |
|
- make treatment of optional namelists (EXF_NML_04, EXF_NML_SGRUNOFF, |
| 187 |
|
EXF_NML_OBCS) more consistent, so that you need not have these lists |
| 188 |
|
in data.exf when they are not used. |
| 189 |
|
o optim/optim_readdata.F: |
| 190 |
|
- add some output about obcs variables that might help debugging |
| 191 |
|
- add a check about nvarlength |
| 192 |
|
o pkg/ctrl: once again, fix computation of endrec in s/r ctrl_init_rec |
| 193 |
|
o pkg/diagnostics (velocity potential): |
| 194 |
|
- fix mask for OBCS (still problems in stream-function with OBCS); |
| 195 |
|
- add specific parameter (default = main code CG2D params) for solver; |
| 196 |
|
- in case of poor convergence, use solution corresponding to lowest residual. |
| 197 |
|
o pkg/seaice/seaice_growth.F: small changes that permit vectorizable |
| 198 |
|
adjoint code. In general statements like tmpscal=fld(i,j) are problematic |
| 199 |
|
and should be avoided where possible. |
| 200 |
|
o model/src (rStar): |
| 201 |
|
- include r* effect on reference-state (PhiRef) in totPhiHyd |
| 202 |
|
(so that it's really the potential anomaly at the cell center); |
| 203 |
|
affect solution only if z* with full pressure in EOS. |
| 204 |
|
- test-exp global_ocean.90x40x15: reduce viscA4 (was not stable) and |
| 205 |
|
update output (because of z* in totPhiHyd and viscA4 changes); |
| 206 |
|
- keep the previous diagnostics as "PHIHYDcR" (closer to potential anomaly |
| 207 |
|
@ constant r) until interpolation at constant r is implemented. |
| 208 |
|
o pkg/diagnostics |
| 209 |
|
- fix RCENTER diagnostics in case cell center is not at interface middle |
| 210 |
|
o add code for TEOS-10 (www.teos-10.org, McDougall et al. 2011). Use |
| 211 |
|
this eos with eosType = 'TEOS10', in data (PARM01). This eos implies |
| 212 |
|
that THETA and SALT are "conservative temperature" and "absolute |
| 213 |
|
salinity" |
| 214 |
|
o pkg/ctrl |
| 215 |
|
- fix computation of endrec in s/r ctrl_init_rec |
| 216 |
|
- ctrl_pack/unpack_*: remove unused variables and some obsolete code |
| 217 |
|
o pkg/seaice/seaice_growth.F |
| 218 |
|
- fix recomputation with FENTY_AREA_EXPANSION_CONTRACTION code, |
| 219 |
|
remove then obsolete STORE directives |
| 220 |
|
- fix STORE directives for tice and tices, so that only the |
| 221 |
|
appropriate part of the fields are stored (and not the entire |
| 222 |
|
field); update global_ocean.cs32x15.seaice AD output. |
| 223 |
|
- cosmetic changes: adjust CPP flags and indentation for better |
| 224 |
|
legibilty |
| 225 |
|
o pkg/autodiff: add mnc-output of adBottomDragFld (code contributed by |
| 226 |
|
Chris Wilson) |
| 227 |
|
o pkg/gmredi: |
| 228 |
|
Add Sub-Meso Eddies parameterisation from Baylor (for now, disabled with TAF), |
| 229 |
|
and test it in verification exp. "front_relax.bvp" (update output.bvp.txt) |
| 230 |
|
|
| 231 |
|
checkpoint63 (2011/07/11) |
| 232 |
|
o pkg/exch2: |
| 233 |
|
- rename+move: nTiles in W2_EXCH2_PARAMS.h --> exch2_nTiles in W2_EXCH2_TOPOLOGY.h |
| 234 |
|
- exch2 version of Cumul-Sum routine: S/R W2_SET_MAP_CUMSUM set up the |
| 235 |
|
(2*nTiles)x(nTiles) Cumul-Sum matrix and Master Proc collects all tiles X,Y |
| 236 |
|
increments, calculates Cumul-Sum @ tile origin and sends results to each tile. |
| 237 |
|
o pkg/diagnostics: |
| 238 |
|
- deal with pairs of post-processed diags which are computed together; in |
| 239 |
|
particular, re-order list of diags within 1 file to save 2nd computation. |
| 240 |
|
- add S/R to compute velocity stream-function (called after velocity-potential |
| 241 |
|
calculation) from divergence free transport -> new diagnostics "PsiVEL". |
| 242 |
|
- add parameter to select grid-point location where PsiVEL == 0. |
| 243 |
|
o eesupp/src: |
| 244 |
|
new S/R to calculate cumulated sum on tiled array, corner grid-cell location |
| 245 |
|
(specific S/R needed with exch2 topology - for now, just print an error msg) |
| 246 |
|
o pkg/shelfice |
| 247 |
|
- Implement friction velocity-dependent transfer coefficients following |
| 248 |
|
Holland and Jenkins, JPO, 1999 |
| 249 |
|
Original code by M. Losch with small modifs. |
| 250 |
|
Enable with CPP option SHI_ALLOW_GAMMAFRICT |
| 251 |
|
and runtime flag SHELFICEuseGammaFrict |
| 252 |
|
o verification lab_sea & global_ocean.cs32x15: |
| 253 |
|
- use tools/xmakedepend (instead of cyrus_makedepend, in genmake_local): |
| 254 |
|
does not miss some dependencies (but is slower) and also works with many files |
| 255 |
|
(default makedepend limitation for AD built). |
| 256 |
|
- use cyrus_makedepend in adjustment.128x64x1 (so that it still tested). |
| 257 |
|
o pkg/diagnostics: |
| 258 |
|
- add run-time parameter "useMissingValue" (def=False) to fill land-point |
| 259 |
|
(i.e., where mask=0) with MissingValue ; used only in MNC output file. |
| 260 |
|
This replaces CPP-option DIAGNOSTICS_MISSING_VALUE. |
| 261 |
|
- Extend the use of "missing_value" setting from data.diagnostics for the case |
| 262 |
|
of counter-diagnostics (where fraction-weight is zero). This applies both to |
| 263 |
|
MNC and MDS output files. |
| 264 |
|
o pkg/seaice/seaice_growth.F: |
| 265 |
|
- add 'a priori' cap of latent heat flux based on actual mass considerations |
| 266 |
|
when SEAICE_ADD_SUBLIMATION_TO_FWBUDGET is used in conjunction with the |
| 267 |
|
non-legacy seaice_solve4temp. |
| 268 |
|
- add diagnostic of the residual sublimation freshwater flux (the freshwater |
| 269 |
|
flux taken from the ocean if the total actual latent heat fluxes exceed |
| 270 |
|
the maximum) |
| 271 |
|
- remove the CPP switch around the *_FWbySublim unit change and move diagnostic |
| 272 |
|
fill for SIatmQnt, SIfwSubl, SIatmFW to the end of seaice_growth.F. |
| 273 |
|
- add diagnostic of the actual sublimation freshwater flux (that is 0. ifndef |
| 274 |
|
SEAICE_ADD_SUBLIMATION_TO_FWBUDGET) and of the latent heat flux (evap+sublim). |
| 275 |
|
- bug fix : frWtrAtm needs a_FWbySublim ifdef SEAICE_ADD_SUBLIMATION_TO_FWBUDGET. |
| 276 |
|
- bug fix : missing part to Martin's r_FWbySublim pathological case fix. |
| 277 |
|
o verification 1D_ocean_ice_column: |
| 278 |
|
made model domain truly 1x1x23 instead of 5x5x23 with only center cell (3,3,:) wet |
| 279 |
|
update results |
| 280 |
o pkg/diagnostics: |
o pkg/diagnostics: |
| 281 |
- Implement setting of "Post-Processed" diagnostics (corresponding to gdiag(5)='P') |
- Implement setting of "Post-Processed" diagnostics (corresponding to gdiag(5)='P') |
| 282 |
which are not filled-up but computed from other diags ; In this case, the mate |
which are not filled-up but computed from other diags ; In this case, the mate |
| 287 |
of fields for this time period, then the next time reccord (it used to be |
of fields for this time period, then the next time reccord (it used to be |
| 288 |
one field with the full set of time records then the next field). |
one field with the full set of time records then the next field). |
| 289 |
Note: this is now similar to NetCDF (MNC) output file. <== |
Note: this is now similar to NetCDF (MNC) output file. <== |
|
- Extend the use of "missing_value" setting from data.diagnostics for the case |
|
|
of counter-diagnostics (where fraction-weight is zero). This applies both to |
|
|
MNC and MDS output files. |
|
| 290 |
|
|
| 291 |
checkpoint62z (2011/06/21) |
checkpoint62z (2011/06/21) |
| 292 |
o adjoint monitor: |
o adjoint monitor: |
Parent Directory
| 
Revision Log
| 
Revision Graph
| 
Patch