| 1 |
C $Header: /u/gcmpack/MITgcm/model/src/calc_gw.F,v 1.27 2006/06/20 20:57:37 baylor Exp $ |
| 2 |
C !DESCRIPTION: \bv |
| 3 |
C $Name: $ |
| 4 |
|
| 5 |
#include "CPP_OPTIONS.h" |
| 6 |
|
| 7 |
CBOP |
| 8 |
C !ROUTINE: CALC_GW |
| 9 |
C !INTERFACE: |
| 10 |
SUBROUTINE CALC_GW( |
| 11 |
I bi, bj, KappaRU, KappaRV, |
| 12 |
I myTime, myIter, myThid ) |
| 13 |
C !DESCRIPTION: \bv |
| 14 |
C *==========================================================* |
| 15 |
C | S/R CALC_GW |
| 16 |
C | o Calculate vert. velocity tendency terms ( NH, QH only ) |
| 17 |
C *==========================================================* |
| 18 |
C | In NH and QH, the vertical momentum tendency must be |
| 19 |
C | calculated explicitly and included as a source term |
| 20 |
C | for a 3d pressure eqn. Calculate that term here. |
| 21 |
C | This routine is not used in HYD calculations. |
| 22 |
C *==========================================================* |
| 23 |
C \ev |
| 24 |
|
| 25 |
C !USES: |
| 26 |
IMPLICIT NONE |
| 27 |
C == Global variables == |
| 28 |
#include "SIZE.h" |
| 29 |
#include "EEPARAMS.h" |
| 30 |
#include "PARAMS.h" |
| 31 |
#include "GRID.h" |
| 32 |
#include "DYNVARS.h" |
| 33 |
#include "NH_VARS.h" |
| 34 |
|
| 35 |
C !INPUT/OUTPUT PARAMETERS: |
| 36 |
C == Routine arguments == |
| 37 |
C bi,bj :: current tile indices |
| 38 |
C KappaRU :: vertical viscosity at U points |
| 39 |
C KappaRV :: vertical viscosity at V points |
| 40 |
C myTime :: Current time in simulation |
| 41 |
C myIter :: Current iteration number in simulation |
| 42 |
C myThid :: Thread number for this instance of the routine. |
| 43 |
INTEGER bi,bj |
| 44 |
_RL KappaRU(1-OLx:sNx+OLx,1-OLy:sNy+OLy,Nr) |
| 45 |
_RL KappaRV(1-OLx:sNx+OLx,1-OLy:sNy+OLy,Nr) |
| 46 |
_RL myTime |
| 47 |
INTEGER myIter |
| 48 |
INTEGER myThid |
| 49 |
|
| 50 |
#ifdef ALLOW_NONHYDROSTATIC |
| 51 |
|
| 52 |
C !LOCAL VARIABLES: |
| 53 |
C == Local variables == |
| 54 |
C iMin, iMax, |
| 55 |
C jMin, jMax |
| 56 |
C flx_NS :: Temp. used for fVol meridional terms. |
| 57 |
C flx_EW :: Temp. used for fVol zonal terms. |
| 58 |
C flx_Up :: Temp. used for fVol vertical terms. |
| 59 |
C flx_Dn :: Temp. used for fVol vertical terms. |
| 60 |
INTEGER iMin,iMax,jMin,jMax |
| 61 |
_RL flx_NS(1-OLx:sNx+OLx,1-OLy:sNy+OLy) |
| 62 |
_RL flx_EW(1-OLx:sNx+OLx,1-OLy:sNy+OLy) |
| 63 |
_RL flx_Dn(1-OLx:sNx+OLx,1-OLy:sNy+OLy) |
| 64 |
_RL flx_Up(1-OLx:sNx+OLx,1-OLy:sNy+OLy) |
| 65 |
_RL fZon (1-OLx:sNx+OLx,1-OLy:sNy+OLy) |
| 66 |
_RL fMer (1-OLx:sNx+OLx,1-OLy:sNy+OLy) |
| 67 |
_RL del2w (1-OLx:sNx+OLx,1-OLy:sNy+OLy) |
| 68 |
C i,j,k - Loop counters |
| 69 |
INTEGER i,j,k, kp1 |
| 70 |
_RL wOverride |
| 71 |
_RS hFacWtmp |
| 72 |
_RS hFacStmp |
| 73 |
_RS hFacCtmp |
| 74 |
_RS recip_hFacCtmp |
| 75 |
_RL slipSideFac |
| 76 |
_RL tmp_VbarZ, tmp_UbarZ, tmp_WbarZ |
| 77 |
_RL viscLoc |
| 78 |
_RL Half |
| 79 |
PARAMETER(Half=0.5D0) |
| 80 |
CEOP |
| 81 |
|
| 82 |
C Catch barotropic mode |
| 83 |
IF ( Nr .LT. 2 ) RETURN |
| 84 |
|
| 85 |
iMin = 1 |
| 86 |
iMax = sNx |
| 87 |
jMin = 1 |
| 88 |
jMax = sNy |
| 89 |
|
| 90 |
C Lateral friction (no-slip, free slip, or half slip): |
| 91 |
IF ( no_slip_sides ) THEN |
| 92 |
slipSideFac = -1. _d 0 |
| 93 |
ELSE |
| 94 |
slipSideFac = 1. _d 0 |
| 95 |
ENDIF |
| 96 |
CML half slip was used before ; keep the line for now, but half slip is |
| 97 |
CML not used anywhere in the code as far as I can see. |
| 98 |
C slipSideFac = 0. _d 0 |
| 99 |
|
| 100 |
C-- Initialise gW to zero |
| 101 |
DO k=1,Nr |
| 102 |
DO j=1-OLy,sNy+OLy |
| 103 |
DO i=1-OLx,sNx+OLx |
| 104 |
gW(i,j,k,bi,bj) = 0. |
| 105 |
ENDDO |
| 106 |
ENDDO |
| 107 |
ENDDO |
| 108 |
|
| 109 |
C-- Boundaries condition at top |
| 110 |
DO j=jMin,jMax |
| 111 |
DO i=iMin,iMax |
| 112 |
flx_Up(i,j)=0. |
| 113 |
ENDDO |
| 114 |
ENDDO |
| 115 |
|
| 116 |
C--- Sweep down column |
| 117 |
DO k=2,Nr |
| 118 |
kp1=k+1 |
| 119 |
wOverRide=1. |
| 120 |
IF (k.EQ.Nr) THEN |
| 121 |
kp1=Nr |
| 122 |
wOverRide=0. |
| 123 |
ENDIF |
| 124 |
C-- horizontal bi-harmonic dissipation |
| 125 |
IF (momViscosity .AND. viscA4W.NE.0. ) THEN |
| 126 |
C- calculate the horizontal Laplacian of vertical flow |
| 127 |
C Zonal flux d/dx W |
| 128 |
DO j=1-Oly,sNy+Oly |
| 129 |
fZon(1-Olx,j)=0. |
| 130 |
DO i=1-Olx+1,sNx+Olx |
| 131 |
C- Problem here: drF(k)*_hFacC & fZon are not at the same Horiz.&Vert. location |
| 132 |
fZon(i,j) = drF(k)*_hFacC(i,j,k,bi,bj) |
| 133 |
& *_dyG(i,j,bi,bj) |
| 134 |
& *_recip_dxC(i,j,bi,bj) |
| 135 |
& *(wVel(i,j,k,bi,bj)-wVel(i-1,j,k,bi,bj)) |
| 136 |
#ifdef COSINEMETH_III |
| 137 |
& *sqcosFacU(j,bi,bj) |
| 138 |
#endif |
| 139 |
ENDDO |
| 140 |
ENDDO |
| 141 |
C Meridional flux d/dy W |
| 142 |
DO i=1-Olx,sNx+Olx |
| 143 |
fMer(i,1-Oly)=0. |
| 144 |
ENDDO |
| 145 |
DO j=1-Oly+1,sNy+Oly |
| 146 |
DO i=1-Olx,sNx+Olx |
| 147 |
C- Problem here: drF(k)*_hFacC & fMer are not at the same Horiz.&Vert. location |
| 148 |
fMer(i,j) = drF(k)*_hFacC(i,j,k,bi,bj) |
| 149 |
& *_dxG(i,j,bi,bj) |
| 150 |
& *_recip_dyC(i,j,bi,bj) |
| 151 |
& *(wVel(i,j,k,bi,bj)-wVel(i,j-1,k,bi,bj)) |
| 152 |
#ifdef ISOTROPIC_COS_SCALING |
| 153 |
#ifdef COSINEMETH_III |
| 154 |
& *sqCosFacV(j,bi,bj) |
| 155 |
#endif |
| 156 |
#endif |
| 157 |
ENDDO |
| 158 |
ENDDO |
| 159 |
|
| 160 |
C del^2 W |
| 161 |
C Difference of zonal fluxes ... |
| 162 |
DO j=1-Oly,sNy+Oly |
| 163 |
DO i=1-Olx,sNx+Olx-1 |
| 164 |
del2w(i,j)=fZon(i+1,j)-fZon(i,j) |
| 165 |
ENDDO |
| 166 |
del2w(sNx+Olx,j)=0. |
| 167 |
ENDDO |
| 168 |
|
| 169 |
C ... add difference of meridional fluxes and divide by volume |
| 170 |
DO j=1-Oly,sNy+Oly-1 |
| 171 |
DO i=1-Olx,sNx+Olx |
| 172 |
C First compute the fraction of open water for the w-control volume |
| 173 |
C at the southern face |
| 174 |
hFacCtmp=max( _hFacC(i,j,k-1,bi,bj)-Half,0. _d 0 ) |
| 175 |
& + min( _hFacC(i,j,k ,bi,bj) ,Half ) |
| 176 |
IF (hFacCtmp .GT. 0.) THEN |
| 177 |
recip_hFacCtmp = 1./hFacCtmp |
| 178 |
ELSE |
| 179 |
recip_hFacCtmp = 0. _d 0 |
| 180 |
ENDIF |
| 181 |
del2w(i,j)=recip_rA(i,j,bi,bj) |
| 182 |
& *recip_drC(k)*recip_hFacCtmp |
| 183 |
& *( |
| 184 |
& del2w(i,j) |
| 185 |
& +( fMer(i,j+1)-fMer(i,j) ) |
| 186 |
& ) |
| 187 |
ENDDO |
| 188 |
ENDDO |
| 189 |
C-- No-slip BCs impose a drag at walls... |
| 190 |
CML ************************************************************ |
| 191 |
CML No-slip Boundary conditions for bi-harmonic dissipation |
| 192 |
CML need to be implemented here! |
| 193 |
CML ************************************************************ |
| 194 |
ELSE |
| 195 |
C- Initialize del2w to zero: |
| 196 |
DO j=1-Oly,sNy+Oly |
| 197 |
DO i=1-Olx,sNx+Olx |
| 198 |
del2w(i,j) = 0. _d 0 |
| 199 |
ENDDO |
| 200 |
ENDDO |
| 201 |
ENDIF |
| 202 |
|
| 203 |
C Flux on Southern face |
| 204 |
DO j=jMin,jMax+1 |
| 205 |
DO i=iMin,iMax |
| 206 |
C First compute the fraction of open water for the w-control volume |
| 207 |
C at the southern face |
| 208 |
hFacStmp=max(_hFacS(i,j,k-1,bi,bj)-Half,0. _d 0) |
| 209 |
& + min(_hFacS(i,j,k ,bi,bj),Half) |
| 210 |
tmp_VbarZ=Half*( |
| 211 |
& _hFacS(i,j,k-1,bi,bj)*vVel( i ,j,k-1,bi,bj) |
| 212 |
& +_hFacS(i,j, k ,bi,bj)*vVel( i ,j, k ,bi,bj)) |
| 213 |
flx_NS(i,j)= |
| 214 |
& tmp_VbarZ*Half*(wVel(i,j,k,bi,bj)+wVel(i,j-1,k,bi,bj)) |
| 215 |
& -(viscAh_W(i,j,k,bi,bj)+viscAh_W(i,j-1,k,bi,bj))*Half |
| 216 |
& *_recip_dyC(i,j,bi,bj) |
| 217 |
& *(hFacStmp*(wVel(i,j,k,bi,bj)-wVel(i,j-1,k,bi,bj)) |
| 218 |
C- Problem here: No-slip bc CANNOT be written in term of a flux |
| 219 |
& +(1. _d 0 - hFacStmp)*(1. _d 0 - slipSideFac) |
| 220 |
& *wVel(i,j,k,bi,bj)) |
| 221 |
& +(viscA4_W(i,j,k,bi,bj)+viscA4_W(i,j-1,k,bi,bj))*Half |
| 222 |
& *_recip_dyC(i,j,bi,bj)*(del2w(i,j)-del2w(i,j-1)) |
| 223 |
#ifdef ISOTROPIC_COS_SCALING |
| 224 |
#ifdef COSINEMETH_III |
| 225 |
& *sqCosFacV(j,bi,bj) |
| 226 |
#else |
| 227 |
& *CosFacV(j,bi,bj) |
| 228 |
#endif |
| 229 |
#endif |
| 230 |
C The last term is the weighted average of the viscous stress at the open |
| 231 |
C fraction of the w control volume and at the closed fraction of the |
| 232 |
C the control volume. A more compact but less intelligible version |
| 233 |
C of the last three lines is: |
| 234 |
CML & *( (1 _d 0 - slipSideFac*(1 _d 0 - hFacStmp)) |
| 235 |
CML & *wVel(i,j,k,bi,bi) + hFacStmp*wVel(i,j-1,k,bi,bj) ) |
| 236 |
ENDDO |
| 237 |
ENDDO |
| 238 |
C Flux on Western face |
| 239 |
DO j=jMin,jMax |
| 240 |
DO i=iMin,iMax+1 |
| 241 |
C First compute the fraction of open water for the w-control volume |
| 242 |
C at the western face |
| 243 |
hFacWtmp=max(_hFacW(i,j,k-1,bi,bj)-Half,0. _d 0) |
| 244 |
& + min(_hFacW(i,j,k ,bi,bj),Half) |
| 245 |
tmp_UbarZ=Half*( |
| 246 |
& _hFacW(i,j,k-1,bi,bj)*uVel( i ,j,k-1,bi,bj) |
| 247 |
& +_hFacW(i,j, k ,bi,bj)*uVel( i ,j, k ,bi,bj)) |
| 248 |
flx_EW(i,j)= |
| 249 |
& tmp_UbarZ*Half*(wVel(i,j,k,bi,bj)+wVel(i-1,j,k,bi,bj)) |
| 250 |
& -(viscAh_W(i,j,k,bi,bj)+viscAh_W(i-1,j,k,bi,bj))*Half |
| 251 |
& *_recip_dxC(i,j,bi,bj) |
| 252 |
& *(hFacWtmp*(wVel(i,j,k,bi,bj)-wVel(i-1,j,k,bi,bj)) |
| 253 |
C- Problem here: No-slip bc CANNOT be written in term of a flux |
| 254 |
& +(1 _d 0 - hFacWtmp)*(1 _d 0 - slipSideFac) |
| 255 |
& *wVel(i,j,k,bi,bj) ) |
| 256 |
& +(viscA4_W(i,j,k,bi,bj)+viscA4_W(i-1,j,k,bi,bj))*Half |
| 257 |
& *_recip_dxC(i,j,bi,bj)*(del2w(i,j)-del2w(i-1,j)) |
| 258 |
#ifdef COSINEMETH_III |
| 259 |
& *sqCosFacU(j,bi,bj) |
| 260 |
#else |
| 261 |
& *CosFacU(j,bi,bj) |
| 262 |
#endif |
| 263 |
C The last term is the weighted average of the viscous stress at the open |
| 264 |
C fraction of the w control volume and at the closed fraction of the |
| 265 |
C the control volume. A more compact but less intelligible version |
| 266 |
C of the last three lines is: |
| 267 |
CML & *( (1 _d 0 - slipSideFac*(1 _d 0 - hFacWtmp)) |
| 268 |
CML & *wVel(i,j,k,bi,bi) + hFacWtmp*wVel(i-1,j,k,bi,bj) ) |
| 269 |
ENDDO |
| 270 |
ENDDO |
| 271 |
C Flux on Lower face |
| 272 |
DO j=jMin,jMax |
| 273 |
DO i=iMin,iMax |
| 274 |
C Interpolate vert viscosity to W points |
| 275 |
viscLoc = ( KappaRU(i,j,k) +KappaRU(i+1,j,k) |
| 276 |
& +KappaRU(i,j,kp1)+KappaRU(i+1,j,kp1) |
| 277 |
& +KappaRV(i,j,k) +KappaRV(i,j+1,k) |
| 278 |
& +KappaRV(i,j,kp1)+KappaRV(i,j+1,kp1) |
| 279 |
& )*0.125 _d 0 |
| 280 |
tmp_WbarZ = Half*( wVel(i,j, k ,bi,bj) |
| 281 |
& +wVel(i,j,kp1,bi,bj)*wOverRide ) |
| 282 |
flx_Dn(i,j)= |
| 283 |
& tmp_WbarZ*tmp_WbarZ |
| 284 |
& -viscLoc*recip_drF(k) |
| 285 |
& *( wVel(i,j, k ,bi,bj) |
| 286 |
& -wVel(i,j,kp1,bi,bj)*wOverRide ) |
| 287 |
ENDDO |
| 288 |
ENDDO |
| 289 |
C Divergence of fluxes |
| 290 |
DO j=jMin,jMax |
| 291 |
DO i=iMin,iMax |
| 292 |
gW(i,j,k,bi,bj) = 0. |
| 293 |
& -( |
| 294 |
& +_recip_dxF(i,j,bi,bj)*( |
| 295 |
& flx_EW(i+1,j)-flx_EW(i,j) ) |
| 296 |
& +_recip_dyF(i,j,bi,bj)*( |
| 297 |
& flx_NS(i,j+1)-flx_NS(i,j) ) |
| 298 |
& +recip_drC(k) *( |
| 299 |
& flx_Up(i,j) -flx_Dn(i,j) ) |
| 300 |
& ) |
| 301 |
caja * recip_hFacU(i,j,k,bi,bj) |
| 302 |
caja NOTE: This should be included |
| 303 |
caja but we need an hFacUW (above U points) |
| 304 |
caja and an hFacUS (above V points) too... |
| 305 |
C-- prepare for next level (k+1) |
| 306 |
flx_Up(i,j)=flx_Dn(i,j) |
| 307 |
ENDDO |
| 308 |
ENDDO |
| 309 |
|
| 310 |
C---+----1----+----2----+----3----+----4----+----5----+----6----+----7-|--+----| |
| 311 |
|
| 312 |
C- Compute effective gW_[n+1/2] terms (including Adams-Bashforth weights) |
| 313 |
C and save gW_[n] into gwNm1 for the next time step. |
| 314 |
c#ifdef ALLOW_ADAMSBASHFORTH_3 |
| 315 |
c CALL ADAMS_BASHFORTH3( |
| 316 |
c I bi, bj, k, |
| 317 |
c U gW, gwNm, |
| 318 |
c I momStartAB, myIter, myThid ) |
| 319 |
c#else /* ALLOW_ADAMSBASHFORTH_3 */ |
| 320 |
CALL ADAMS_BASHFORTH2( |
| 321 |
I bi, bj, k, |
| 322 |
U gW, gwNm1, |
| 323 |
I myIter, myThid ) |
| 324 |
c#endif /* ALLOW_ADAMSBASHFORTH_3 */ |
| 325 |
|
| 326 |
C- end of the k loop |
| 327 |
ENDDO |
| 328 |
|
| 329 |
#endif /* ALLOW_NONHYDROSTATIC */ |
| 330 |
|
| 331 |
RETURN |
| 332 |
END |
Parent Directory
| 
Revision Log
| 
Revision Graph