| 4 |
Notes on tags used in MITgcmUV |
Notes on tags used in MITgcmUV |
| 5 |
============================== |
============================== |
| 6 |
|
|
| 7 |
|
o pkg/offline: |
| 8 |
|
- change where Offline-Fields are loaded, from load_fields_driver.F to |
| 9 |
|
directly within FORWARD_STEP, just after updating iter number and time; |
| 10 |
|
- move setting of GM and KPP diffusivity out of OFFLINE_FIELDS_LOAD into new |
| 11 |
|
S/R OFFLINE_GET_DIFFUS that is called towards the end of DO_OCEANIC_PHYS; |
| 12 |
|
These changes allow: |
| 13 |
|
a) to recover exactly on-line results using deltaToffline=deltaTClock, |
| 14 |
|
tested (using offlineIter0= nIter0 -1 and offlineTimeOffset= -deltaT/2) |
| 15 |
|
both with Lin & Non-Lin Free-Surf, both with staggerTimeStep & synchronous |
| 16 |
|
time-step, both with GM & KPP files and with recomputing these fields; |
| 17 |
|
b) no needs to switch on staggerTimeStep anymore; |
| 18 |
|
c) simplify rescaling of offline horiz. velocity with Non-Lin Free-Surf. |
| 19 |
|
- update output of experiments tutorial_cfc_offline (11 to 13 digits left) |
| 20 |
|
and tutorial_dic_adjoffline (only 3 digits left). |
| 21 |
|
o pkg/offline: |
| 22 |
|
- change (rarely used) offset-iteration parameter to an offset time (in s): |
| 23 |
|
can easily recover previous offset-iter effect by setting offset-time |
| 24 |
|
(offlineTimeOffset) to offset-iter x deltaToffline; |
| 25 |
|
- do not turn-off Temp & Salinity surface relaxation when using KPP (used |
| 26 |
|
in KPP as part of surface buoyancy forcing) in offline run. |
| 27 |
|
- fix ptracers KPP non-local term in the case where KPP is re-computed. |
| 28 |
|
o OpenAD |
| 29 |
|
- Adding preliminary ggl90 test to OpenAD verification. |
| 30 |
|
Config. with ggl90 currently diverges after 64 time steps |
| 31 |
|
so will need more work to become more stable |
| 32 |
|
- Adding preliminary kpp test to OpenAD verification. |
| 33 |
|
required small tweakes in kpp_routines.F |
| 34 |
|
- Merge and update from Krishna Narayanan's contrib area: |
| 35 |
|
* Add one variable for preliminary OpenAD DIVA |
| 36 |
|
* Add replacement time-stepping for OpenAD DIVA |
| 37 |
|
* genmake2 flag -diva (but only for OpenAD; preliminary) |
| 38 |
|
* required modifs for OAD_support |
| 39 |
|
successful preliminary test for tutorial_global_oce_biogeo/ |
| 40 |
|
o pkg/seaice: |
| 41 |
|
- initialise deltaC in seaice_lsr, cosmetic changes in seaice_lsr |
| 42 |
|
- add cpp-flag SEAICE_DELTA_SMOOTHREG for option of regularising |
| 43 |
|
delta with a smooth function in s/r seaice_calc_viscosity |
| 44 |
|
(no effect on EVP) |
| 45 |
|
- rename local suffix "sqr" into "sq" for more consistent variable names |
| 46 |
|
o pkg/seaice: |
| 47 |
|
add new parameter SEAICE_deltaMin that is used ***only*** |
| 48 |
|
for regularizing Delta (and nothing else, like the parameter |
| 49 |
|
which is also used for all sorts of things SEAICE_EPS). |
| 50 |
|
Defaults to SEAICE_EPS for backward compatibility only. |
| 51 |
|
o pkg/seaice: replace erroneously storing e12Csqr on the wrong tape with |
| 52 |
|
the wrong key by inialising it before use -> fixes a recomputation |
| 53 |
|
problem but does not fix the gradient |
| 54 |
|
o pkg/layers: |
| 55 |
|
- move resetting of flux arrays to outside the iLa loop |
| 56 |
|
(this caused layers_diapycnal to not work with more than one layer coord.) |
| 57 |
|
- fix bug in calculation of layers_recip_delta in layers_init_fixed.F |
| 58 |
|
o pkg/layers: |
| 59 |
|
- remove from common block 3D layers fields (used to diagnose transport |
| 60 |
|
in tracer coords) and declare them as local arrays in layers_calc.F |
| 61 |
|
o pkg/seaice: |
| 62 |
|
- Rename SEAICE_VECTORIZE_LSR_ZEBRA to SEAICE_LSR_ZEBRA and define it |
| 63 |
|
in lab_sea forward experiment (changes results -> updated) |
| 64 |
|
- update adjoint experiment lab_sea after cleaning up seaice_lsr_tridiagu/v |
| 65 |
|
o pkg/seaice: modify seaice_lsr.F in order to improve the gradient |
| 66 |
|
computations; for SEAICE_VECTORIZE_LSR |
| 67 |
|
- move the loops over which the tridiagonal solvers (seaice_lsr_tridiagu/v) |
| 68 |
|
sweep to the outside, |
| 69 |
|
- remove store directives and add "CADJ loop sequential" directives |
| 70 |
|
in analogy to model/src/solve_tridiagonal.F |
| 71 |
|
- replace many "#ifdef SEAICE_VECTORIZE_LSR_ZEBRA" by variable loop |
| 72 |
|
boundaries to yield more readable code. This has also the charming side |
| 73 |
|
effect that your can use SEAICE_VECTORIZE_LSR_ZEBRA without |
| 74 |
|
SEAICE_VECTORIZE_LSR (i.e. adjoint without recomputations in |
| 75 |
|
seaice_lsr_tridiagu/v now requires either of these flags, vectorization |
| 76 |
|
still requires SEAICE_VECTORIZE_LSR with SEAICE_VECTORIZE_LSR_ZEBRA as |
| 77 |
|
an additional option) |
| 78 |
|
The tridiagonal solvers are now completely analoguous to solve_tridiagonal.F |
| 79 |
|
but the adjoint gradients (which are affected by this change) still explode. |
| 80 |
|
o coupler (atm & ocn _compon_interf): |
| 81 |
|
- avoid using Exch2 IO mapping if not suitable. |
| 82 |
|
|
| 83 |
|
checkpoint65m (2015/06/15) |
| 84 |
|
o pkg/layers: |
| 85 |
|
- clean-up old parameters setting (from Sep 2012): |
| 86 |
|
+ remove old parameters setting from common block (declared locally); |
| 87 |
|
+ stop if any old parameter is mixed with new parameters; |
| 88 |
|
+ check for valid or consistent parameter setting (e.g., layers_name) |
| 89 |
|
- check that layers_bounds for Pot.Density are in the right range; |
| 90 |
|
this is supposed to catch un-updated setting after recently shifting |
| 91 |
|
potential density field "prho" by -1000 (now in "sigma" scale). |
| 92 |
|
o pkg/layers: |
| 93 |
|
- Added missing call to layers_fill for keeping track of tottend in |
| 94 |
|
pkg/diagnostics/diagnostics_fill_state.F |
| 95 |
|
o pkg/layers: |
| 96 |
|
- added CPP flux LAYERS_FINEGRID_DIAPYCNAL to disable fine-grid binning |
| 97 |
|
of diapycnal fluxes |
| 98 |
|
o pkg/layers: |
| 99 |
|
- reverted changes to interpolation to previous treatment; there |
| 100 |
|
was never actually a problem |
| 101 |
|
o pkg/layers: |
| 102 |
|
- fix interpolation of layers tracers in the presence of partial cells |
| 103 |
|
- add lots of new functionality to compute water mass transformation |
| 104 |
|
(enabled by CPP flag LAYERS_THERMODYNAMICS) |
| 105 |
|
- add new calls from main code (diags_oceanic_surf_flux.F, |
| 106 |
|
thermodynamics.F, impldiff.F) and gad to save the necessary fields |
| 107 |
|
for layers_thermodynamics calculations |
| 108 |
|
o pkg/ptracers: |
| 109 |
|
- add capability of stepping a passive tracer forward without advection |
| 110 |
|
by setting PTRACERS_advScheme to 0. |
| 111 |
|
o pkg/diagnostics: |
| 112 |
|
- add few (threads) _BARRIER around any update of "diag_pkgStatus". |
| 113 |
o pkg/layers: |
o pkg/layers: |
| 114 |
- shift potential density field "prho" by -1000 for density layer averaging so |
- shift potential density field "prho" by -1000 for density layer averaging so |
| 115 |
that it corresponds to usual "sigma" definition (i.e., rho(p=const)-1000); |
that it corresponds to usual "sigma" definition (i.e., rho(p=const)-1000); |
Parent Directory
| 
Revision Log
| 
Revision Graph
| 
Patch