| 4 |
Notes on tags used in MITgcmUV |
Notes on tags used in MITgcmUV |
| 5 |
============================== |
============================== |
| 6 |
|
|
| 7 |
|
o pkg/seaice/autodiff: |
| 8 |
|
- add CPP flag to include clipping of zeta in EVP code; |
| 9 |
|
although clipping of zeta is not necessary for EVP and not recommended |
| 10 |
|
it was done by default, now clipping is disabled by default |
| 11 |
|
- modify computation and store zetaC/Z in seaice_evp for TAF |
| 12 |
|
- add CPP-flag SEAICE_ALLOW_CLIPVELS (undefined by default) |
| 13 |
|
- change seaice_calc_viscosities/strainraties for better |
| 14 |
|
vectorization after AD by TAF |
| 15 |
|
- rearrange common blocks in SEAICE.h once again for easier dumps of |
| 16 |
|
adjoint output, adjust adcommon.h and g_common.h |
| 17 |
|
- change flux form of default advection scheme for better vectorization |
| 18 |
|
of adjoint code |
| 19 |
|
- modify lab_sea/input.lsr to test the flux form of default advection |
| 20 |
|
scheme |
| 21 |
|
o autodiff: fix seaice fields in g_common.h (left over from 3 to 1 level |
| 22 |
|
transition in checkpoint61r) |
| 23 |
|
|
| 24 |
|
checkpoint61x |
| 25 |
|
o adjoint: adding two new verification experiments |
| 26 |
|
* tutorial_global_biogeo: an online dic sensitivity experiment |
| 27 |
|
(work with G. Forget) |
| 28 |
|
* tutorial_dic_adjoffline: an offline dic sensitivity experiment |
| 29 |
|
(an update of the Dutkiewicz et al., 2006 paper) |
| 30 |
|
Both use the same CO2FLUX cost function, but different controls |
| 31 |
|
Gradient checks look good, but some recomp. could still be traightened out |
| 32 |
|
Should add xx_dic control variable to test at some point |
| 33 |
|
o model/src: |
| 34 |
|
add run-time parameters to check and/or mask initial Temp & Salt |
| 35 |
|
o pkg/ptracers: |
| 36 |
|
apply zonal filter (if used) to all passive tracers |
| 37 |
|
|
| 38 |
|
checkpoint61w |
| 39 |
|
o adjoint: |
| 40 |
|
* update and finish work on adjoint of dic package (G. Forget) |
| 41 |
|
* start adjoint of offline package |
| 42 |
|
* bug fix for cube-sphere TLM code (active_file_g.F) |
| 43 |
|
o tutorial_global_oce_optim: |
| 44 |
|
now tested with 4 tiles (change data.grdchk and update output_adm.txt) |
| 45 |
|
and add SIZE.h_mpi for MPI testing |
| 46 |
|
o pkg autodiff |
| 47 |
|
- fix bug in autodiff_restore |
| 48 |
|
- make genmake2 remove comments and empty lines from ad_input_code.f in |
| 49 |
|
order to reduce file size (new script file remove_comments_sed) |
| 50 |
|
o allows to specify vertical profile of vertical viscosity (viscArNr) |
| 51 |
|
o pkg autodiff |
| 52 |
|
autodiff_ini_model_io: comment out writing of grid information to netcdf |
| 53 |
|
files, because at this time the grid fields are a bunch of zeros |
| 54 |
|
(this is analogous to ini_mnc_vars.F) |
| 55 |
|
o pkg seaice/autodiff: |
| 56 |
|
left-over from removing the 3-time-level fields: |
| 57 |
|
- small rearrangement of common blocks (SEAICE.h) |
| 58 |
|
- fix field declaration in order to get meaningful output in |
| 59 |
|
adseaice.*.nc (adcommon.h) |
| 60 |
|
o pkg obcs: |
| 61 |
|
- go through NH code only if nonHydrostatic=T |
| 62 |
|
- fix restart for Orlanski (Hydrostatic; still to fix NH+Orlanski restart) |
| 63 |
|
- move call to OBCS_CALC before SEAICE_MODEL which needs seaice-obc fields |
| 64 |
|
=> Update output of experiment seaice_obcs |
| 65 |
|
- useOBCSbalance: code is wrong with multi-tile / proc: add a stop. |
| 66 |
|
- fix NH+Orlanski restart (add wvel OBC value in pickup file) |
| 67 |
|
- fix restart for useOBCSprescribe + useEXF |
| 68 |
|
o pkg ctrl |
| 69 |
|
- make obcs as control parameter work also with useSingleCPUio |
| 70 |
|
- replace a few sny and snx by Ny and Nx to be consistent with |
| 71 |
|
ctrl_set_globfld_x/yz.F |
| 72 |
|
o pkg seaice |
| 73 |
|
- re-introduce global field frWtrAtm for the case of |
| 74 |
|
ALLOW_MEAN_SFLUX_COST_CONTRIBUTION defined (SEAICE.h, seaice_growth.F) |
| 75 |
|
- catch the case of an empty filename in seaice_cost_init_fixed.F |
| 76 |
|
o pkg ecco: |
| 77 |
|
- small bug fixes in the_main_loop |
| 78 |
|
. fix order of monitor arguments |
| 79 |
|
. change SEAICE_MULTILEVEL to SEAICE_MULTICATEGORY |
| 80 |
|
. move include DYNVARS.h up outside the ALLOW_AUTODIFF_TAMC block so |
| 81 |
|
that ALLOW_ECCO_FORWARD_RUN works with NONLIN_FRSURF |
| 82 |
|
- replace cos(yc*deg2rad) by _rA in cost_mean_heat/saltflux.F, fix the |
| 83 |
|
imbalance diagnostics for wmean_h/sflux = 0 |
| 84 |
|
- add actual values to f_obcsn/s/w/e in ecco_cost_final.F |
| 85 |
|
- catch the case of empty filenames in ecco_cost_init_fixed.F |
| 86 |
|
and ecco_readparms.F |
| 87 |
|
- fix the logic of wsalt=0.and.wsalt2=0 to wsalt=0.or.wsalt2=0 (and theta) |
| 88 |
|
to make this statement meaningfull in ecco_cost_weights.F |
| 89 |
|
o pkg zonal_filt: |
| 90 |
|
- change ZONAL_FILTER S/R interface (allows to filter 2-D fields) |
| 91 |
|
- when using rStar, filter uVel*hFacW instead of uVel |
| 92 |
|
(tested without topography: fix a Pb of growing instability with NLFS) |
| 93 |
|
o pkg/aim/thsice: |
| 94 |
|
-Add capacity to read a Q-Flux to be applied in slab-mixed layer thsice_slab_ocean.F |
| 95 |
|
(works as SST restoring) |
| 96 |
|
o pkg/thsice: |
| 97 |
|
-add specific parameter for salt restoring in slab mixed-layer |
| 98 |
|
(by default equal to the temperature restoring) |
| 99 |
|
|
| 100 |
|
checkpoint61v |
| 101 |
|
o pkg/kpp: |
| 102 |
|
-add double diffusive contributions as a hack, turned off by default |
| 103 |
|
for now, and the code can be excluded with a CPP-flag |
| 104 |
|
EXCLUDE_KPP_DOUBLEDIFF just as EXCLUDE_KPP_SHEAR_MIX |
| 105 |
|
-add a verification experiments to "vermix" to test this code |
| 106 |
|
o pkg/mom_common: |
| 107 |
|
Move computation of length scales to mom_init_fixed.F (avoid recomputation |
| 108 |
|
at each time-steps*levels). |
| 109 |
|
o pkg/layers: |
| 110 |
|
Created a new package for computing volume fluxes in temperature or |
| 111 |
|
[not yet implemented] density layers, e.g. for calculating residual |
| 112 |
|
overturning circulations. |
| 113 |
|
o pkg/diagnostics: |
| 114 |
|
implement a RS type version for the set of DIAGNOSTICS-FILL subroutines |
| 115 |
|
o fix few RS/RL type mismatch (check with ifort 11.1 "-warn all" option) |
| 116 |
|
o pkg/mdsio (& pkg/rw): rework high level S/R interface: |
| 117 |
|
To avoid mixing type (RS/RL) of input/output array argument, |
| 118 |
|
replace single mixed-type array with a pair from each type (RS/RL). |
| 119 |
|
o setting pickupSuff force to start from a pickup (even if nIter0=0): this |
| 120 |
|
is now implemented for all pkgs (previously only done for the main pickup) |
| 121 |
|
|
| 122 |
|
checkpoint61u |
| 123 |
|
o pkg/obcs: |
| 124 |
|
- change OB?eta from _RS to _RL and put them into a separate commen block |
| 125 |
|
- add code to read OB?eta from a file via obcs_prescribe_read: the code |
| 126 |
|
compiles and does not destroy any other experiments, the data is read |
| 127 |
|
properly, but the code is not tested in actual applications |
| 128 |
|
o ini_mnc_vars.F: if available (ALLOW_CAL), use startTime_1 and 2 to |
| 129 |
|
create a meaningful unit attribute for the time coordinate variable |
| 130 |
|
o pkg/ebm: add plenty of missing "_d 0" ; |
| 131 |
|
update results of experiment global_ocean_ebm (fwd + adjoint tests) |
| 132 |
|
Notes: several pieces of code only works for (very) specific set up |
| 133 |
|
(domain size, lat range, etc ...); should be documented & stop if one |
| 134 |
|
of those assumptions is not meet. |
| 135 |
|
o pkg/seaice: add missing "_d 0" in HSNOW initialisation |
| 136 |
|
update results of experiments: lab_sea (3 fwd tests + 2 adjoint tests) |
| 137 |
|
and offline_exf_seaice (seaicetd + adjoint). |
| 138 |
|
o add missing "_d 0" in ShortWave flux penetration (S/R SWFRAC): |
| 139 |
|
update results of experiments: lab_sea (4 fwd tests + 3 adjoint), |
| 140 |
|
seaice_obcs, global_ocean.cs32x15 (icedyn & thsice) and natl_box (x2). |
| 141 |
|
o few changes to pass when compiling with strick checking of arguments |
| 142 |
|
across S/R (allow to find few bugs in S/R calls). |
| 143 |
|
o pkg/seaice: |
| 144 |
|
- seaice_evp: move if-statement out of loop (TEM) at the cost of 11 new |
| 145 |
|
2d-fields, for slightly better performance |
| 146 |
|
|
| 147 |
|
checkpoint61t |
| 148 |
|
o pkg/seaice: |
| 149 |
|
- remove unused variables in seaice_evp |
| 150 |
|
- reduce number of SQRT in seaice_evp and seaice_calc_viscosities |
| 151 |
|
o eeset_parms.F, open_copy_data_file.F & ini_parms.F: |
| 152 |
|
named file replacing SCRATCH files (for TARGET_BGL & TARGET_CRAYXT): |
| 153 |
|
each processor opens a different file (a fix from Matt). |
| 154 |
|
o cg2d: |
| 155 |
|
use simple EXCH (overlap size = 1 and ignore corners), like in cg3d.F : |
| 156 |
|
this reduces number of EXCH calls by 2 when using exch2. |
| 157 |
|
For now, keep previous size array for gc2d_nsa.F to avoid Pb with store |
| 158 |
|
directives and missing simple EXCH (exch_s3d_rx) adjoint code. |
| 159 |
|
o pkg/thsice: |
| 160 |
|
add code for Near Infra-Red albedo (from Jeff) |
| 161 |
|
(turned on with run-time flag: thSIce_calc_albNIR, in data.ice). |
| 162 |
|
o ini_model_io: |
| 163 |
|
move (from packages_readparms.F & initialise_fixed.F) to ini_model_io.F |
| 164 |
|
MNC & MONITOR initialisation calls; move MNC calls from {pkg}_readparms.F |
| 165 |
|
to {pkg}_init_fixed.F for {pkg}= kpp, seaice & shelfice. |
| 166 |
|
|
| 167 |
|
checkpoint61s |
| 168 |
|
o pkg/exch2 (exch2_rxN_cube*): |
| 169 |
|
remove last BARRIER (no need to synchronise after getting data from shared |
| 170 |
|
buffer (get) as long as any change to buffer (put,recv) is between BARRIER) |
| 171 |
|
o change tiling of natl_box (from 2 to 4 tiles) to get it tested with MPI+MTH |
| 172 |
|
and generate new output (for the 2 tests) using default rotationPeriod. |
| 173 |
|
o pkg/exch2: |
| 174 |
|
- always call exch2_*_cube, not exch-1 anymore, if useCubedSphereExchange=F |
| 175 |
|
(was already the case with AUTODIFF_EXCH2 defined) |
| 176 |
|
- add bj in exch2 arrays and S/R. |
| 177 |
|
- can now use pkg/exch2 for regular (non CS) set-up; switch test experiment |
| 178 |
|
MLAdjust to exch2 (identical results except with MPI -> truncation diff); |
| 179 |
|
o add pkg/longstep for ptracer timestep a multiple of U/V/T/S timestep |
| 180 |
|
- adds deltaT argument to the subroutines TIMESTEP_TRACER DWNSLP_APPLY |
| 181 |
|
GAD_ADVECTION GAD_CALC_RHS GAD_IMPLICIT_R GAD_SOM_ADVECT |
| 182 |
|
- verification experiment natl_box.longstep |
| 183 |
|
o build_options: ifort+mpi_aces optfile (following Constantinos advise): |
| 184 |
|
allow different versions of ifort (8 or 9) and different versions |
| 185 |
|
of mpi (mpich or mpich2) to be used with the same optfile. |
| 186 |
|
requires to load a netcdf module and set MPI_INC_DIR (see comments inside |
| 187 |
|
optfile). Change testing scripts (example_scripts/ACESgrid) accordingly. |
| 188 |
|
o pkg/seaice: |
| 189 |
|
fix previous modif (3-level field removed) for old-pickup reading |
| 190 |
|
|
| 191 |
|
checkpoint61r |
| 192 |
|
o pkg/seaice: |
| 193 |
|
- clean up the 3-time levels of UICE,VICE,HEFF,AREA in three steps: |
| 194 |
|
1. introduce fields for time step (N-1) |
| 195 |
|
2. do not use levels 2 and 3 anymore (but keep them) |
| 196 |
|
3. replace 3D versions of UICE,VICE,HEFF,AREA by 2D versions. This |
| 197 |
|
includes changes to pkg/thsice/thsice_get_velocity.F, |
| 198 |
|
model/src/external_forcing_surf.F, pkg/obcs/obcs_apply_seaice.F, and |
| 199 |
|
pkg/obcs/obcs_apply_uvice.F. |
| 200 |
|
- completely remove code for leap-frog time stepping for second order |
| 201 |
|
advection |
| 202 |
|
- adjusted adjoint accordingly; |
| 203 |
|
gradient checks have improved by an order of magnitude! |
| 204 |
|
|
| 205 |
|
checkpoint61q |
| 206 |
|
o pkg/seaice: fix bug in free slip boundary conditons for LSOR solver |
| 207 |
|
- the boundary conditions along open boundaries are unclear, because the |
| 208 |
|
associated masks are set to zero and are thus treated as closed when |
| 209 |
|
applying the slip-boundary conditions. |
| 210 |
|
- move initialisation of grid variables from seaice_init_varia to _fixed |
| 211 |
|
o adding doc/diags_changes.txt |
| 212 |
|
o model/src: |
| 213 |
|
do_the_model_io: rename few pkg S/R which are called from do_the_model_io, |
| 214 |
|
to avoid "_diags" naming if no relation to pkg/diagnostics. |
| 215 |
|
ini_theta,ini_salt: fix (at least compile the code) reading from NetCDF |
| 216 |
|
file and only call PLOT_FIELD_XYZ if debugMode=T. |
| 217 |
|
remove few PLOT_FIELD_XY calls outside initialisation. |
| 218 |
|
remove few unnecessary BARRIER. |
| 219 |
|
o pkg/salt_plume: add SaltPlumeSouthernOcean to allow disabling |
| 220 |
|
of salt_plume package in Southern Ocean |
| 221 |
|
o pkg/thsice: |
| 222 |
|
- fix hOceMxL in overlap (needed for sea-ice advection); |
| 223 |
|
update output of experiment global_ocean.cs32x15.icedyn. |
| 224 |
|
o pkg/seaice: |
| 225 |
|
- make the (placeholder-)tracer IceAge a little more meaningful |
| 226 |
|
by avoiding ridging of age due to convergent ice flow |
| 227 |
|
- add diagnostics for surface temperature, atmospheric and oceanic |
| 228 |
|
heat flux, and atmospheric and oceanic ice growth |
| 229 |
|
|
| 230 |
|
checkpoint61p |
| 231 |
|
o eesupp (global_max & global_sum): |
| 232 |
|
- add one more element to buffer (start at index 0) for GL_MAX/SUM output; |
| 233 |
|
- remove starting & endding barrier (no longer needed) |
| 234 |
|
o pkg/mdsio, writelocal: |
| 235 |
|
works also in multi-threaded when LOCBIN_IO_THREAD_SAFE is undefined |
| 236 |
|
(remove the STOP): uses shared buffer IO to store data from all threads |
| 237 |
|
and then let master-thread write nThreads tiles. |
| 238 |
|
o pkg/mdsio (read/write field): |
| 239 |
|
- tiled IO done in 1 piece (all levels at a time); |
| 240 |
|
- new header file "MDSIO_BUFF_3D.h" with 3-D buffers. |
| 241 |
|
- multi-threaded: allow to read/write local (non-shared) array |
| 242 |
|
(was already working with singleCpuIO ; now works also without); |
| 243 |
|
=> no longer needs barrier call after reading a file (added inside |
| 244 |
|
mds_read/write field). |
| 245 |
|
- move barrier calls outside gather/scatter_2d to mds_read/write_field |
| 246 |
|
o pkg/obcs/exf |
| 247 |
|
- exf_getffieldrec.F: add a few ". _d 0" and |
| 248 |
|
add code to catch the case fldperiod = 0. |
| 249 |
|
- obcs_precribe_read.F: change if statements so that they match |
| 250 |
|
the variable type (_RL) |
| 251 |
|
- add a separate exf_iprec_obcs to exf_set_obcs (by default equal to |
| 252 |
|
exf_iprec) as part of namelist EXF_NML_01 |
| 253 |
|
o pkg/mdsio: |
| 254 |
|
- read/write tiled (local) files: read/write 1-level tile chunk at a time |
| 255 |
|
(instead of segment of length sNx); expected to speed up tiled IO. |
| 256 |
|
o pkg/exch2: |
| 257 |
|
- take buffer copy from/to array out of S/R exch2_send/recv into new |
| 258 |
|
S/R exch2_put/get ; Exch of local variable now works with MPI+MTH |
| 259 |
|
(tested by removing commom block statement in SOLVE_FOR_PRESSURE.h). |
| 260 |
|
- ad version of send/recv no longer needed (but needs exch2_ad_put/ad_get) |
| 261 |
|
- simplify argument list of S/R exch2_get_uv_bounds and exch2_get_scal_bounds |
| 262 |
|
(which replaces exch2_get_recv_bounds & exch2_get_send_bounds) by |
| 263 |
|
including Topology header file (and common blocks). |
| 264 |
|
- implement EXCH_IGNORE_CORNERS in scalar exchange (rx1); |
| 265 |
|
remove 2nd exch call in exch2_s3d_rx (no longer needed). |
| 266 |
|
- document S/R arg. list (+ update some comments) |
| 267 |
|
o pkg/seaice: |
| 268 |
|
- added SEAICE_CLIM_AIR code, which permits modulation of surface air |
| 269 |
|
temperature and humidity over sea ice based on climatological values. |
| 270 |
|
- clean up computation of Hibler+Bryan (1987) stress coupling for the case |
| 271 |
|
of LSR (this change is expected to change the results slightly because now |
| 272 |
|
it uses slightly difference moduli from the second last LSR solution, but |
| 273 |
|
that is more consistent with the stress computations; this part of the code |
| 274 |
|
is not tested in the verification experiments) |
| 275 |
|
- fix diagnostics for flux of ice in case of B-grid |
| 276 |
|
- turn strain rates eij and press into global fields within global |
| 277 |
|
common block in SEAICE.h for more straighforward diagnostics |
| 278 |
|
- fix diagnostics for principal stress components, press, zeta, eta |
| 279 |
|
to give something meaning full |
| 280 |
|
o change tiling of global_ocean.cs32x15 (12 tiles, 32x16) and generate |
| 281 |
|
new output (all 4 tests) using default rotationPeriod. |
| 282 |
|
o pkg KPP: |
| 283 |
|
- add flag (LimitHblStable) to avoid limiting BL depth under stable conditions |
| 284 |
|
- fix vertical indices in RI_IWMIX (match between KPP grid from 0 to Nr+1 |
| 285 |
|
and normal grid from 1 to Nr) |
| 286 |
|
o Fix computations of L4rdt in mom_calc_visc.F |
| 287 |
|
o eesupp: |
| 288 |
|
new S/R ALL_PROC_DIE (like EEDIE, without irrelevant threads err msg): |
| 289 |
|
should be called before a "stop" if we know that all process are going |
| 290 |
|
to stop; if one Proc does call this S/R, will hang in MPI_FINALISE. |
| 291 |
|
|
| 292 |
|
checkpoint61o |
| 293 |
|
o pkg/exch2: change check_depths to issue just a warning (instead of error+stop) |
| 294 |
|
o model/src/ini_curvilinear_grid.F: |
| 295 |
|
compute angleSin/CosC when they are not read from a file |
| 296 |
|
o pkg/seaice: |
| 297 |
|
add pseudo time stepping for LSOR |
| 298 |
|
o pkg/exch2: |
| 299 |
|
temporary fix for exch2_s3d_rx; update results of exp. hs94.cs-32x32x5.impIGW |
| 300 |
|
o pkg/seaice: |
| 301 |
|
Adding separate sea ice albedo, drag coefficients, and HO in Southern |
| 302 |
|
Ocean from those in Arctic Ocean. See *_south in SEAICE_PARAMS.h |
| 303 |
|
o SingleCpuIO: |
| 304 |
|
fix bug put before chkp61n (May 11), when more than 2 MPI cpu are used. |
| 305 |
|
o pkg/seaice: |
| 306 |
|
put exchanged variable in (local) common block for MPI+Multi-Threads |
| 307 |
|
to work (necessary with exch1, and for now with exch2 too). |
| 308 |
|
o verification: to get more experiments tested with mpi, mth and mpi+mth, |
| 309 |
|
change tiling of: |
| 310 |
|
aim.5l_Equatorial_Channel, aim.5l_LatLon, dome, exp4, global_ocean.90x40x15, |
| 311 |
|
hs94.cs-32x32x5, MLAdjust, offline_exf_seaice & tutorial_plume_on_slope |
| 312 |
|
update output of exp: |
| 313 |
|
exp4, global_ocean.90x40x15(+.dwslp), MLAdjust(all 5), |
| 314 |
|
tutorial_plume_on_slope and hs94.cs-32x32x5 (use default rotationPeriod) |
| 315 |
|
+ ideal_2D_ocean: use default rotation period + test Global_Sum_SINGLECPU |
| 316 |
|
(note: output.txt generated on faulks using only 1 tile) |
| 317 |
|
o eesupp & mdsio (SingleCpuIO): |
| 318 |
|
- remove MISSING_TILE_IO (from MDSIO_OPTIONS.h), replaced by run-time |
| 319 |
|
parameter "W2_useE2ioLayOut" (read from file "data.exch2"). |
| 320 |
|
Note: was not effective for globalFile & useSingleCpuIO=F; now fixed. |
| 321 |
|
- move MDSIO_SCPU.h (pkg/mdsio) to EEBUFF_SCPU.h (eesupp/inc). |
| 322 |
|
- fix GLOBAL_SUM_SINGLECPU when using Exch2; re-use same buffers and same |
| 323 |
|
gather/scatter S/R as with SingleCpuIO (=> 1 less 2D global RL array). |
| 324 |
|
o pkg/exch2: |
| 325 |
|
- new code to set-up W2-Exch2 topology (replace matlab-topology-generator): |
| 326 |
|
read parameter file "data.exch2" if it exists; otherwise try default regular |
| 327 |
|
cube without blank-tile; new header file "W2_EXCH2_SIZE.h"; |
| 328 |
|
- update matlab-topology-generator (exch2_setup_cs6_print.m) to be compatible |
| 329 |
|
with new code (not to be used, just as a backup solution). |
| 330 |
|
- add examples (e.g., for lat-lon-cap_120) of parameter file "data.exch2" |
| 331 |
|
in utils/exch2/input. |
| 332 |
|
|
| 333 |
|
checkpoint61n |
| 334 |
|
o eesupp & mdsio (SingleCpuIO): |
| 335 |
|
- move mapping to global io-buffer inside gather_2d/scater_2d: this save |
| 336 |
|
memory (1 less 2D global RL array); skip 1 buffer copy; and only send/ |
| 337 |
|
receive real*4 array (instead of real*8) to read/write 32.bit files. |
| 338 |
|
(new template for gather_2d/scater_2d, new set of S/R MDS_PASS_R4,8toRL,S |
| 339 |
|
and remove new mapping S/R) |
| 340 |
|
o pkg/mdsio: in preparation for topology-generator within pkg/exch2: |
| 341 |
|
- Change declaration of SingleCpuIO buffer, reduced to only 1 dim.: |
| 342 |
|
xyBuffer_size, which is set to W2_ioBufferSize from W2_EXCH2_PARAMS.h |
| 343 |
|
when using exch2. Add S/R to map global io-buffer to global model array. |
| 344 |
|
- uses exch2_global_Nx,y instead of exch2_domain_nxt,nyt. |
| 345 |
|
|
| 346 |
|
checkpoint61m |
| 347 |
|
o write_utils : change WRITE_1D_R8 to WRITE_1D_RL |
| 348 |
|
and add 3 other S/R for other types (R4,R8,RS). |
| 349 |
|
Change calls to WRITE_0D/1D_R[4,8,S,L] to match the type of argument. |
| 350 |
|
o eesupp/inc/CPP_EEMACROS.h (EXCH & GLOBAL_SUM/MAX macros): |
| 351 |
|
- ensure suffix (_R4,_R8,_RS,_RL) correspondance between macro and S/R ; |
| 352 |
|
- reduce number of macro definitions (keep only those which are used); |
| 353 |
|
can always call directly the corrresponding S/R if needed. |
| 354 |
|
- everywhere in the code: change macros (EXCH & GLOBAL_SUM/MAX) suffix |
| 355 |
|
_R4/_R8 to _RS/_RL when applied to _RS/_RL variable. |
| 356 |
|
Fixed wrong type of arg in: |
| 357 |
|
atm2d/put_ocnvars.F cheapaml/cheapaml.F ocn_compon_interf/ocn_apply_import.F |
| 358 |
|
ctrl/ctrl_map_ini_ecco.F seaice/seaice_do_diags.F seaice/seaice_model.F |
| 359 |
|
To fix: global_sum of var in common block is wrong (multi-threaded) in: |
| 360 |
|
dic/dic_atmos.F ebm/ebm_area_t.F ebm/ebm_zonalmean.F sbo/sbo_calc.F |
| 361 |
|
To fix: aim_v23/aim_do_co2.F (BARRIER & GLOBAL_SUM within bi,bj loops) |
| 362 |
|
o adjustment.cs-32x32x1: test "blank tiles" when compiling with mpi; 2nd |
| 363 |
|
"nlfs" mpi test is going to stop when checking for unconnected tile egdes. |
| 364 |
|
o eesupp/src: new S/R to stop everyone when 1 (or more) Proc. find an error |
| 365 |
|
o pkg/exch2: new S/R EXCH2_CHECK_DEPTHS to check that unconnected tile |
| 366 |
|
edges are closed (zero depth); might happen with "blank tiles"; |
| 367 |
|
for now, disabled if using OBCS. |
| 368 |
|
o tools/gemake2: delete modifs from March 27-29 2009 (with option "-mpi": |
| 369 |
|
link *_mpi files from "MODS" dir) to return to a strait-forward behavior. |
| 370 |
|
o move read_write.F to write_utils.F ; new parameter (maxLengthPrt1D) |
| 371 |
|
to reduce length of 1.D array print in STDOUT. |
| 372 |
|
o adjustment.cs-32x32x1: changed to oceanic test with 1 continent |
| 373 |
|
+ use 48 tiles (16x8): this will allow to test "blank tiles". |
| 374 |
|
o start to test MPI + multi-threaded (change testreport to skip tests |
| 375 |
|
where number of tiles do not match). |
| 376 |
|
o genmake2: add option "-omp" to turn on Open-MP ; |
| 377 |
|
passed to testreport (if using -mth) ; merge few (*+mth) optfiles. |
| 378 |
|
o pkg/exch2 (+ utils/exch2): ignore blank tiles (instead of removing them) |
| 379 |
|
and keep the original tile numbering. Note: compatible with old |
| 380 |
|
W2_EXCH2_TOPOLOGY.h & w2_e2setup.F topology files. |
| 381 |
|
|
| 382 |
|
checkpoint61l |
| 383 |
|
o Overlaps had been forgotten in calculating ijk keys : fixed |
| 384 |
|
o fix cyrus makedepend for gcc 4.4 |
| 385 |
|
o experiment fizhi-cs-aqualev20: |
| 386 |
|
switching to new version of S/R GETPWHERE (#define TRY_NEW_GETPWHERE) |
| 387 |
|
and generate new output. Note: new output is much closer to the old |
| 388 |
|
one (v1.8) generated with pgf77 -O2 on Mar 14 2007 than it is from |
| 389 |
|
the previous one (v1.9) generated with pgf77 -O0 on Jun 13 2007. |
| 390 |
|
o pkg/fizhi: |
| 391 |
|
- change to avoid indices going deliberately over array-bounds |
| 392 |
|
- remove unnecessary MPI stuff |
| 393 |
|
- perpetual spring equinox with "#define FIZHI_USE_FIXED_DAY" |
| 394 |
|
- new version of S/R GETPWHERE available with #define TRY_NEW_GETPWHERE: |
| 395 |
|
does not go over array-bounds and therefore returns reliable output. |
| 396 |
|
- in fizhi_driver.F & fizhi_turb.F, fix "get_alarm" calls: |
| 397 |
|
this S/R has 4 output arguments, which are used inside "get_alarm"; |
| 398 |
|
when it is called with the same variable "ndum" in place of several |
| 399 |
|
output arguments, this produces weird side effect inside "get_alarm" |
| 400 |
|
(e.g., datein=dates(n) is overwritten by timein=times(n) since both |
| 401 |
|
datein & timein, resp. 2nd and 3rd arg., are pointed to the same "ndum" |
| 402 |
|
address) and produces error. |
| 403 |
|
o tools/gemake2: (March 27-29 2009) |
| 404 |
|
with option "-mpi": link *_mpi files from "MODS" dir (this is what |
| 405 |
|
testreport does). |
| 406 |
|
o lab_sea experiment: change from 2x1 tiles to 2x2 tiles and regenerate output |
| 407 |
|
o pkg/seaice: fix bug in new version of S/R SEAICE_LSR.F: |
| 408 |
|
update lab_sea.lsr & lab_sea.salt_plume results |
| 409 |
|
o pkg/seaice: |
| 410 |
|
Major overhaul of the LSOR solver: change discretization from original |
| 411 |
|
finite differences to finite volume. The old routine is still in the |
| 412 |
|
file seaice_lsr.F and can be turned on with the CPP-flag |
| 413 |
|
SEAICE_OLD_AND_BAD_DISCRETIZATION |
| 414 |
|
Along with this change, the discretization of the strain rates is changed. |
| 415 |
|
The same flag turns on the old discretization in seaice_calc_strainrates.F |
| 416 |
|
All verification experiments with seaice are affected (except for |
| 417 |
|
seaice_exf_offline); global_ocean.cs32x15/input.icedyn recovers the old |
| 418 |
|
results by setting the above flag. |
| 419 |
|
Further changes involve changes in defaults: |
| 420 |
|
- ALLOW_FLOODING is defined and turned on by default |
| 421 |
|
- SEAICEadvSnow = .true. is now the default |
| 422 |
|
- SEAICEadvSalt = .true. is now the default |
| 423 |
|
- SEAICEadvAge = .true. is now the default |
| 424 |
|
- SEAICE_clipVelocities = .false. is now the default |
| 425 |
|
(as per J. Zhang's recommendation) |
| 426 |
|
- B-grid, and thus not tested: SEAICE_TEST_ICE_STRESS_1/EXPLICIT_SSH_SLOPE |
| 427 |
|
is defined, SEAICE_TEST_ICE_STRESS_1 is renamed into SEAICE_BICE_STRESS |
| 428 |
|
- seaice_growth: replace computation of UG by a simple copy from wspeed |
| 429 |
|
|
| 430 |
|
checkpoint61k |
| 431 |
|
o pkg/seaice: a few bug fixes mostly concerning the metric terms: |
| 432 |
|
seaice_lsr: fixed a few metric terms, |
| 433 |
|
affects lab_sea.lsr, lab_sea.salt_plume, seaice_obcs |
| 434 |
|
add more debugging diagnostics |
| 435 |
|
seaice_evp: removed superfluous many metric terms of FV discretization |
| 436 |
|
of stress divergence |
| 437 |
|
affects lab_sea, lab_sea.hb87 |
| 438 |
|
seaice_ocean_stress: removed superfluous many metric terms of FV discretization |
| 439 |
|
of stress divergence for non-evp solver |
| 440 |
|
(this part is not tested anywhere) |
| 441 |
|
seaice_calc_strainrates: fixed a bug in no-slip boundary conditions, |
| 442 |
|
they were effectively only half slip |
| 443 |
|
affects lab_sea.hb87 |
| 444 |
|
updated all relevant verification experiments (except lab_sea//output_adm.txt |
| 445 |
|
which was left over, but fails at level 7 & 6) |
| 446 |
|
|
| 447 |
|
checkpoint61j |
| 448 |
|
o pkg/seaice: turn a few hard wired parameters into run time parameters |
| 449 |
|
o pkg/seaice/exf/monitor: turn netcdf coordinate variable T into model |
| 450 |
|
time in seconds as promised by the usints in seaice_monitor and exf_monitor |
| 451 |
o pkg/mnc: fix the coordinate variables for multi-tile curvi-linear grids |
o pkg/mnc: fix the coordinate variables for multi-tile curvi-linear grids |
| 452 |
that are not cubed-sphere grids |
that are not cubed-sphere grids |
| 453 |
o pkg/seaice: fix a bug for the rare case of SEAICE_VECTORIZE_LSR |
o pkg/seaice: fix a bug for the rare case of SEAICE_VECTORIZE_LSR |
Parent Directory
| 
Revision Log
| 
Revision Graph
| 
Patch