| 4 |
Notes on tags used in MITgcmUV |
Notes on tags used in MITgcmUV |
| 5 |
============================== |
============================== |
| 6 |
|
|
| 7 |
|
o pkg/autodiff |
| 8 |
|
- Add inAdMode switch for SEAICEuseDYNAMICS: |
| 9 |
|
SEAICEuseDYNAMICSinAdMode |
| 10 |
|
(extend logic in ADAUTODIFF_INADMODE_SET) |
| 11 |
|
o pkg/seaice |
| 12 |
|
- Step 1 of merging ice-thickness distribution (ITD) code from |
| 13 |
|
MITgcm_contrib/torge/itd/ (author: Torge Martin) |
| 14 |
|
The code implements concept laid out in Thorndike et al. (1975); |
| 15 |
|
code is enabled via CPP flag SEAICE_ITD |
| 16 |
|
nITD (in SEAICE_SIZE.h) defines number of ice thickness categories; |
| 17 |
|
code is not complete yet!! |
| 18 |
|
- SEAICE_SIZE.h now included in almost all seaice routines |
| 19 |
|
o verification (using seaice): |
| 20 |
|
- switch back seaice-dynamics (+ lower SOLV_MAX_ITERS) in ADM exp. |
| 21 |
|
global_ocean.cs32x15.seaice ; update results. |
| 22 |
|
- compile B-grid dynamics in FW test exp 1D_ocean_ice_column ; |
| 23 |
|
- compile C-grid dynamics with #define SEAICE_ALLOW_JFNK in FW test exp |
| 24 |
|
offline_exf_seaice (previously compiling B-grid). |
| 25 |
|
o pkg/exf: |
| 26 |
|
- fix some STOP conditions in exf_check.F (from Aug 28, modifs did not allow |
| 27 |
|
to specify neither just 1 comp of wind or wind-stress, nor no surface momentum |
| 28 |
|
forcing): just replace CPP option ALLOW_ATM_WIND with run-time param useAtmWind. |
| 29 |
|
o pkg/layers (adjust changes from Sept 19): |
| 30 |
|
- do snap-shot output and fill diagnostics from inside iLa loop (layers_calc.F) |
| 31 |
|
this allows to remove "layers_maxNum" dimension from all layers arrays. |
| 32 |
|
o pkg/layers (adjust changes from Sept 19): |
| 33 |
|
- remove ALLOW_LAYERS_OUTPUT ; fix snap-shot and time-ave output ; |
| 34 |
|
- snap-shot extended to all type of layers (not just the 1rst type); |
| 35 |
|
- remove layers variables when corresponding CPP option is undef. |
| 36 |
|
o pkg/layers: |
| 37 |
|
- rename pkg/layers variable and add new output (layer exist & un-weighted |
| 38 |
|
velocity in layer) |
| 39 |
|
o pkg/seaice: |
| 40 |
|
- simplify (?) ocean-drag coefficent computations by using |
| 41 |
|
the same routine in all three solvers (saves some code lines) |
| 42 |
|
- remove global fields DRAGA and DRAGS from C-grid code and |
| 43 |
|
declare them locally in lsr.F (the only place where they are |
| 44 |
|
still used) |
| 45 |
|
- update AD output of lab_sea.evp (very sensitive anyway). |
| 46 |
|
o pkg/seaice: add Jacobian-free Newton Krylov solver for ice dynamics |
| 47 |
|
(Lemieux et al. 2010, 2012). Turn on with the usual SEAICE_ALLOW_JFNK |
| 48 |
|
and SEAICEuseJFNK=.true. A few things to note: |
| 49 |
|
- involves a FGMRES, that is parallel but very inefficient, |
| 50 |
|
because of many scalar products (global sums) in the Gram-Schmidt |
| 51 |
|
orthogonalization method. Needs to be replaced. Original code |
| 52 |
|
from J.-F. Lemieux |
| 53 |
|
- preconditioner for FGMRES is a stripped down version of LSOR |
| 54 |
|
- SOLV_MAX_ITERS is now used for LSOR as a solver and as a |
| 55 |
|
preconditioner. For the preconditioner, the default is set to 10. |
| 56 |
|
- Jacobian times vector is computed with finite differences |
| 57 |
|
- marginally related to JFNK-solver: add a smooth regularization for |
| 58 |
|
zeta with tanh that will improve convergence (Lemieux and Tremblay, |
| 59 |
|
2009) for any implicit solver (not implemented for EVP because EVP |
| 60 |
|
does not need this regularization in the first place). Define |
| 61 |
|
SEAICE_ZETA_SMOOTHREG in SEAICE_OPTIONS.h to use. |
| 62 |
|
o eesupp/src: |
| 63 |
|
- check in genmake2 that intrinsic S/R flush is available. |
| 64 |
|
- add S/R to flush IO unit (if intrinsic S/R flush is available) |
| 65 |
|
- flush standard & error msg unit before MPI termination |
| 66 |
|
- if debugMode, flush IO unit after printing msg |
| 67 |
|
o pkg/ctrl: |
| 68 |
|
- ctrl_unpack.F, ALLOW_GENARR3D_CONTROL code: fix missing argument in call |
| 69 |
|
to ctrl_set_unpack_xyz. |
| 70 |
|
|
| 71 |
|
checkpoint64 (2012/10/09) |
| 72 |
|
o pkg/shelfice |
| 73 |
|
- in mom_fluxform.F, mom_vecinv.F: |
| 74 |
|
CALL SHELFICE_V_DRAG(bi,bj,k,vFld,KE,KappaRU,vF,myThid) |
| 75 |
|
Bug fix KappaRU -> KappaRV |
| 76 |
|
does not change verification/isomip but may change other setups |
| 77 |
|
(presumably constant KappaRU=KappaRV) |
| 78 |
|
- fix diagnostics SHIU/Vdrag (thanks to Tom Millgate) |
| 79 |
|
o pkg/obcs: |
| 80 |
|
- fixed some indexing bugs in obcs_seaice_sponge.F |
| 81 |
|
- added obcs relaxation subroutines for HSALT and HSNOW |
| 82 |
|
- changes verification/seaice_obcs/results/obcs_seaice_sponge.F |
| 83 |
|
o pkg/seaice: |
| 84 |
|
- add "multicategory" parameterization for snow in analogy to ice, |
| 85 |
|
default is off for now. |
| 86 |
|
o pkg/obcs: |
| 87 |
|
- added obcs_seaice_sponge.F for obcs relaxation of seaice variables; |
| 88 |
|
see verification/seaice_obcs/input.seaiceSponge for example application. |
| 89 |
|
- remove local commom bloc "OBCS_GLOBAL" in obcs_readparms.F and do all the OB |
| 90 |
|
index setting by master-thread. This fix multi-threads Pb in previous modifs. |
| 91 |
|
o pkg/layers : |
| 92 |
|
- add diagnostics to pkg/layers |
| 93 |
|
- move the meat of pkg/layers from layers_calc.F to layers_fluxcalc.F |
| 94 |
|
(from R.Tulloch). Layers_calc.F now calls layers_fluxcalc.F accordingly. |
| 95 |
|
o pkg/generic_advdiff/gad_implicit_r.F : |
| 96 |
|
- fix ADVr diagnostics in case of both implicit vertical advection and diffusion. |
| 97 |
|
o pkg/seaice/seaice_growth.F: |
| 98 |
|
- fix SItflux diagnostics in virtual salt flux + balanceEmpmr case. |
| 99 |
|
o pkg/obcs: |
| 100 |
|
- change default setting of OBCS_UVICE_OLD from defined to undef; |
| 101 |
|
- use new parameter OB_indexNone for null index value (instead of hard-coded 0) |
| 102 |
|
in all pkg/obcs code and in other pkgs; remove k loop in obcs_check_depths.F; |
| 103 |
|
- pass OB_indexNone as (new) argument to S/R CTRL_MASK_SET_[X,Y]Z; |
| 104 |
|
- change OB null index value (OB_indexNone) to -99 ; this fix the problem |
| 105 |
|
with Southern & Western OB at a tile edge (at least in a simple set-up). |
| 106 |
|
o pkg/obcs: |
| 107 |
|
- change settings of OB index in case OB coincides with tile edge (shared by |
| 108 |
|
2 tiles), so that common OB location is registered by both tiles (instead of |
| 109 |
|
only 1). However, for Southern and Western OB, index value = 0 is not |
| 110 |
|
recognized as valid OB index (=default null value) and needs to be fixed. |
| 111 |
|
- modify accordingly OBCS_BALANCE_FLOW and OBCS_MONITOR (avoid double counting) |
| 112 |
|
- new parameter "OB_indexNone" for null index value (for now, still zero) |
| 113 |
|
to replace hard-coded null index value of zero. |
| 114 |
|
o pkg/exch2: |
| 115 |
|
- improve W2_USE_E2_SAFEMODE (faster) and W2_E2_DEBUG_ON: |
| 116 |
|
always check buffer-size (but outside the loop); move checking for valid |
| 117 |
|
index from W2_USE_E2_SAFEMODE to W2_E2_DEBUG_ON; in W2_E2_DEBUG_ON, |
| 118 |
|
print each tile and point connection only if |W2_printMsg|>= 2 and 3. |
| 119 |
|
o pkg/ctrl (pack & unpack): |
| 120 |
|
- in ctrl_pack.F & ctrl_unpack.F, use the same local file-name variable |
| 121 |
|
(locally defined) for different variables (+ remove the variable-specific |
| 122 |
|
file name in CTRL_GENARR.h). For now, only implemented for GENARR2/3D_CONTROL |
| 123 |
|
(to be extended to other local file names in ctrl.h). |
| 124 |
|
o pkg/flt: |
| 125 |
|
- add a FLT_Iter0 option (similar to PTRACERS_Iter0). |
| 126 |
|
o pkg/mdsio: |
| 127 |
|
- fix missing STOP in mdsio_readvec_loc.F (case ioUnit=0) when file is missing |
| 128 |
|
o tools/genmake2: |
| 129 |
|
- add similar dependences for target EXE_AD & EXE_FTL as we have for EXECUTABLE |
| 130 |
|
(added by Alistair in version 1.51 for building without make depend); |
| 131 |
|
- remove Header files list (H_SRC_FILES) from the dependence of targets |
| 132 |
|
ad_input_code.f & ftl_input_code.f (not clear why it was added in the first |
| 133 |
|
place in version 1.23): this save going through staf when unrelated header |
| 134 |
|
file has been modified. |
| 135 |
|
Note: for an intermediate target (e.g., ad_input_code.f or adtaf) without |
| 136 |
|
"make depend" step, still need to do "make links" before "make adtaf". |
| 137 |
|
o eesupp/src (+ few pkg src files) & genmake2: |
| 138 |
|
- finish to remove ALWAYS_USE_MPI in source code that TAF does not see; |
| 139 |
|
- no longer set ALWAYS_USE_MPI in Makefile (was done by genmake2). |
| 140 |
|
- improve genmake2 checking for ALLOW_${PKG} in CPP_OPTIONS.h files (distinguish |
| 141 |
|
upper/lower case ; account for space between "#" and undef/define) |
| 142 |
|
o pkg/flt (with exch2): |
| 143 |
|
- cleaning: fix so that it compiles; move remaining MPI calls out of flt_exch2.F; |
| 144 |
|
- simplify threads synchronisation; fix (many issues) multi-threaded with MPI. |
| 145 |
|
|
| 146 |
checkpoint63s (2012/09/05) |
checkpoint63s (2012/09/05) |
| 147 |
o pkg/ecco: |
o pkg/ecco: |
| 148 |
- remove #ifdef ALLOW_SMOOTH_CORREL3D brackets. |
- remove #ifdef ALLOW_SMOOTH_CORREL3D brackets. |
Parent Directory
| 
Revision Log
| 
Revision Graph
| 
Patch