3 |
Notes on tags used in MITgcmUV |
Notes on tags used in MITgcmUV |
4 |
============================== |
============================== |
5 |
|
|
6 |
next patch: |
checkpoint45a_post |
7 |
o bug fix in packages_readparms (missing ALLOW_CALENDAR) |
o compute Rho over the whole domain (extend iMin,jMin indices) |
8 |
o bug fix in CG2D.h (correct array size for LETS_MAKE_JAM |
o specific exchange for X-slice domain (case Ny=1) |
9 |
o Adjoint-related bug fixes in kpp: |
o argument futureIter of S/R obcs_calc missing in exp4 & internal_wave /code |
10 |
- kpp_calc: sore of kpphbl avoids recomputation/call to S/R kppmix |
o solve_for_pressure : set cg2d_x to zero where OBCS are applied |
11 |
- kpp_routines: store of Rib avoids partial recomputation bug of TAF. |
|
12 |
o added Eliassen-Palm flux hooks (dynamics,ini_fields) |
checkpoint44h_post = checkpoint45 |
13 |
o removed unused TAF keys iikey, idkey (dynamics,thermodynamics) |
o Fix to restore differentiability: |
14 |
o flow directives: |
place at which packages_init_variables is called in |
15 |
- new for S/R do_the_model_io |
initialise_varia is crucial for reverse control flow. |
16 |
- enhanced for S/R checkpoint |
o modifications in GMRedi to enable stable adjoint |
17 |
o added E-P fluxes to gradient check package |
(corresponding modif's in thermodynamics, dynamics) |
18 |
|
o added missing hook for sponge layer code in external_forcing |
19 |
release1_p4 |
o modified test cost functions |
20 |
o Removed f90 comments ("!") |
o storing in gad_advection (plus removal of write(0,*)! |
21 |
o Modified initialisations to enable exf package for MITgcm |
o missing headers TR1.h added in convective_adjustment routines |
22 |
(without invoking ECCO_PACKAGE). |
o errorMessageUnit set non-equal zero in eeboot |
23 |
o modifications to gradient check package (Martin Losch) |
(conflict on some platforms, e.g. IBM SP3) |
24 |
- enable centered differences |
o modified carbon verif. |
25 |
- modified format of standard output |
|
26 |
|
checkpoint44h_pre |
27 |
release1_p3 |
o vertical grid option: allow to put the Interface at the middle between |
28 |
o added tangent linear setup to verification/carbon/ |
2 cell-centers ; replace delZ in pkg/kpp by drF. |
29 |
o added adjoint sensitivity setup to verification/global_ocean... |
o GM Advective form: Tracers are advected using the residual transport (= |
30 |
o moved packages_init_variables before convective_adjustment |
Euler+GM-bolus); set param GM_AdvSeparate=T to return to previous form. |
31 |
o output of verif. which use GMRedi brought up-to-date with |
|
32 |
MAIN trunk (those were re-created after changes in GMRedi) |
checkpoint44g_post |
33 |
o included CPP option SINGLE_LAYER_MODE |
o fix surface correction term in multiDimAdvection (affects plume_on_slope |
34 |
to configure barotropic setup (Martin Losch). |
results) ; ifdef missing in thermodynamics.F |
35 |
o Removed auxiliary variable in impldiff |
|
36 |
(caused numerical differences with MAIN trunk routine). |
checkpoint44f_pre,checkpoint44f_post |
37 |
o Introduced CPP option ALLOW_AUTODIFF_MONITOR to |
o added PTRACERS package |
38 |
disable adcommon.h/g_common.h by default. |
This allows an arbitrary number of passive tracers to be integrated |
39 |
o Cleaned exchange calls in cg2d. |
forward simultaneously with the dynamicaly model. |
40 |
|
+ Implemented so far: |
41 |
release1_p2 |
- basic forward algorithm (time-stepping, advection, diffusion, convection) |
42 |
o fixed array boundary problem for adjoint |
- I/O and checkpointing |
43 |
(related to partial recomputation within subroutine |
- GM/Redi *but* using the GM/Redi coefficient of Salt |
44 |
related to dynamics -> calc_viscosity -> mom_... |
+ Not implemented so far: |
45 |
o fixed recomputation in thermodynamics |
- KPP |
46 |
o prepare (but not complete) recomputation problem in impldiff |
- OBCS |
47 |
o fixed problem for adjoint of global_max |
+ No specific example supplied (yet) but global_ocean.90x40x15 has the |
48 |
o switched off sponge layer code in external_forcing |
necessary data.ptracer file. Simply use -enable=ptracers and uncomment |
49 |
|
line in data.pkg. |
50 |
release1_p1 |
+ This package is disabled by default since it increases storage. |
51 |
o merged necessary patches: |
|
52 |
- AD-related changes for GMRedi |
|
53 |
- fixes i KPP (delZ -> drF) |
o testscript: added -cleanup option |
54 |
- hook to OBCS songe layer code in external_forcing |
o Shap_filt: a) new shap_filt S/R to use no-slip BC with S2 filter |
55 |
- errorMessageUnit non-zero in eeboot.F |
b) enable to filter 2D fields. |
56 |
- modified test cost function and carbon verif. |
|
57 |
|
checkpoint44e_pre,post |
58 |
release1_chkpt44d_post = chkpt44d_post |
o re-wrote convective adjustment to use pre-calculated weights |
59 |
|
Affects: model/src/convective_adjustment.F, convective_adjustment_ini.F |
60 |
|
Adds: model/src/convective_weights.F, convectively_mixtracer.F |
61 |
|
Original calls to convect() still in place. |
62 |
|
|
63 |
chkpt44d_pre,post |
chkpt44d_pre,post |
64 |
o added missing EXCLUDE_MONITOR flags |
o added missing EXCLUDE_MONITOR flags |
69 |
plus resetting for case SunOS |
plus resetting for case SunOS |
70 |
o added timer_stats.c routine for IBM SP3 |
o added timer_stats.c routine for IBM SP3 |
71 |
o removed variables in dynamics |
o removed variables in dynamics |
72 |
|
!!! I forgot to create tag chkpt44d_pre !!! |
73 |
|
!!! But chkpt44d_pre = chkpt44c_post !!! |
74 |
|
|
75 |
o real fresh water flux implemented with non-linear free-surface. |
o real fresh water flux implemented with non-linear free-surface. |
76 |
|
|