| 4 |
Notes on tags used in MITgcmUV |
Notes on tags used in MITgcmUV |
| 5 |
============================== |
============================== |
| 6 |
|
|
| 7 |
|
o eesupp: |
| 8 |
|
- add CPP option to activate some pieces of code for coupling to GEOS AGCM; |
| 9 |
|
- shift commonly used IO unit (in eeboot.F) for coupling to GEOS AGCM. |
| 10 |
|
o pkg/bling: |
| 11 |
|
- bling_production: Fix typo in calcium carbonate uptake. Also rearranging |
| 12 |
|
and cleaning |
| 13 |
|
o model/src: |
| 14 |
|
- move k loop inside correction_step.F (+ update calling S/R) |
| 15 |
|
- collect implicit pressure gradient tendencies in local 2-D array; |
| 16 |
|
affect results (at machine truncation level) when using 3-D solver; update |
| 17 |
|
output of exp.: exp4 (x1), global_ocean.cs32x15 (x2),hs94.cs-32x32x5.impIGW, |
| 18 |
|
rotating_tank, short_surf_wave,tutorial_deep_convection (x2) |
| 19 |
|
& tutorial_plume_on_slope. |
| 20 |
|
|
| 21 |
|
checkpoint66j (2017/08/15) |
| 22 |
|
o pkg/ptracers: |
| 23 |
|
- fix surface forcing under ice-shelf (useShelfIce=T). |
| 24 |
|
o pkg/atm_phys: |
| 25 |
|
- use hFac weighted average when applying A-grid atm-phys momentum tendency |
| 26 |
|
to C-grid dynamics; this ensures wind-stress match mom tendency. |
| 27 |
|
o eesupp/src: |
| 28 |
|
- change default for opening scratch files to avoid STATUS='SCRATCH' |
| 29 |
|
- close scratch files with STATUS='DELETE' |
| 30 |
|
- changes affect eeset_parms.F and open_copy_data_file.F and many routines |
| 31 |
|
that read namelists (because they close the scratch files) |
| 32 |
|
- you can revert to old default by setting USE_FORTRAN_SCRATCH_FILES |
| 33 |
|
in CPP_EEOPTIONS.h (tested in ideal_2D_oce/code & lab_sea/code_ad) |
| 34 |
|
o pkg/atm_phys: |
| 35 |
|
- fix unset arguments in (unused) S/R surface_flux_0d. |
| 36 |
|
o eesupp/src: |
| 37 |
|
- add argument "procId" to S/R EESET_PARMS to fix scratchFile1 & 2 setting |
| 38 |
|
(case TARGET_BGL or TARGET_CRAYXT) for early call (from eeboot_minimal.F): |
| 39 |
|
called with myProcId in eeboot.F and with mpiMyWId in eeboot_minimal.F |
| 40 |
|
o pkg/streamice: |
| 41 |
|
- remove blank between _EXCH_XY_RL and "()", which my Apple Xcode |
| 42 |
|
preprocessor /usr/bin/cpp cannot interpret; 3 instances in total in |
| 43 |
|
streamice_driving_stress_fd.F and streamice_advect_2dtracer.F |
| 44 |
|
o pkg/diagnostics: |
| 45 |
|
- new option to allow to cumulate thickness-factor (hFac) weighted field |
| 46 |
|
(if field permitted); implemented by changing sign of "jdiag" pointer |
| 47 |
|
(> 0: cumlulate field, < 0: cumulate hFac weighted field). |
| 48 |
|
a) restricted to 3-d (Nr levels), level centered diagnostics at U,V or Cell |
| 49 |
|
center position (gcode(2)='U','V','M' & gcode(3,5,9:10)='R',' ','MR') |
| 50 |
|
b) for now, activated by setting fileFlags 3rd character to 'h'. |
| 51 |
|
|
| 52 |
|
checkpoint66i (2017/07/18) |
| 53 |
|
o pkg/cheapaml: |
| 54 |
|
- add parameter "useRelativeWind", default set to False; this change the |
| 55 |
|
previous setting which was always using relative wind. |
| 56 |
|
o add hooks for the Parallel Data Assimilation Framework PDAF |
| 57 |
|
(http://pdaf.awi.de/) in eeboot_minimal.F and the_main_loop.F; |
| 58 |
|
to use set CPP-flag USE_PDAF in CPP_EEOPTIONS.h |
| 59 |
|
o pkg/exf: |
| 60 |
|
- interpolation of vector fields (exf_interp_uv.F): fix bug in calculation |
| 61 |
|
of South-Pole mean value. |
| 62 |
|
- interpolation near the poles, case where second additional row is at the |
| 63 |
|
pole (or beyond the pole and moved to the pole): change first addition row |
| 64 |
|
value to a linear interpolation between pole and 1rst (S.pole)/last (N.pole) |
| 65 |
|
row (instead of just a copy of it). |
| 66 |
|
o pkg/seaice: |
| 67 |
|
- prevent JFNK and KRYLOV solvers from trying to reduce a residual of 0 |
| 68 |
|
- add the option to compute no-slip du/dy, dv/dx (i.e. |
| 69 |
|
the offdiagonal strain rate e12) by a second order approximation |
| 70 |
|
on the boundary; works only with JFNK, KRYLOV, or EVP solvers, because |
| 71 |
|
it is too messy to implement for the implicit LSR matrix. |
| 72 |
|
SEAICE_2ndOrderBC = .FALSE. by default |
| 73 |
|
|
| 74 |
|
checkpoint66h (2017/06/02) |
| 75 |
|
o pkg/seaice: |
| 76 |
|
- add new routines to compute stress tensor components and stress tensor |
| 77 |
|
divergence to be reused in s/r seaice_lsr and s/r seaice_calc_lhs |
| 78 |
|
- change names of diagnostics SIsigI/II (commonly used for invariants) |
| 79 |
|
to SIsig1/sig2 (more common for principal components) to avoid confusion |
| 80 |
|
o pkf/exf: |
| 81 |
|
- add a stop if useRelativeWind with un-updated uwind/vwind (e.g., constant); |
| 82 |
|
similarly, stop if rotateStressOnAgrid with constant wind-stress. |
| 83 |
|
o pkg/seaice: |
| 84 |
|
- remove (comment out) diagnostics SIuwind,SIvwind,SIfu,SIfv (duplicate of |
| 85 |
|
EXFuwind,EXFvwind,oceTAUX,oceTAUY) and point them to their equivalent. |
| 86 |
|
- add diagnostics of wind-stress over seaice (SItaux,SItauy) and seaice |
| 87 |
|
fraction weighted wind stress over ocean+seaice (SIatmTx,SIatmTy). |
| 88 |
|
o pkg/diagnostics: |
| 89 |
|
- small piece of code to rename (old) diagnostics, when namelist is processed. |
| 90 |
|
o model/src: |
| 91 |
|
- new run-time param (useMin4hFacEdges) to select method for setting hFacW,S: |
| 92 |
|
originally, set as minimum of adjacent hFacC factor ; now (new default) |
| 93 |
|
computed from rSurfW,S and rLowW,S by applying same rules as for hFacC. |
| 94 |
|
Only matters when useShelfIce=T with particular ice-shelf cavity geometry. |
| 95 |
|
o pkg/seaice: |
| 96 |
|
- add momemtum advection, off by default, still in test phase |
| 97 |
|
most of the momentum advection code is adapted and reused from mom_vecinv |
| 98 |
|
o pkg/mom_vecinv: |
| 99 |
|
- pass these runtime flags: |
| 100 |
|
selectVortScheme, highOrderVorticity, upwindVorticity, useJamartMomAdv |
| 101 |
|
as formal parameters to s/r mom_vi_u/v_coriolis, mom_vi_u/v_coriolis_c4, |
| 102 |
|
so that these routines can also be used in pkg/seaice. |
| 103 |
|
|
| 104 |
|
checkpoint66g (2017/04/24) |
| 105 |
|
o model/src: |
| 106 |
|
- more robust ordering of column thickness integral (=Sum_k(drF*hFac) ) |
| 107 |
|
to get same expression (regarding machine truncation) in different places |
| 108 |
|
in ini_masks_etc.F (e.g., rSurfC,W,S is now exactly zero when it should). |
| 109 |
|
Update forward output of experiments: global_ocean.cs32x15 (all 5 except |
| 110 |
|
viscA4), isomip (x3), tutorial_global_oce_in_p & tutorial_plume_on_slope. |
| 111 |
|
o pkg/shelfice: |
| 112 |
|
- ALLOW_SHIFWFLX_CONTROL: move setting of maskSHI from shelfice_update_masks.F |
| 113 |
|
to shelfice_init_fixed.F |
| 114 |
|
- change S/R SHELFICE_UPDATE_MASKS to SHELFICE_INIT_DEPTHS (still called from |
| 115 |
|
ini_masks_etc.F for now) to update ocean upper boundary position according |
| 116 |
|
to ice-shelf topography (instead of directly updating hFacC). |
| 117 |
|
o model/src: |
| 118 |
|
- move call to CTRL_INIT & CTRL_ADMTLM from the middle to the bottom of |
| 119 |
|
packages_init_fixed.F to allow other pkgs to set mask and weight fields. |
| 120 |
|
|
| 121 |
|
checkpoint66f (2017/04/06) |
| 122 |
|
o model/src: |
| 123 |
|
- add specific run-time param to select level of printed plot-field-maps, |
| 124 |
|
set by default to debugLevel. Remove internal param "printDomain". |
| 125 |
|
o pkg/seaice: |
| 126 |
|
- add flag to add snow mass to seaiceMassC/U/V, the default is |
| 127 |
|
SEAICEaddSnowMass=.FALSE., not to change previous results |
| 128 |
|
o pkg/bling |
| 129 |
|
- Bug fix in bling_production regarding unit conversion for diagnostics |
| 130 |
|
being put in the wrong place |
| 131 |
|
o model/src: |
| 132 |
|
- add back recomputation of rSurfW & S to ensure consistent column thickness |
| 133 |
|
(needed for rStar) between Sum_k(hFac*drF) and rSurf-rLow, for now only |
| 134 |
|
when using pkg/shelfice; was initially added on 2012/06/01 (ini_masks_etc.F |
| 135 |
|
revision 1.51) but was removed accidentally on 2014/02/08 (v1.53). |
| 136 |
|
o pkg/autodiff: |
| 137 |
|
- clean-up "checkpoint.flow" |
| 138 |
|
o pkg/rw (Contribution from Jody Klymak): |
| 139 |
|
option to set mds file suffix as function of time in simulation: |
| 140 |
|
*) new input parameter "rwSuffixType" (read from PARM03 namelist) |
| 141 |
|
to select file suffix format; by default (=0), use iteration number; |
| 142 |
|
*) add simple routine to return file suffix according to rwSuffixType; |
| 143 |
|
*) in model/src and in most pkgs (except: atm2d, fizhi, regrid & streamice), |
| 144 |
|
use new S/R to get file suffix, except for debug output file (diagFreq) |
| 145 |
|
and pkgs offline & rbcs input files. |
| 146 |
|
o pkg/rw: |
| 147 |
|
- uniform processing of "I10" suffix across writing fld routines; used to |
| 148 |
|
simplify debug writing of cg2/3d_b/x in solve_for_pressure.F & post_cg3d.F |
| 149 |
|
(similar to cg3d.F, cg3d_ex0.F) |
| 150 |
|
o pkg/mom_vecinv: |
| 151 |
|
- highOrderVorticity in selectVortScheme=2 case: fix c4 coeff in splitted |
| 152 |
|
vorticity flux (keep same sum in case vBarXm=vBaXp, uBarYm=uBarYp). |
| 153 |
|
o pkg/bling: |
| 154 |
|
- Many small changes intended to clean the code. |
| 155 |
|
Unnecessary ADJOINT_SAFEs removed ; |
| 156 |
|
removed BLING_COST.h BLING_CTRL.h bling_cost.F ; |
| 157 |
|
diagnostics added. |
| 158 |
|
|
| 159 |
|
checkpoint66e (2017/03/14) |
| 160 |
|
o pkg/exf: |
| 161 |
|
- interpolation with #undef EXF_INTERP_USE_DYNALLOC: move buffer size |
| 162 |
|
definition outside EXF_PARAM.h in new header file: EXF_INTERP_SIZE.h; |
| 163 |
|
use fixed size buffer (passed to S/R EXF_INTERP & EXF_INTERP_UV) to |
| 164 |
|
avoid any dynamic allocation ; |
| 165 |
|
- S/R EXF_CHECK: add new check & stop when file-name is defined but not |
| 166 |
|
read-in ; add warning when file is read-in but not used within pkg/exf ; |
| 167 |
|
+ check for buffer-size (if #undef EXF_INTERP_USE_DYNALLOC). |
| 168 |
|
- update exf_summary.F (print {FLD}RepCycle). |
| 169 |
|
o verification/global_with_exf: |
| 170 |
|
- switch on exf interpolation in AD test (just like the forward test) |
| 171 |
|
o pkg/mom_vecinv: |
| 172 |
|
- add storage directives to avoid recomputations within mom_vecinv.F |
| 173 |
|
which were generating seg. fault with gfortran -devel |
| 174 |
|
o pkg/ctrl, pkg/smooth: |
| 175 |
|
- initialize array halos which were generating seg. fault with |
| 176 |
|
gfortran -devel in adjoint mode |
| 177 |
|
o pkg/exf: |
| 178 |
|
- for each input field, enable to specify a field specific repeat-cycle |
| 179 |
|
({FLD}RepCycle) which defaults to pkg/exf main "repeatPeriod". |
| 180 |
|
o verification/lab_sea: |
| 181 |
|
- update AD results of primary test and noseaicedyn test after switching |
| 182 |
|
to new version of TAF (5.3.13 vs 5.3.2 before); agreement with gradient |
| 183 |
|
check stays about the same. |
| 184 |
|
o pkg/exf: |
| 185 |
|
- fix bug in exf_zenithangle.F for cases select_ZenAlbedo > 0 with useCAL=T. |
| 186 |
|
- exf_check: MAX_LAT_INC need only be "loc_nlat-1", not equal to loc_nlat |
| 187 |
|
o pkg/bling, USE_EXFCO2 code: |
| 188 |
|
- add parameter "apco2RepCycle" (set to pkg/exf "repeatPeriod" by default); |
| 189 |
|
- update calls to EXF S/R (use new version: EXF_INIT_FLD & EXF_SET_FLD). |
| 190 |
|
o pkg/exf: |
| 191 |
|
- rename S/R EXF_INIT to EXF_INIT_VARIA ; |
| 192 |
|
- add arguments: fldName (for debug msg) and fldRepeatCycle (set to |
| 193 |
|
repeatPeriod) to S/R EXF_SET_UV and to S/R EXF_SET_GEN (renamed EXF_SET_FLD) |
| 194 |
|
and update calls in pkg/exf calling S/R (exf_getffields.F & exf_getclim.F) ; |
| 195 |
|
- add argument: fldName (for debug msg) to S/R EXF_INIT_GEN (renamed |
| 196 |
|
EXF_INIT_FLD) and update calls in exf_init_varia.F |
| 197 |
|
- For now, for backward compatibility, keep a simplified version of S/Rs |
| 198 |
|
EXF_SET_GEN & EXF_INIT_GEN in file exf_set_gen.F |
| 199 |
|
- add and pass new arguments: fldName (for debug msg) and fldRepeatCycle |
| 200 |
|
to S/R EXF_GetFFieldRec ; update calls in pkg/obcs/obcs_exf_load.F |
| 201 |
|
o tools/genmake2: |
| 202 |
|
- replace "bc -l" (only used once) by "expr" (already used). |
| 203 |
|
o verification/global_ocean.cs32x15: |
| 204 |
|
- turn off pkg/cal (useCAL=F) in secondary FWD exp. "icedyn". |
| 205 |
|
Note: exf setting (data.exf) can be compared to other exf secondary exp |
| 206 |
|
"seaice" (kept with useCAL=T) which has same time dependent forcing. |
| 207 |
|
|
| 208 |
|
checkpoint66d (2017/02/13) |
| 209 |
|
o pkg/exf: |
| 210 |
|
- remove setting of useCAL=T when using pkg/exf (useEXF=T); add warning |
| 211 |
|
early on (in packages_boot.F) to stress this change in pkg setting. |
| 212 |
|
- In (most) verification exp. that use pkg/exf, add useCAL=.TRUE. to data.pkg |
| 213 |
|
o model/src: |
| 214 |
|
- move call to ecco_phys to end of time step; this may induce minor |
| 215 |
|
cost function changes by shifting time averages by one time step |
| 216 |
|
for some variables; this revision resulted in changed adjoint results |
| 217 |
|
in MITgcm_contrib/verification_other/global_oce_cs32 |
| 218 |
|
o pkg/exf: |
| 219 |
|
- Change EXF_READ_EVAP to only activate code within #ifdef ALLOW_ATM_TEMP ; |
| 220 |
|
this prevents to read evap when ALLOW_ATM_TEMP is undef and other fields, |
| 221 |
|
e.g. precip, are not available. |
| 222 |
|
o pkg/flt: |
| 223 |
|
- flt_init_varia.F: use pickupSuff if specified |
| 224 |
|
- add code (within #ifdef DEVEL_FLT_EXCH2) and test case as a first, |
| 225 |
|
intermediate, step towards completing pkg/exch2 support within pkg/flt; |
| 226 |
|
ifdef DEVEL_FLT_EXCH2 then: |
| 227 |
|
- flt_init_fixed.F: comment out stop |
| 228 |
|
- flt_exch2.F: conserve ipart, jpart (grid coordinates) |
| 229 |
|
- exch2_recv_get_vec.F: transform ipart, jpart (grid coordinates) |
| 230 |
|
- flt_mapping.F: attempt to fix longitude diagnostic near date-line |
| 231 |
|
- flt_init_varia.F: hack initialization (for testing purposes w. LLC90) |
| 232 |
|
o tools/genmake2 & pkg/pkg_depend: |
| 233 |
|
- define new type of pkg dependency rule ("=pkg_x"), less strict than |
| 234 |
|
the "+" rule, that add the dependent pkg (pkg_x) unless it's disabled |
| 235 |
|
(-pkg_x in packages.conf). |
| 236 |
|
- change pkg/exf dependency rule from "exf +cal" to "exf =cal"; |
| 237 |
|
- add "+cal" to dependency list of pkg/ecco (needed to compile). |
| 238 |
|
- simplify FWD experiment "offline_exf_seaice" by removing pkg/cal. |
| 239 |
|
o utils/matlab: |
| 240 |
|
- adding ocean_basin routines for defining basin masks. |
| 241 |
|
o pkg/exf: |
| 242 |
|
- new S/R EXF_GETFFIELD_START replacing EXF_GETFIELD_START, but with more |
| 243 |
|
arguments, allowing to check (and stop) if parameter setting of |
| 244 |
|
{FLD}startdate1 & date2 and of {FLD}StartTime are consistent with |
| 245 |
|
useCAL= T or F. For now, kept a simplified version of EXF_GETFIELD_START |
| 246 |
|
for backward compatibility. |
| 247 |
|
- change accordingly calls in exf_init_fixed.F + default {FLD}StartTime in |
| 248 |
|
exf_readparms.F ; apply same changes to pkg/icefront (for SGRunOff field). |
| 249 |
|
o pkg/bling (USE_EXFCO2 code): |
| 250 |
|
- fix missing setting of apco2StartTime according to apco2startdate1 & date2 |
| 251 |
|
by adding a call to S/R EXF_GETFFIELD_START. |
| 252 |
|
o pkg/exf: |
| 253 |
|
- allows to compile without pkg/cal (but disabled by current dependency |
| 254 |
|
rules, i.e., default "pkg/pkg_depend"). Minimal changes to enable to |
| 255 |
|
run with useCAL=F ; this involves setting directly {FLD}StartTime instead |
| 256 |
|
of {FLD}startdate1 & date2 ; consistency check not yet coded. |
| 257 |
|
o verification/global_ocean.cs32x15: |
| 258 |
|
- fix all data.exf for a shift of 1 day in all field startdate1 so that it now |
| 259 |
|
corresponds to middle of 1rst month; update outputs (2 fwd + 3 adm + 3 tlm). |
| 260 |
|
o pkg/exf: |
| 261 |
|
- allow to use EXF_ZENITHANGLE without pkg/cal + simplify (saving some |
| 262 |
|
calculations); change FWD experiment global_ocean.cs32x15 |
| 263 |
|
(#define ALLOW_ZENITHANGLE) to test/check that this code compiles. |
| 264 |
|
o pkg/exf: |
| 265 |
|
- In preparation for addition to NAMELIST, rename internal (_RL) parameter |
| 266 |
|
{FIELD}startdate to {FIELD}StartTime |
| 267 |
|
Note: new name also fit better the content, i.e, a start time in second. |
| 268 |
|
o eesupp/src: |
| 269 |
|
- add 1 argument to S/R EEBOOT & EEBOOT_MINIMAL for interface with ESMF |
| 270 |
|
o pkg/exf: |
| 271 |
|
- add option ALLOW_READ_TURBFLUXES (and corresponding set of parameters) |
| 272 |
|
to allow to read-in Sensible & Latent Heat flux ; |
| 273 |
|
- simplify pkg/exf main options: |
| 274 |
|
1) move small piece of code (case ALLOW_ATM_TEMP undef) out of |
| 275 |
|
exf_bulkformulae.F into exf_wind.F so that exf_bulkformulae.F is |
| 276 |
|
compiled only when both ALLOW_ATM_TEMP & ALLOW_BULKFORMULAE are defined; |
| 277 |
|
2) make call to EXF_RADIATION and to EXF_WIND independent of CPP option |
| 278 |
|
ALLOW_BULKFORMULAE which now only applies to S/R EXF_BULKFORMULAE ; |
| 279 |
|
3) improve options documentation in EXF_OPTIONS.h. |
| 280 |
|
o pkg/ecco: |
| 281 |
|
- cost_gencost_customize.F,ecco.h,ecco_phys.F,ecco_readparms.F, and |
| 282 |
|
ecco_summary.F: add passive tracer cost functions to gencost; |
| 283 |
|
m_ptracer and m_boxmean_ptracer |
| 284 |
|
o pkg/ctrl: |
| 285 |
|
- ctrl_map_forcing.F: pass myIter and myTime to ctrl_map_forcing as input |
| 286 |
|
parameters; move myIter .EQ. nIter0 inside ctrl_map_forcing; add to list |
| 287 |
|
of implemented gentim2d contol parameters: xx_qnet, xx_empmr, xx_qsw, |
| 288 |
|
xx_sst, xx_sss, xx_pload, xx_saltflux, xx_fu, xx_fv, xx_fe, xx_fn |
| 289 |
|
o model/src: |
| 290 |
|
- forward_step.F: pass myIter and myTime to ctrl_map_forcing as input |
| 291 |
|
parameters; move myIter .EQ. nIter0 inside ctrl_map_forcing |
| 292 |
|
o pkg/exf: |
| 293 |
|
- exf_getffields.F: add to list of implemented gentim2d contol parameters: |
| 294 |
|
xx_lwflux, xx_swflux, xx_evap, xx_apressure, xx_areamask |
| 295 |
|
|
| 296 |
|
checkpoint66c (2017/01/19) |
| 297 |
|
o verification: |
| 298 |
|
- add new option "-src" to testreport to only produce source code (*.f), |
| 299 |
|
including taf output src code (if using '-adm'/'-tlm' option). |
| 300 |
|
o pkg/exf: |
| 301 |
|
- add saltflx :: Net upward salt flux in psu.kg/m^2/s |
| 302 |
|
o pkg/diagnostics: |
| 303 |
|
- add run-time variable diagMdsDir to specify a subdirectory for |
| 304 |
|
diagnostic output when mds is being used and mdsioLocalDir is not. |
| 305 |
|
o pkg/mnc: |
| 306 |
|
- mnc_utils.F: edit stop message |
| 307 |
|
o pkg/autodiff: |
| 308 |
|
- autodiff_ini_model_io.F: PRINT_MESSAGE, clean-up comments, add _*_MASTER, |
| 309 |
|
add HAVE_SYSTEM bracket around system call to mkdir |
| 310 |
|
o pkg/profiles: |
| 311 |
|
- profiles_findunit.F, profiles_init_fixed.F, profiles_interp.F, |
| 312 |
|
profiles_readparms.F: call ALL_PROC_DIE before 'stop' |
| 313 |
|
- profiles_ini_io.F: use PRINT_MESSAGE, clean-up comments, add _*_MASTER, |
| 314 |
|
add HAVE_SYSTEM bracket around system call to mkdir |
| 315 |
|
o pkg/smooth: |
| 316 |
|
- smooth_filtervar2d.F, smooth_filtervar3d.F: call ALL_PROC_DIE before 'stop' |
| 317 |
|
- smooth_init2d.F, smooth_init3d.F: move print statements to readparms |
| 318 |
|
- smooth_readparms.F: add cleaner print statements from smooth_init?d.F |
| 319 |
|
o pkg/exf: |
| 320 |
|
- check: count errors and stop cleanly at the end (call to ALL_PROC_DIE); |
| 321 |
|
issue error message when climsst_interpMethod.GT.0 and climsst_lon_inc > 500 |
| 322 |
|
generalize tests on interpMethod, nlat, and lon_inc to all exf variables. |
| 323 |
|
- summary: skip detailed field report if not used (empty file-name). |
| 324 |
|
o tools/xmakedepend: |
| 325 |
|
- In case line editor "ed" is missing, switch to "sed" command to clean-up |
| 326 |
|
Makefile from previous dependencies (anything after "# DO NOT DELETE" line). |
| 327 |
|
o pkg/exf: |
| 328 |
|
- change definition of "hflux" input file to always include Short-Wave; |
| 329 |
|
this is now consistent with hflux computation (if using bulk-formulae) |
| 330 |
|
and hflux diagnostic & documentation (in EXF_FIELDS.h). |
| 331 |
|
|
| 332 |
|
checkpoint66b (2016/12/01) |
| 333 |
|
o model & momentum pkgs: |
| 334 |
|
- implement fully implicit bottom friction combined with implicit surf. |
| 335 |
|
pressure, for now, only in hydrostic formulation: |
| 336 |
|
+ to use: compile with #define ALLOW_SOLVE4_PS_AND_DRAG in CPP_OPTIONS.h |
| 337 |
|
and run with selectImplicitDrag=2 ; |
| 338 |
|
+ currently missing diagnostics of bottom drag and vert. viscosity tendency; |
| 339 |
|
- partial implicit time-stepping (bot. friction + vert. visc first, then |
| 340 |
|
surf. press) corresponding to selectImplicitDrag=1 should be applicable |
| 341 |
|
to NH formulation but not available until we solve for surf. pressure |
| 342 |
|
increment (instead of full etaN). |
| 343 |
|
- change exp. global_ocean.cs32x15.thsice to test implicit bottom drag |
| 344 |
|
with implicit vertical viscosity. |
| 345 |
|
o pkg/seaice: add linear remapping scheme (Lipscomb 2001) to ITD model |
| 346 |
|
- this is a first version that includes many regularisations that may |
| 347 |
|
not all be necessary, to be sorted out, |
| 348 |
|
- turned on by runtime parameter SEAICEuseLinRemapITD (true by default) |
| 349 |
|
- change one verification experiment in seaice_itd to use and test new code |
| 350 |
|
- seaice_growth: catch a case where the argument of log can be zero |
| 351 |
|
or smaller (important for code stability) |
| 352 |
|
o model/src: |
| 353 |
|
- modify SOLVE_DIAGONAL_LOWMEMORY version of solve_tridiagonal.F |
| 354 |
|
& solve_pentadiagonal.F to enable to re-used inverse matrix for solving |
| 355 |
|
similar A.X'=Y' system (with same matrix A) in subsequent calls; |
| 356 |
|
switch based on errCode (In/Out) argument value; in calling routines, |
| 357 |
|
initialise errCode before first call. |
| 358 |
|
|
| 359 |
|
checkpoint66a (2016/10/18) |
| 360 |
|
o pkg/bling: |
| 361 |
|
- BLING package now loads in atmospheric pCO2 file using exf package routines |
| 362 |
|
Enable with #define USE_EXFCO2 in BLING_OPTIONS.h |
| 363 |
|
o model/src & pkg/ptracers: |
| 364 |
|
- with INCLUDE_IMPLVERTADV_CODE defined, also call MOM_U,V_IMPLICIT_R & |
| 365 |
|
GAD_IMPLICIT_R (instead of IMPLDIFF) when just implicitViscosity and |
| 366 |
|
implicitDiffusion (respectively) are used (even without momImplVertAdv |
| 367 |
|
or temp,salt,PTRACERS_ImplVertAdv). |
| 368 |
|
o model/src & generic_advdiff: |
| 369 |
|
- add parenthesis in main diagonal calculation of implicit vertical |
| 370 |
|
diffusion matrix (impldiff.F & gad_implicit_r.F). |
| 371 |
|
This affects results at machine truncation level; update output of many |
| 372 |
|
verification experiments (31 fwd ; 10 adm ; 8 tlm ; 3 oad). |
| 373 |
|
o pkg/mom_commom & generic_advdiff: |
| 374 |
|
- add deep atmosphere and anelastic scaling factor in implicit vertical |
| 375 |
|
advection routines (mom_u,v_implicit_r.F & gad*_impl*_r.F). |
| 376 |
|
o model/src: |
| 377 |
|
- remove few unnecessary temp array copies in solve_tridiagonal.F |
| 378 |
|
& solve_pentadiagonal.F |
| 379 |
|
|
| 380 |
|
checkpoint65z (2016/09/28) |
| 381 |
|
o pkg/seaice: fix multithreading for seaice_fgmres.F |
| 382 |
|
- remove unecessary variables from common blocks |
| 383 |
|
- use extra dimension MAX_NO_THREADS for all variables of the Krylov space |
| 384 |
|
basis, that are stored in commom blocks (including integer variable i1) |
| 385 |
|
- clean up code a little by removing commented lines that refer to |
| 386 |
|
original code |
| 387 |
|
- turn on multi-threaded test for experiment offline_exf_seaice.dyn_jfnk. |
| 388 |
|
o MITgcm manual: |
| 389 |
|
- add uptodate pkg/ecco documentation to the manual (section 8.1; ecco.tex), |
| 390 |
|
preliminary documentation for generic controls (8.3; ctrl.tex), |
| 391 |
|
a placeholder for smooth (8.4; smooth.tex), and comment out the deprecated |
| 392 |
|
ecco_costfunction.tex |
| 393 |
|
o pkg/ecco: |
| 394 |
|
- update and cleanup list of CCP options |
| 395 |
|
- re-assign gencost_mask to specify masks rather than gencost_errfile; 'c' |
| 396 |
|
is now always passed to cost_generic; gencost_mask is now MAX_LEN_FNAM |
| 397 |
|
- add checks that retired run-time options are not used, and in relation |
| 398 |
|
to gencost_mask and gencost_errfile |
| 399 |
|
- finish deprecating gencost_scalefile, gencost_smooth2Ddiffnbt, and |
| 400 |
|
gencost_nrecperiod |
| 401 |
|
- fix m_eta computation for virtual salt flux case |
| 402 |
|
- add m_horflux_vol case (selected via gencost_barfile) that computes |
| 403 |
|
transports via cost_gencost_boxmean (gencost_flag=-3) |
| 404 |
|
- use gencost_flag more systematically (i.e., avoid using gencost_name) |
| 405 |
|
for code switches (aside from physical variable selection) |
| 406 |
|
o pkg/seaice: make seaice_fgmres.F fit for multithreading; this first |
| 407 |
|
attempt appears to be a bit of a hack and my need some repair down the road |
| 408 |
|
o pkg/exf: |
| 409 |
|
- interpolation near South Pole: fix criteria on input field latitude. |
| 410 |
|
o pkg/icefront & pkg/exf: |
| 411 |
|
- move setting of icefront Sub-Glacial RunOff forcing (currently unused) |
| 412 |
|
from pkg/exf (read from data.exf) to pkg/icefront (read from data.icefront) |
| 413 |
|
o pkg/exf: |
| 414 |
|
- replace CPP option ALLOW_EXF_OUTPUT_INTERP with run-time parameter |
| 415 |
|
"exf_output_interp" (in namelist: EXF_NML_04); change output file name |
| 416 |
|
(add suffix "_out" instead of a prefix); keep same record number as in |
| 417 |
|
input file but always write record 1 first (to avoid a STOP in I/O S/R). |
| 418 |
|
Change output file content by writing interpolated field before applying |
| 419 |
|
land-sea mask (helpful for debug). |
| 420 |
|
- 2 compon. vector interpolation with rotation: apply scaling factor on |
| 421 |
|
each compon. (Eastward and Northward) just before rotation. |
| 422 |
|
o pkg/bling: |
| 423 |
|
- Cleaned package. Removed not needed code and comments. |
| 424 |
|
- adjoint example now works |
| 425 |
|
(MITgcm_contrib/verification_other/global_oce_biogeo_bling/code_ad) |
| 426 |
|
o pkg/ptracers: |
| 427 |
|
- add Bruno Deremble's code to complete the linear surface correction |
| 428 |
|
for tracer conservation; |
| 429 |
|
- enabled for individual ptracers by PTRACERS_linFSConserve(iTr)=.TRUE. |
| 430 |
|
|
| 431 |
|
checkpoint65y (2016/07/29) |
| 432 |
o pkg/debug (stats): |
o pkg/debug (stats): |
| 433 |
- change GLOBAL_SUM to GLOBAL_SUM_TILE (result independent of tile to |
- change GLOBAL_SUM to GLOBAL_SUM_TILE (result independent of tile to |
| 434 |
proc/thread mapping); |
proc/thread mapping); |
| 441 |
writing delayed until next monitor output. |
writing delayed until next monitor output. |
| 442 |
o pkg/obcs: |
o pkg/obcs: |
| 443 |
- fix bug in OBCS_BALANCE_FLOW (OBCS_balanceFacN misspelled OBCS_balanceFacW) |
- fix bug in OBCS_BALANCE_FLOW (OBCS_balanceFacN misspelled OBCS_balanceFacW) |
| 444 |
|
o pkg/cal: |
| 445 |
|
- calandar-dump: treat case myTime < timePhase in the same way as in |
| 446 |
|
eesupp/src/diff_phase_multiple.F (when not using calandar-dump). |
| 447 |
o pkg/dic & seaice: |
o pkg/dic & seaice: |
| 448 |
- fix compilation of ocean component for coupled atm+ocn model with |
- fix compilation of ocean component for coupled atm+ocn model with |
| 449 |
seaice dynamics & dic. |
seaice dynamics & dic. |
Parent Directory
| 
Revision Log
| 
Revision Graph
| 
Patch