/[MITgcm]/MITgcm/doc/tag-index

Diff of /MITgcm/doc/tag-index

Parent Directory | Revision Log | View Revision Graph Revision Graph | View Patch Patch

-revision 1.1462 by jmc,
Wed Jun 22 19:17:36 2011 UTC
+revision 1.1728 by jmc,
Thu Aug 30 23:13:49 2012 UTC
 Line 4 
 $Name$
      Notes on tags used in MITgcmUV
      ==============================
+ o pkg/ggl90:
+   - fix bug in filling corner mask array "mskCor" (FILL_CS_CORNER_TR_RL call)
+     used in ALLOW_GGL90_SMOOTH code.
+ o pkg/flt:
+   - fix typo in FLT_WITHOUT_Y_PERIODICITY (the effect was to ignore stopping
+     floats at the Southern end of the domain).
+ o eesupp/inc:
+   - add a list of CPP-options which can be set by Makefile (DEF_IN_MAKEFILE.h);
+     but not used (and not to be used) anywhere, just for documentation.
+ o pkg/autodiff, exf, ptracers & seaice (AD & TL output):
+   - comment out the very nasty "writeBinaryPrec = writeStatePrec" line, since:
+     a) can damage previously written records in any other multi-record files.
+     b) if needed, can easily do it cleanly, by calling S/R WRITE_REC_3D_RL/S.
+ o tools/genmake2:
+   - define variable MAKE within Makefile and use it in place of 'make';
+   - update AD_CONFIG.h (if needed) before building any of the executable
+     (forward, adjoint or tangent-lin); this involves a 2nd invocation of make;
+     and add the usual #ifndef/#define AD_CONFIG_H in AD_CONFIG.h;
+   - always print usual warning comment 'this file is automatically generated'
+     from script convert_cpp_cmd2defines (+ remove this warning from genmake2).
+ o pkg/cheapaml:
+   - updated to correspond to the code used in accepted paper (Nico Wienders).
+ o pkg/exf:
+     added run time switch useAtmWind to replace ALLOW_ATM_WIND cpp switch.
+     ALLOW_ATM_WIND now just sets the useAtmWind default (see exf_readparms.F)
+     and force defines ALLOW_BULKFORMULAE (EXF_OPTIONS.h).
+ o pkg/exf, autodiff, ctrl, ecco and seaice:
+     remove ALLOW_ATM_WIND brackets, or replace them with useAtmWind ones.
+ o pkg/ctrl, ecco:
+     allow to compile both ALLOW_U/VSTRESS_CONTROL and ALLOW_U/VWIND_CONTROL.
+     Depending on useAtmWind, one is inactive, and the other is active (see
+     exf_getffields.F/exf_getsurfacefluxes.F).
+ o tools/genmake2:
+   - improve error msg when failing in 2nd test of check_for_broken_Ff
+   - fix option '-makefile' short-name (-mf according to genmake2 -help)
+   - source local config file "genmake_local" just after processing the genmake2
+     command arguments (previously done before processing arg).
+ o pkg/grdchk:
+   - add a call to TURNOFF_MODEL_IO for Tangent-Linear run (in this case, call
+     to this routine from cost_final has been dropped in g_cost_final)
+   - improve printed information (more explicit msg, no longer using "print *,"
+      fix fcpertminus printed value).
+ o model/src:
+   - move call to COST_INIT_VARIA out of packages_init_variables.F directly
+     into initialise_varia.F (just after calling PACKAGES_INIT_VARIABLES):
+     this prevents getting cost-funct reset in adinitialise_varia recomputations.
+ o pkg/dic:
+   - move initialisation of "freefe" array from from dic_biotic_forcing.F to
+     fe_chem.F: this prevents TAF from dropping this initialisation in TLM code.
+ o pkg/grdchk:
+   - grdchk_main.F : added CALL GLOBAL_SUM_INT( ierr , myThid )
+   - grdchk_getadxx.F etc. : omit I/O if ierr.EQ.0
+ o pkg/autodiff/ctrl/grdchk :
+   - attempt to sort out what works and what does not,
+     with regard to I/O, and to warn/stop accordingly.
+ o pkg/grdchk:
+   - attempt to fix grdchk in the case of useSingleCpuIO.
+   - "if ( myProcId .EQ. grdchkwhichproc .AND. ierr .EQ. 0 )"
+    statements were moved from grdchk_main.F and into
+    grdchk_getadxx.F grdchk_getxx.F grdchk_setxx.F.
+   - a global sum is applied to record numbers icvrec/layer
+    out of grdchk_loc so that they be passed to I/O routines.
+ o pkg/smooth:
+   - fix recomputation that appeared with nlfs : use smooth_recip_hFacC,
+     smooth_hFacW, smooth_hFacS (copied at init_fixed) in 3d case.
+ o verification/testreport:
+   - add option '-tlm' for Tangent-Linear test (-adm,-oad,-tlm are all exclusive);
+   - can also use compressed (gzip) reference output file;
+   - change format (in pkg/grdchk) of TLM output (+ add forward gradient)
+     to be easier to collect by testreport.
+   - update existing TLM output (half out-of-date by more than 3 yrs).
+   - report NETCDF & LAPACK settings (from genmake2) in output dir.
+   - report run-time in each individual summary (previously only for AD test);
+     for TLM test, add also TAF version.
+   - add daily TLM test on (old) aces cluster + weekly test on baudelaire.
+ checkpoint63r (2012/08/15)
+ o tools/genmake2:
+   - fix Makefile for target ftlall (+ update files to clean after make ftlall)
+ o pkg/ecco/ecco_cost_final.F :
+   - fix fc print statements (to stdout and costfunction0000) so
+     that the printed value is always after the global sum.
+ o pkg/seaice/seaice_dynsolver.F :
+   - fix ice strength assymptote at SEAICE_area_max<1
+     (no change for the standard case of SEAICE_area_max=1)
+ o some cleaning:
+   - pkg source files: ensure that all *.F files from each pkg included the
+     corresponding ${PKG}_OPTIONS.h file (except stand-alone src file which
+     only includes CPP_EEOPTIONS.h); fix S/R CTRL_SMOOTH ; update AD experiment
+     code_ad/*.F src files; add new option-file in pkg/admtlm + corresponding
+     includes;
+   - model/src: add PACKAGES_CONFIG.h wherever ALLOW_AUTODIFF[_TAMC] is used.
+   - ALLOW_DIFFKR_CONTROL: add a stop in S/R CTRL_CHECK in case ALLOW_3D_DIFFKR
+     is undef; also define /DYNVARS_DIFFKR/ common block in DYNVARS.h when
+     ALLOW_DIFFKR_CONTROL is defined, to allow to compile and reach this stop.
+   - update option-file CPP_OPTIONS.h of all AD experiments:
+     * no longer include PACKAGES_CONFIG.h ;
+     * includes ECCO_CPPOPTIONS.h if PACKAGES_CONFIG.h has been included ;
+     * def/undef ALLOW_3D_DIFFKR independently of ALLOW_DIFFKR_CONTROL (before
+       including ECCO_CPPOPTIONS.h).
+ o pkg/cost, pkg/ctrl, pkg/ecco option-file:
+   - clean-up and split content of old pkg/ecco/ECCO_CPPOPTIONS.h file into
+     separated pkg option-file (done according to old doc from ECCO_CPPOPTIONS.h,
+     not necessary accurate).
+   - add a template for option-file ECCO_CPPOPTIONS.h, mostly empty, in
+     pkg/autodiff, i.e., the 1rst pkg to compile for adjoint built, with lines to
+     deliberately break compilation (a remainder that it needs to be customized)
+   - in pkg/ecco, includes ECCO_OPTIONS.h instead of COST_OPTIONS.h
+   - rename COST_CPPOPTIONS.h to COST_OPTIONS.h
+   - rename CTRL_CPPOPTIONS.h to CTRL_OPTIONS.h
+ o tools/genamke2:
+   - check for lapack.
+ o pkg/cost and pkg/ecco : attempt to sort out some of the limiting intricacies.
+   - pkg/cost.cost_final.F : allow for use of pkg/ecco and pkg/cost cost functions
+     jointly. The exclusive ifdefs were removed, to add costs terms (when defined)
+     from both packages.
+   - the fc global sum in pkg/ecco/ecco_cost_final.F was removed accordingly.
+   - for objf_atl we use either the pkg/ecco or the pkg/cost term depending on CPPs.
+   - pkg/ecco and model/src : add CPP brackets around pkg/cost and pkg/ctrl includes
+     of "CTRL_SIZE.h" "ctrl.h" "ctrl_dummy.h" "optim.h" / "cost.h" accordingly.
+ o pkg/cost : add comments to explain special and/or untested setups.
+ o pkg/exf:
+   - use EXF_OPTIONS.h to set pkg/exf options except if done in ECCO_CPPOPTIONS.h
+     (previously, ALLOW_AUTODIFF_TAMC was controling this switch).
+ o verification/*/code_ad:
+   mimimal cleaning of AD experiments ECCO_CPPOPTIONS.h option-file:
+   - reccord when this file is included (define ECCO_CPPOPTIONS_H)
+   - remove the most obvious obsolete features (e.g., #define ALLOW_COST not
+     allowed since Oct 2003 in any option-file except in PACKAGES_CONFIG.h)
+ o pkg/seaice :
+   - added possibility to change ice strength/PRESS0 relationship to
+     HEFF, according to the following run time parameters
+     SEAICEpresPow0 :: HEFF exponent for ice strength below SEAICEpresH0 (integer, >0)
+     SEAICEpresPow1 :: HEFF exponent for ice strength above SEAICEpresH0 (integer, >0)
+     SEAICEpresH0   :: HEFF threshold for ice strength            (m)
+ o pkg/ggl90 :
+   - added run time flag mxlSurfFlag to include the code that ensure
+     mixing between first and second level (previously included as a comment)
+ o verification/1D_ocean_ice_column/input_ad/data.autodiff :
+   - add test of newly added dumpAdByRec=T option where
+     we write all records to one file per variable.
+ o verification/global_ocean.cs32x15/code_ad/CPP_OPTIONS.h :
+   - switch to #define ALLOW_BALANCE_RELAX and ALLOW_BALANCE_FLUXES
+     to check that this code is ok with taf and compiles in adjoint.
+ o pkg/autodiff :
+   - added run time param dumpAdByRec : ad dump format switch.
+     If true then write all records to one file per variable.
+     Else write one file per record (remains the default).
+   - revise seaice_ad_dump.F and exf_adjoint_snapshots_ad.F accordingly.
+ o pkg/seaice:
+  - Added SEAICEuseTILT run time switch : if set to
+    false then omit the tilt term in seaice dynamics.
+  - Fix free drift calculation.
+ o pkg/shelfice:
+   When initializing SHELFICEDragQuadratic, set default to be
+   consistent with shiCdrag in case SHELFICEuseGammaFrict is used,
+   but retain option to set it independently
+   (inconsistenty spotted by Veronique Dansereau)
+ o eesupp/inc:
+   - in EEPARAMS.h, add real-type constant for some frequently used simple number
+      (0,1,2,1/2), with name explicitly referring to RS/RL type.
+ o model/src:
+   - move balancing of relaxation terms to new S/R balance_relax.F.
+ o pkg/autodiff:
+   re-write active I/O high-level (control) S/R:
+   - rename  S/R ACTIVE_READ/WRITE_RL to ACTIVE_READ/WRITE_3D_RL
+   - add 1 more argument (useCurrentDir) to replace the "_LOC" 2nd set of S/R.
+   - use standard pkg/mdsio I/O routines (MDS_READ/WRITE_SEC_XZ/YZ  and
+     MDS_READ/WRITE_FIELD); change loops to multi-threads.
+   - update and merge active_file_control_slice.flow into active_file_control.flow
+   - update the calls in lower level S/R (active_file.F, active_file_g.F and
+     active_file_ad.F) keeping the argument list and function unchanged.
+   - also update lower-level "_LOC" version (call same S/R with "useCurrentDir
+     =.TRUE.") and remove high-level (control) _loc version (not called anymore).
+   - new low-level active I/O routines, one for each type (RL/RS), for active var
+     of various (XY,XZ,YZ) shape and level number (arguments: arrShape, myNr);
+     fix ctrl_init_wet.F & ecco_cost_weights.F by calling the new S/R with the
+     right type.
+ o autodiff and RS=real*4:
+   - call the right type (_RS instead of _RL) of S/R in few output S/R (pkg/exf
+     ad-monitor and snapshots); disable arrType='RS' in mdsio_gl*.F ;
+   - change argument type of 3-D mask argument to _RS to match type in the call
+     (in seaice_cost*.F, ctrl_*_gen.F and *ctrl_bound.F); to be continued.
+ o pkg/thsice:
+   - re-anable balanceEmPmR when useSEAICE=T with useThSIce=T.
+   - option to correct ocean fresh-water flux for imbalance of Atmos+Land
+     fresh-water flux: uniform correction or proportional to local precip.
+ o pkg/ctrl:
+   - new header file CTRL_SIZE.h for parameters maxCtrlArr2D, maxCtrlArr3D
+   - redefine CPP option ALLOW_BOTTOMDRAG_CONTROL to
+                         ALLOW_BOTTOMDRAG_CONTROL_NONGENERIC
+   - verification/hs94.1x64x5/ tests ALLOW_GENARR{2D,3D}_CONTROL
+ checkpoint63q (2012/07/30)
+ o pkg/autodiff:
+   note about monitor and snap-shot of ad-variables:
+   a) need to apply ADEXCH before monitor or output of ad-vars to include
+    overlap contributions (was done for adjDump but missing for monitor).
+   b) since ADEXCH implies adding overlap values to interior array, machine
+    truncation error is expected when adding an ADEXCH call in the middle of
+    the time-stepping (although mathematically identical).
+   modifs:
+   - fix some incorrect type (RL/RS, UV/scalar) in ADEXCH calls and ad-vars
+     declaration (adEddyPsiX,adEddyPsiY in adcommon.h).
+   - add new run-time params (dumpAdVarExch & mon_AdVarExch) to select how  to
+     apply ADEXCH before monitor or snap-shot of ad-variables:
+        =0 : no ADEXCH ; =1 : do ADEXCH ; =2 : apply ADEXCH on a local copy.
+   - implement new method (copy+adexch) for snap-shot and monitor of main model
+     ad-vars (initially not used by default -> output unchanged).
+   - change default (set to 2) to use new method and update few output_adm.txt
+     with accurate ad-monitor (global_ocean.90x40x15, tutorial_global_oce_biogeo
+     and both tutorial_tracer_adjsens tests).
+   - switch 2 global_ocean.cs32x15 AD-tests to AB-2 (seaice_dynmix & thsice)
+     and turn off momDissip_In_AB in the 2 other (needed for stable AB-3).
+ o pkg/ctrl:
+   enable the use of a variable number (maxCtrlArr2D, maxCtrlArr3D) of
+   generic control variables via
+   #define ALLOW_GENARR2D_CONTROL and/or #define ALLOW_GENARR3D_CONTROL
+   Only required customization should be S/R ctrl_map_ini_genarr.F
+   (routine in repository contains example)
+ o pkg/exf:
+   - fix and refine checking of surface-relaxation setting when using EXF
+ o verification (adjoint exp):
+   - reduce length of simulation by half for experiment global_ocean.cs32x15
+     (std, seaice & seaice_dynmix) and exp tutorial_dic_adjoffline; update output.
+ o pkg/mom_common (mom_u/v_bottomdrag), pkg/shelfice (shelfice_u/v_drag)
+   - replace if statements KE.ne.0 with KE.gt.0 to avoid sqrt of negative numbers
+     of KE that can occur with aggressive optimization
+   + put quadratic bottom drag calculation in separated loops (only done if
+     using bottomDragQuadratic)
+ o model/src: CG2D_NSA
+   - clean-up:  stop if cg2dNormaliseRHS=T and AUTODIFF
+      and replace GLOBAL_SUM calls with GLOBAL_SUM_TILE.
+   - in test-exp that uses CG2D_NSA (bottom_ctrl & tutorial_tracer_adjsens),
+      change cg2d residual target to be expressed in Wunits (=> do not normalise
+      RHS ; this fix global_max Pb with TAF); update output.
+ o verification/tutorial_tracer_adjsens:
+    improve parameter consistency (and get rid of warnings):
+    - use deltaTFreeSurf=deltaTtracer (better with NonLin FreeSurf);
+    - switch Temp Adv Scheme from 81 to 80 (no limiter since theta < 0 is physical)
+    - use staggerTimeStep (for stability) in som81; use default rotation period.
+ o pkg/exch2:
+   - with empty facet: fix compact-format (W2_mapIO=1) definition (was previously
+     reset to 0 = 1 long line in X)
+ o pkg/mdsio:
+   - fix reading of dimensions from meta file for large-size domain
+ o model/src:
+   - fix saltIsActiveTr setting for IdealGas eosType (eosType is 6 character long);
+   - to refine CS-grid check, add logical flag "hasWetCSCorners": true if using
+     Cubed-Sphere Exch with CS-corners inside the domain;
+     make new S/R INI_GLOBAL_DOMAIN from code in ini_linear_phisurf.F to calculate
+     globalArea and to set hasWetCSCorners.
+   - add warning if usingCurvilinearGrid and momAdvection in flux-form (missing
+     metric terms); stop if hasWetCSCorners=T and momAdvection in flux-form.
+   - re-activate the stop if useCDscheme and hasWetCSCorners=T (previously if
+     useCDscheme and useCubedSphereExchange).
+ o pkg/seaice/seaice_growth.F :
+   - compute the SEAICE_DISABLE_HEATCONSFIX term accordingly
+     for the (.NOT.useRealFreshWaterFlux).OR.(nonlinFreeSurf.EQ.0)
+     when the previous code was not correct.
+   - in all cases diagnoze the boundary condition mismatch to SIaaflux
+   - in the case of useRealFreshWaterFlux.AND.(nonlinFreeSurf.GT.0) add
+     that term to QNET (essentially to set the melt/freeze water to 0degC).
+     when SEAICEheatConsFix is true, so that the ocean+ice budget heat is closed.
+ o verification/global_ocean.cs32x15/input.seaice :
+   - add SEAICEheatConsFix in data.seaice, ensuring closed ocean+ice heat budget.
+   - add diagnostics (commented out) for budget test using my matlab codes (gcmfaces).
+   - update results (output.seaice.txt) accordingly.
+ o verification/lab_sea/input/data.seaice :
+   - remove SEAICEheatConsFix, which is now be tested
+     in global_ocean.cs32x15. The presently coded fix
+     is only consistent with real fresh water + nlfs.
+   - update results (output.txt) accordingly.
+ o pkg/salt_plume: added diagnotic for salt tendency (oceSPtnd).
+ o AD_CONFIG.h file:
+   - in genmake2: remove "#undef ALLOW_ECCO_OPTIMIZATION" from AD_CONFIG.h;
+   - remove #include "AD_CONFIG.h" from all included OPTIONS files and
+     add it explicitly in each fortran src file where it is needed.
+     motivation: header file AD_CONFIG.h is generated/modified after "make depend"
+     thus introducing potentially wrong dependency; therefore, better to include
+     it only where it's really needed, and after any other included OPTIONS file
+     (less tempting to bring conditional includes according to ALLOW_ADJOINT_RUN
+     or ALLOW_TANGENTLINEAR_RUN).
+ checkpoint63p (2012/07/06)
+ o pkg/autodiff:
+   - rename S/R INI_AUTODIFF to AUTODIFF_INIT_VARIA ;
+   - create new header file "AUTODIFF_PARAMS.h" for pkg/autodiff parameters
+     which are now read from new parameter file "data.autodiff" (inAdExact
+     previously in "data", use{KPP/GMRedi/SEAICE}inAdMode previously in data.pkg)
+   - add file "data.autodiff" in all verification/*/input_ad/ directories.
+   - add run-time parameter to turn On/Off SEAICEuseFREEDRIFT in adjoint/forward
+     mode (used in global_ocean.cs32x15.seaice_dynmix test exp.)
+ o verification/tutorial_tracer_adjsens/
+   - add extra sub-verification which uses "SOM" advection scheme
+     (2nd Order-Moment Advection Scheme, Prather Limiter) with
+       #define GAD_ALLOW_TS_SOM_ADV in GAD_OPTIONS.h
+     and in data:
+       multiDimAdvection=.TRUE.,
+       tempAdvScheme=81,
+       saltAdvScheme=81,
+ o model/src:
+   - introduce ALLOW_BALANCE_RELAX which allow the removal
+    of the global mean of relaxation terms by setting
+    balanceThetaClimRelax and balanceSaltClimRelax
+   - disable balanceEmPmR and balanceQnet in the case when useSeaice.
+    This case is now treated appropriately in seaice_growth.F
+ o pkg/seaice:
+   - do balanceEmPmR and balanceQnet based on the net fluxes
+     (SItflux) that enter the ocean+ice system rather than QNET
+     and EMPMR. Works best in real fresh water.
+   - SIatmFW, SIatmQnet, and SItflux computations had to be moved
+     out of the diagnotics bloc, and stored in global arrays.
+ checkpoint63o (2012/06/29)
+ o vertical mixing schemes:
+   - remove unused S/R CALC_BUOYANCY (left from branch-atmos merging)
+   - replace computation of density (in Function STATE1, opps_calc.F) by a call
+     to S/R FIND_RHO_SCALAR
+   - pass sigmaR to S/R PP81_CALC, MY82_CALC (but not yet used) and to GGL90_CALC
+     and use it for N^2 calculation in ggl90_calc.F (save 2 density computations);
+     affects output (vermix.ggl90) at truncation level.
+ o testreport:
+   - change adjustment of SIZE.h for mpi+mth testing: preserve enough tiles
+     (nSx,nSy) for the number of threads (from eedata.mth)
+ o model/src:
+   - remove 2nd set of calls at the top of initialise_varia.F (Adjoint part)
+     related to Non_Lin Free-Surf or Bottom-Control (but not longer necessary).
+ o verification/tutorial_tracer_adjsens/
+   - compile with ALLOW_CG2D_NSA
+   - run with nonlinFreeSurf=3
+   - update reference output_adm.txt
+ o tools/genmake2:
+   - add "-DHAVE_NETCDF" to the DEFINES list when NetCDF is available ;
+     refine usePROFILES switching off (only when NetCDF is not available).
+ o verification (atmospheric set-up):
+   - change eosType='IDEALGAS' to eosType='IDEALG' to match variable declaration
+ o pkg/profiles : refine profiles and files handling.
+ o model/src and pkg/ecco : introduce run-time switch usePROFILES.
+ o model/src/the_main_loop.F, forward_step.F and pkg/exf/EXF_OPTIONS.h :
+   - changes to use model/src/the_main_loop.F rather than the ecco version
+     in ecco type adjoint runs.
+ o model/src:
+   - move computation of recip of grid distance & area from INI_MASKS_ETC to INI_GRID
+   - move setting of h0Fac[C,W,S] from INI_LINEAR_PHISURF to INI_MASKS_ETC
+   - update_masks_etc.F: rename SMOOTH*_R4,R8 function to the corresponding type (RS,RL)
+ o model/src:
+   - fix S/R UPDATE_CG2D for ALLOW_DEPTH_CONTROL case (was empty before);
+     update output of test experiment bottom_ctrl_5x5.
+ o pkg/generic_advdiff:
+   - implement vertical bi-harmonic diffusion (for now, only for T & S and
+     only explicitly).
+ o verification/bottom_ctrl_5x5/
+   update reference output (results changed quite some time ago,
+   not sure when; but new reference results look ok in terms of grdchk)
+ o verification/1D_ocean_ice_column/
+   For verification purposes, the exact adjoint rather than an approx.
+   should be used (e.g., to comare against F.D. or TLM).
+   removed autodiff_inadmode*_ad.F routines in code_ad/
+   Update results/ accordingly.
+ o pkg/cal:
+   - change cal_getdate.F so that output date is function of input time only
+     (less ambiguous, not mixed up with input iteration number).
+   - fix calendarDumps for non-zero phase-lag (S/R CAL_TIME2DUMP).
+ o pkg/diagnostics:
+   - implement calendarDumps also for stats-diags
+   - implement calendarDumps also for snap-shot output (negative freq)
+   - for now, ignores calendarDumps if non-zero phase-lag.
+ checkpoint63n (2012/06/04)
+ o model/src:
+   - recompute rSurfW & S to ensure consistent column thickness (needed for rStar)
+     between Sum_k(hFac*drF) and rSurf-rLow, for now only when using pkg/shelfice;
+   - use h0FacC, W & S to compute hFac_surfC, W & S in calc_surf_dr.F (more
+     similar to calc_r_star.F); update results of experiment internal_wave and
+     tutorial_global_oce_in_p (affected @ truncation level).
+ o pkg/shelfice:
+   - add parameterisation of advective-diffusive flux into the ice shelf
+     according to Holland and Jenkins (1999), eq.22-33, as suggested by Jan De Ryd
+   - simply eps3/3a computation
+ o eesupp:
+  - add new CPP option (DISCONNECTED_TILES) to disconnect tiles (no exchange
+    between tiles, just fill-in edges assuming locally periodic subdomain) and
+    add corresponding exchange subroutines (EXCH0);
+  - new CG-solver version (_EX0) for disconnected-tiles special case.
+ o model/src (CG-solvers):
+  - modify Single-Reduction version (CG2D_SR) in case cg2dMaxIters is reached
+    (similar final residual output and max-Iter number as with standard version);
+  - allows to store and use lowest-residual solution (standard & _SR versions);
+  - cleaning up (remove old, commented out code, LETS_MAKE_JAM ...).
+  - by default, also compile CG2D_SR version (define ALLOW_SRCG in CPP_OPTIONS.h)
+ o eesup/src:
+  - global_vec_sum.F cleaning: add a BARRIER ; update comments ; remove ndef
+    ALWAYS_USE_MPI (no longer needed after fixing usingMPI settings - March 2012)
+ o pkg/salt_plume added to lab_sea/input_ad.noseaicedyn
+ checkpoint63m (2012/04/27)
+ o pkg/shelfice:
+  - account for partical cells in surface level below ice-shelf for hydrostatic
+    pressure calculion (set uniformFreeSurfLev to F); update isomip experiments.
+ o ad test experiments:
+  - update and rename to "ad_optfile.local" local AD-optfile specific to a
+    particular adjoint verification experiment (in verification/*/code_ad/)
+ o pkg/seaice & pkg/exf:
+  - Part two of ECCO-CLIVAR code merges.
+    Completes merge of seaice area relax. code.
+    Requires unified CPP option EXF_ALLOW_SEAICE_RELAX
+    to be set in EXF_OPTIONS.h since this pkg controls its I/O
+    (and avoids second CPP option in SEAICE_OPTIONS.h)
+  - change default value of costIceStart & costIceEnd to a valid date.
+  - add snowPrecip to HSNOW in seaice_growth
+ o pkg/ptracers:
+  - move ptracers summary and setup check into separated S/R ; add a check for
+    valid PTRACERS_Iter0 value (error+stop or just warning).
+ o pkg/cal (+ pkg/ecco):
+  - avoid calling a S/R with 2 times the same arg (FWD: unless both are only used
+    as input; and for AD, unless both are not differentiable (e.g., k index)).
+ o model/src:
+  - define internal flag "uniformFreeSurfLev" (=T when free-surface level-index
+    is everywhere k=1) to speed-up some calculations (by pass kSurf reference);
+    temporary set to usingZCoords value (to be changed to F when using shelfice).
+  - hydrostatic pressure: implement partial cell near the surface also for
+    Z-coordinates (for the case where uniformFreeSurfLev=F); replace test on
+    usingZCoords with test on uniformFreeSurfLev where needed (in NonHyd code).
+ o pkg/ctrl/ctrl_init.F:
+  - add a formal parameter to s/r ctrl_init_rec to print
+    xx_$(ctrl_valiable)_file for easier debugging
+  - move call of cal_* routines into the if-block where they are really
+    needed so that xx_$(ctrl_variable)_period can be 0 or -12 without having
+    to specify a startdate
+ o pkg/seaice:
+  - update comments in SEAICE_OPTIONS.h (+ similar update of lab_sea/code and
+    global_ocean.cs32x15/code versions).
+  - turn on seaice Multi-category (in data.seaice, SEAICE_multDim=7) for
+    verification experiment global_ocean.cs32x15.seaice and update results.
+ o pkg/exf:
+  - new S/R EXF_GETFIELD_START to return starting-time of a forcing field
+    (according to useExfYearlyFields).
+ o pkg/cal:
+  - move calendar setting & summary (previously in cal_readparams.F) into new
+    S/R CAL_INIT_FIXED to be called later in the initialisation process, once
+    length of the simulation is set (after SET_PARMS call); This is an attempt
+    to fix Pb when, e.g.: pkg/longstep, modifies the length of the run.
+  - For safety: keep a record of pkg/cal parameter settings status
+            to allow to STOP if one pkg/cal S/R is called too early.
+  - add a stop in all basic pkg/cal subroutines if called before setting is done.
+  - modify cal_getdate.F: stop when nothing was returned;
+    if called with myIter=-1, simply retrun modelBaseDate (1,2 = startDate_1,_2);
+    this allows to remove "cal.h" from seaice_read_parms.F & init_mnc_vars.F.
+ checkpoint63l (2012/04/04)
+ o pkg/bbl:
+   - fix call to FIND_RHO_2D in bbl_calc_rho.F (case multi-tiles per proc);
+     update global_with_exf output.yearly.txt file.
+ o pkg/cal:
+   - start to clean-up cal_set.F ; fix for the case where baseTime is non-zero.
+   - remove 2 unsused and very short (1 line only) functions ;
+   - remove unsused subroutines from AD-Diff list.
+ o model/src:
+   - start to implement clean wait to stop in initialization S/R (config_check.F,
+      packages_check.F, ini_parms.F and pkg/obcs/obcs_check.F + pkg/mypackage)
+ o folowing PRINT_MESSAGE changes:
+   - remove reminder multiple print to the same file (STDOUT) when multi-threads.
+ o pkg/bbl:
+   - modification so it can work with z*
+   - replace deltaT with dTtracerLev(k) for asynchronous time stepping
+   - global_with_exf/input.yearly verification; changes output.yearly.txt
+ o eesupp/src:
+   - PRINT_MESSAGE: use OpenMP critical instruction (similar to PRINT_ERROR);
+     writing to unit 0 (if unit=errorMessageUnit) extended to the case myThid=1
+   - CHECK_THREADS: specific code if using OpenMP threading: improve checking of
+     # of threads across processors (e.g., different env OMP_NUM_THREADS setting)
+   - improve special preliminary call to EESET_PARMS (if eedata is read 2 times).
+   - when using MPI, always set usingMPI=T ; set default accordingly and stop
+     if not consistent.
+   - ALL_PROC_DIE: add a BARRIER after MPI_FINALIZE call ; skip this when thread
+     argument is zero (in case called within single-thread section).
+ o pkg/fizhi:
+     clean-up turbulence cold-start switch: decided in fizhi_init_vars.F, stored
+     in common bloc (fizhi_coms.h) and then passed as argument up to S/R TURBIO.
+ o pkg/exch2:
+   - add argument "updateCorners" to S/R exch2_get_uv_bounds (enable to
+     activate EXCH_IGNORE_CORNERS in vector EXCH S/R).
+   - switch 1rst EXCH call to IGNORE_CORNERS (instead of UPDATE_CORNERS)
+     but keep the 2nd with UPDATE_CORNERS: this prevents overwriting good data
+     with bad ones coming from a not-yet-updated halo.
+     This should fix the problem when sNx <> sNy and GCD(sNx,sNy) < 2*Olx.
+ o pkg/frazil:
+     move FrzForcT diag from frazil_diagnostics_state.F to frazil_calc_rhs.F
+     http://mitgcm.org/pipermail/mitgcm-devel/2012-March/005384.html
+ o model/src/config_check.F: add a check for OLx>sNx/OLy>sNy
+ o verification/lab_sea
+   * #undef SEAICE_EXCLUDE_FOR_EXACT_AD_TESTING
+   * #define ALLOW_SST_CONTROL & ALLOW_SST_CONTROL
+   * nTimeSteps 3 -> 4
+   * changed reference output accordingly
+ o pkg/ctrl:
+   * fix ctrl_get_gen for xx_sst, xx_sss
+ o pkg/fizhi:
+   - fix 2 calls to S/R GRD2MSC (argument igrd) in update_earth_exports.F
+     update results of exp. fizhi-cs-32x32x40 (fail with only 6 digits for cg2d)
+   - by default, use new version of S/R GETPWHERE (#define TRY_NEW_GETPWHERE);
+     fix few out-of-bounds indices in referencing vars (initialise igrd to 1);
+     clean-up fizhi_turb.F (remove unused & uninitialised qdiaglsm array).
+ o pkg/gmredi:
+   - add diagnostic of Redi (main diagnonal) vertical diffusive flux of Temp
+ o pkg/fizhi:
+   - fix few initialisation and multi-tile Pb: aqualev & gridalt now pass the
+     (old)aces open64 tuv test.
+ o pkg/icefront: fix bug and reorder loop in icefront_init_varia.F
+ o pkg/mom_vecinv:
+   - fix 1 loop range in mom_vecinv.F to be like all the other loops which
+     set/update gu,gv/Diss (fix Pb of uninitialised fVerUkp in admom_vecinv
+     with non-lin free-surf).
+ o model/src: move k-loop outside of i/j-loops when (re-)computing R_low
+   in ini/update_masks_etc.F in order to avoid -O3 optimization problems with
+   some compilers (gfortran 4.6.0 on MacOS)
+ o momentum:
+   - separate fVer?(:,:,kUp) & fVer?(:,:,kDown) in argument list of MOM_FLUXFORM
+     & MOM_VECINV subroutines (to help TAF). This fixes Pb of uninitialised
+     fVer?(kUp) in admom_fluxform with non-lin free-surf or bottom control.
+ checkpoint63k (2012/03/17)
+ o model/src:
+    - selectAddFluid (ALLOW_ADDFLUID): add missing term in horizontal momentum
+      flux-form (mom_fluxform.F) and vertical momentum (calc_gw.F) equations.
+    - SOLVE_DIAGONAL: switch default to use adjoinable & vectorized version
+      (#undef SOLVE_DIAGONAL_LOWMEMORY); update output of exp. front_relax.bvp.
+ o model/src: solve_*diagonal.F
+    - replace division with multiplication by inverse (similar to original version)
+      update output of AD exp. global_ocean.90x40x15.
+    - fix pkg/gmredi/gmredi_calc_psi_bvp.F & pkg/ggl90/ggl90_calc.F
+      to enable the use of the vectorized & differentiable solve_tridiagonal.F
+    - test exp. global_ocean.90x40x15/input_ad: switch to advection scheme 20 for
+      vertical advection (to get solve_tridiagonal tested in AD mode).
+ o pkg/seaice: ECCO-CLIVAR merge:
+. seaice AREA relaxation
+      * CPP option: SEAICE_ALLOW_AREA_RELAXATION
+      * runtime flag: SEAICE_tauAreaObsRelax
+. modified global mean cost imbalance code: CPP option
+       ALLOW_SSH_GLOBMEAN_COST_CONTRIBUTION
+ o another iteration on solve_*diagonal.F
+    Make choices independent of AUTODIFF
+    Default will be the adjoinable vectorized version.
+    Alternative choices are:
+    * non-adjoinable original version with low memory footprint:
+      #define SOLVE_DIAGONAL_LOWMEMORY
+    * still adjoinable but does not vectorize (k as inner-most loop):
+      #define SOLVE_DIAGONAL_KINNER (keeping SOLVE_DIAGONAL_LOWMEMORY undef)
+ o testreport
+  - add option to use a home made wrapper around "make"
+  - remove MITGCM_IEEE env. variable (no longer active)
+ o pkg/shelfice
+  - improve computation of friction velocity for velocity dependent
+    exchange coefficients: include velocities into SHELFICEBoundaryLayer
+    block to average velocities vertically
+  - add TAF store directives for SHI_ALLOW_GAMMAFRICT
+ o Modify diagonal solvers, now three versions (upon request):
+   #ifndef SOLVE_DIAGONAL_AUTODIFF
+. default version (dating back a while ago)
+   #else /* def SOLVE_DIAGONAL_AUTODIFF */
+   # ifdef SOLVE_DIAGONAL_KINNER
+. adjoinable version with k-innermost loop (G. Forget)
+   # else
+. adjoinable vectorized version of 2.
+   # endif
+   #endif /* SOLVE_DIAGONAL_AUTODIFF */
+   All flags are currently set within routines, but could/should be moved to CPP-options.
+   Array bounds for 1. are as before, for 2.,3. cover the entire domain
+   (could be changed for 1. too).
+ o verification/lab_sea (fwd):
+   - update the 4 fwd lab_sea test experiment: no longer SEAICE_GROWTH_LEGACY,
+     use 7 categories (except hb87), new solve4temp params (except salt_plume),
+     constant salinity (=4) or variable salinity (salt_plume) done through
+     SItracer.
+ o pkg/seaice:
+   - rename parameters: SIsalFrac to SEAICE_saltFrac & SIsal0 to SEAICE_salt0;
+   - remove old parameters ( SEAICE_freeze, SEAICEturbFluxFormula, SEAICE_gamma_t,
+      SEAICE_gamma_t_frz, SEAICE_availHeatTaper & SEAICE_availHeatFracFrz)
+     from SEAICE_PARMS.h ; still need to fix seaice_cost_sst.F (broken when
+     SEAICE_freeze got retired) but already better than previously with
+     SEAICE_freeze=123456.7 deg.C;
+   - move setting of facOpenGrow/Melt from SEAICE_INIT_FIXED to SEAICE_READPARMS
+     (safer multi-threaded setting); always set SEAICEuseEVP;
+   - setting of ocean-ice turb. flux coeff: moved from SEAICE_CHECK to
+     SEAICE_READPARMS & SEAICE_INIT_FIXED (this fixes wrong summary report);
+     stop if multiple specifications for the same coeff; make sure default
+     SEAICE_mcPheePiston is compatible with drF(1) & deltaT.
+ o pkg/obcs:
+   - allow to switch to upwind 1rst order advection scheme for the advective
+     flux computation at the open-boundary; disabled for adjoint built.
+ o pkg/ptracers:
+   - rename PTRACERS_RESTART.h -> PTRACERS_START.h
+   - rename ptracers_fields_blocking_exchanges.F -> ptracers_fields_blocking_exch.F
+   - add parameter to decide when to start a tracer; use internal flag to
+     know which tracer to step-forward; switch this internal flag at the top
+     of forward_step.F (new S/R PTRACERS_SWITCH_ONOFF); for now, not enabled
+     for adjoint built.
+ o pkg/seaice:
+   - add SEAICE_SIZE.h anywhere in the code where SEAICE.h is included ;
+   - move MULTDIM declaration and setting from SEAICE.h to SEAICE_SIZE.h ;
+   - keep experiment offline_exf_seaice truly single category (set MULTDIM=1).
+ o pkg/ecco:
+   - fix writing of output file DiagnosCost_ClimSalt/Theta
+     and DiagnosCost_DriftSalt/Theta (if multi-tiles per proc).
+ o pkg/seaice:
+   - add missing initialisation of COR_ICE array in dynsolver.F
+   - merging SEAICE_MULTICATEGORY define and undef cases. Now the same code is
+     compiled either way, and is tested in all exps (multidim was only tested
+     in lab_sea ad before).
+   - cosmetic change to prepare for future addition of ITD. no change to results.
+     no defaults change.
+   - added run time param SEAICE_multDim (7 for SEAICE_MULTICATEGORY, 1 otherwise).
+   - by lack of adequate verification experiments, I further used custom versions
+     of global_ocean.cs32x15 to test multicat or not, with pickup or not. All
+     should be correct.
+   - avoid recomputations in ad.
+ o array initialisation in adjoint mode:
+) pkg/mom_common/mom_calc_ke.F:
+     add full array initialisation of KE inside mom_calc_ke.F (ALLOW_AUTODIFF_TAMC):
+     The S/R which calls MOM_CALC_KE initialises KE but in some recomputation
+     part this initialisation is gone, resulting in Floating Point Exception
+     (caught by open64 compiler with debug option).
+) model/src/convective_weights.F
+     set output convectCount over the full array range (instead of doing
+     an update): this fix a similar issue when called within recomputation (adjoint)
+     which could by-pass the initialisation (normally done in the caller S/R).
+ o pkg/seaice:
+   - final version of turbulent flux params, as agreed at MIT ice meet:
+        SEAICE_mcPheePiston (default is ~ 0.000875)
+        SEAICE_mcPheeTaper (default is 0.)
+        SEAICE_frazilFrac (defualt is 1.)
+        SEAICE_mcPheeStepFunc (default is false)
+   - changes results at trunction error level
+   - ifdef SEAICE_EXTERNAL_FLUXES, disallow SEAICE_waterAlbedo
+ o added pkg/frazil:
+   - see frazil_description.tex for details
+   - verification experiment is global_with_exf/input.yearly; changes output.yearly.txt
+ o pkg/shelfice/shelfice_thermodynamics.F
+   - fix exponents for velocity dependent exchange coefficients: change
+     from 2/3(=0) to 0.66666666666666667
+   - move computation of velocity dependent exchange coefficients into
+     separate loop, so that it applies to both thermodynamics
+   - move some divisions by constants out of loops
+ checkpoint63j (2012/02/16)
+ o pkg/seaice:
+   - overhaul of seaice_summary.F after feb 2012 clean-up of seaice_growth.F etc.
+   - SItracer:  added run time params, and allow coupling to ocn salinity.
+ o pkg/seaice: remove (almost) all divisions by constants within loops
+   in seaice_growth.F and seaice_solve4temp.F, this changes (only) the cg2d
+   results of global_ocean.cs32x15.seaice from 16 to 11 digits (but only
+   those, caused by 1/QI -> recip_QI) -> update output.seaice.txt,
+   output_adm.seaice_dynmix.txt and also lab_sea: output_adm.evp.txt
+ o pkg/seaice :
+   - simplify McPhee tapering, and allow to use it with SEAICEturbFluxFormula
+     = 1 or 2, using newly added run time parameter SEAICE_availHeatTaper.
+   - fix d_AREAbyATM, d_AREAbyOCN, d_AREAbyICE diags I broke in r1.148.
+   - add permanent SItflux diag that corresponds to TFLUX but includes
+     ice+snow. Hence SItflux-TFLUX should match the ice+snow heat budget.
+   - allow activation/testing of a fix for suspected missing term in
+     ocn-ice heat budget (to be confirmed). To test this, you want to
+     undef SEAICE_DISABLE_HEATCONSFIX, and then set the run time param
+     SEAICEheatConsFix to .TRUE. that also allows the 'SIaaflux' diagnostic.
+   - no change in results, fwd or ad.
+ o global_ocean.cs32x15/results: 3 changes following from the pkg/seaice commit below:
+   - switch to variable freezing point.
+        Results change : cg2d -> 1 digits; ad -> 3-6-6 digits.
+   - bug fix : the sublimation term that was missing is now activated.
+        Results change : cg2d -> 2 digits; ad -> 8-6-6 digits.
+   - bug fix : flooding sign term for SIsal0.NE.0.
+        Result change : cg2d -> 8 digits; ad -> 11-12-9 digits.
+ o pkg/seaice:
+   - unified freezing point treatment : the old SEAICE_VARIABLE_FREEZING_POINT
+     is now the default, and the old default constant freezing point is recovered
+     with SEAICE_tempFrz0    = -1.96,
+          SEAICE_dTempFrz_dS = 0.,
+   - retire SEAICE_freeze (= the old way of specifycing the constant freezing point)
+   - remove ALLOW_SEAICE_FLOODING brackets; run time switch is already there.
+   - bug fix (thanks to M. Losch) : the sublimation term that was missing is now
+     activated. To reproduce old results with this bug define SEAICE_DISABLE_SUBLIM.
+   - bug fix (silly me) : flooding sign term (seaice_growth.F r 1.149line 1211)
+     was wrong for SIsal0.NE.0. Changes cs32x15 results (see above item for detail).
+   - bug fix (thanks to O. Jahn) : area loss for melting  got messed up in
+     seaice_growth r149 for legacy branch (no results change).
+   - introduce SEAICE_CAP_SUBLIM : caps sublimation heat flux in solve4temp (code
+     from I. Fenty).
+   - results did not change except for global_ocean.cs32x15, mostly due to the
+     switch to variable freezing point (see above item for detail).
+ o pkg/thsice:
+   - fix and simplify interface to salt_plume pkg ; diagnose directly amount of
+     seawater freezing in thsice_calc_thickn (not used).
+ o pkg/seaice: turbulent flux term specification via run time params.
+   - replace CPP switches with run time switches to facilitate testing
+   - retired CPPs : MCPHEE_OCEAN_ICE_HEAT_FLUX,
+                    GRADIENT_MIXED_LAYER_TURBULENCE_FACTOR
+   - added runtime parameter : SEAICEturbFluxFormula
+   - move Mcphee constants to SEAICE_PARAMS.h
+   - seaice_readparms.F : read turbulent flux term params (SEAICEturbFluxFormula;
+              SEAICE_availHeatFrac/Frz, or SEAICE_gamma_t/_frz if apply);
+              remove seaice_summary call (moved to seaice_init_fixed)
+   - seaice_init_fixed.F : set SEAICE_availHeatFrac/Frz depending on
+     SEAICEturbFluxFormula, and seaice_summary call (moved from seaice_readparms)
+   - seaice_summary.F : add missing report of read turbulent flux term set up
+                        and SITRACER.
+   - seaice_check.F : check range; inform and stop if retired CPPs are used
+                      (incl. SEAICE_AGE).
+   - seaice_growth.F : unified use of SEAICE_availHeatFrac/Frz, without
+          CPP switch, same computation as before for the different options.
+   - change results at the truncation level.
+ o pkg/seaice:
+   - remove un-used parameters LAD & MAX_TICE ; add new run-time parameters:
+     useMaykutSatVapPoly (default=F), postSolvTempIter (default=2) and
+     SEAICE_wetAlbTemp (default=-1e-3). Remove option SEAICE_SOLVE4TEMP_LEGACY.
+   - update verification/*/input*/data.seaice:
+     to reproduce previous results with former #defined SEAICE_SOLVE4TEMP_LEGACY
+     code, set useMaykutSatVapPoly=T & postSolvTempIter=0 ;
+     and to reproduce previous results with former non-SOLVE4TEMP_LEGACY code,
+     set SEAICE_wetAlbTemp=0 & SEAICE_snowThick=0.
+ o pkg/seaice/seaice_solve4temp.F (SOLVE4TEMP_LEGACY):
+   - remove LEGACY code for solving for tsurf (A1,A2,A3) but maintain the same
+     algorithm (same choice: useMaykutPolySatVap=T, postSolvTempIter=0);
+     difference in results only due to machine truncation.
+   - remove MAX_TICE (tsurf is always =< TMELT anyway); keep MIN_TICE if using
+     MaykutPolySatVap; keep MIN_LWDOWN.
+   - adapt SEAICE_MODIFY_GROWTH_ADJ code (untested) to non-legacy formulation.
+   - update output of verification exp. lab_sea (4 fwd: 11 cg2d digits except hb87,
+     + 3 AD: 6 digits except evp) and offline_exf_seaice.seaice (12 digits).
+ o pkg/salt_plume:
+   - isolate saltPlumeFlux array exch into a specific S/R called from
+     do_oceanic_phys.F
+   - add interface to pkg/salt_plume (filling of saltPlumeFlux) in pkg/thsice.
+ o verification:
+   - move untested set-up to MITgcm_contrib/verification_other.
+ o pkg/seaice:
+   - retire old way of SEAICE_AGE. Now replaced with ALLOW_SITRACER.
+ o pkg/seaice:
+   - seaice_solve4temp.F (keeping results unchanged):
+     * add 2 future run-time params to select which saturation vapor pressure
+       formula is used and how post-iteration fluxes are set.
+       Temporary set these 2 params according to option SOLVE4TEMP_LEGACY;
+     * New piece of code (not used) to update surf. atmos. heat flux F_ia using
+       the linearized expression to stay consistent with tsurf finding.
+   - seaice_freedrift.F avoid floating exception in sqrt(y) since y=sqrt(x*x)-x
+     can be < 0.
+ o pkg/seaice/seaice_solve4temp.F (SOLVE4TEMP_LEGACY undef):
+   - remove from F_ia (and from it's derivative dFiDTs1) the contribution of
+     conductive heat flux F_c , and add it explicitly when updating tsurf as
+     solution of: Fc = Fia + d/dT(Fia - Fc) *delta.tsurf
+     so that now F_ia has a consistent meaning through the entire routine.
+     This causes differences at truncation level: update output of fwd exp.
+     global_ocean.cs32x15.seaice.
+ o pkg/seaice:
+   - lots of editing/cleaning in seaice_solve4temp.F + better documentation
+     of arrays, sign/direction of fluxes; no effect on results.
+   - replace hard coded 273.16 (=previous value of celsius2K) by celsius2K
+     in seaice_budget_ocean.F & seaice_solve4temp.F (SOLVE4TEMP_LEGACY part);
+     note: this should have been done before changing celsius2K default.
+   - to keep results unchanged, add back old value of celsius2K in verification
+     exp. input*/data : lab_sea (AD), offline_exf_seaice (fwd+AD) & seaice_obcs
+   - replace hard coded seawater freezing temp in seaice_solve4temp.F (part
+     SOLVE4TEMP_LEGACY) of 271.20 (=previous value of celsius2K + default
+     value of SEAICE_freeze) by celsius2K+SEAICE_freeze so that seawater
+     freezing temp is consistent with value in seaice_growth.F ; this causes
+     differences at machine truncation level: update lab_sea AD output (3/4).
+ o pkg/seaice:
+   - mask (in exf_init_fixed.F) atemp,aqh,evap,precip,runoff,lwflx,areamask
+     and climsst/sss even when using seaice (to prevent calculation that might
+     use SST on land); does not affect any verification exp.
+   - apply mask to taux,tauy (in seaice_get_dynforcing.F); update output
+     of lab_sea.fd & lab_sea.hb87 and global_ocean.cs32x15.seaice_dynmix.
+ o pkg/seaice/seaice_growth.F: add mask to avoid non-zero ice thickness
+   over land (non-legacy code)
+ o vectorize salt_plume package:
+   - add extra code to salt_plume_frac and salt_plume_tendency_apply_s
+   - this code computes plumebk everywhere (also over land), thus requires
+     code to avoid divisions by zero and catch the case of kLev=Nr.
+   - efficiency of salt_plume_frac is still sub-optimal
+ o tools (genmake2):
+   - fix makefile NAME specification (option: -makefile NAME) by adding
+     "-f $MAKEFILE" in few places where it was missing including makedepend
+     command (contribution from Paul).
+ checkpoint63i (2012/01/24)
+ o model/src:
+   - add factorized versions of some EOS code to find_rho.F,
+     define USE_FACTORIZED_EOS to turn it on (default for TARGET_NEC_SX)
+ o model/src:
+   - add celsius2K to namelist PARM01; remove unused recip_rhoNil from PARAMS.h;
+   - change default celsius2K from 273.16 to 273.15
+   - to keep results unchanged, add back old value of celsius2K in verification
+     exp. input*/data : aim.5l_cs, 1D_ocean_ice_column, lab_sea (fwd),
+      global_ocean.cs32x15(seaice) & cpl_aim+ocn(input_atm).
+ o model/src: add a little initialisation trick to find_rho.F and find_alpha.F
+   to make TAF generate vectorizable code
+ o pkg/seaice/seaice_evp.F: add code (copied from CICE) that avoids underflows.
+    Turn on by defining SEAICE_EVP_ELIMINATE_UNDERFLOWS. Most compilers
+    have flags that do this more efficiently.
+ o pkg/diagnostics/diagstats_calc.F
+   - for TARGET_NEC_SX fix the treatment of the scaling factor tmpFac=scaleFact
+ o pkg/obcs (with exch2):
+   - derive OB indices in overlap region from OB-InsideMask (in obcs_init_fixed.F)
+     and remove code (in obcs_readparms.F) which was based on EXCH of indices
+     (through buffer) and did not account for exch2 (e.g., CS-grid) topology.
+ o pkg/exf and seaice exp:
+   - change offline_exf_seaice.seaicetd test exp. to use default DIFF1 value (=0)
+     so that switching off area & heff advection flags does not affect results.
+   - change default EXF_LWDOWN_WITH_EMISSIVITY to #define in EXF_OPTIONS.h
+   - keep exp. seaice_obcs & 1D_ocean_ice_column(ad) unchanged (using #undef
+     EXF_LWDOWN_WITH_EMISSIVITY) and, using pkg/exf default emissivities, update
+     results of verification experiments: 1D_ocean_ice_column(fwd),
+     global_ocean.cs32x15.seaice(fwd+ad) & seaice_dynmix(ad), lab_sea (all 8)
+     and offline_exf_seaice (all 4).
+ o pkg/seaice:
+   - remove array YNEG, RIVER, TMIX (replaced by local var) from common block;
+     remove also areaNm1 & hEffNm1 when SEAICE_GROWTH_LEGACY is undef.
+   - fix bug: areaNm1 & hEffNm1 are needed in seaice_growth (SEAICE_GROWTH_LEGACY)
+     but were not filled when resp. SEAICEadvArea=F & SEAICEadvHeff=F and
+     not multi-dim advection;
+   - fix seaice_growth.F to compile with both SEAICE_GROWTH_LEGACY and
+     SHORTWAVE_HEATING #undef.
+ o optim: make Makefile more robust (?)
+ o pkg/obcs:
+   * fix treatment of obcs_read_pickup for TAF
+ o pkg/shelfice:
+   * add missing initialisation of num_shifwflx
+   * new adjoint verification exp. for isomip
+   * change Prandtl and Schmidt numbers from generic formulations
+     (in terms of model diffus. and visc.) to separate runtime
+     parameters, with Holland and Jenkins (1999) default values.
+ o pkg/exf (interpolation):
+   - extended input field (2 rows) near the N & S poles:
+     * fill in with the symetric value (when even Nb of data in longitude)
+     * add average value at the poles only for scalar quantities ; for vector
+       component interpolation, skip the averaging and keep duplicated values,
+       unless the 2 components have same location (uvInterp=T) in which case
+       interpolate the 2 components together using S/R EXF_INTERP_UV.
+   - new S/R EXF_INTERP_UV to interpolate the 2 components of a vector field
+     together: this allows to account properly for local orientation when
+     averaging at N & S pole.
+   - add CPP option EXF_USE_OLD_INTERP_POLE to recover old results regarding
+     exf-interpolation near N & S poles. Undef by default.
+   - add debug check/print on 1rst iter if debugLevel >=2.
+ o pkg/exf:
+   - change masking flags definition (no longer "parameter"), put them in
+     common block and set them in S/R EXF_INIT_FIXED.
+   - change recognized masking flag from s,u,v to c,w,s since the mask
+     to use (maskC,W,S) depends on position on model grid and has little to
+     do with scalar/vector difference; fix masking if using P-coordinates.
+   - fix masking of wind-stress (was wrong if interpolated or read on A-grid)
+     and update results of exp. global_with_exf (x2).
+   - fix setting of zenith-angle table when SHORTWAVE_HEATING is undef.
+ o pkg/obcs
+   - S/R obcs_calc_stevens: fix a small bug (gammat -> gammas where appropriate)
+ o pkg/exf (interpolation):
+   - assume periodicity in X only if input field cover full longitude range.
+   - improve search for lat. index (supposed to be faster, in ~log2(ny) steps,
+     and should vectorise).
+   - fix input lat of the 2 added row (in case we provide N.pole data).
+ checkpoint63h (2011/12/30)
+ o pkg/seaice:
+   - change declaration of UVM from _RL to _RS in order to get B-grid
+     uIce,vIce monitor when _RS = real*4
+ o pkg/offline:
+   - new S/R OFFLINE_RESET_PARMS for resetting main-model params (switch off flags);
+   - compute vertical velocity if not read from file (switch on exactConserv
+     & staggerTimeStep);
+   - fix output of ptracer monitor iter & time for default PTRACERS_monitorFreq.
+ o verification:
+   - update grid-files (in tutorial_held_suarez_cs/input and fizhi-cs-32x32x40/input)
+     with grid-angle generated using "utils/matlab/cs_grid/cubeCalcAngle.m":
+     fix Pb in yG definition @ 2 missing corners (previously yG was averaged);
+   - update output of exp. global_ocean.cs32x15.viscA4 after changing
+     grid-file angles (used here for coriolis-3d);
+   - remove (or gzip) few files not needed to run fizhi-cs experiments.
+ o pkg/exf (vector interpolation):
+   - rotate vector component toward local grid-axis orientation using grid-angles
+     (stored in GRID.h) instead of locally computed rotation angle; keep old code
+     available with exf option EXF_USE_OLD_VEC_ROTATION (undef by default).
+ o pkg/seaice:
+   - allow to compile with exf options ALLOW_ATM_TEMP or ALLOW_DOWNWARD_RADIATION
+     undefined. Fix and remove the stop when ALLOW_RUNOFF is undefined.
+ o pkg/PKG_OPTIONS.h files:
+   - move #include "CPP_OPTIONS.h" just after #include "PACKAGES_CONFIG.h",
+     outside of the #ifdef ALLOW_THISPKG / #endif block.
+   - update verification/*/code/PKG_OPTIONS.h files.
+ o add a verification experiment (offline_exf_seaice/input.seaice) that test
+   the B-grid LSOR solver
+ o pkg/seaice: sort out emissivity and long wave radiation step 2:
+   - add emissivity for snow and ice as new run-time parameters
+     (unfortunately) that default to the corresponding exf values; this
+     changes results in most verification experiments so that for backward
+     compatibility these parameter have to be set explicitly to inconsistent
+     values (ugly);
+   - the exf-flag EXF_LWDOWN_WITH_EMISSIVITY then gets rid off the hard-wired
+     emissivities of 0.97 associated with lwdownloc in S/R seaice_solve4temp
+     until we agree how to handle this bug properly
+ o pkg/exf (interpolation):
+   - add a set of run-time param ({inputfield}_interpMethod), one for
+     each interpolated input field, to select the interpolation method
+     with a value of zero switching off the interpolation;
+   - print interpolation parameters to standard output;
+   - refine exf_check.F accordingly ; stop if USE_NO_INTERP_RUNOFF is defined
+     (to use instead runoff_interpMethod=0 in data.exf).
+   - simplify default setting of interpolation parameters (for all fields,
+     set position to match model grid cell-center position, leading to trivial
+     interpolation in case of regular Lat-Lon grid) and fix for the case of
+     non-uniform delY.
+ o pkg/exf/exf_radiation.F: add the factor ocean_emissivity to lwdown, for
+    within #ifdef EXF_LWDOWN_WITH_EMISSIVITY/#endif. Unless ocean_emissivity=1.
+    this factor is absolutely required, because the lw exitance (= out-going
+    long wave radiation) is emissivity*stefanBoltzmann*T^4 + rho*lwdown,
+    where the reflectivity rho = 1-emissivity for conservation reasons. Once
+    this is established, we can remove the CPP-Flag (if we want to).
+ o pkg/seaice:
+   - qualify a stop statement in S/R seaice_check: EXF_READ_EVAP and
+     ALLOW_BULKFORMULAE are only required if SEAICE_EXTERNAL_FLUXES is defined
+   - sort out emissivity and long wave radiation, step 1: make SEAICE_emissivity
+     what the name implies (and not emissivity*BoltzmannConstant), this means
+     the old default value of 5.5e-8 now means
+.5e-8/5.670e-8 = 0.97001763668430343479 as ocean_emissivity in
+     S/R exf_readparms
+   - FV-discretization of the B-grid LSOR solver, turn on with
+     SEAICE_LSRBNEW. This new solver also involves a slight re-organization
+     of the calling routine dynsolver, so that B and C-grid code look a
+     little more alike.
+   - fix index bugs in ostres.F (B-grid code)
+ o pkg/seaice/seaice_ocean_budget.F: change (unused) code to make it look better
+ o pkg/seaice/seaice_readparms.F: make ocean_emissivity*stefanBoltzmann the
+   default for SEAICE_emissivity, if useEXF; requires setting a wrong
+   parameter value in global_ocean.cs32x15/input.seaice/data.seaice for
+   backward compatiblity
+ o model/src: use sigmaR in calc_ivdc
+ checkpoint63g (2011/12/16)
+ o pkg/seaice/seaice_growth.F:
+   fix heat conservation of the coupled ocean-seaice system.
+   when SEAICE_HEAT_CONSERV_FIX is defined, the mismatch of
+   solid vs liquid water treatment is resolved via QNET.
+ o pkg/seaice/seaice_growth.F:
+   add missing term for sublimation in saltFlux, which was breaking conservation.
+ o pkg/seaice/seaice_model.F:
+   Remove this patholigical case treatment in evolution branch,
+   where it is un-necessary and breaks conservation.
+   Update global_ocean.cs32x15/results accordingly for
+   output.seaice.txt output_adm.seaice.txt and output_adm.seaice_dynmix.txt
+ o model/src:
+   useRealFreshWaterFlux: remove masking of EmPmP (using maskInC) in
+   external_forcing_surf.F and apply mask directly in S/R SOLVE_FOR_PRESSURE.
+ o pkg/obcs:
+   modify Stevens boundary conditions so that S/R obcs_calc_stevens computes
+   tendency for temperature and salinity that is then applied to update
+   theta and salt in S/R obcs_apply_ts; this requires proper masking of all
+   horizontal advection and diffusion contributions on the boundary (in
+   S/Rs gad_advection, gad_calc_rhs, gmredi_rtransport, gmredi_calc_diff), and
+   removing the mask in S/R timestep_tracer. The advantage of this approach is
+   that the updated tracer values include contributions from the surface
+   forcing and vertical diffusion.
+   This modification changes exp4.stevens because of vertical diffusion.
+ o model/src:
+   change non-curvilinear grid setting to also work with exch2 + blank tiles:
+   - move delX,delY to new header file (SET_GRID.h) and adjust length to 1rst
+     face dimensions.
+   - new S/R INI_LOCAL_GRID to define model tile-local grid (using same units
+     as delX,delY) and called from S/R INI_CARTESIAN_GRID, INI_CYLINDER_GRID
+     and INI_SPHERICAL_POLAR_GRID.
+   change test-experiment global_ocean.90x40x15:
+   - split the domain in many more tiles (9x4 tiles of 10x10 each);
+   - switch to exch2 and remove 1 blank-tile (but only for MPI test).
+   - update results of both test (std & dwnslp).
+ o pkg/diagnostics/diagstat_calc.F: fix problem with non-initialised halos in
+    tmp-fields for TARGET_NEC_SX at the cost of extra loops and less vectorisation
+ o model/src:
+   - take sigma-coords related code out of update_etah.F into new S/R UPDATE_ETAWS
+     and fix initialisation when using sigma-coords.
+ o model/src:
+  - improve Gael's changes from Jan 14, 2011 "addition of an 'update' of hfacc etc.
+    to their current value at the beginning of forward_step.F": add a new run-time
+    switch "doResetHFactors" (default=false) and use it to do (or not) the resetting.
+  - set doResetHFactors=T in the following test exp:
+     global_ocean.90x40x15/input (rStar) & input.dwnslp (linFS),
+     global_ocean.cs32x15/input.seaice (rStar), hs94.cs-32x32x5/input.impIGW (linFS)
+     and tutorial_global_oce_in_p (nonLinFS in p)
+ o pkg/ecco
+   - added IES cost term
+ o verification/advect_xz
+   - scale down the depth by 1/100 (from 200 km to just 2.km)
+   - add secondary test using divergent flow, r-star and implicit vertical advection.
+   - add note (check_conserve_TS.txt) and matlab script to check global
+     conservation of T & S with NonLin-FreeSurf and Adams-Bashforth.
+ o pkg/rbcs:
+   - add missing relaxUFile & relaxVFile in namelist (in rbcs_readparms.F).
+ o pkg/obcs (Stevens BCs):
+   - fix various problems in the phase speed computations of the Stevens
+     boundary conditions:
+     - sign error for western and southern boundaries,
+     - time tendencies were underestimated by factor of dTracerLev and some just
+       zero (for multidim advection schemes) -> now estimate time tendencies
+       from difference between previous (n) time level and the one before (n-1);
+       this requires storing a field per tracer per boundary, also in the pickup
+       files.
+     - replace deltaTmom in cflMer/Zon computation with dTracerLev
+     - set default phase velocity to cflMer/Zon
+   - update verification experiment exp4.stevens
+ o pkg/generic_advdiff (implicit_r):
+   - fix implicit vertical advection conservation with AB and/or NonLin-FreeSurf
+     (remove Tr*d/dz(w) in gad_implicit_r.F and add it in gad_calc_rhs.F);
+   - fix implicit vertical diffusion or advection with NonLin FreeSurf
+     (was using current recip_hFacC instead of future recip_hFacC): set future
+     recip_hFac in thermodynamics.F and pass & use it in all vertical implicit
+     solver S/R ;
+   - fill diagnostics of vertical advective fluxes (computed from tendency)
+     when using implicit vert. advection;
+   - update output of exp that uses NonLin-FreeSurf & implicit vertical diffusion:
+      global_ocean.90x40x15 (fwd+AD), global_ocean.cs32x15 (4 fwd +4 AD),
+      tutorial_global_oce_in_p (fwd), tutorial_tracer_adjsens (AD)
+      and cpl_aim+ocn (used in ocean); to be fixed: bottom_ctrl_5x5.
+ o pkg/diagnostics:
+   - add diags for T & S tendency which goes through Adams-Bashforth.
+ checkpoint63f (2011/11/30)
+ o pkg/obcs/obcs_calc_stevens.F:
+   - join some loops to reduce number of if-statements
+   - fix a bug in code for northern and southern boundaries.
+ o pkg/seaice/seaice_init_varia.F
+   - undo resetting UVM=0 that has slipped in unnoticed in r1.49 and
+     caused zero ice-velocities for the B-grid code
+ o pkg/seaice/seaice_init_fixed.F
+   - fix a bug in the initialisation of the second metric terms
+     k2AtC and k2AtZ: k2AtC was set to the value of k2AtZ and k2AtZ was 0.
+   - this affects all seaice configurations with useCurviLinearGrid = .true.
+     and SEAICEuseDynamics = .true. (default) and SEAICEuseMetricTerms
+     = .true. (default)
+   - the only verification experiment affected is global_ocean.cs32x15:
+     icedyn, seaice, and seaice_dynmix (AD)
+ o pkg/thsice/thsice_diagnostics_state.F
+   - only for TARGET_NEC_SX: initialise all of tmpFld, so that overlap does
+     not contain garbage that can lead to surprises in diagstats_calc.F
+ o model/src:
+   - move bi,bj loops inside INTEGR_CONTINUITY ;
+   - call INTEGR_CONTINUITY directly from forward_step.F
+     (previously called from momentum_correction_step.F)
+   - call UPDATE_ETAH from integr_continuity.F and update initialise_varia.F,
+     forward_step.F and pkg/ecco/the_main_loop.F
+ o model/src/do_oceanic_phys.F:
+   - reset addMass to zero only if useICEFRONT ;
+   - separated k-loop for rhoInSitu calculation ;
+   - add 2 diagnostics for wVel*delta^k(Rho) at constant pressure (= WdRHO_P)
+     and at constant T,S (= WdRHOdP);
+   NOTE: change diagnostics WRHOMASS : this is now simply the Vertical transport
+     of In-Situ Density (Anomaly) (=wVel*rho^bar-k) (previously was Vertical
+     transport of locally-referenced potential density);
+     to recover old diag: (WRHOMASS)_old = (WRHOMASS)_new -1/2 * (WdRHOdP)
+ o gluemnc:
+   - replace ncrcat with ncks when adding variables to *.glob.nc
+   - change the way variables without record dimensions are handled so that
+     files that have variables with and without record variables can be glued:
+     add a record dimension with a new unique name to be able to differentiate
+     between variable with and without a-prior record dimensions. Remove
+     this record dimension before adding the variable to the *.glob.nc file.
+     This procedure automatically works also for grid.*.nc files that do not
+     have any record dimension.
+ o verification/cpl_aim+ocn:
+   - generate new reference output on baudelaire (gfortran, mpich2) (previously
+     from old aces-grid cluster, ao + ifort) and update 'run_cpl_test' script.
+ o pkg/obcs/obcs_calc_stevens.F:
+   - add code for northern and southern boundaries.
+   - reorder some loops to save some operations and so that drFBar can be scalar
+ o model/src:
+   - re-arrange extrapolation calculation in Adams-Bashforth S/R (affects machine
+     truncation) and add diagnostics for Adams-Bashforth tendency increment.
+   - update test-exp. AD output:
+D_ocean_ice_column, bottom_ctrl_5x5, lab_sea.noseaicedyn, obcs_ctrl;
+   and FWD output: exp4 (x3), fizhi-cs-32x32x40, fizhi-cs-aqualev20,
+     front_relax (x3), global_ocean.90x40x15 (x2), global_ocean.cs32x15 (x5),
+     global_ocean_ebm, ideal_2D_oce, internal_wave, isomip (x3), lab_sea (x4),
+     matrix_example, MLAdjust (x5), natl_box (x2), tutorial_advection_in_gyre,
+     tutorial_global_oce_in_p, tutorial_global_oce_latlon and
+     tutorial_plume_on_slope.
+ o pkg/diagnostics/diagstats_calc.F: add special code for TARGET_NEC_SX
+   that vectorises. This code uses 3 extra 2D fields, F90 array assignments
+   and F90 intrinsic routines SUM, MINVAL, MAXVAL, but since TARGET_NEC_SX
+   does not even come with an F77 compiler anymore, this should not be
+   a limitation
+ checkpoint63e (2011/11/06)
+ o pkg/seaice adjoint of LSR:
+   - substantially rework STORE dirs.
+   - add new flag SEAICE_LSR_ADJOINT_ITER and parameter SOLV_MAX_FIXED
+     this flag will do SOLV_MAX_ITERS but only store SOLV_MAX_FIXED fields
+     and only perform SOLV_MAX_FIXED adjoint iters.
+     If flag SEAICE_LSR_ADJOINT_ITER is undef-ined, the store in the iteration
+     always goes to same record, meaning that the last forward field will be
+     the one used in the adjoint - an approximation.
+   - now produces non-zero LSR-related gradients, and acceptable gradients for
+     An's zonal (x-z) test case (to be checked in)
+   - lab_sea: produces acceptable gradients for lab_sea with forward/backward LSR
+     (but these are screwed up by the thermodynamics in the first place);
+     reduce from 4 to 3 timesteps to avoid t.d. issue as temporary solution;
+     output_adm.txt also reflects changes to seaice_lsr
+   - global_ocean.cs32x15:
+     * the "default" input_ad.seaice has been re-defined to always exclude dynamics
+       (output_adm.seaice.txt was updated)
+     * new input_ad.seaice_dynmix tests a forward LSR / backward free-drift
+       (output_adm.seaice_dynmix.txt)
+ o test-exp. exp4+stevens:
+   - switch off pTracers and nonHydrostatic (restart problems with NH); update output
+ o pkg/seaice (lsr solver):
+   - put coriolis term into FORCEX,Y (as in seaice_evp.F; previously in rhsU,V only);
+     update output (change truncation) of FWD test exp lab_sea & global_ocean.cs32x15
+   - fix SEAICEuseFREEDRIFT for new OBCS version (ndef OBCS_UVICE_OLD)
+   - new parameter "LSR_mixIniGuess" allows (if=2,4) to mix free-drift solution
+     into initial guess for linearised solver; use weight ~ (1/err)^2,4.
+   - compute (with new S/R) and print initial (+ free-driff sol. if available)
+     and final residual of linearised system.
+ o pkg/pp81: add diagnostics for vertical eddy viscosity and diffusivity
+   (requires new subroutines pp81_init_fixed.F/pp81_diagnostics_init.F)
+ o pkg/obcs: Stevens BCs (still only for east and west)
+   - fix Stevens BCs by using intermediate velocities of previous
+     time step as "baroclinic" contribution rather velocity of one grid cell
+     inside the boundaries; more consistent but makes the entire treatment
+     more complicated, requires new pickup files and two new S/Rs:
+     obcs_copy_uv_n and obcs_save_uv_n
+     keep old version with flag OBCS_STEVENS_USE_INTERIOR_VELOCITY
+     (undefined by default)
+   - silently move computation of some grid factors into separate loops to
+     save some recomputations
+   - add a hack for non-linear free surface
+   - add a verification experiment to exp4 that tests this code
+ o make gluemnc/xplodemnc work with (moron) non-GNU shell tools
+ checkpoint63d (2011/10/25)
+ o pkg/obcs + seaice-dynamics:
+   - new implementation calls OBCS_APPLY_UVICE (in seaice_dynsolver.F) before
+     solving for uIce,vIce and prevent SEAICE_LSR or SEAICE_EVP to modify OB
+     values (consistent with an implicit method).
+   - add option OBCS_UVICE_OLD (in OBCS_OPTIONS.h, defined by default) to
+     return to older implementation (same results as before); note that none of
+     the "experimental" OBCS_SEAICE options are implemented in the new version.
+   - switch test exp. seaice_obcs to use new version (update output).
+ o pkg/seaice:
+   - add few stop (B-grid & CS-exch, B-grid & obcs, AUTODIFF & SEAICE_maskRHS)
+   - add parenthesis in Water-Ice drag calculations (evp & lsr);
+   - in seaice_lsr.F, simplify and fix AA3 setting in case sNx or sNy=1;
+     and fix case SEAICE_waterDrag=0 (but unstable if no or little seaice)
+   - update output of test exp lab_sea that uses evp or lsr (fwd+ad)
+     and global_ocean.cs32x15 (icedyn & seaice fwd+ad).
+ o pkg/layers:
+   - Include potential density as new coordinate (Thanks to David Munday)
+ o pkg/dic:
+   - Remove subroutine CALC_PCO2_APPROX_CO3 from carbon_chem.F and add
+     carbonate computation/output to CALC_PCO2_APPROX
+ o pkg/generic_advdiff
+   - make TAF generate code that vectorizes:
+     gad_dst3fl_adv_*.F, gad_u3c4_impl_r.F
+ o model/src:
+   - improve vectorization also for adjoint code of
+     solve_tridiagonal and solve_pentadiagonal. As this requires up to
+new 3D-fields, this code is only turned on for the combination of
+     ALLOW_AUTODIFF_TAMC and TARGET_NEC_SX. Also remove error message
+     for singular matrix check (only for TARGET_NEC_SX).
+   - remove hard-wiring of CPP-flag ALLOW_SOLVERS_KLOOPINSIDE to
+     ALLOW_AUTODIFF. This flag now needs to be set outside of
+     solve_tri/pentadiagonal.F
+ o pkg/exf:
+   - fix previous modif in exf_readparms.F (revision 1.66: treatment of
+     optional namelist EXF_NML_OBCS).
+   - add many debug-print (to trace down where pkg/cal error msg comes from)
+ o pkg/mom_vecinv/mom_vi_hdissip.F
+   - rearrange loops so that vectorization is improved and in particular
+     the adjoint routine vectorizes. The code is a little longer now, but I
+     think easier to read.
+     Note: CPP-flag MOM_VI_ORIGINAL_VISCA4 is not tested.
+ o pkg/seaice:
+   - allow to specify initial seaice velocity from files (uIceFile,vIceFile)
+ o pkg/exch2:
+   - alternative code (without tile x tile matrix) for cumulated (global) sum;
+     used as default; keep using matrix version in adjustment.cs-32x32x1.
+ o tools/build_options:
+   - after merging some MPI and/or multi-threads specific optfiles so that
+     the modified default optfile support both built (using genmake2 -mpi
+     for the MPI built, using genmake2 -omp for OpenMP built),
+     move now obsolete optfiles to tools/build_options/unsupported/
+    (add copy of email sent to support list in doc/optfiles_changes.txt)
+ checkpoint63c (2011/09/28)
+ o pkg/ebm:
+   - fix ebm_zonalmean.F for several procs in Ydir (nPy > 1); update output of
+     test exp. global_ocean_ebm (affects machine truncation, 12 digits for cg2d)
+ o pkg/seaice (SItracer pickup)
+ * Simple code for pickup of SItracer fields (#define ALLOW_SITRACER)
+   To be done:
+   - thorough testing of missing fields in S/R seaice_read_pickup
+   - thorough matching between generic tracers and specific fields used
+     (e.g. volume-weighted age, area-weighted age, salinity, ...)
+ o pkg/grdchk:
+   - all procs calculate and print finite-difference gradient (for testreport)
+   - change description output used by testreport;
+   - AD testing: also compare and report Finite-difference gradient;
+     => update all output_adm*.txt
+ o test exp. global_ocean_ebm:
+   - switch back useRealFreshWaterFlux (no problem anymore with run-off);
+     and turn on exactConserve; update (fwd) output.
+ o pkg/seaice (LSR):
+   - print to STDOUT number of solver iterations at SEAICE_monFreq interval.
+   - modify solver to deal cleanly with u/v switch at face edges (CS-grid):
+     * identical results with non-CS topology (but save up to half of EXCH calls)
+     * no much changes in cs-32 verification tests since no ice at face edges
+      (differences due to convergence criteria, since nb of iterations changed)
+     * expect some differences on lat-lon-cap grid.
+     update FWD results of global_ocean.cs32x15 test exp. (icedyn & seaice)
+ o pkg/exch2:
+   - add new CPP option to disable Cumul-Sum matrix (to save storage when using
+     many tiles); until alternative coding is added, just add a stop.
+ o pkg/kpp:
+   - move ptracer/salt correction to KPP diffusivity from calc_3d_diffusivity.F
+     to S/R KPP_CALC_DIFF_PTR (+ add tracer-index to argument list).
+ o optfiles:
+   - merged few MPI specific (*+mpi*) and multi-threads specific (*+mth) optfiles
+     into corresponding standard one, which now support different build:
+     (when using "genmake2 -mpi" for MPI built, "genmake2 -omp" for OpenMP built)
+   - add "darwin_amd64_gfortran" optfile (initially from Constantinos, tested by Gael)
+ o pkg/obcs (seaice):
+   - add missing EXCH for uIce,vIce after applying OBCS;
+   - remove S/R EXF_SWAPFFIELDS_X/Y , disable S/R EXF_SET_OBCS_X/Y by adding
+     argument = number of levels to S/R EXF_SWAPFFIELDS_XZ/YZ and S/R
+     EXF_SET_OBCS_XZ/YZ ; update obcs_exf_load.F;
+   - fix EXF_SET_OBCS_XZ/YZ for multi-threaded (assuming loaded arrays are shared)
+   - verification exp. seaice_obcs: split domain in 2 tiles ; add SIZE.h_mpi
+     and eedata.mth for MPI and mutlti-threaded testing.
+ o pkg/flt:
+   - make FLT EXCH working with multi-threads (and MPI + multi-threads);
+   - use buffer for floats IO (trajectories & profiles); size set in FLT_SIZE.h;
+   - fix pickup and trajectories + profiles output with multi-threads;
+   - add parameter to select group of fields to write along trajectory and
+     profile (default unchanged);
+   - only call PORT_RAND* function when flt_noise <> 0;
+   - add multi-threaded test for verification exp. flt_example.
+ o eesupp/src:
+   - fix S/R ALL_PROC_DIE for MPI+multi-threads.
+   - USE_OMP_THREADING: check that number of threads (from eedata) matches
+     actual number (from env.var. setting).
+ o pkg/ebm:
+   - fix multi-threading in ebm_area_t.F ; clean-up EBM.h header file.
+   - add multi-threaded test for verification exp. global_ocean_ebm
+ checkpoint63b (2011/08/29)
+ o pkg/ebm:
+   - fix sign of runoff;
+   - update directly forcing fields (Qnet,EmPmR) to use standard
+     external_forcing_surf.F (remove ebm_forcing_surf.F);
+   - avoid loading forcing fields a 2nd time (1rst time from ini_forcing.F);
+   - set either tauThetaClimRelax (in data) or tauThetaZonRelax (in data.ebm)
+     and apply relaxation towards corresponding SST field (2-D map or ZonAver).
+   - update global_ocean_ebm parameters file data and update output (fwd+ad).
+ o verification/matrix_example:
+   - split domain in 8 tiles (2x4); add SIZE.h_mpi and eedata.mth for
+     MPI and mutlti-threaded testing.
+ o pkg/matrix:
+   - change name of S/R MATRIX_INIT to S/R MATRIX_INIT_VARIA ;
+   - fix for multi-tiles (add bi,bj indices to exp/impMatrixCounter);
+   - fix for multi-threaded (remove DATA statement);
+   - use standard DIFFERENT_MULTIPLE function;
+   - add MATRIX_OPTIONS.h
+ o verification/matrix_example:
+    To allow to split the domain in several tiles, change the domain size
+     from 31x31 to 32x32 by adding 1 row (on Southern side) + 1 column (on
+     Western side) of land points.
+ o (*_)write_pickup:
+   - fix few bugs in mypackage_read/write_pickup.F
+   - in MDS_WR_METAFILES calls, replace "myTime" by a local copy into
+     array of length 1 (was caught by compiler syntax-check).
+ o pkg/diagnostics (velocity potential solver):
+   - fix initialisation of main diagonal matrix (for overlap values)
+ o tools/build_options:
+   - towards single optfile for MPI and non-MPI built: merge few +mpi optfile
+     into previously non MPI one (linux_am64 pgf77,open64,pathf90,ifort11)
+   - linux_amd64_gfortran: merge +mpi_generic version into standard optfile:
+     * move IEEE flags to DEVEL section; take IEEE flags from +mpi_generic ;
+     * FAST case: for now, keep -O3 with ini_masks_etc.F in NOOPTFILES list;
+     and use "-devel" testreport option for standard testing on baudelaire.
+ o pkg/grdchk:
+   - Assume that all gradient-check position are within the same tile
+     corresponding to global indices (iGloTile,jGloTile) and that
+     positions (i,j indices or nbeg) are local to this tile;
+   - set tile local-indices and proc. number according to tile global-indices;
+   - remove grdchkwhichproc from namelist (retired param);
+   - send adjoint gradient to all procs so that all procs print it to STDOUT;
+   - switch on MPI test for global_ocean.90x40x15 AD test experiment.
+ o adding package bbl (Bottom Boundary Layer)
+   description in MITgcm/pkg/bbl/bbl_description.tex
+   example/test experiment in MITgcm_contrib/bbl
+ o adjoint verifications
+   - changed 5 experiments to multi-tiles
+     * global_ocean.90x40x15
+     * obcs_ctrl
+     * offline_exf_seaice
+     * tutorial_dic_adjoffline
+     * tutorial_tracer_adjsens
+     changed reference output for those that crossed tile boundaries
+     (changed choices of gradient check positions)
+   - with same tiling, switch on MPI test for the last 4 (add SIZE.h_mpi)
+   - temporary bug fix in grdchk_loc.F for obcs
+ o genmake2 & testreport
+   - add target "obj" (just for OBJFILES files) to Makefile;
+   - add option to specify FEXTRAFLAGS on command line.
+ o pkg/mypackage:
+   - remove bi,bj arguments from argument list of S/R MYPACKAGE_CALC_RHS
+ checkpoint63a (2011/08/02)
+ o pkg/diagnostics (velocity potential):
+   - fix velocity-potential calculation when using OBCS: balance net inflow
+     at each level (new S/R OBCS_DIAG_BALANCE).
+   - fix some of the problems with OBCS in stream-function calculation.
+ o pkg/exf/exf_readparms.F
+   - make treatment of optional namelists (EXF_NML_04, EXF_NML_SGRUNOFF,
+     EXF_NML_OBCS) more consistent, so that you need not have these lists
+     in data.exf when they are not used.
+ o optim/optim_readdata.F:
+   - add some output about obcs variables that might help debugging
+   - add a check about nvarlength
+ o pkg/ctrl: once again, fix computation of endrec in s/r ctrl_init_rec
+ o pkg/diagnostics (velocity potential):
+   - fix mask for OBCS (still problems in stream-function with OBCS);
+   - add specific parameter (default = main code CG2D params) for solver;
+   - in case of poor convergence, use solution corresponding to lowest residual.
+ o pkg/seaice/seaice_growth.F: small changes that permit vectorizable
+   adjoint code. In general statements like tmpscal=fld(i,j) are problematic
+   and should be avoided where possible.
+ o model/src (rStar):
+   - include r* effect on reference-state (PhiRef) in totPhiHyd
+     (so that it's really the potential anomaly at the cell center);
+     affect solution only if z* with full pressure in EOS.
+   - test-exp global_ocean.90x40x15: reduce viscA4 (was not stable) and
+     update output (because of z* in totPhiHyd and viscA4 changes);
+   - keep the previous diagnostics as "PHIHYDcR" (closer to potential anomaly
+     @ constant r) until interpolation at constant r is implemented.
+ o pkg/diagnostics
+   - fix RCENTER diagnostics in case cell center is not at interface middle
+ o add code for TEOS-10 (www.teos-10.org, McDougall et al. 2011). Use
+   this eos with eosType = 'TEOS10', in data (PARM01). This eos implies
+   that THETA and SALT are "conservative temperature" and "absolute
+   salinity"
+ o pkg/ctrl
+   - fix computation of endrec in s/r ctrl_init_rec
+   - ctrl_pack/unpack_*: remove unused variables and some obsolete code
+ o pkg/seaice/seaice_growth.F
+   - fix recomputation with FENTY_AREA_EXPANSION_CONTRACTION code,
+     remove then obsolete STORE directives
+   - fix STORE directives for tice and tices, so that only the
+     appropriate part of the fields are stored (and not the entire
+     field); update global_ocean.cs32x15.seaice AD output.
+   - cosmetic changes: adjust CPP flags and indentation for better
+     legibilty
+ o pkg/autodiff: add mnc-output of adBottomDragFld (code contributed by
+   Chris Wilson)
+ o pkg/gmredi:
+    Add Sub-Meso Eddies parameterisation from Baylor (for now, disabled with TAF),
+    and test it in verification exp. "front_relax.bvp" (update output.bvp.txt)
+ checkpoint63  (2011/07/11)
+ o pkg/exch2:
+  - rename+move: nTiles in W2_EXCH2_PARAMS.h --> exch2_nTiles in W2_EXCH2_TOPOLOGY.h
+  - exch2 version of Cumul-Sum routine: S/R W2_SET_MAP_CUMSUM set up the
+    (2*nTiles)x(nTiles) Cumul-Sum matrix and Master Proc collects all tiles X,Y
+    increments, calculates Cumul-Sum @ tile origin and sends results to each tile.
+ o pkg/diagnostics:
+  - deal with pairs of post-processed diags which are computed together; in
+    particular, re-order list of diags within 1 file to save 2nd computation.
+  - add S/R to compute velocity stream-function (called after velocity-potential
+    calculation) from divergence free transport -> new diagnostics "PsiVEL".
+  - add parameter to select grid-point location where PsiVEL == 0.
+ o eesupp/src:
+    new S/R to calculate cumulated sum on tiled array, corner grid-cell location
+    (specific S/R needed with exch2 topology - for now, just print an error msg)
+ o pkg/shelfice
+  - Implement friction velocity-dependent transfer coefficients following
+    Holland and Jenkins, JPO, 1999
+    Original code by M. Losch with small modifs.
+    Enable with CPP option SHI_ALLOW_GAMMAFRICT
+    and runtime flag SHELFICEuseGammaFrict
+ o verification lab_sea & global_ocean.cs32x15:
+  - use tools/xmakedepend (instead of cyrus_makedepend, in genmake_local):
+    does not miss some dependencies (but is slower) and also works with many files
+    (default makedepend limitation for AD built).
+  - use cyrus_makedepend in adjustment.128x64x1 (so that it still tested).
+ o pkg/diagnostics:
+  - add run-time parameter "useMissingValue" (def=False) to fill land-point
+     (i.e., where mask=0) with MissingValue ; used only in MNC output file.
+    This replaces CPP-option DIAGNOSTICS_MISSING_VALUE.
+  - 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.
+ o pkg/seaice/seaice_growth.F:
+   - add 'a priori' cap of latent heat flux based on actual mass considerations
+     when SEAICE_ADD_SUBLIMATION_TO_FWBUDGET is used in conjunction with the
+     non-legacy seaice_solve4temp.
+   - add diagnostic of the residual sublimation freshwater flux (the freshwater
+     flux taken from the ocean if the total actual latent heat fluxes exceed
+     the maximum)
+   - remove the CPP switch around the *_FWbySublim unit change and move diagnostic
+    fill for SIatmQnt, SIfwSubl, SIatmFW to the end of seaice_growth.F.
+   - add diagnostic of the actual sublimation freshwater flux (that is 0. ifndef
+    SEAICE_ADD_SUBLIMATION_TO_FWBUDGET) and of the latent heat flux (evap+sublim).
+   - bug fix : frWtrAtm needs a_FWbySublim ifdef SEAICE_ADD_SUBLIMATION_TO_FWBUDGET.
+   - bug fix : missing part to Martin's r_FWbySublim pathological case fix.
+ o verification 1D_ocean_ice_column:
+    made model domain truly 1x1x23 instead of 5x5x23 with only center cell (3,3,:) wet
+    update results
  o pkg/diagnostics:
   - Implement setting of "Post-Processed" diagnostics (corresponding to gdiag(5)='P')
     which are not filled-up but computed from other diags ; In this case, the mate
     diag points to the primary (filled-up) diag to used for post processing.
   - Update Velocity-potential computation using post-processed diag framework.
-  Change the ordering of fields and time-periods in MDS output file
+  ==> Change the ordering of fields and time-periods in MDS output file
     when using periodic averaging: now writes one time record with the full set
     of fields for this time period, then the next time reccord (it used to be
     one field with the full set of time records then the next field).
-    Note: this is now similar to NetCDF (MNC) output file.
+      Note: this is now similar to NetCDF (MNC) output file. <==
  checkpoint62z (2011/06/21)
  o adjoint monitor:

 Legend:



Removed from v.1.1462
 


changed lines


 
Added in v.1.1728
 Legend:



Removed from v.1.1462
 


changed lines


 
Added in v.1.1728
-Removed from v.1.1462
+Added in v.1.1728

	ViewVC Help
Powered by ViewVC 1.1.22