| 4 |
Notes on tags used in MITgcmUV |
Notes on tags used in MITgcmUV |
| 5 |
============================== |
============================== |
| 6 |
|
|
| 7 |
|
o various mnc cleanups and improvements: |
| 8 |
|
+ shrink lookup tables by factor of ~4 |
| 9 |
|
+ better error reporting when running out of lookup space |
| 10 |
|
+ able to handle longer path/file names (up to 500 chars) |
| 11 |
|
o adjoint: |
| 12 |
|
- Another overhaul of store dirs. for NLFS to eliminate "hidden" recomputations. |
| 13 |
|
- TBD: "hidden" mom_vecinv recomp. in dynamics |
| 14 |
|
o change forcing_In_AB to affects both T,S forcing and Momentum forcing |
| 15 |
|
(allow to differentiate between forcing components using new integer flags: |
| 16 |
|
momForcingOutAB=1/0 & tracForcingOutAB=1/0) |
| 17 |
|
and add new flag to put Dissipation tendency out of Adams-Bashforth. |
| 18 |
|
o ecco |
| 19 |
|
- enable trend removal from SSH anomaly obs. |
| 20 |
|
o adjoint |
| 21 |
|
- a fix for PmEpR when #undef EXACT_CONSERV |
| 22 |
|
- fix of "hidden" recomp. in thermodynamics (no TAF warnings) |
| 23 |
|
- try to fix "hidden" recomp. in forward_step for NLFS |
| 24 |
|
(NLFS seems broken sometimes after I checked it in) |
| 25 |
|
o seaice: |
| 26 |
|
- add c-grid version of the dynamic solver: |
| 27 |
|
+ three new routines that are the c-grid substitute lsr, dynsolver |
| 28 |
|
and ostres |
| 29 |
|
+ put a few fields that were local to dynsolver into global common |
| 30 |
|
blocks, so that I can move the computation of stresses etc into |
| 31 |
|
seaice_lsr (saves coding but may break the adjoint). |
| 32 |
|
- replace more hardwired parameters by runtime parameters |
| 33 |
|
- add ice masks that mask the rhs of the implicit solvers where there |
| 34 |
|
is no ice (commented out in seaice_dynsolver, because i am not sure |
| 35 |
|
if this works properly), eventually this should replace the clipping |
| 36 |
|
of ice velocities in seaice_dynsolver to +/-40cm/s. |
| 37 |
|
o remove the BUILD_INFO.h include from CPP_EEOPTIONS.h and explicitly |
| 38 |
|
include it only where its needed |
| 39 |
|
o move balancing of fluxes before the call of external_forcing_surf |
| 40 |
|
o seaice: |
| 41 |
|
- correct units of turning angles (now they are really in degrees) |
| 42 |
|
- move computation of GWATX/Y out of main routine into dynsolver |
| 43 |
|
- initialize GWATX/Y in seaice_init |
| 44 |
|
o Apply mask on Fresh-Water flux (needed for SSH forcing). |
| 45 |
|
update output.txt of exp. global_ocean_ebm & cpl_aim+ocn (affects cg2d residual) |
| 46 |
|
o seaice: add runtime flags for turning angles (default to zero) |
| 47 |
|
o Adding runtime parameter to exf/, ctrl/, to remove on-the-fly global mean |
| 48 |
|
and trend from exf forcing and ctrl control adjustments |
| 49 |
|
o fix variable aliasing in gad_calc_rhs that cannot be handled by TAF |
| 50 |
|
(since checkpoint57x_post) |
| 51 |
|
o change the default rotation period to 86164.s = the Earth rotation |
| 52 |
|
o 1rst implementation of Implicit IGW using the 3-D solver (use3Dsolver=T) |
| 53 |
|
and based on a reference stratification. Tested within hs94.cs-32x32x5. |
| 54 |
|
o seaice advection: add code for choosing between the default (2nd order |
| 55 |
|
central differences) and multiDim advection schemes provided by |
| 56 |
|
pkg/generic_advdiff |
| 57 |
|
o seaice: |
| 58 |
|
- removed all references to SEAICE_GRID.h fields from |
| 59 |
|
dynsolver, advect, diffus, lsr, without changing the results of lab_sea |
| 60 |
|
(cubed-sphere configuration will change a little) |
| 61 |
|
- grid computations in seaice_init are obsolete now. |
| 62 |
|
- cleaned up code a little so that is easier to read (hopefully). |
| 63 |
|
- added new (better?) discretizations of default advection and diffusion |
| 64 |
|
(turn on with SEAICEuseFluxForm) |
| 65 |
|
- replace 1 in 3d global fields by kSurface in growth.F |
| 66 |
|
|
| 67 |
|
checkpoint58a_post |
| 68 |
|
o seaice: |
| 69 |
|
- bug fix in advect.F |
| 70 |
|
o shelfice: |
| 71 |
|
- separate shelfice load anomaly from pload (breaks with time |
| 72 |
|
dependent forcing), => introduce constant field shelficeLoadAnomaly. |
| 73 |
|
Its default is 0., but it may be computed more cleverly from (unknown) |
| 74 |
|
t- and s-profiles (tRef, sRef) and actual EOS. For now this has to be |
| 75 |
|
done offline. A good approximation of the pressure load anomaly is |
| 76 |
|
necessary to avoid large initial adjustment processes underneath |
| 77 |
|
deep-reaching shelfice. |
| 78 |
|
- small fix in ISOMIP shelfice_thermodynamics |
| 79 |
|
- add more sophisticated thermodynamics of Hellmer (1989), this is |
| 80 |
|
now the default |
| 81 |
|
- give a few constants more intuitive names |
| 82 |
|
o add code to balance EmPmP and Qnet at the end of do_ocean_physics. Useful |
| 83 |
|
if bulk formulae are used in long integration (especially EmPmR). Turn |
| 84 |
|
on with balanceEmPmR = .true. or balanceQnet = .true. in data, PARM01 |
| 85 |
|
if balancePrintMean, the imbalance that is substracted is printed to |
| 86 |
|
STDOUT. |
| 87 |
|
o fix bug (typo) in obcs_apply_eta.F |
| 88 |
|
o add a new package shelfice which enables us to model the circulation |
| 89 |
|
underneath shelf-ice with draughts of hundreds of meters. |
| 90 |
|
- works |
| 91 |
|
- to do: + include improved thermodynamics |
| 92 |
|
+ check whether kSurfC (from model/inc/SURFACE.h) can be used |
| 93 |
|
instead of new field kTopC |
| 94 |
|
+ can we use topoFile, Ro_surf and existing code in ini_depth.F |
| 95 |
|
instead of new field R_shelfIce and new code in ini_depth.F |
| 96 |
|
and ini_mask_etc.F? |
| 97 |
|
+ adjust vertical mixing schemes? |
| 98 |
|
o seaice pkg: |
| 99 |
|
- add a cpp-flag to include a simple formula for a salinity dependent |
| 100 |
|
freezing point of sea water at the surface |
| 101 |
|
- change the contents of SEAICE_TEST_ICE_STRESS_1 so that zonal and |
| 102 |
|
meridional stresses are now different (and hopefully correct) |
| 103 |
|
o AIM pkg: change threshold on large-scale condensation to conserve Heat. |
| 104 |
|
update results of aim.5l_LatLon test ; fix title & units of AIM diagnostics |
| 105 |
|
and change 3 diagnostics names (OSR,OSRCLR,KM -> TSR,TSRCLR,DRAG) |
| 106 |
|
o pkg/diagnostics: add options and parameters for regional-statistics |
| 107 |
|
diagnostics ; define regions by reading a region-mask from a file |
| 108 |
|
(might add other options later) |
| 109 |
|
o pkg/bulk_force: |
| 110 |
|
- add option and parameters to use AIM surface-flux formulae. |
| 111 |
|
- Change loading part: S/R BULKF_FIELDS_LOAD only take care of bulkf_fields ; |
| 112 |
|
others forcing fields (defined in PARM05, parameter file "data") are loaded |
| 113 |
|
from S/R EXTERNAL_FIELDS_LOAD, as usually (with useBulkforce = T or F). |
| 114 |
|
- initialise all bulkf_fields in bulkf_init.F ; do in-lining of exf_bulkcdn.F |
| 115 |
|
- use the right EXCH call for uwind,vwind (to work on CS-grid). |
| 116 |
|
- re-arrange header files (move parameters from BULKF.h to BULKF_PARAMS.h) |
| 117 |
|
and parameters (note: calcWindStress replaces .NOT.readwindstress). |
| 118 |
|
- cleaner interface with pkg/thSIce (new S/R THSICE_GET_PRECIP). |
| 119 |
|
|
| 120 |
|
checkpoint58 |
| 121 |
|
o add "--embed-source" option to genmake2 which, when the supporting |
| 122 |
|
tools can be compiled (as determined by genmake2) will turn on the |
| 123 |
|
embed_files package which then embeds the entire MITgcm source code |
| 124 |
|
(*.[fFhc] + Makefile) used for the build within the executable |
| 125 |
|
- requested by CNH but off by default |
| 126 |
|
- adds a paltry <9MB to the mitgcmuv executable in most cases |
| 127 |
|
- only writes the output when useEMBEDSRC in eedata is true |
| 128 |
|
o clean up 3-dim. relaxation codes: |
| 129 |
|
* Fix I/O inconsistency in pkg/rbcs: replace precFloat32 by readBinaryPrec |
| 130 |
|
* Remove 3-dim. relaxation code from pkg/exf (now use only pkg/rbcs) |
| 131 |
|
* Thanks to Tom Haine for testing! |
| 132 |
|
|
| 133 |
|
checkpoint57z_post |
| 134 |
|
o make CG3D solver compatible with Free-surface at k > 1 (p-coordinate): |
| 135 |
|
affects output (@ truncation level) of exp. plume_on_slope ; updated. |
| 136 |
|
o fix the pickup behavior for obcs in combination with ptracers |
| 137 |
|
o fix a bug in ptracers_external_fields_load |
| 138 |
|
o no AB-2 at the first iteration for Gw (consistent with Gu,Gv,Gt,Gs ...) |
| 139 |
|
affects results of NH experiments (but all our NH tests have Gw[it=1]==0 !!) |
| 140 |
|
o transfered surface NH pressure to eta field (if exactConserv). |
| 141 |
|
o gmredi: fix bugs in slope tapering ldd97 (depthZ had the wrong sign) |
| 142 |
|
and update lab_sea output.txt file. |
| 143 |
|
o completing 3-dim. relaxation code: additional term |
| 144 |
|
1/tau*( theta(i,j,k) - thetaStar(i,j,k) ) |
| 145 |
|
in S/R EXTERNAL_FORCING_T (file external_forcing.F) |
| 146 |
|
* At compile time, in "EXF_OPTIONS.h" set |
| 147 |
|
#define ALLOW_CLIMTEMP_RELAXATION |
| 148 |
|
#define ALLOW_CLIMSALT_RELAXATION |
| 149 |
|
* At runtime in "data", under &PARM03 set (e.g., units in seconds) |
| 150 |
|
tauThetaClimRelax3Dim = 15552000.0, |
| 151 |
|
tauSaltClimRelax3Dim = 15552000.0, |
| 152 |
|
* 3 cases: |
| 153 |
|
1) Use a constant-in-time 3-dim. field |
| 154 |
|
2) Provide 12 3-dim. fields corresponding to Jan., ... , Dec. |
| 155 |
|
3) Real-time relax. to time-evolving prescribed field |
| 156 |
|
o Adding Laplacian-type smoothness constraint on |
| 157 |
|
initial and boundary (forcing) controls (ic, bc) |
| 158 |
|
* enable at compile-time via |
| 159 |
|
ALLOW_SMOOTH_IC_COST_CONTRIBUTION, ALLOW_SMOOTH_BC_COST_CONTRIBUTION |
| 160 |
|
* enable at run-time via mult_smooth_ic = 1, mult_smooth_bc = 1 |
| 161 |
|
* Absolute scale somewhat unclear so far |
| 162 |
|
* no area weighting was included since no other cost term does so far |
| 163 |
|
|
| 164 |
|
checkpoint57y_post |
| 165 |
|
First step for a NLFS adjoint |
| 166 |
|
o initially suppress rStar (new flag DISABLE_RSTAR_CODE) |
| 167 |
|
o new init. routines for calc_r_star, calc_surf_dr |
| 168 |
|
o still need to deal with ini_masks_etc |
| 169 |
|
o testreport seemed happy |
| 170 |
|
checkpoint57y_pre |
| 171 |
|
|
| 172 |
|
o obcs: always reset wVel & eta (to zero or prescribe it) at the OBs. |
| 173 |
|
o add new package rbcs (for relaxing ptracers at boundaries/depth/surface etc) |
| 174 |
|
o add new offset feature to offline pkg (plus minor cleanup) |
| 175 |
|
o cfc fixes for final year of integration |
| 176 |
|
o clean up GCHEM_OPTIONS.h/DIC_OPTIONS.h (all options should |
| 177 |
|
be in GCHEM_OPTIONS). |
| 178 |
|
|
| 179 |
|
checkpoint57x_post |
| 180 |
|
o fix testreport Pb in multiple test experiment. |
| 181 |
|
o cleanup our namelist syntax in verification: |
| 182 |
|
- make it more F95 standards compliant (some work still needed) |
| 183 |
|
- allows the vast majority of our tests to work with gfortran |
| 184 |
|
o Non-Hydrostatic: |
| 185 |
|
- put all NH variables (formerly in DYNVARS.h & GW.h) in NH_VARS.h |
| 186 |
|
- fix exchange calls for CS-grid using the new EXCH_S3D_RL. |
| 187 |
|
o Allow to apply AB on T,S rather than on AB(gT,gS): |
| 188 |
|
- implemented within #ifdef ALLOW_ADAMSBASHFORTH_3 |
| 189 |
|
- use the same arrays (gtNm,gsNm) to hold tracer field at previous |
| 190 |
|
time-steps (if AB(T,S)) and tendencies (if AB(gT,gS)). |
| 191 |
|
- need a "clever" read_pickup to restart from the other AB formulation |
| 192 |
|
o move local commom bloc /GlobalLo/ (used for singleCpuIO) in EESUPPORT.h |
| 193 |
|
o remove some unused variables (reduces number of compiler warnings) |
| 194 |
|
o adjoint |
| 195 |
|
* Disable useVariableViscosity for adjoint for time being |
| 196 |
|
(there are a number of extensive recomputations) |
| 197 |
|
* Initialize certain variables needed for adjoint |
| 198 |
|
|
| 199 |
|
checkpoint57w_post |
| 200 |
|
o pkg/mdsio: use MAX_LEN_FNAM (instead of hard coded 80) for file-names |
| 201 |
|
fix the singleCpuIO RS version of MDSWRITEFIELD |
| 202 |
|
and add a WARNING if using globalFiles=T with multi-processors. |
| 203 |
|
o write grid-file(s) in only one format (MDS or MNC): default=MNC (if |
| 204 |
|
using MNC); can force MDS output by setting writegrid_mnc=F in data.mnc |
| 205 |
|
o tracer advection scheme: |
| 206 |
|
add DST-2 (=Lax-Wendroff) & 1rst order upwind schemes (the way to get |
| 207 |
|
those 2 schemes before, changing GAD_FLUX_LIMITER.h + to recompile, |
| 208 |
|
was not very convenient). |
| 209 |
|
add implicit version for DST-3 & DST3_Flux_Limit vertical advection; |
| 210 |
|
(further tests needed for the implicit DST & Flux-Limiter schemes). |
| 211 |
|
o PTRACERS: added namelist parameter PTRACERS_ref in analogy to tRef |
| 212 |
|
and sRef for convenience, handy, when you want to initialize |
| 213 |
|
ptracers in a large domain with a constant non-zero value. Also, |
| 214 |
|
untangle ptracers_init.F a little. |
| 215 |
|
|
| 216 |
|
checkpoint57v_post |
| 217 |
|
o change dome exp. to use variable viscosity in side-drag terms. |
| 218 |
|
o remove all the files from the cfc_offline test since they are now |
| 219 |
|
incorporated into tutorial_cfc_offline which has been producing |
| 220 |
|
identical results for a few days of testing |
| 221 |
|
o add new parameter: sideDragFactor to enable half-slip-side BC |
| 222 |
|
and change side-drag S/R to use viscosity from mom_calc_visc |
| 223 |
|
(can switch back by setting sideDragFactor=0.); affects results at |
| 224 |
|
truncation level; regenerate output for exp.: global_ocean.cs32x10(x3), |
| 225 |
|
global_ocean_ebm, global_ocean_pressure, matrix_example & MLAdjust(x5). |
| 226 |
|
o vorticity & strain are now computed without masking, to account for |
| 227 |
|
no-slip-side BC in variable viscosity calculation ; mask is applied |
| 228 |
|
anyway later on, before computing advection/coriolis terms. |
| 229 |
|
o modify exp4 to test new obcs support for reading from external files |
| 230 |
|
and for passive tracers |
| 231 |
|
o fix behavior of ptracers_monitor when monitor output is written to a |
| 232 |
|
netcdf-file; still not independent of monitor.F (writes in the same |
| 233 |
|
file). |
| 234 |
|
o change default value of tRef (put a uniform, meaningful value for ocean |
| 235 |
|
& atmos) and initialize sRef (was missing, as Martin mentioned). |
| 236 |
|
o Redefine viscC?Leith*. Change the coefficients of Leith viscosity |
| 237 |
|
to be proportional to enstrophy dissipation length. Thus, |
| 238 |
|
viscC?leith?_new=pi*(viscC?leith?_old)^(1/3) |
| 239 |
|
This is now consistent with Smagorinsky and with the manual. |
| 240 |
|
|
| 241 |
|
checkpoint57u_post |
| 242 |
|
o OBCS and PTRACERS: add open boundary support for passive tracers |
| 243 |
|
- either use homogenous (pseudo) v.Neumann conditions or prescribe |
| 244 |
|
OB-values from file; this is not different from the way theta and salinity |
| 245 |
|
are treated |
| 246 |
|
- however, Orlanski-radiation conditions are not supported, and the model |
| 247 |
|
will stop if you use pTracers and Orlanski at the same time. |
| 248 |
|
- beefed up the rountine obcs_external_fields_load: now only those open |
| 249 |
|
boundary values are overwritten with values from files for which there |
| 250 |
|
are really files, otherwise the OB-fields remain untouched. This makes |
| 251 |
|
it possible to use different OBs at different ends of the domain (as |
| 252 |
|
with EXF) |
| 253 |
|
o reduce number of exch calls in shap_filt (s2). |
| 254 |
o CS-grid: fix divergence contribution to Leith viscosity. Affects (@ |
o CS-grid: fix divergence contribution to Leith viscosity. Affects (@ |
| 255 |
truncation level) 4/5 MLAdjust test results (11&12 digits match). |
truncation level) 4/5 MLAdjust test results (11&12 digits match). |
| 256 |
|
|
Parent Directory
| 
Revision Log
| 
Revision Graph
| 
Patch