| 4 |
Notes on tags used in MITgcmUV |
Notes on tags used in MITgcmUV |
| 5 |
============================== |
============================== |
| 6 |
|
|
| 7 |
o pkg/seaice: revise EVP code (seaice_evp.F) in preparation for a more |
checkpoint65q (2015/11/18) |
| 8 |
|
o pkg/seaice/seaice_evp: |
| 9 |
|
- fix residual computations so that they no longer affect the actual |
| 10 |
|
simulation code |
| 11 |
|
- fix seaice_readparams to be able to actually set new parameters |
| 12 |
|
SEAICEaEVPcStar and SEAICEaEVPalphaMin, also fix logic a little |
| 13 |
|
o model/src: |
| 14 |
|
- fix wrong truncation (single prec) from "FLOAT" function in expressions |
| 15 |
|
like: deltaT*FLOAT(myIter) when myIter is large: |
| 16 |
|
a) replaced by DFLOAT (if not seen by TAF); |
| 17 |
|
b) remove "FLOAT" and use implicit type conversion (if seen by TAF). |
| 18 |
|
o pkg/aim_v23: |
| 19 |
|
- write/read the expected CO2 pickups (from S/R AIM_DO_CO2): fix reading |
| 20 |
|
when pickupSuff is set ; do write rolling pickups + at the end of the run. |
| 21 |
|
o pkg/atm_ocn_coupler + ocn_compon_interf: |
| 22 |
|
- in prep for using pkg/seaice Dyn in coupled set-up, implement sequential |
| 23 |
|
coupling option-1: ATM -(cpl)-> OCN and then OCN -(cpl)-> ATM ; |
| 24 |
|
add corresponding call in DO_OCEANIC_PHYS. |
| 25 |
|
o coupler pkgs: |
| 26 |
|
Start a new way to set-up optionally-exchanged variables (previously, only |
| 27 |
|
there for exchange_DIC): switch on/off from 1 place, at run-time, in coupler |
| 28 |
|
data.cpl (new) instead of at compile time in ATMIDS.h & OCNIDS.h ; |
| 29 |
|
send switch to ATM & OCN compon before reading their own "data.cpl" (so that |
| 30 |
|
compononent param default can be set accordingly). |
| 31 |
|
Note: this new switch (cpl_exchange_[xxx]) is not a per variable switch |
| 32 |
|
but rather applies to a group a fields (going both ways like for DIC) |
| 33 |
|
and corresponds to an option of coupling config. |
| 34 |
|
coupler (pkg/atm_ocn_coupler): |
| 35 |
|
- change parameter file from "data" to "data.cpl" (in cpl_read_params.F), |
| 36 |
|
and read namelist directly (without eessup S/R calls) instead of |
| 37 |
|
in the very un-flexible old way, 1 param per line, in fixed order; |
| 38 |
|
- add error-counter to track errors (e.g., missing "data.cpl") and to stop |
| 39 |
|
cleanly in CPL_CHECK_CPLCONFIG |
| 40 |
|
- add new coupling-exchange selectors (some not yet active) including |
| 41 |
|
cpl_exchange_DIC (to set to 3 to replace atmCpl_exchange_DIC=T in ATMIDS.h |
| 42 |
|
and ocnCpl_exchange_DIC=T in OCNIDS.h); |
| 43 |
|
- add 2 new S/R (CPL_SEND_ATM/OCN_CPLPARMS) to send coupler-params to each |
| 44 |
|
component ; called directly from coupler.F, after comp. registrations; |
| 45 |
|
- rename few variable-ID names (in ATMIDS.h & OCNIDS.h): |
| 46 |
|
atm/ocnSeaIceName --> atm/ocnSIceMassName ; atm/ocnFCO2Name -> |
| 47 |
|
-> atm/ocnCO2FluxName ; atm/ocnWSpdName --> atm/ocnWindSpdName |
| 48 |
|
components (pkg/atm_ & ocn_compon_interf): |
| 49 |
|
- add ATM/OCN_CPL_OPTIONS.h in component pkgs and include the corresponding |
| 50 |
|
CPP header file in all pkg source files. |
| 51 |
|
- move checking of comp. configs out of cpl_exch_configs.F into new S/R |
| 52 |
|
ATM/OCN_CHECK_CPLCONFIG; |
| 53 |
|
- add error-counter to track errors (e.g., in "data.cpl") and to stop |
| 54 |
|
cleanly in ATM/OCN_CHECK_CPLCONFIG; |
| 55 |
|
- add new S/R CPL_IMPORT_CPLPARMS to import coupling-exchange selectors |
| 56 |
|
from coupler ; called from initialise_fixed.F, after PACKAGES_BOOT and |
| 57 |
|
before PACKAGES_READPARMS; |
| 58 |
|
verification/cpl_aim+ocn/: |
| 59 |
|
- update ATMIDS.h & OCNIDS.h (in shared_code) and use new parameter file |
| 60 |
|
"data.cpl" for coupler (in input_cpl). |
| 61 |
|
o eesupp/inc: |
| 62 |
|
- change the MPI-global-sum default (in CPP_EEOPTIONS.h) to: |
| 63 |
|
#define GLOBAL_SUM_ORDER_TILES (only affects MPI built of global_sum_tile.F) |
| 64 |
|
to always cumulate tile local-sum in the same order (still uses |
| 65 |
|
MPI_Allreduce like standard code). |
| 66 |
|
- in verification experiment code, remove local CPP_EEOPTIONS.h_mpi files |
| 67 |
|
that are now identical to standard version (in eesupp/inc). |
| 68 |
|
o pkg/shelfice: |
| 69 |
|
- add SHELFICETransCoeffTFile for initializing spatially varying |
| 70 |
|
heat and salt transfer coefficients |
| 71 |
|
o pkg/atm_compon_interf & ocn_compon_interf: |
| 72 |
|
- combine into 1 (new) S/R CPL_EXPORT_IMPORT_DATA the two previous |
| 73 |
|
S/R CPL_EXPORT_MY_DATA & CPL_IMPORT_EXTERNAL_DATA |
| 74 |
|
- move bi,bj loops inside S/R OCN_STORE_MY_DATA |
| 75 |
|
o pkg/atm_ocn_coupler: |
| 76 |
|
- initialise to zero ATMVARS.h & OCNVARS.h arrays. |
| 77 |
|
o pkg/rbcs: |
| 78 |
|
- select RBCS-mask for Temp & Salt the same way as for pTracer ; this allows |
| 79 |
|
to compile with maskLEN=1 when all tracers use the same RBCS-mask. |
| 80 |
|
- add a short report (to STDOUT) about RBCS-mask settings. |
| 81 |
|
|
| 82 |
|
checkpoint65p (2015/10/23) |
| 83 |
|
o pkg/seaice/seaice_evp.F: fix store directives and key computations, since |
| 84 |
|
there is no adjoint test for EVP, we don't know what it does to the adjoint |
| 85 |
|
solution, but at least there are no more recomputation warnings left |
| 86 |
|
o pkg/seaice/seaice_evp.F: slight modification of averaging so that EVP stands |
| 87 |
|
a chance of giving the same results as a fully converged VP (JFNK) solver |
| 88 |
|
o tools/adjoint_options/adjoint_diva: |
| 89 |
|
- fix bug (typo) in Tang-Linear FLAGS setting for MPI + DIVA |
| 90 |
|
o pkg/streamice: |
| 91 |
|
- spotted bug in streamice_driving_stress using dyG where should have been dxG |
| 92 |
|
o pkg/exf: |
| 93 |
|
- remove STOP statement that prevented to use wind or (wind-)stress |
| 94 |
|
interpolation for constant-in-time forcing fields (wind/stress period = 0). |
| 95 |
|
o pkg/seaice: |
| 96 |
|
- replace SEAICE_EVP_USE_ADAPTATION by run time parameters, requires |
| 97 |
|
5 additionals 2D fields and 6 new store directives |
| 98 |
|
- change logic in seaice_readparms.F: EVP code can now be turned on in |
| 99 |
|
various ways, setting SEAICE_deltaTev is no longer required, |
| 100 |
|
alternatively set SEAICE_evpAlpha/Beta, SEAICEuseEVPstar, SEAICEuseEVPrev, |
| 101 |
|
or SEAICEaEVPcoeff |
| 102 |
|
o pkg/shelfice/shelfice_thermodynamics.F |
| 103 |
|
- redefining uLoc/vLoc to u/v, instead of 2u/v |
| 104 |
|
o pkg/seaice/seaice_evp.F: introduce "adaptive EVP" by Madlen Kimmritz |
| 105 |
|
- this is an intermediate state of a truely converging EVP that has |
| 106 |
|
the potential of outperforming implicit VP methods, I do not recommend |
| 107 |
|
using it yet. |
| 108 |
|
- for now, "adaptive EVP" is turned on by setting SEAICE_EVP_USE_ADAPTATION |
| 109 |
|
and all parameters are hardwired |
| 110 |
|
- for convergence it still requires very smooth regularisations, that means, |
| 111 |
|
if SEAICE_DELTA_SMOOTHREG is defined, deltaCreg=deltaC+deltaMin, which is |
| 112 |
|
different from the VP case with deltaCreg=sqrt(deltaC**2+deltaMin**2); also |
| 113 |
|
specific averaging appears to be required, still subject to tuning |
| 114 |
|
- still needs cleaning and I need to turn parameter into runtime parameters |
| 115 |
|
- add code to compute and print residuals of iteration, compile with |
| 116 |
|
defining ALLOW_SEAICE_EVP_RESIDUAL to enable it |
| 117 |
|
o pkg/seaice/seaice_evp.F: fix a bug (factor of four in front of e12Csq) |
| 118 |
|
that entered with r1.54 on Sep04, 2015, -> update results |
| 119 |
|
|
| 120 |
|
checkpoint65o (2015/09/14) |
| 121 |
|
o pkg/mom_vecinv & mom_common: |
| 122 |
|
- add deep-atmosphere geometry factor in mom_calc_visc.F + in routines |
| 123 |
|
called from there. |
| 124 |
|
- add anelastic and deep-atmosphere geometry factor in pkg/mom_vecinv ; this |
| 125 |
|
allows to use Vector-Invariant form in deep atmos and anelastic formulation |
| 126 |
|
o pkg/seaice: introduce fast ice parameterisation following Itkin et al. (2015) |
| 127 |
|
and Koenig-Beatty+Holland (2012) |
| 128 |
|
- two new parameters SEAICE_tensilFac and SEAICE_tensilDepth |
| 129 |
|
- global 2D field tensileStrength, computed in seaice_calc_ice_strength |
| 130 |
|
- requires an extra input argument for S/R seaice_calc_viscosities |
| 131 |
|
- tensileStrength's depth dependence is different from Itkin et al. (2015) |
| 132 |
|
(to be changed and played with, for now exponential) |
| 133 |
|
o pkg/seaice: introduce new parameter SEAICEscaleSurfStress (default = .FALSE.) |
| 134 |
|
- if TRUE scale ice-ocean and ice-atmosphere stress acting on sea ice |
| 135 |
|
by AREA according to Connelly et al. (2004), JPO. |
| 136 |
|
- For EVP and the LSR solver, the implicit part of the drag term is |
| 137 |
|
the only non-zero term in the denominator when there is no ice (which |
| 138 |
|
prevented running the model with zero ice-ocean drag). If |
| 139 |
|
SEAICEscaleSurfStress = .TRUE., this results in a division by zero |
| 140 |
|
(or zero main diagonals BU/BV) which need to be caught. A practical |
| 141 |
|
consequence is that for open water, the momentum equation reduce |
| 142 |
|
to trivially 0 = 0 + 0 (for EVP). For LSR, BU/BV are reset to 1, if |
| 143 |
|
they happen to be zero, often leading to a non-zero solution over |
| 144 |
|
open water. |
| 145 |
|
o pkg/seaice: revise EVP code (seaice_evp.F) in preparation for a more |
| 146 |
efficient EVP method. |
efficient EVP method. |
| 147 |
- introduce area-weighted averages for e12 (analogously to |
- introduce area-weighted averages for e12 (analogously to |
| 148 |
seaice_calc_viscosities.F); this totally changes lab_sea.hb87 (2 digits |
seaice_calc_viscosities.F); this totally changes lab_sea.hb87 (2 digits |
| 149 |
remain, but also changing the order of terms in the previous average |
remain, but also changing the order of terms in the previous average |
| 150 |
changed the results as much) -> update experiment |
changed the results as much) -> update experiment |
| 151 |
- add SEAICE_DELTA_SMOOTHREG code |
- add SEAICE_DELTA_SMOOTHREG code |
| 152 |
- disentangle computation of zetaC/deltaC and zetaZ/deltaZ, also for |
- disentangle computation of zetaC/deltaC and zetaZ/deltaZ, also for |
| 153 |
TEM-code |
TEM-code |
| 154 |
- adjust loop bounds so that only required fields are computed |
- adjust loop bounds so that only required fields are computed |
| 155 |
- add new averaging code for zetaZ/deltaZ (again similar to |
- add new averaging code for zetaZ/deltaZ (again similar to |
| 156 |
seaice_calc_viscosities.F), inactive by default for now (will change |
seaice_calc_viscosities.F), inactive by default for now (will change |
| 157 |
results again), but will replace old code eventually |
results again), but will replace old code eventually |
| 158 |
o eesupp/src: |
o eesupp/src: |
| 163 |
and should be faster than GLOBAL_SUM_SEND_RECV code. |
and should be faster than GLOBAL_SUM_SEND_RECV code. |
| 164 |
- switch almost all verification experiments that use GLOBAL_SUM_SEND_RECV |
- switch almost all verification experiments that use GLOBAL_SUM_SEND_RECV |
| 165 |
(defined in CPP_EEOPTIONS.h_mpi) to GLOBAL_SUM_ORDER_TILES instead. |
(defined in CPP_EEOPTIONS.h_mpi) to GLOBAL_SUM_ORDER_TILES instead. |
| 166 |
- add CPP_EEOPTIONS.h_mpi with GLOBAL_SUM_ORDER_TILES defined in 13 more exp.: |
- add CPP_EEOPTIONS.h_mpi with GLOBAL_SUM_ORDER_TILES defined in 14 more exp.: |
| 167 |
aim.5l_cs, cheapAML_box, deep_anelastic, dome, fizhi-cs-32x32x40, |
aim.5l_cs, cheapAML_box, deep_anelastic, dome, fizhi-cs-32x32x40, |
| 168 |
fizhi-cs-aqualev20, global_ocean_ebm, matrix_example, natl_box, |
fizhi-cs-aqualev20, fizhi-gridalt-hs, global_ocean_ebm, matrix_example, |
| 169 |
solid-body.cs-32x32x1, tutorial_advection_in_gyre, tutorial_baroclinic_gyre |
natl_box, solid-body.cs-32x32x1, tutorial_advection_in_gyre, |
| 170 |
and tutorial_held_suarez_cs. |
tutorial_baroclinic_gyre and tutorial_held_suarez_cs. |
| 171 |
|
|
| 172 |
checkpoint65n (2015/07/29) |
checkpoint65n (2015/07/29) |
| 173 |
o pkg/thsice: |
o pkg/thsice: |
Parent Directory
| 
Revision Log
| 
Revision Graph
| 
Patch