| 4 |
Notes on tags used in MITgcmUV |
Notes on tags used in MITgcmUV |
| 5 |
============================== |
============================== |
| 6 |
|
|
| 7 |
|
o pkg/autodiff: |
| 8 |
|
note about monitor and snap-shot of ad-variables: |
| 9 |
|
a) need to apply ADEXCH before monitor or output of ad-vars to include |
| 10 |
|
overlap contributions (was done for adjDump but missing for monitor). |
| 11 |
|
b) since ADEXCH implies adding overlap values to interior array, machine |
| 12 |
|
truncation error is expected when adding an ADEXCH call in the middle of |
| 13 |
|
the time-stepping (although mathematically identical). |
| 14 |
|
modifs: |
| 15 |
|
- add new run-time params (dumpAdVarExch & mon_AdVarExch) to select how to |
| 16 |
|
apply ADEXCH before monitor or snap-shot of ad-variables: |
| 17 |
|
=0 : no ADEXCH ; =1 : do ADEXCH ; =2 : apply ADEXCH on a local copy. |
| 18 |
|
- implement new method (copy+adexch) for snap-shot and monitor of main model |
| 19 |
|
ad-vars (but not yet used by default). |
| 20 |
|
- fix some incorrect type (RL/RS, UV/scalar) in ADEXCH calls and ad-vars |
| 21 |
|
declaration (adEddyPsiX,adEddyPsiY in adcommon.h). |
| 22 |
|
o pkg/ctrl: |
| 23 |
|
enable the use of a variable number (maxCtrlArr2D, maxCtrlArr3D) of |
| 24 |
|
generic control variables via |
| 25 |
|
#define ALLOW_GENARR2D_CONTROL and/or #define ALLOW_GENARR3D_CONTROL |
| 26 |
|
Only required customization should be S/R ctrl_map_ini_genarr.F |
| 27 |
|
(routine in repository contains example) |
| 28 |
|
o pkg/exf: |
| 29 |
|
- fix and refine checking of surface-relaxation setting when using EXF |
| 30 |
|
o verification (adjoint exp): |
| 31 |
|
- reduce length of simulation by half for experiment global_ocean.cs32x15 |
| 32 |
|
(std, seaice & seaice_dynmix) and exp tutorial_dic_adjoffline; update output. |
| 33 |
|
o pkg/mom_common (mom_u/v_bottomdrag), pkg/shelfice (shelfice_u/v_drag) |
| 34 |
|
- replace if statements KE.ne.0 with KE.gt.0 to avoid sqrt of negative numbers |
| 35 |
|
of KE that can occur with aggressive optimization |
| 36 |
|
+ put quadratic bottom drag calculation in separated loops (only done if |
| 37 |
|
using bottomDragQuadratic) |
| 38 |
|
o model/src: CG2D_NSA |
| 39 |
|
- clean-up: stop if cg2dNormaliseRHS=T and AUTODIFF |
| 40 |
|
and replace GLOBAL_SUM calls with GLOBAL_SUM_TILE. |
| 41 |
|
- in test-exp that uses CG2D_NSA (bottom_ctrl & tutorial_tracer_adjsens), |
| 42 |
|
change cg2d residual target to be expressed in Wunits (=> do not normalise |
| 43 |
|
RHS ; this fix global_max Pb with TAF); update output. |
| 44 |
|
o verification/tutorial_tracer_adjsens: |
| 45 |
|
improve parameter consistency (and get rid of warnings): |
| 46 |
|
- use deltaTFreeSurf=deltaTtracer (better with NonLin FreeSurf); |
| 47 |
|
- switch Temp Adv Scheme from 81 to 80 (no limiter since theta < 0 is physical) |
| 48 |
|
- use staggerTimeStep (for stability) in som81; use default rotation period. |
| 49 |
|
o pkg/exch2: |
| 50 |
|
- with empty facet: fix compact-format (W2_mapIO=1) definition (was previously |
| 51 |
|
reset to 0 = 1 long line in X) |
| 52 |
|
o pkg/mdsio: |
| 53 |
|
- fix reading of dimensions from meta file for large-size domain |
| 54 |
|
o model/src: |
| 55 |
|
- fix saltIsActiveTr setting for IdealGas eosType (eosType is 6 character long); |
| 56 |
|
- to refine CS-grid check, add logical flag "hasWetCSCorners": true if using |
| 57 |
|
Cubed-Sphere Exch with CS-corners inside the domain; |
| 58 |
|
make new S/R INI_GLOBAL_DOMAIN from code in ini_linear_phisurf.F to calculate |
| 59 |
|
globalArea and to set hasWetCSCorners. |
| 60 |
|
- add warning if usingCurvilinearGrid and momAdvection in flux-form (missing |
| 61 |
|
metric terms); stop if hasWetCSCorners=T and momAdvection in flux-form. |
| 62 |
|
- re-activate the stop if useCDscheme and hasWetCSCorners=T (previously if |
| 63 |
|
useCDscheme and useCubedSphereExchange). |
| 64 |
|
o pkg/seaice/seaice_growth.F : |
| 65 |
|
- compute the SEAICE_DISABLE_HEATCONSFIX term accordingly |
| 66 |
|
for the (.NOT.useRealFreshWaterFlux).OR.(nonlinFreeSurf.EQ.0) |
| 67 |
|
when the previous code was not correct. |
| 68 |
|
- in all cases diagnoze the boundary condition mismatch to SIaaflux |
| 69 |
|
- in the case of useRealFreshWaterFlux.AND.(nonlinFreeSurf.GT.0) add |
| 70 |
|
that term to QNET (essentially to set the melt/freeze water to 0degC). |
| 71 |
|
when SEAICEheatConsFix is true, so that the ocean+ice budget heat is closed. |
| 72 |
|
o verification/global_ocean.cs32x15/input.seaice : |
| 73 |
|
- add SEAICEheatConsFix in data.seaice, ensuring closed ocean+ice heat budget. |
| 74 |
|
- add diagnostics (commented out) for budget test using my matlab codes (gcmfaces). |
| 75 |
|
- update results (output.seaice.txt) accordingly. |
| 76 |
|
o verification/lab_sea/input/data.seaice : |
| 77 |
|
- remove SEAICEheatConsFix, which is now be tested |
| 78 |
|
in global_ocean.cs32x15. The presently coded fix |
| 79 |
|
is only consistent with real fresh water + nlfs. |
| 80 |
|
- update results (output.txt) accordingly. |
| 81 |
|
o pkg/salt_plume: added diagnotic for salt tendency (oceSPtnd). |
| 82 |
|
o AD_CONFIG.h file: |
| 83 |
|
- in genmake2: remove "#undef ALLOW_ECCO_OPTIMIZATION" from AD_CONFIG.h; |
| 84 |
|
- remove #include "AD_CONFIG.h" from all included OPTIONS files and |
| 85 |
|
add it explicitly in each fortran src file where it is needed. |
| 86 |
|
motivation: header file AD_CONFIG.h is generated/modified after "make depend" |
| 87 |
|
thus introducing potentially wrong dependency; therefore, better to include |
| 88 |
|
it only where it's really needed, and after any other included OPTIONS file |
| 89 |
|
(less tempting to bring conditional includes according to ALLOW_ADJOINT_RUN |
| 90 |
|
or ALLOW_TANGENTLINEAR_RUN). |
| 91 |
|
|
| 92 |
|
checkpoint63p (2012/07/06) |
| 93 |
|
o pkg/autodiff: |
| 94 |
|
- rename S/R INI_AUTODIFF to AUTODIFF_INIT_VARIA ; |
| 95 |
|
- create new header file "AUTODIFF_PARAMS.h" for pkg/autodiff parameters |
| 96 |
|
which are now read from new parameter file "data.autodiff" (inAdExact |
| 97 |
|
previously in "data", use{KPP/GMRedi/SEAICE}inAdMode previously in data.pkg) |
| 98 |
|
- add file "data.autodiff" in all verification/*/input_ad/ directories. |
| 99 |
|
- add run-time parameter to turn On/Off SEAICEuseFREEDRIFT in adjoint/forward |
| 100 |
|
mode (used in global_ocean.cs32x15.seaice_dynmix test exp.) |
| 101 |
|
o verification/tutorial_tracer_adjsens/ |
| 102 |
|
- add extra sub-verification which uses "SOM" advection scheme |
| 103 |
|
(2nd Order-Moment Advection Scheme, Prather Limiter) with |
| 104 |
|
#define GAD_ALLOW_TS_SOM_ADV in GAD_OPTIONS.h |
| 105 |
|
and in data: |
| 106 |
|
multiDimAdvection=.TRUE., |
| 107 |
|
tempAdvScheme=81, |
| 108 |
|
saltAdvScheme=81, |
| 109 |
|
o model/src: |
| 110 |
|
- introduce ALLOW_BALANCE_RELAX which allow the removal |
| 111 |
|
of the global mean of relaxation terms by setting |
| 112 |
|
balanceThetaClimRelax and balanceSaltClimRelax |
| 113 |
|
- disable balanceEmPmR and balanceQnet in the case when useSeaice. |
| 114 |
|
This case is now treated appropriately in seaice_growth.F |
| 115 |
|
o pkg/seaice: |
| 116 |
|
- do balanceEmPmR and balanceQnet based on the net fluxes |
| 117 |
|
(SItflux) that enter the ocean+ice system rather than QNET |
| 118 |
|
and EMPMR. Works best in real fresh water. |
| 119 |
|
- SIatmFW, SIatmQnet, and SItflux computations had to be moved |
| 120 |
|
out of the diagnotics bloc, and stored in global arrays. |
| 121 |
|
|
| 122 |
|
checkpoint63o (2012/06/29) |
| 123 |
|
o vertical mixing schemes: |
| 124 |
|
- remove unused S/R CALC_BUOYANCY (left from branch-atmos merging) |
| 125 |
|
- replace computation of density (in Function STATE1, opps_calc.F) by a call |
| 126 |
|
to S/R FIND_RHO_SCALAR |
| 127 |
|
- pass sigmaR to S/R PP81_CALC, MY82_CALC (but not yet used) and to GGL90_CALC |
| 128 |
|
and use it for N^2 calculation in ggl90_calc.F (save 2 density computations); |
| 129 |
|
affects output (vermix.ggl90) at truncation level. |
| 130 |
|
o testreport: |
| 131 |
|
- change adjustment of SIZE.h for mpi+mth testing: preserve enough tiles |
| 132 |
|
(nSx,nSy) for the number of threads (from eedata.mth) |
| 133 |
|
o model/src: |
| 134 |
|
- remove 2nd set of calls at the top of initialise_varia.F (Adjoint part) |
| 135 |
|
related to Non_Lin Free-Surf or Bottom-Control (but not longer necessary). |
| 136 |
|
o verification/tutorial_tracer_adjsens/ |
| 137 |
|
- compile with ALLOW_CG2D_NSA |
| 138 |
|
- run with nonlinFreeSurf=3 |
| 139 |
|
- update reference output_adm.txt |
| 140 |
|
o tools/genmake2: |
| 141 |
|
- add "-DHAVE_NETCDF" to the DEFINES list when NetCDF is available ; |
| 142 |
|
refine usePROFILES switching off (only when NetCDF is not available). |
| 143 |
|
o verification (atmospheric set-up): |
| 144 |
|
- change eosType='IDEALGAS' to eosType='IDEALG' to match variable declaration |
| 145 |
|
o pkg/profiles : refine profiles and files handling. |
| 146 |
|
o model/src and pkg/ecco : introduce run-time switch usePROFILES. |
| 147 |
|
o model/src/the_main_loop.F, forward_step.F and pkg/exf/EXF_OPTIONS.h : |
| 148 |
|
- changes to use model/src/the_main_loop.F rather than the ecco version |
| 149 |
|
in ecco type adjoint runs. |
| 150 |
|
o model/src: |
| 151 |
|
- move computation of recip of grid distance & area from INI_MASKS_ETC to INI_GRID |
| 152 |
|
- move setting of h0Fac[C,W,S] from INI_LINEAR_PHISURF to INI_MASKS_ETC |
| 153 |
|
- update_masks_etc.F: rename SMOOTH*_R4,R8 function to the corresponding type (RS,RL) |
| 154 |
|
o model/src: |
| 155 |
|
- fix S/R UPDATE_CG2D for ALLOW_DEPTH_CONTROL case (was empty before); |
| 156 |
|
update output of test experiment bottom_ctrl_5x5. |
| 157 |
|
o pkg/generic_advdiff: |
| 158 |
|
- implement vertical bi-harmonic diffusion (for now, only for T & S and |
| 159 |
|
only explicitly). |
| 160 |
|
o verification/bottom_ctrl_5x5/ |
| 161 |
|
update reference output (results changed quite some time ago, |
| 162 |
|
not sure when; but new reference results look ok in terms of grdchk) |
| 163 |
|
o verification/1D_ocean_ice_column/ |
| 164 |
|
For verification purposes, the exact adjoint rather than an approx. |
| 165 |
|
should be used (e.g., to comare against F.D. or TLM). |
| 166 |
|
removed autodiff_inadmode*_ad.F routines in code_ad/ |
| 167 |
|
Update results/ accordingly. |
| 168 |
|
o pkg/cal: |
| 169 |
|
- change cal_getdate.F so that output date is function of input time only |
| 170 |
|
(less ambiguous, not mixed up with input iteration number). |
| 171 |
|
- fix calendarDumps for non-zero phase-lag (S/R CAL_TIME2DUMP). |
| 172 |
|
o pkg/diagnostics: |
| 173 |
|
- implement calendarDumps also for stats-diags |
| 174 |
|
- implement calendarDumps also for snap-shot output (negative freq) |
| 175 |
|
- for now, ignores calendarDumps if non-zero phase-lag. |
| 176 |
|
|
| 177 |
checkpoint63n (2012/06/04) |
checkpoint63n (2012/06/04) |
| 178 |
o model/src: |
o model/src: |
| 179 |
- recompute rSurfW & S to ensure consistent column thickness (needed for rStar) |
- recompute rSurfW & S to ensure consistent column thickness (needed for rStar) |
Parent Directory
| 
Revision Log
| 
Revision Graph
| 
Patch