| 4 |
Notes on tags used in MITgcmUV |
Notes on tags used in MITgcmUV |
| 5 |
============================== |
============================== |
| 6 |
|
|
| 7 |
o take bi,bj loops outside calc_gw to fix Pb with local array KappaRU,V |
o pkg/exf: changes to properly interpolate/mask/scale initial fields |
| 8 |
|
o utils/matlab/cs_grid/read_cs: routines to speed up reading of CS510 fields |
| 9 |
|
and portions thereof. See read_cs.txt for description and examples. |
| 10 |
|
o utils/matlab/cs_grid/bk_line: to compute transport through passages, |
| 11 |
|
add a script (grt_circ_bkl.m) to generate broken-line between 2 points, |
| 12 |
|
following the great-circle arc between those 2 points. |
| 13 |
|
o trap un-initialized variable with pathscale compiler on weddell.mit.edu |
| 14 |
|
(thanks to Constantinos): fix few problems and get (almost) all exp to pass. |
| 15 |
|
o 2nd-Order Moment Advection Scheme (Prather, 1986): first check-in |
| 16 |
|
- enable by setting #define GAD_ALLOW_SOM_ADVECT (in GAD_OPTIONS.h) |
| 17 |
|
- used without limiter (AdvScheme=80) or with Prather limiter (AdvScheme=81) |
| 18 |
|
- still needs work (not working with some options ; efficiency to improve) |
| 19 |
|
and serious testing. |
| 20 |
|
- use this scheme for temperature & salinity in 2 advection test exp.: |
| 21 |
|
advect_xy & advect_xz, respectively. |
| 22 |
|
|
| 23 |
|
checkpoint58u_post |
| 24 |
|
o new test-exp: fizhi-cs-32x32x40 (40 levels) to replace the 10 levels. |
| 25 |
|
o move call to INI_FORCING from PACKAGES_INIT_VARIABLES to INITIALISE_VARIA. |
| 26 |
|
o testreport: add option "-skipdir" to skip some test. |
| 27 |
|
o exf: when input wind-stress (#undef ALLOW_ATM_WIND), by-pass turbulent |
| 28 |
|
momentum calculation. |
| 29 |
|
o gad_advection: fix vertAdvecScheme (if different from advectionScheme) |
| 30 |
|
o some cleaning: usePickupBeforeC35 no longer supported ; remove this option. |
| 31 |
|
remove checkpoint.F and the_correction_step.F (no longer used); |
| 32 |
|
do the k loop inside CYCLE_TRACER (supposed to be more efficient). |
| 33 |
|
o add option (linFSConserveTr) to correct for tracer source/sink due to |
| 34 |
|
Linear Free surface |
| 35 |
|
o pkg/seaice: fix a bug in the flooding algorithm: turn off the snow machine |
| 36 |
|
o pkg/thsice: fix reading mnc-pickups |
| 37 |
|
|
| 38 |
|
checkpoint58t_post |
| 39 |
|
o pkg/diagnostics: fix multi-threaded output (when singleCpuIO=F) |
| 40 |
|
o pkg/mdsio: |
| 41 |
|
- clean-up mdsio_readfield.F & mdsio_writefield_new.F and |
| 42 |
|
move to new files: mdsio_read_field.F (S/R MDS_READ_FIELD) & |
| 43 |
|
mdsio_write_field.F (S/R MDS_WRITE_FIELD); |
| 44 |
|
- fix multi-threaded singleCpuIO option in MDS_WRITE_FIELD |
| 45 |
|
- collect old S/R in one file: mdsio_rw_field.F (now just a call to |
| 46 |
|
either MDS_READ_FIELD or MDS_WRITE_FIELD ) |
| 47 |
|
- new S/R (to replace MDSWRITEMETA), which allows to write more |
| 48 |
|
informations (variable names, time ...) to meta files. |
| 49 |
|
- modify rdmds.m to give back the meta file content. |
| 50 |
|
o pkg/diagnostics: change vertical interpolation to be safer, and more |
| 51 |
|
flexible (define p-levels to interpolate to in data.diagnostics); |
| 52 |
|
add argument to S/R DIAGNOSTICS_GET_POINTERS to catch the right |
| 53 |
|
diagnostic pointer. |
| 54 |
|
o pkg/seaice: more cleaning up of seaice_growth, added many comments |
| 55 |
|
o pkg/exf: put runoff0 and runoff1 into the appropriate common block |
| 56 |
|
o fix multi-category seaice: |
| 57 |
|
- change cpp flag SEAICE_MULTILEVEL to more meaningful name: |
| 58 |
|
SEAICE_MULTICATEGORY |
| 59 |
|
- fix short wave heat flux |
| 60 |
|
o pkg/seaice: change default (SEAICE_OPTIONS.h): use C-grid dynamics |
| 61 |
|
and allow to use EVP solver. Change forward lab_sea: test both EVP |
| 62 |
|
solver and LSR solver, and use C-grid dynamics (like lab_sea adjoint) |
| 63 |
|
o Initial check-in of lab_sea adjoint |
| 64 |
|
(very similar to natl_box_adjoint/code_seaice) |
| 65 |
|
* used C-grid and EVP solver |
| 66 |
|
* used KPP, GM/Redi |
| 67 |
|
* uses inAdExact=.TRUE. |
| 68 |
|
* gradient checks fail for both ADM and TLM |
| 69 |
|
due to sensitive snow melt code in seaice_grwoth (look for "JZ") |
| 70 |
|
f.d. gradients actually seem more problematic than ADM, TLM gradients; |
| 71 |
|
issue is related to forward mismatch between g77 an ifort |
| 72 |
|
o pkg/seaice: overhaul of thermodynamics |
| 73 |
|
- growth.F and groatb.F are replaced by new routine seaice_growth.F |
| 74 |
|
- budget.F is replaced by two new routines seaice_budget_ocean/ice.F |
| 75 |
|
- move a few global fields out of SEAICE.h into seaice_growth.F and |
| 76 |
|
make them 2D (FICE/QNETO/ ...) |
| 77 |
|
- remove FO (it is the same as QNETO) |
| 78 |
|
- introduce a few local fields to avoid modifying external fields such |
| 79 |
|
as atemp, etc. |
| 80 |
|
|
| 81 |
|
checkpoint58s_post |
| 82 |
|
o pkg/exf: moved exf_init_runoff.F to exf_init.F and exf_set_gen.F |
| 83 |
|
changed exf_set_gen.F so that |
| 84 |
|
genperiod=0 means input file is one time-constant field and |
| 85 |
|
genperiod=-12 means input file contains 12 monthly means |
| 86 |
|
o CG2D: use pre-computed solver main-diagonal (stored in common block); |
| 87 |
|
this affects truncation error ; generate new output file for exp. which |
| 88 |
|
fails: exp4(12d), fizhi-cs-aqualev20(12d), front_relax(10d), |
| 89 |
|
global_ocean.90x40x15(10d), global_ocean.cs32x15(10d).thsice(11d).viscA4(10d), |
| 90 |
|
ideal_2D_oce(10d), internal_wave(09d), isomip.htd(11d), matrix_example(07d), |
| 91 |
|
MLAdjust(4x11d,12d), natl_box(11d), rotating_tank(12d), |
| 92 |
|
tutorial_global_oce_in_p(10d), tutorial_plume_on_slope(11d), |
| 93 |
|
vermix(12d).ggl90(12d).my82(11d).opps(12d).pp81(11d) |
| 94 |
|
o start implementation of deep-atmosphere and anelastic formulation |
| 95 |
|
o verification: remove 2nd copy of input files and use script "prepare_run" |
| 96 |
|
to make a symbolic link to the 1rst copy (reduce downloaded size by ~81.M) |
| 97 |
|
o mom_common/vecinv/fluxform: fix a few initialisation isssues |
| 98 |
|
o profiles: fix for multithreaded case |
| 99 |
|
|
| 100 |
|
o diagnostics: rename few diags: |
| 101 |
|
s/'PRESSURE'/'RCENTER '/g |
| 102 |
|
s/'TICE '/'oceFreez'/g |
| 103 |
|
s/'TAUX '/'oceTAUX '/g |
| 104 |
|
s/'TAUY '/'oceTAUY '/g |
| 105 |
|
s/'SWFLUX '/'oceQsw '/g |
| 106 |
|
s/DIFx/DFxE/g |
| 107 |
|
s/DIFy/DFyE/g |
| 108 |
|
and add few ocean-forcing diagnostics: |
| 109 |
|
atmPload, sIceLoad, oceQnet, oceFWflx, oceSflux, surForcT & surForcS |
| 110 |
|
o modifications to pkg/seaice/growth.F and budget.F, which change results of |
| 111 |
|
lab_sea experiment: HSNOW was sometimes treated as effective and sometimes |
| 112 |
|
as actual snow thickness. It now refers everywhere to effective snow |
| 113 |
|
thickness, meaning that for non-zero AREA, actual thickness is HEFF/AREA. |
| 114 |
|
|
| 115 |
|
checkpoint58r_post |
| 116 |
|
o modifications to pkg/seaice/growth.F, which change results of |
| 117 |
|
lab_sea experiment: fixes pertain to melting snow over open water, to |
| 118 |
|
masking EVAP under sea ice, and to using the correct AREA variable for |
| 119 |
|
computations of EVAP, PRECIP, HSNOW, and GHEFF |
| 120 |
|
o add ability to read bathy, salt, and theta using MNC (off by def) |
| 121 |
|
o add flooding algorithm and advection of HSNOW for pkg/seaice; |
| 122 |
|
both are turned off by default and do no change any results |
| 123 |
|
o seaice advection with pkg/thsice: a step toward the use of non-linear |
| 124 |
|
advection scheme, but #ifdef or commented out for now (requires more EXCH). |
| 125 |
|
o use standard version of PTRACERS_INI_VARIA in tutorial_tracer_adjsens |
| 126 |
|
(identical results) and the standard version of EXTERNAL_FIELDS_LOAD |
| 127 |
|
in natl_box (affects results @ truncation level: update output.txt) |
| 128 |
|
o multi-threading and IO: |
| 129 |
|
- replace some call to MDSIO S/R by the corresponding READ/WRITE_FLD S/R. |
| 130 |
|
- remove most of BEGIN_MASTER/ END_MASTER around READ/WRITE_FLD S/R calls |
| 131 |
|
(needed to work with SingleCpuIO): tested in all verif. exp. |
| 132 |
|
o GAD: use simpler DST3 formulation as default (old version left within #ifdef) |
| 133 |
|
affects the results (but none of the verification test uses this scheme). |
| 134 |
|
o Adding some pkg/diagnostics capability to pkg/seaice |
| 135 |
|
(will certainly be subject to changes) |
| 136 |
|
o ctrl: adding new control variables: |
| 137 |
|
lwflux, lwdown, evap, snowprecip, apressure, runoff |
| 138 |
|
|
| 139 |
|
checkpoint58q_post |
| 140 |
|
o added CPP_EEOPTIONS.h flag DISABLE_MPI_READY_TO_RECEIVE. Speeds up |
| 141 |
|
1/6-deg Southern Ocean configuration by 8% on IBM Datastar and |
| 142 |
|
high_res_cube configuration by 4% on Columbia. |
| 143 |
|
o clean-up multi-threaded problems (reported by debugger tcheck on ACES); |
| 144 |
|
add (in global_with_exf/code) a version of EXF_INTERP_READ without |
| 145 |
|
dynamical allocation (seems to work in multi-threaded). |
| 146 |
|
o new function "MASTER_CPU_IO" (=T if master thread & master CPU, if using |
| 147 |
|
single CPU io) so that TAF does not have to see EESUPPORT.h |
| 148 |
|
o pkg/seaice: implement sea-ice loading (when using RealFreshWaterFlux) |
| 149 |
|
|
| 150 |
|
checkpoint58p_post |
| 151 |
|
o obcs: stop if obcs is used with useCDscheme = .true. |
| 152 |
|
o shelfice: |
| 153 |
|
- fix a sign error so that no slip at shelfice interface works |
| 154 |
|
- remove stop for combination non-hydrostatic code and shelfice. That does |
| 155 |
|
not mean, that the results are correct, but at least they are numerically |
| 156 |
|
stable |
| 157 |
|
o diagnoctics: modify UVELPHI and VVELPHI to include hfac (i.e. mass-weighted) |
| 158 |
|
o added adjoint exch routined for all exch*3d* routines |
| 159 |
|
and updated flow directives |
| 160 |
|
|
| 161 |
|
checkpoint58o_post |
| 162 |
|
o replace calls to EXCH_Z & EXCH_UV_AGRID: use the new _3D_ version ; |
| 163 |
|
remove old subroutines (_XY_ & _XYZ_); update the adjoint version. |
| 164 |
|
o change controling S/R for writing pickup file (new S/R do_write_pickup) |
| 165 |
|
and clean-up reading/writing of pickup files. |
| 166 |
|
o fix local version of read/write pickup (NH part) in rotating_tank/code |
| 167 |
|
o simplify EXCH interface: |
| 168 |
|
- change _XYZ_ interface to _3D_ subroutine (with 3rd dim in argument list) |
| 169 |
|
- not often used EXCH S/R (exch_z, exch_uv_agrid): keep only the _3D_ version |
| 170 |
|
o added the regrid package |
| 171 |
|
o shelfice: |
| 172 |
|
- add a simple boundary layer scheme to reduce noise |
| 173 |
|
- fix the indexing in shelfice_u/v_drag so that there will be actually drag |
| 174 |
|
|
| 175 |
|
checkpoint58n_post |
| 176 |
|
o fix globalArea multi-threaded calculation: need to apply GLOBAL_SUM to |
| 177 |
|
local variable (not in common block). |
| 178 |
|
o lot of small changes for multi-threaded. Most tests (using OMP & ifort) pass. |
| 179 |
|
- still problems with pkgs: mnc, exp2, dic, cal, exf (and probably more); |
| 180 |
|
- still BEGIN_MASTER/END_MASTER to remove around READ/WRITE_FLD S/R calls |
| 181 |
|
to work with SingleCpuIO. |
| 182 |
|
o add "_d 0" to real constant of pkg/cfc subroutines ; |
| 183 |
|
update exp. cfc_exemple & tutorial_cfc_offline output files. |
| 184 |
|
o multi-threading and IO: |
| 185 |
|
- all threads call READ_FLD S/R (needed if using SingleCpuIO): |
| 186 |
|
(no BEGIN_MASTER/ END_MASTER around the call, but inside MDSIO S/R. |
| 187 |
|
- barrier before and after reading. |
| 188 |
|
- stop if other than master is calling MDS_WRITELOCAL (if #undef IO_THREAD_SAFE) |
| 189 |
|
- avoid using MDS_WRITELOCAL for LAND pkg output. |
| 190 |
|
- new S/R AIM_WRITE_PHYS (replaces AIM_WRITE_LOCAL) to write AIM physics |
| 191 |
|
common-block variables ; Allows multi-threading with master-thread IO. |
| 192 |
|
o change default MAX_NO_THREADS from 32 to 4 ; |
| 193 |
|
and stop cleanly when MAX_NO_THREADS is too small. |
| 194 |
|
o add S/R BAR_CHECK to check multi-threaded barrier synchronization |
| 195 |
|
|
| 196 |
|
checkpoint58m_post |
| 197 |
|
o add a Non-Hydroatatic test (with 3-D Coriolis, NHMTerms, biharmonic Visc. |
| 198 |
|
and side-drag) on CS-grid : global_ocean.cs32x15.viscA4 |
| 199 |
|
o make "tutorial_global_oce_latlon" from standard "global_ocean.90x40x15" exp. |
| 200 |
|
keep global_ocean.90x40x15 (for AD testing) with former secondary test |
| 201 |
|
(".qshyd") as the main test (mv output.qshyd.txt to output.txt) |
| 202 |
|
o CVS modules: update MITgcm_verif_basic; remove MITgcm_verif_ocean & _verif_atmos; |
| 203 |
|
add MITgcm_tutorials (= code + all the tutorials) |
| 204 |
|
o ecco, option to compute cost for total ssh (mean+anom) |
| 205 |
|
o consistent with addition of ptracer and ssh in pkg/profiles |
| 206 |
|
o profiles: adding ptracer and ssh |
| 207 |
|
o testreport: change multi-test output-file name |
| 208 |
|
(from output.txt.{name} to output.{name}.txt) |
| 209 |
|
o profiles: now ok for mutli-tiles case |
| 210 |
|
o seaice: add 4 parameters for different advection schemes for area, |
| 211 |
|
(effective) seaice thickness, enthalphy, and snow height. Their default |
| 212 |
|
is the standard parameter SEAICEadvScheme |
| 213 |
|
o calc_gw: |
| 214 |
|
- add cosFacU in zonal viscous flux (harmonic.visc.). |
| 215 |
|
- put back side-drag (call new S/R MOM_W_SIDEDRAG) and output.txt. |
| 216 |
|
of tutorial_plume_on_slope. |
| 217 |
|
o obcs_calc.F : fix index Pb in OBSeta & OBNeta (Thanks to Gianmaria). |
| 218 |
|
o calc_gw: true flux-form, account for horizontal grid spacing. |
| 219 |
|
update rotating_tank results, but not plume_on_slope (affected at |
| 220 |
|
truncation level, 11 digits). |
| 221 |
|
|
| 222 |
|
checkpoint58l_post |
| 223 |
|
o calc_gw: |
| 224 |
|
- take bi,bj loops outside calc_gw to fix Pb with local array KappaRU,V |
| 225 |
|
- remove side-drag for now (need to fix it); update plume_on_slope results. |
| 226 |
|
- fix for partial-cell & non-uniform dz; update plume_on_slope results. |
| 227 |
o commit code to make SST/SSS weights spatially variable |
o commit code to make SST/SSS weights spatially variable |
| 228 |
(based on either variable in-situ or variable initial T/S weights) |
(based on either variable in-situ or variable initial T/S weights) |
| 229 |
o commit code to relax deviations of xx_theta, xx_salt to 5*sigma, |
o pkg/ecco: |
| 230 |
now based on horizontally varying weights wthetaLev, wsaltLev |
* commit code to relax deviations of xx_theta, xx_salt to 5*sigma, |
| 231 |
|
now based on horizontally varying weights wthetaLev, wsaltLev |
| 232 |
|
* cost_averagesflags now returns year indices in addition to day, month |
| 233 |
|
* sflux balance is performed on annual basis |
| 234 |
|
(sflxumm, sfluxmm2, bal_sfluxmm) |
| 235 |
|
* for SSH anomaly, separate more cleanly contrib from T/P, ERS, GFO, |
| 236 |
|
including separate counting and multipliers |
| 237 |
|
o pkg/rbcs |
| 238 |
|
* add all necessary infrastructure for pkg/rbcs levN adjoint checkpointing |
| 239 |
|
|
| 240 |
checkpoint58k_post |
checkpoint58k_post |
| 241 |
o testreport: |
o testreport: |
Parent Directory
| 
Revision Log
| 
Revision Graph
| 
Patch