4 |
Notes on tags used in MITgcmUV |
Notes on tags used in MITgcmUV |
5 |
============================== |
============================== |
6 |
|
|
7 |
|
o vectorize salt_plume package: |
8 |
|
- add extra code to salt_plume_frac and salt_plume_tendency_apply_s |
9 |
|
- this code computes plumebk everywhere (also over land), thus requires |
10 |
|
code to avoid divisions by zero and catch the case of kLev=Nr. |
11 |
|
- efficiency of salt_plume_frac is still sub-optimal |
12 |
|
o tools (genmake2): |
13 |
|
- fix makefile NAME specification (option: -makefile NAME) by adding |
14 |
|
"-f $MAKEFILE" in few places where it was missing including makedepend |
15 |
|
command (contribution from Paul). |
16 |
|
|
17 |
|
checkpoint63i (2012/01/24) |
18 |
|
o model/src: |
19 |
|
- add factorized versions of some EOS code to find_rho.F, |
20 |
|
define USE_FACTORIZED_EOS to turn it on (default for TARGET_NEC_SX) |
21 |
|
o model/src: |
22 |
|
- add celsius2K to namelist PARM01; remove unused recip_rhoNil from PARAMS.h; |
23 |
|
- change default celsius2K from 273.16 to 273.15 |
24 |
|
- to keep results unchanged, add back old value of celsius2K in verification |
25 |
|
exp. input*/data : aim.5l_cs, 1D_ocean_ice_column, lab_sea (fwd), |
26 |
|
global_ocean.cs32x15(seaice) & cpl_aim+ocn(input_atm). |
27 |
|
o model/src: add a little initialisation trick to find_rho.F and find_alpha.F |
28 |
|
to make TAF generate vectorizable code |
29 |
|
o pkg/seaice/seaice_evp.F: add code (copied from CICE) that avoids underflows. |
30 |
|
Turn on by defining SEAICE_EVP_ELIMINATE_UNDERFLOWS. Most compilers |
31 |
|
have flags that do this more efficiently. |
32 |
|
o pkg/diagnostics/diagstats_calc.F |
33 |
|
- for TARGET_NEC_SX fix the treatment of the scaling factor tmpFac=scaleFact |
34 |
|
o pkg/obcs (with exch2): |
35 |
|
- derive OB indices in overlap region from OB-InsideMask (in obcs_init_fixed.F) |
36 |
|
and remove code (in obcs_readparms.F) which was based on EXCH of indices |
37 |
|
(through buffer) and did not account for exch2 (e.g., CS-grid) topology. |
38 |
|
o pkg/exf and seaice exp: |
39 |
|
- change offline_exf_seaice.seaicetd test exp. to use default DIFF1 value (=0) |
40 |
|
so that switching off area & heff advection flags does not affect results. |
41 |
|
- change default EXF_LWDOWN_WITH_EMISSIVITY to #define in EXF_OPTIONS.h |
42 |
|
- keep exp. seaice_obcs & 1D_ocean_ice_column(ad) unchanged (using #undef |
43 |
|
EXF_LWDOWN_WITH_EMISSIVITY) and, using pkg/exf default emissivities, update |
44 |
|
results of verification experiments: 1D_ocean_ice_column(fwd), |
45 |
|
global_ocean.cs32x15.seaice(fwd+ad) & seaice_dynmix(ad), lab_sea (all 8) |
46 |
|
and offline_exf_seaice (all 4). |
47 |
|
o pkg/seaice: |
48 |
|
- remove array YNEG, RIVER, TMIX (replaced by local var) from common block; |
49 |
|
remove also areaNm1 & hEffNm1 when SEAICE_GROWTH_LEGACY is undef. |
50 |
|
- fix bug: areaNm1 & hEffNm1 are needed in seaice_growth (SEAICE_GROWTH_LEGACY) |
51 |
|
but were not filled when resp. SEAICEadvArea=F & SEAICEadvHeff=F and |
52 |
|
not multi-dim advection; |
53 |
|
- fix seaice_growth.F to compile with both SEAICE_GROWTH_LEGACY and |
54 |
|
SHORTWAVE_HEATING #undef. |
55 |
|
o optim: make Makefile more robust (?) |
56 |
|
o pkg/obcs: |
57 |
|
* fix treatment of obcs_read_pickup for TAF |
58 |
|
o pkg/shelfice: |
59 |
|
* add missing initialisation of num_shifwflx |
60 |
|
* new adjoint verification exp. for isomip |
61 |
|
* change Prandtl and Schmidt numbers from generic formulations |
62 |
|
(in terms of model diffus. and visc.) to separate runtime |
63 |
|
parameters, with Holland and Jenkins (1999) default values. |
64 |
|
o pkg/exf (interpolation): |
65 |
|
- extended input field (2 rows) near the N & S poles: |
66 |
|
* fill in with the symetric value (when even Nb of data in longitude) |
67 |
|
* add average value at the poles only for scalar quantities ; for vector |
68 |
|
component interpolation, skip the averaging and keep duplicated values, |
69 |
|
unless the 2 components have same location (uvInterp=T) in which case |
70 |
|
interpolate the 2 components together using S/R EXF_INTERP_UV. |
71 |
|
- new S/R EXF_INTERP_UV to interpolate the 2 components of a vector field |
72 |
|
together: this allows to account properly for local orientation when |
73 |
|
averaging at N & S pole. |
74 |
|
- add CPP option EXF_USE_OLD_INTERP_POLE to recover old results regarding |
75 |
|
exf-interpolation near N & S poles. Undef by default. |
76 |
|
- add debug check/print on 1rst iter if debugLevel >=2. |
77 |
|
o pkg/exf: |
78 |
|
- change masking flags definition (no longer "parameter"), put them in |
79 |
|
common block and set them in S/R EXF_INIT_FIXED. |
80 |
|
- change recognized masking flag from s,u,v to c,w,s since the mask |
81 |
|
to use (maskC,W,S) depends on position on model grid and has little to |
82 |
|
do with scalar/vector difference; fix masking if using P-coordinates. |
83 |
|
- fix masking of wind-stress (was wrong if interpolated or read on A-grid) |
84 |
|
and update results of exp. global_with_exf (x2). |
85 |
|
- fix setting of zenith-angle table when SHORTWAVE_HEATING is undef. |
86 |
|
o pkg/obcs |
87 |
|
- S/R obcs_calc_stevens: fix a small bug (gammat -> gammas where appropriate) |
88 |
|
o pkg/exf (interpolation): |
89 |
|
- assume periodicity in X only if input field cover full longitude range. |
90 |
|
- improve search for lat. index (supposed to be faster, in ~log2(ny) steps, |
91 |
|
and should vectorise). |
92 |
|
- fix input lat of the 2 added row (in case we provide N.pole data). |
93 |
|
|
94 |
|
checkpoint63h (2011/12/30) |
95 |
|
o pkg/seaice: |
96 |
|
- change declaration of UVM from _RL to _RS in order to get B-grid |
97 |
|
uIce,vIce monitor when _RS = real*4 |
98 |
|
o pkg/offline: |
99 |
|
- new S/R OFFLINE_RESET_PARMS for resetting main-model params (switch off flags); |
100 |
|
- compute vertical velocity if not read from file (switch on exactConserv |
101 |
|
& staggerTimeStep); |
102 |
|
- fix output of ptracer monitor iter & time for default PTRACERS_monitorFreq. |
103 |
|
o verification: |
104 |
|
- update grid-files (in tutorial_held_suarez_cs/input and fizhi-cs-32x32x40/input) |
105 |
|
with grid-angle generated using "utils/matlab/cs_grid/cubeCalcAngle.m": |
106 |
|
fix Pb in yG definition @ 2 missing corners (previously yG was averaged); |
107 |
|
- update output of exp. global_ocean.cs32x15.viscA4 after changing |
108 |
|
grid-file angles (used here for coriolis-3d); |
109 |
|
- remove (or gzip) few files not needed to run fizhi-cs experiments. |
110 |
|
o pkg/exf (vector interpolation): |
111 |
|
- rotate vector component toward local grid-axis orientation using grid-angles |
112 |
|
(stored in GRID.h) instead of locally computed rotation angle; keep old code |
113 |
|
available with exf option EXF_USE_OLD_VEC_ROTATION (undef by default). |
114 |
|
o pkg/seaice: |
115 |
|
- allow to compile with exf options ALLOW_ATM_TEMP or ALLOW_DOWNWARD_RADIATION |
116 |
|
undefined. Fix and remove the stop when ALLOW_RUNOFF is undefined. |
117 |
|
o pkg/PKG_OPTIONS.h files: |
118 |
|
- move #include "CPP_OPTIONS.h" just after #include "PACKAGES_CONFIG.h", |
119 |
|
outside of the #ifdef ALLOW_THISPKG / #endif block. |
120 |
|
- update verification/*/code/PKG_OPTIONS.h files. |
121 |
|
o add a verification experiment (offline_exf_seaice/input.seaice) that test |
122 |
|
the B-grid LSOR solver |
123 |
|
o pkg/seaice: sort out emissivity and long wave radiation step 2: |
124 |
|
- add emissivity for snow and ice as new run-time parameters |
125 |
|
(unfortunately) that default to the corresponding exf values; this |
126 |
|
changes results in most verification experiments so that for backward |
127 |
|
compatibility these parameter have to be set explicitly to inconsistent |
128 |
|
values (ugly); |
129 |
|
- the exf-flag EXF_LWDOWN_WITH_EMISSIVITY then gets rid off the hard-wired |
130 |
|
emissivities of 0.97 associated with lwdownloc in S/R seaice_solve4temp |
131 |
|
until we agree how to handle this bug properly |
132 |
|
o pkg/exf (interpolation): |
133 |
|
- add a set of run-time param ({inputfield}_interpMethod), one for |
134 |
|
each interpolated input field, to select the interpolation method |
135 |
|
with a value of zero switching off the interpolation; |
136 |
|
- print interpolation parameters to standard output; |
137 |
|
- refine exf_check.F accordingly ; stop if USE_NO_INTERP_RUNOFF is defined |
138 |
|
(to use instead runoff_interpMethod=0 in data.exf). |
139 |
|
- simplify default setting of interpolation parameters (for all fields, |
140 |
|
set position to match model grid cell-center position, leading to trivial |
141 |
|
interpolation in case of regular Lat-Lon grid) and fix for the case of |
142 |
|
non-uniform delY. |
143 |
|
o pkg/exf/exf_radiation.F: add the factor ocean_emissivity to lwdown, for |
144 |
|
within #ifdef EXF_LWDOWN_WITH_EMISSIVITY/#endif. Unless ocean_emissivity=1. |
145 |
|
this factor is absolutely required, because the lw exitance (= out-going |
146 |
|
long wave radiation) is emissivity*stefanBoltzmann*T^4 + rho*lwdown, |
147 |
|
where the reflectivity rho = 1-emissivity for conservation reasons. Once |
148 |
|
this is established, we can remove the CPP-Flag (if we want to). |
149 |
|
o pkg/seaice: |
150 |
|
- qualify a stop statement in S/R seaice_check: EXF_READ_EVAP and |
151 |
|
ALLOW_BULKFORMULAE are only required if SEAICE_EXTERNAL_FLUXES is defined |
152 |
|
- sort out emissivity and long wave radiation, step 1: make SEAICE_emissivity |
153 |
|
what the name implies (and not emissivity*BoltzmannConstant), this means |
154 |
|
the old default value of 5.5e-8 now means |
155 |
|
5.5e-8/5.670e-8 = 0.97001763668430343479 as ocean_emissivity in |
156 |
|
S/R exf_readparms |
157 |
|
- FV-discretization of the B-grid LSOR solver, turn on with |
158 |
|
SEAICE_LSRBNEW. This new solver also involves a slight re-organization |
159 |
|
of the calling routine dynsolver, so that B and C-grid code look a |
160 |
|
little more alike. |
161 |
|
- fix index bugs in ostres.F (B-grid code) |
162 |
|
o pkg/seaice/seaice_ocean_budget.F: change (unused) code to make it look better |
163 |
|
o pkg/seaice/seaice_readparms.F: make ocean_emissivity*stefanBoltzmann the |
164 |
|
default for SEAICE_emissivity, if useEXF; requires setting a wrong |
165 |
|
parameter value in global_ocean.cs32x15/input.seaice/data.seaice for |
166 |
|
backward compatiblity |
167 |
|
o model/src: use sigmaR in calc_ivdc |
168 |
|
|
169 |
|
checkpoint63g (2011/12/16) |
170 |
|
o pkg/seaice/seaice_growth.F: |
171 |
|
fix heat conservation of the coupled ocean-seaice system. |
172 |
|
when SEAICE_HEAT_CONSERV_FIX is defined, the mismatch of |
173 |
|
solid vs liquid water treatment is resolved via QNET. |
174 |
|
o pkg/seaice/seaice_growth.F: |
175 |
|
add missing term for sublimation in saltFlux, which was breaking conservation. |
176 |
|
o pkg/seaice/seaice_model.F: |
177 |
|
Remove this patholigical case treatment in evolution branch, |
178 |
|
where it is un-necessary and breaks conservation. |
179 |
|
Update global_ocean.cs32x15/results accordingly for |
180 |
|
output.seaice.txt output_adm.seaice.txt and output_adm.seaice_dynmix.txt |
181 |
|
o model/src: |
182 |
|
useRealFreshWaterFlux: remove masking of EmPmP (using maskInC) in |
183 |
|
external_forcing_surf.F and apply mask directly in S/R SOLVE_FOR_PRESSURE. |
184 |
|
o pkg/obcs: |
185 |
|
modify Stevens boundary conditions so that S/R obcs_calc_stevens computes |
186 |
|
tendency for temperature and salinity that is then applied to update |
187 |
|
theta and salt in S/R obcs_apply_ts; this requires proper masking of all |
188 |
|
horizontal advection and diffusion contributions on the boundary (in |
189 |
|
S/Rs gad_advection, gad_calc_rhs, gmredi_rtransport, gmredi_calc_diff), and |
190 |
|
removing the mask in S/R timestep_tracer. The advantage of this approach is |
191 |
|
that the updated tracer values include contributions from the surface |
192 |
|
forcing and vertical diffusion. |
193 |
|
This modification changes exp4.stevens because of vertical diffusion. |
194 |
|
o model/src: |
195 |
|
change non-curvilinear grid setting to also work with exch2 + blank tiles: |
196 |
|
- move delX,delY to new header file (SET_GRID.h) and adjust length to 1rst |
197 |
|
face dimensions. |
198 |
|
- new S/R INI_LOCAL_GRID to define model tile-local grid (using same units |
199 |
|
as delX,delY) and called from S/R INI_CARTESIAN_GRID, INI_CYLINDER_GRID |
200 |
|
and INI_SPHERICAL_POLAR_GRID. |
201 |
|
change test-experiment global_ocean.90x40x15: |
202 |
|
- split the domain in many more tiles (9x4 tiles of 10x10 each); |
203 |
|
- switch to exch2 and remove 1 blank-tile (but only for MPI test). |
204 |
|
- update results of both test (std & dwnslp). |
205 |
|
o pkg/diagnostics/diagstat_calc.F: fix problem with non-initialised halos in |
206 |
|
tmp-fields for TARGET_NEC_SX at the cost of extra loops and less vectorisation |
207 |
|
o model/src: |
208 |
|
- take sigma-coords related code out of update_etah.F into new S/R UPDATE_ETAWS |
209 |
|
and fix initialisation when using sigma-coords. |
210 |
|
o model/src: |
211 |
|
- improve Gael's changes from Jan 14, 2011 "addition of an 'update' of hfacc etc. |
212 |
|
to their current value at the beginning of forward_step.F": add a new run-time |
213 |
|
switch "doResetHFactors" (default=false) and use it to do (or not) the resetting. |
214 |
|
- set doResetHFactors=T in the following test exp: |
215 |
|
global_ocean.90x40x15/input (rStar) & input.dwnslp (linFS), |
216 |
|
global_ocean.cs32x15/input.seaice (rStar), hs94.cs-32x32x5/input.impIGW (linFS) |
217 |
|
and tutorial_global_oce_in_p (nonLinFS in p) |
218 |
|
o pkg/ecco |
219 |
|
- added IES cost term |
220 |
|
o verification/advect_xz |
221 |
|
- scale down the depth by 1/100 (from 200 km to just 2.km) |
222 |
|
- add secondary test using divergent flow, r-star and implicit vertical advection. |
223 |
|
- add note (check_conserve_TS.txt) and matlab script to check global |
224 |
|
conservation of T & S with NonLin-FreeSurf and Adams-Bashforth. |
225 |
|
o pkg/rbcs: |
226 |
|
- add missing relaxUFile & relaxVFile in namelist (in rbcs_readparms.F). |
227 |
|
o pkg/obcs (Stevens BCs): |
228 |
|
- fix various problems in the phase speed computations of the Stevens |
229 |
|
boundary conditions: |
230 |
|
- sign error for western and southern boundaries, |
231 |
|
- time tendencies were underestimated by factor of dTracerLev and some just |
232 |
|
zero (for multidim advection schemes) -> now estimate time tendencies |
233 |
|
from difference between previous (n) time level and the one before (n-1); |
234 |
|
this requires storing a field per tracer per boundary, also in the pickup |
235 |
|
files. |
236 |
|
- replace deltaTmom in cflMer/Zon computation with dTracerLev |
237 |
|
- set default phase velocity to cflMer/Zon |
238 |
|
- update verification experiment exp4.stevens |
239 |
|
o pkg/generic_advdiff (implicit_r): |
240 |
|
- fix implicit vertical advection conservation with AB and/or NonLin-FreeSurf |
241 |
|
(remove Tr*d/dz(w) in gad_implicit_r.F and add it in gad_calc_rhs.F); |
242 |
|
- fix implicit vertical diffusion or advection with NonLin FreeSurf |
243 |
|
(was using current recip_hFacC instead of future recip_hFacC): set future |
244 |
|
recip_hFac in thermodynamics.F and pass & use it in all vertical implicit |
245 |
|
solver S/R ; |
246 |
|
- fill diagnostics of vertical advective fluxes (computed from tendency) |
247 |
|
when using implicit vert. advection; |
248 |
|
- update output of exp that uses NonLin-FreeSurf & implicit vertical diffusion: |
249 |
|
global_ocean.90x40x15 (fwd+AD), global_ocean.cs32x15 (4 fwd +4 AD), |
250 |
|
tutorial_global_oce_in_p (fwd), tutorial_tracer_adjsens (AD) |
251 |
|
and cpl_aim+ocn (used in ocean); to be fixed: bottom_ctrl_5x5. |
252 |
|
o pkg/diagnostics: |
253 |
|
- add diags for T & S tendency which goes through Adams-Bashforth. |
254 |
|
|
255 |
|
checkpoint63f (2011/11/30) |
256 |
|
o pkg/obcs/obcs_calc_stevens.F: |
257 |
|
- join some loops to reduce number of if-statements |
258 |
|
- fix a bug in code for northern and southern boundaries. |
259 |
|
o pkg/seaice/seaice_init_varia.F |
260 |
|
- undo resetting UVM=0 that has slipped in unnoticed in r1.49 and |
261 |
|
caused zero ice-velocities for the B-grid code |
262 |
|
o pkg/seaice/seaice_init_fixed.F |
263 |
|
- fix a bug in the initialisation of the second metric terms |
264 |
|
k2AtC and k2AtZ: k2AtC was set to the value of k2AtZ and k2AtZ was 0. |
265 |
|
- this affects all seaice configurations with useCurviLinearGrid = .true. |
266 |
|
and SEAICEuseDynamics = .true. (default) and SEAICEuseMetricTerms |
267 |
|
= .true. (default) |
268 |
|
- the only verification experiment affected is global_ocean.cs32x15: |
269 |
|
icedyn, seaice, and seaice_dynmix (AD) |
270 |
|
o pkg/thsice/thsice_diagnostics_state.F |
271 |
|
- only for TARGET_NEC_SX: initialise all of tmpFld, so that overlap does |
272 |
|
not contain garbage that can lead to surprises in diagstats_calc.F |
273 |
|
o model/src: |
274 |
|
- move bi,bj loops inside INTEGR_CONTINUITY ; |
275 |
|
- call INTEGR_CONTINUITY directly from forward_step.F |
276 |
|
(previously called from momentum_correction_step.F) |
277 |
|
- call UPDATE_ETAH from integr_continuity.F and update initialise_varia.F, |
278 |
|
forward_step.F and pkg/ecco/the_main_loop.F |
279 |
|
o model/src/do_oceanic_phys.F: |
280 |
|
- reset addMass to zero only if useICEFRONT ; |
281 |
|
- separated k-loop for rhoInSitu calculation ; |
282 |
|
- add 2 diagnostics for wVel*delta^k(Rho) at constant pressure (= WdRHO_P) |
283 |
|
and at constant T,S (= WdRHOdP); |
284 |
|
NOTE: change diagnostics WRHOMASS : this is now simply the Vertical transport |
285 |
|
of In-Situ Density (Anomaly) (=wVel*rho^bar-k) (previously was Vertical |
286 |
|
transport of locally-referenced potential density); |
287 |
|
to recover old diag: (WRHOMASS)_old = (WRHOMASS)_new -1/2 * (WdRHOdP) |
288 |
|
o gluemnc: |
289 |
|
- replace ncrcat with ncks when adding variables to *.glob.nc |
290 |
|
- change the way variables without record dimensions are handled so that |
291 |
|
files that have variables with and without record variables can be glued: |
292 |
|
add a record dimension with a new unique name to be able to differentiate |
293 |
|
between variable with and without a-prior record dimensions. Remove |
294 |
|
this record dimension before adding the variable to the *.glob.nc file. |
295 |
|
This procedure automatically works also for grid.*.nc files that do not |
296 |
|
have any record dimension. |
297 |
|
o verification/cpl_aim+ocn: |
298 |
|
- generate new reference output on baudelaire (gfortran, mpich2) (previously |
299 |
|
from old aces-grid cluster, ao + ifort) and update 'run_cpl_test' script. |
300 |
|
o pkg/obcs/obcs_calc_stevens.F: |
301 |
|
- add code for northern and southern boundaries. |
302 |
|
- reorder some loops to save some operations and so that drFBar can be scalar |
303 |
|
o model/src: |
304 |
|
- re-arrange extrapolation calculation in Adams-Bashforth S/R (affects machine |
305 |
|
truncation) and add diagnostics for Adams-Bashforth tendency increment. |
306 |
|
- update test-exp. AD output: |
307 |
|
1D_ocean_ice_column, bottom_ctrl_5x5, lab_sea.noseaicedyn, obcs_ctrl; |
308 |
|
and FWD output: exp4 (x3), fizhi-cs-32x32x40, fizhi-cs-aqualev20, |
309 |
|
front_relax (x3), global_ocean.90x40x15 (x2), global_ocean.cs32x15 (x5), |
310 |
|
global_ocean_ebm, ideal_2D_oce, internal_wave, isomip (x3), lab_sea (x4), |
311 |
|
matrix_example, MLAdjust (x5), natl_box (x2), tutorial_advection_in_gyre, |
312 |
|
tutorial_global_oce_in_p, tutorial_global_oce_latlon and |
313 |
|
tutorial_plume_on_slope. |
314 |
|
o pkg/diagnostics/diagstats_calc.F: add special code for TARGET_NEC_SX |
315 |
|
that vectorises. This code uses 3 extra 2D fields, F90 array assignments |
316 |
|
and F90 intrinsic routines SUM, MINVAL, MAXVAL, but since TARGET_NEC_SX |
317 |
|
does not even come with an F77 compiler anymore, this should not be |
318 |
|
a limitation |
319 |
|
|
320 |
|
checkpoint63e (2011/11/06) |
321 |
|
o pkg/seaice adjoint of LSR: |
322 |
|
- substantially rework STORE dirs. |
323 |
|
- add new flag SEAICE_LSR_ADJOINT_ITER and parameter SOLV_MAX_FIXED |
324 |
|
this flag will do SOLV_MAX_ITERS but only store SOLV_MAX_FIXED fields |
325 |
|
and only perform SOLV_MAX_FIXED adjoint iters. |
326 |
|
If flag SEAICE_LSR_ADJOINT_ITER is undef-ined, the store in the iteration |
327 |
|
always goes to same record, meaning that the last forward field will be |
328 |
|
the one used in the adjoint - an approximation. |
329 |
|
- now produces non-zero LSR-related gradients, and acceptable gradients for |
330 |
|
An's zonal (x-z) test case (to be checked in) |
331 |
|
- lab_sea: produces acceptable gradients for lab_sea with forward/backward LSR |
332 |
|
(but these are screwed up by the thermodynamics in the first place); |
333 |
|
reduce from 4 to 3 timesteps to avoid t.d. issue as temporary solution; |
334 |
|
output_adm.txt also reflects changes to seaice_lsr |
335 |
|
- global_ocean.cs32x15: |
336 |
|
* the "default" input_ad.seaice has been re-defined to always exclude dynamics |
337 |
|
(output_adm.seaice.txt was updated) |
338 |
|
* new input_ad.seaice_dynmix tests a forward LSR / backward free-drift |
339 |
|
(output_adm.seaice_dynmix.txt) |
340 |
|
o test-exp. exp4+stevens: |
341 |
|
- switch off pTracers and nonHydrostatic (restart problems with NH); update output |
342 |
|
o pkg/seaice (lsr solver): |
343 |
|
- put coriolis term into FORCEX,Y (as in seaice_evp.F; previously in rhsU,V only); |
344 |
|
update output (change truncation) of FWD test exp lab_sea & global_ocean.cs32x15 |
345 |
|
- fix SEAICEuseFREEDRIFT for new OBCS version (ndef OBCS_UVICE_OLD) |
346 |
|
- new parameter "LSR_mixIniGuess" allows (if=2,4) to mix free-drift solution |
347 |
|
into initial guess for linearised solver; use weight ~ (1/err)^2,4. |
348 |
|
- compute (with new S/R) and print initial (+ free-driff sol. if available) |
349 |
|
and final residual of linearised system. |
350 |
|
o pkg/pp81: add diagnostics for vertical eddy viscosity and diffusivity |
351 |
|
(requires new subroutines pp81_init_fixed.F/pp81_diagnostics_init.F) |
352 |
|
o pkg/obcs: Stevens BCs (still only for east and west) |
353 |
|
- fix Stevens BCs by using intermediate velocities of previous |
354 |
|
time step as "baroclinic" contribution rather velocity of one grid cell |
355 |
|
inside the boundaries; more consistent but makes the entire treatment |
356 |
|
more complicated, requires new pickup files and two new S/Rs: |
357 |
|
obcs_copy_uv_n and obcs_save_uv_n |
358 |
|
keep old version with flag OBCS_STEVENS_USE_INTERIOR_VELOCITY |
359 |
|
(undefined by default) |
360 |
|
- silently move computation of some grid factors into separate loops to |
361 |
|
save some recomputations |
362 |
|
- add a hack for non-linear free surface |
363 |
|
- add a verification experiment to exp4 that tests this code |
364 |
|
o make gluemnc/xplodemnc work with (moron) non-GNU shell tools |
365 |
|
|
366 |
|
checkpoint63d (2011/10/25) |
367 |
|
o pkg/obcs + seaice-dynamics: |
368 |
|
- new implementation calls OBCS_APPLY_UVICE (in seaice_dynsolver.F) before |
369 |
|
solving for uIce,vIce and prevent SEAICE_LSR or SEAICE_EVP to modify OB |
370 |
|
values (consistent with an implicit method). |
371 |
|
- add option OBCS_UVICE_OLD (in OBCS_OPTIONS.h, defined by default) to |
372 |
|
return to older implementation (same results as before); note that none of |
373 |
|
the "experimental" OBCS_SEAICE options are implemented in the new version. |
374 |
|
- switch test exp. seaice_obcs to use new version (update output). |
375 |
|
o pkg/seaice: |
376 |
|
- add few stop (B-grid & CS-exch, B-grid & obcs, AUTODIFF & SEAICE_maskRHS) |
377 |
|
- add parenthesis in Water-Ice drag calculations (evp & lsr); |
378 |
|
- in seaice_lsr.F, simplify and fix AA3 setting in case sNx or sNy=1; |
379 |
|
and fix case SEAICE_waterDrag=0 (but unstable if no or little seaice) |
380 |
|
- update output of test exp lab_sea that uses evp or lsr (fwd+ad) |
381 |
|
and global_ocean.cs32x15 (icedyn & seaice fwd+ad). |
382 |
|
o pkg/layers: |
383 |
|
- Include potential density as new coordinate (Thanks to David Munday) |
384 |
|
o pkg/dic: |
385 |
|
- Remove subroutine CALC_PCO2_APPROX_CO3 from carbon_chem.F and add |
386 |
|
carbonate computation/output to CALC_PCO2_APPROX |
387 |
|
o pkg/generic_advdiff |
388 |
|
- make TAF generate code that vectorizes: |
389 |
|
gad_dst3fl_adv_*.F, gad_u3c4_impl_r.F |
390 |
|
o model/src: |
391 |
|
- improve vectorization also for adjoint code of |
392 |
|
solve_tridiagonal and solve_pentadiagonal. As this requires up to |
393 |
|
5 new 3D-fields, this code is only turned on for the combination of |
394 |
|
ALLOW_AUTODIFF_TAMC and TARGET_NEC_SX. Also remove error message |
395 |
|
for singular matrix check (only for TARGET_NEC_SX). |
396 |
|
- remove hard-wiring of CPP-flag ALLOW_SOLVERS_KLOOPINSIDE to |
397 |
|
ALLOW_AUTODIFF. This flag now needs to be set outside of |
398 |
|
solve_tri/pentadiagonal.F |
399 |
|
o pkg/exf: |
400 |
|
- fix previous modif in exf_readparms.F (revision 1.66: treatment of |
401 |
|
optional namelist EXF_NML_OBCS). |
402 |
|
- add many debug-print (to trace down where pkg/cal error msg comes from) |
403 |
|
o pkg/mom_vecinv/mom_vi_hdissip.F |
404 |
|
- rearrange loops so that vectorization is improved and in particular |
405 |
|
the adjoint routine vectorizes. The code is a little longer now, but I |
406 |
|
think easier to read. |
407 |
|
Note: CPP-flag MOM_VI_ORIGINAL_VISCA4 is not tested. |
408 |
|
o pkg/seaice: |
409 |
|
- allow to specify initial seaice velocity from files (uIceFile,vIceFile) |
410 |
|
o pkg/exch2: |
411 |
|
- alternative code (without tile x tile matrix) for cumulated (global) sum; |
412 |
|
used as default; keep using matrix version in adjustment.cs-32x32x1. |
413 |
|
o tools/build_options: |
414 |
|
- after merging some MPI and/or multi-threads specific optfiles so that |
415 |
|
the modified default optfile support both built (using genmake2 -mpi |
416 |
|
for the MPI built, using genmake2 -omp for OpenMP built), |
417 |
|
move now obsolete optfiles to tools/build_options/unsupported/ |
418 |
|
(add copy of email sent to support list in doc/optfiles_changes.txt) |
419 |
|
|
420 |
|
checkpoint63c (2011/09/28) |
421 |
|
o pkg/ebm: |
422 |
|
- fix ebm_zonalmean.F for several procs in Ydir (nPy > 1); update output of |
423 |
|
test exp. global_ocean_ebm (affects machine truncation, 12 digits for cg2d) |
424 |
|
o pkg/seaice (SItracer pickup) |
425 |
|
* Simple code for pickup of SItracer fields (#define ALLOW_SITRACER) |
426 |
|
To be done: |
427 |
|
- thorough testing of missing fields in S/R seaice_read_pickup |
428 |
|
- thorough matching between generic tracers and specific fields used |
429 |
|
(e.g. volume-weighted age, area-weighted age, salinity, ...) |
430 |
|
o pkg/grdchk: |
431 |
|
- all procs calculate and print finite-difference gradient (for testreport) |
432 |
|
- change description output used by testreport; |
433 |
|
- AD testing: also compare and report Finite-difference gradient; |
434 |
|
=> update all output_adm*.txt |
435 |
|
o test exp. global_ocean_ebm: |
436 |
|
- switch back useRealFreshWaterFlux (no problem anymore with run-off); |
437 |
|
and turn on exactConserve; update (fwd) output. |
438 |
|
o pkg/seaice (LSR): |
439 |
|
- print to STDOUT number of solver iterations at SEAICE_monFreq interval. |
440 |
|
- modify solver to deal cleanly with u/v switch at face edges (CS-grid): |
441 |
|
* identical results with non-CS topology (but save up to half of EXCH calls) |
442 |
|
* no much changes in cs-32 verification tests since no ice at face edges |
443 |
|
(differences due to convergence criteria, since nb of iterations changed) |
444 |
|
* expect some differences on lat-lon-cap grid. |
445 |
|
update FWD results of global_ocean.cs32x15 test exp. (icedyn & seaice) |
446 |
|
o pkg/exch2: |
447 |
|
- add new CPP option to disable Cumul-Sum matrix (to save storage when using |
448 |
|
many tiles); until alternative coding is added, just add a stop. |
449 |
|
o pkg/kpp: |
450 |
|
- move ptracer/salt correction to KPP diffusivity from calc_3d_diffusivity.F |
451 |
|
to S/R KPP_CALC_DIFF_PTR (+ add tracer-index to argument list). |
452 |
|
o optfiles: |
453 |
|
- merged few MPI specific (*+mpi*) and multi-threads specific (*+mth) optfiles |
454 |
|
into corresponding standard one, which now support different build: |
455 |
|
(when using "genmake2 -mpi" for MPI built, "genmake2 -omp" for OpenMP built) |
456 |
|
- add "darwin_amd64_gfortran" optfile (initially from Constantinos, tested by Gael) |
457 |
|
o pkg/obcs (seaice): |
458 |
|
- add missing EXCH for uIce,vIce after applying OBCS; |
459 |
|
- remove S/R EXF_SWAPFFIELDS_X/Y , disable S/R EXF_SET_OBCS_X/Y by adding |
460 |
|
argument = number of levels to S/R EXF_SWAPFFIELDS_XZ/YZ and S/R |
461 |
|
EXF_SET_OBCS_XZ/YZ ; update obcs_exf_load.F; |
462 |
|
- fix EXF_SET_OBCS_XZ/YZ for multi-threaded (assuming loaded arrays are shared) |
463 |
|
- verification exp. seaice_obcs: split domain in 2 tiles ; add SIZE.h_mpi |
464 |
|
and eedata.mth for MPI and mutlti-threaded testing. |
465 |
|
o pkg/flt: |
466 |
|
- make FLT EXCH working with multi-threads (and MPI + multi-threads); |
467 |
|
- use buffer for floats IO (trajectories & profiles); size set in FLT_SIZE.h; |
468 |
|
- fix pickup and trajectories + profiles output with multi-threads; |
469 |
|
- add parameter to select group of fields to write along trajectory and |
470 |
|
profile (default unchanged); |
471 |
|
- only call PORT_RAND* function when flt_noise <> 0; |
472 |
|
- add multi-threaded test for verification exp. flt_example. |
473 |
|
o eesupp/src: |
474 |
|
- fix S/R ALL_PROC_DIE for MPI+multi-threads. |
475 |
|
- USE_OMP_THREADING: check that number of threads (from eedata) matches |
476 |
|
actual number (from env.var. setting). |
477 |
|
o pkg/ebm: |
478 |
|
- fix multi-threading in ebm_area_t.F ; clean-up EBM.h header file. |
479 |
|
- add multi-threaded test for verification exp. global_ocean_ebm |
480 |
|
|
481 |
|
checkpoint63b (2011/08/29) |
482 |
|
o pkg/ebm: |
483 |
|
- fix sign of runoff; |
484 |
|
- update directly forcing fields (Qnet,EmPmR) to use standard |
485 |
|
external_forcing_surf.F (remove ebm_forcing_surf.F); |
486 |
|
- avoid loading forcing fields a 2nd time (1rst time from ini_forcing.F); |
487 |
|
- set either tauThetaClimRelax (in data) or tauThetaZonRelax (in data.ebm) |
488 |
|
and apply relaxation towards corresponding SST field (2-D map or ZonAver). |
489 |
|
- update global_ocean_ebm parameters file data and update output (fwd+ad). |
490 |
|
o verification/matrix_example: |
491 |
|
- split domain in 8 tiles (2x4); add SIZE.h_mpi and eedata.mth for |
492 |
|
MPI and mutlti-threaded testing. |
493 |
|
o pkg/matrix: |
494 |
|
- change name of S/R MATRIX_INIT to S/R MATRIX_INIT_VARIA ; |
495 |
|
- fix for multi-tiles (add bi,bj indices to exp/impMatrixCounter); |
496 |
|
- fix for multi-threaded (remove DATA statement); |
497 |
|
- use standard DIFFERENT_MULTIPLE function; |
498 |
|
- add MATRIX_OPTIONS.h |
499 |
|
o verification/matrix_example: |
500 |
|
To allow to split the domain in several tiles, change the domain size |
501 |
|
from 31x31 to 32x32 by adding 1 row (on Southern side) + 1 column (on |
502 |
|
Western side) of land points. |
503 |
|
o (*_)write_pickup: |
504 |
|
- fix few bugs in mypackage_read/write_pickup.F |
505 |
|
- in MDS_WR_METAFILES calls, replace "myTime" by a local copy into |
506 |
|
array of length 1 (was caught by compiler syntax-check). |
507 |
|
o pkg/diagnostics (velocity potential solver): |
508 |
|
- fix initialisation of main diagonal matrix (for overlap values) |
509 |
|
o tools/build_options: |
510 |
|
- towards single optfile for MPI and non-MPI built: merge few +mpi optfile |
511 |
|
into previously non MPI one (linux_am64 pgf77,open64,pathf90,ifort11) |
512 |
|
- linux_amd64_gfortran: merge +mpi_generic version into standard optfile: |
513 |
|
* move IEEE flags to DEVEL section; take IEEE flags from +mpi_generic ; |
514 |
|
* FAST case: for now, keep -O3 with ini_masks_etc.F in NOOPTFILES list; |
515 |
|
and use "-devel" testreport option for standard testing on baudelaire. |
516 |
|
o pkg/grdchk: |
517 |
|
- Assume that all gradient-check position are within the same tile |
518 |
|
corresponding to global indices (iGloTile,jGloTile) and that |
519 |
|
positions (i,j indices or nbeg) are local to this tile; |
520 |
|
- set tile local-indices and proc. number according to tile global-indices; |
521 |
|
- remove grdchkwhichproc from namelist (retired param); |
522 |
|
- send adjoint gradient to all procs so that all procs print it to STDOUT; |
523 |
|
- switch on MPI test for global_ocean.90x40x15 AD test experiment. |
524 |
|
o adding package bbl (Bottom Boundary Layer) |
525 |
|
description in MITgcm/pkg/bbl/bbl_description.tex |
526 |
|
example/test experiment in MITgcm_contrib/bbl |
527 |
|
o adjoint verifications |
528 |
|
- changed 5 experiments to multi-tiles |
529 |
|
* global_ocean.90x40x15 |
530 |
|
* obcs_ctrl |
531 |
|
* offline_exf_seaice |
532 |
|
* tutorial_dic_adjoffline |
533 |
|
* tutorial_tracer_adjsens |
534 |
|
changed reference output for those that crossed tile boundaries |
535 |
|
(changed choices of gradient check positions) |
536 |
|
- with same tiling, switch on MPI test for the last 4 (add SIZE.h_mpi) |
537 |
|
- temporary bug fix in grdchk_loc.F for obcs |
538 |
o genmake2 & testreport |
o genmake2 & testreport |
539 |
- add target "obj" (just for OBJFILES files) to Makefile; |
- add target "obj" (just for OBJFILES files) to Makefile; |
540 |
- add option to specify FEXTRAFLAGS on command line. |
- add option to specify FEXTRAFLAGS on command line. |