| 3 |
Notes on tags used in MITgcmUV |
Notes on tags used in MITgcmUV |
| 4 |
============================== |
============================== |
| 5 |
|
|
| 6 |
|
ecco_c44_e20 |
| 7 |
|
o before major merges from the Scripps setup |
| 8 |
|
here are two small updates: |
| 9 |
|
- field swapping for surface flux controls enabled |
| 10 |
|
(remains t.b.d. for atmos. state plus obcs) |
| 11 |
|
- bug fix in active_file.F for xz-/yz- routines |
| 12 |
|
- added missing xz-/yz- routines to active_file_ad.F |
| 13 |
|
|
| 14 |
|
ecco_c44_e19 |
| 15 |
|
o merged necessary patches: |
| 16 |
|
- AD-related changes for GMRedi |
| 17 |
|
- fixes i KPP (delZ -> drF) |
| 18 |
|
- hook to OBCS songe layer code in external_forcing |
| 19 |
|
- errorMessageUnit non-zero in eeboot.F |
| 20 |
|
- included GMREDI, EXACT_CONSERV to natl_box_adjoint verif. |
| 21 |
|
|
| 22 |
|
ecco_c44_e18 |
| 23 |
|
o sponge layer code for obcs was missing |
| 24 |
|
- added hook in external_forcing.F |
| 25 |
|
- added storing for TAF |
| 26 |
|
o new routine cost_init_barfiles to initialise |
| 27 |
|
adjoint fields adtbar, adsbar, adpsbar |
| 28 |
|
(needed in conjunction with corrected ladinit for active I/O) |
| 29 |
|
o cleaned up and corrected order in packages_readparms, ecco_readparms |
| 30 |
|
o corrected and simplified weights for obcs controls |
| 31 |
|
- weight fields simplified |
| 32 |
|
- bug fix in ctrl_pack |
| 33 |
|
- no more write of control vector in ctrl_pack |
| 34 |
|
o i/j index bug fixes in cost_ssh_mean |
| 35 |
|
|
| 36 |
|
ecco_c44_e17 |
| 37 |
|
o bug fixes in cost package |
| 38 |
|
* wrong identifier for active I/O of monthly mean psbar |
| 39 |
|
(cost_averagesfields, cost_ssh) xx_psbar_dummy |
| 40 |
|
* cleaned all identifier for active I/O of monthly fields |
| 41 |
|
* wrong ladinit in various cost routines |
| 42 |
|
o merge with interval (chkpt44c_post - chkpt44d_post) |
| 43 |
|
* added missing EXCLUDE_MONITOR flags |
| 44 |
|
* changed "e" to "_d" in gmredi_slope_limit, gmredi_slope_psi |
| 45 |
|
(incompatible typ in MIN/MAX expressions caused problems |
| 46 |
|
on IBM SP3) |
| 47 |
|
* in genmake added variable MAKEDEPEND |
| 48 |
|
plus resetting for case SunOS |
| 49 |
|
added timer_stats.c routine for IBM SP3 |
| 50 |
|
* removed variables in dynamics |
| 51 |
|
|
| 52 |
|
ecco_c44_e16 |
| 53 |
|
o new ECCO checkpoint, the first since ecco_c43_e15 |
| 54 |
|
which is branched from MITgcm. |
| 55 |
|
All changes were merged in between (see ecco-branch-mod? tags). |
| 56 |
|
o last changes: |
| 57 |
|
* assuming path for TAF and TAMC are known (e.g. in $path) |
| 58 |
|
* genmake: |
| 59 |
|
- added tools/xmakedepend to circumvene shortcoming of |
| 60 |
|
SUN's makedepend (size too small) |
| 61 |
|
- invoking ../tools/xmakedepend instead of makedepend |
| 62 |
|
- changed compiler flags for case SunOS |
| 63 |
|
|
| 64 |
|
ecco-branch-mod5 |
| 65 |
|
o mod's to reproduce open boundary run |
| 66 |
|
o enable field swapping for OB's and climatology |
| 67 |
|
o update store directives for OB control |
| 68 |
|
o added option ALLOW_OBCS_PRESCRIBE |
| 69 |
|
o call of getobcs now out of obcs_calc (param. list modified) |
| 70 |
|
|
| 71 |
|
ecco-branch-mod4 |
| 72 |
|
o this is the second part of changes necessary to reconcile |
| 73 |
|
the natl_box_adjoint verification (with and without grdchk). |
| 74 |
|
>>>!!! CHANGES BETWEEN THE TAGS ecco-branch-mod3 !!!<<< |
| 75 |
|
>>>!!! AND ecco-branch-mod4 !!!<<< |
| 76 |
|
>>>!!! SHOULD BE MERGED INTO THE MAIN TRUNK !!!<<< |
| 77 |
|
- genmake: removed $S64 overwrite for case SunOS |
| 78 |
|
- pkg/exf: update and corrections for field swapping and obcs |
| 79 |
|
- pkg/ecco: parameter lists for the_model_main, the_main_loop |
| 80 |
|
harmonized between ECCO and MITgcm |
| 81 |
|
- pkg/autodiff: added flow directives for obcs, mdsio_gl_slice |
| 82 |
|
updated checkpointing_lev... lists for obcs |
| 83 |
|
- model/src: minor changes in forward_step, plot_field |
| 84 |
|
- pkg/mdsio: added mdsio_gl_slice |
| 85 |
|
|
| 86 |
|
ecco-branch-mod3 |
| 87 |
|
o this is the first part of changes necessary to reconcile |
| 88 |
|
natl_box_adjoint verification (with and without grdchk). |
| 89 |
|
This set of changes is relevant for ECCO only. |
| 90 |
|
The next set/tag of changes (ecco-branch-mod4) |
| 91 |
|
should be merged into the main trunk. |
| 92 |
|
- update some options |
| 93 |
|
- parameter lists of ecco_the_model_main and ecco_the_main_loop |
| 94 |
|
differed from those of the MITgcm counterparts. |
| 95 |
|
|
| 96 |
|
ecco-branch-mod2 |
| 97 |
|
o update ecco-branch-mod1 to checkpoint44 |
| 98 |
|
o exclude from updates: |
| 99 |
|
- verification/ |
| 100 |
|
- adjoint/ |
| 101 |
|
- pkg/cost |
| 102 |
|
- pkg/ctrl |
| 103 |
|
- pkg/ecco |
| 104 |
|
- pkg/obcs |
| 105 |
|
- model/src/the_model_main.F |
| 106 |
|
- model/src/the_main_loop.F |
| 107 |
|
|
| 108 |
|
checkpoint44 |
| 109 |
|
o apply Shapiro filter on T&S before calc_PhiHyd (stagger time stepping) |
| 110 |
|
o fix bug in mom_vi_del2uv |
| 111 |
|
o added some stuff for AD |
| 112 |
|
|
| 113 |
|
ecco-branch-mod1 |
| 114 |
|
o based on c43 |
| 115 |
|
o replaced packages cost, ctrl, ecco, obcs |
| 116 |
|
o removed MITgcm verif.'s, added ECCO verif.'s |
| 117 |
|
o added lsopt, optim for line search optimization. |
| 118 |
|
|
| 119 |
|
checkpoint43 |
| 120 |
|
o added two new advection tests |
| 121 |
|
o added global 4x4 degree ocean test |
| 122 |
|
o added test of H&S94 with free surface on cube |
| 123 |
|
o some final mods for gradient check, adjoint,... |
| 124 |
|
|
| 125 |
|
checkpoint42 |
| 126 |
|
Fixed AD-related problems: |
| 127 |
|
o Store directives up-to-date with re-arranged Adams-Bashforth |
| 128 |
|
(mainly thermodynamics.F) |
| 129 |
|
o New store directives for multi-dim. advection schemes |
| 130 |
|
* new CPP flag ALLOW_MULTI_DIM_ADVECTION |
| 131 |
|
* new common block and key passkey |
| 132 |
|
(mainly gad_advection.F) |
| 133 |
|
o Modified store directives for split of dynamics/thermodynamics |
| 134 |
|
for the case ALLOW_KPP |
| 135 |
|
o Cleaned argument list for timestep_tracer.F |
| 136 |
|
|
| 137 |
|
checkpoint41 |
| 138 |
|
o re-formatted/added comments in prologues of key subroutines |
| 139 |
|
to work with protex |
| 140 |
|
|
| 141 |
|
checkpoint40 |
| 142 |
|
o moved calc_gs, calc_gt and calc_tr1 back to model/src |
| 143 |
|
o added FLT (floats) package from Arne along with demo. experiment |
| 144 |
|
o re-arranged Adams-Bashforth scheme for tracers to allow easier |
| 145 |
|
implementation of forward-in-time forcing terms |
| 146 |
|
o more mods for non-linear free-surface including new variable |
| 147 |
|
o modified multi-dim method to work on cube |
| 148 |
|
o bug fix in KPP (from DM) |
| 149 |
|
|
| 150 |
|
checkpoint40pre9 |
| 151 |
|
o added DST advection schemes |
| 152 |
|
o fix MPI bug on SGI's |
| 153 |
|
o modifications for non-linear free-surface |
| 154 |
|
o fixed conditionals for fixed flow/passive tracer only mode |
| 155 |
|
o added tension/strain form of stress tensor |
| 156 |
|
o changes to commenting of arguments for TAF/TAMC |
| 157 |
|
o added/updated many matlab scripts |
| 158 |
|
o added multi-dimensional advection method for tracers |
| 159 |
|
|
| 160 |
|
checkpoint40pre8 |
| 161 |
|
o new package rw/ |
| 162 |
|
Added method for dumping intermediate local arrays: |
| 163 |
|
mdsio_writetile - same as mdsio_writefield except works from inside |
| 164 |
|
bi,bj loop |
| 165 |
|
mdsio_writelocal - same as mdsio_writetile except works for local |
| 166 |
|
arrays |
| 167 |
|
write_local_r? - higher-level wrapper for mdsio_writelocal |
| 168 |
|
Controlled by diagFreq. Defaults to zero (ie. no dumps) |
| 169 |
|
Example given at end of mom_vecinv.F that dumps some local arrays. |
| 170 |
|
o replaced calc_mom_rhs.F by mom_fluxform.F/mom_vecinc.F to |
| 171 |
|
control flux-form versus vector invariant code. |
| 172 |
|
Switch is runtime flag vectorInvariantMomentum (defaults to false) |
| 173 |
|
or genmake -DISABLE ... |
| 174 |
|
o TAMC-related update to cope with mom_fluxform.F/mom_vecinc.F |
| 175 |
|
(additional storing is now needed for case vecinv; |
| 176 |
|
currently nothing implemented to switch this storing off |
| 177 |
|
in case fluxform is used since no CPP flag available). |
| 178 |
|
|
| 179 |
checkpoint40pre7 |
checkpoint40pre7 |
| 180 |
o bug fix in dynamics.F |
o bug fix in dynamics.F |
| 181 |
calc_viscosity called before k=1,Nr loop. |
calc_viscosity called before k=1,Nr loop. |
Parent Directory
| 
Revision Log
| 
Revision Graph
| 
Patch