| 1 |
C $Id: g_calc.F,v 1.3 1998/06/23 13:36:41 adcroft Exp $ |
| 2 |
#include "CPP_OPTIONS.h" |
| 3 |
#include "CPP_MACROS.h" |
| 4 |
|
| 5 |
#define _DB0 0. |
| 6 |
#define _DB1 1. |
| 7 |
#define _wdvdz 1.D0 |
| 8 |
|
| 9 |
C=================================================================== |
| 10 |
C Procedure name: G_CALC | |
| 11 |
C Function: Caculates the tendency terms for the momentum| |
| 12 |
C equations. | |
| 13 |
C Comments: | |
| 14 |
C=================================================================== |
| 15 |
CStartofinterface |
| 16 |
SUBROUTINE G_CALC( |
| 17 |
I U, V, W, PH, PS, K, |
| 18 |
U RHS2D, divH, |
| 19 |
O GU, GV ) |
| 20 |
IMPLICIT NONE |
| 21 |
C============== Global data ========================================== |
| 22 |
#include "SIZE.h" |
| 23 |
#include "AJAINF.h" |
| 24 |
#include "OPERATORS.h" |
| 25 |
#include "GRID.h" |
| 26 |
#include "PARAMS.h" |
| 27 |
#include "OLDG.h" |
| 28 |
#include "FORCING.h" |
| 29 |
#include "MASKS.h" |
| 30 |
C============= Routine arguments ===================================== |
| 31 |
C /--------------------------------------------------------------\ |
| 32 |
C |U, V, W - X,Y,Z Velocity ( m/s, m/s, Pa/s ). | |
| 33 |
C |PH - Hydrostatic pressure perturbation (m). | |
| 34 |
C |PS - Surface pressure (m). | |
| 35 |
C |GU - d/dt U (m/s/s). | |
| 36 |
C |GU - d/dt V (m/s/s). | |
| 37 |
C |RHS2D - Contribution to righ-hand side of | |
| 38 |
C | pressure eqaution. | |
| 39 |
C |divH - Contribution to barotropic divergence. | |
| 40 |
C \--------------------------------------------------------------/ |
| 41 |
REAL U (nx,ny,nz) |
| 42 |
REAL V (nx,ny,nz) |
| 43 |
REAL W (nx,ny,nz) |
| 44 |
REAL PH(nx,ny,nz) |
| 45 |
REAL PS(nx,ny ) |
| 46 |
REAL GU(nx,ny,nz) |
| 47 |
REAL GV(nx,ny,nz) |
| 48 |
REAL RHS2D(nx,ny) |
| 49 |
REAL divH (nx,ny) |
| 50 |
INTEGER K |
| 51 |
CEndofinterface |
| 52 |
C============= Local variables ====================================== |
| 53 |
C /--------------------------------------------------------------\ |
| 54 |
C |Slice Notation | |
| 55 |
C |phiK - XY slice of variable "phi" at level K. | |
| 56 |
C |phiKP1 - XY slice of variable "phi" at level K+1. | |
| 57 |
C |phiKM1 - XY slice of variable "phi" at level K-1. | |
| 58 |
C |phiNM1 - Variable "phi" at time level N-1. | |
| 59 |
C |phiU - Varaible "phi" interpolated to U grid. | |
| 60 |
C |phiV - Varaible "phi" interpolated to V grid. | |
| 61 |
C \--------------------------------------------------------------/ |
| 62 |
C /--------------------------------------------------------------\ |
| 63 |
C |Geometry terms | |
| 64 |
C | o xa,ya,za - X, Y, Z face areas ( m.Pa, m.Pa, m**2 ). | |
| 65 |
C | o rV - 1/Volume ( m**2.Pa ) | |
| 66 |
C | o ?Grad?Op - Generic gradient operator A[XYZ][uvwp].d/d[xyz] | |
| 67 |
C | uGradxOp: AXu/DXu, uGradyOp: AYu/DYu etc... | |
| 68 |
C \--------------------------------------------------------------/ |
| 69 |
REAL xaK (nx+1,0:ny) |
| 70 |
REAL yaK (0:nx,ny+1) |
| 71 |
REAL uGradxOp (nx ,ny ) |
| 72 |
REAL uGradyOp (nx ,ny ) |
| 73 |
REAL uGradzOpK (nx ,ny ) |
| 74 |
REAL uGradzOpKP1(nx ,ny ) |
| 75 |
REAL vGradxOp (nx ,ny ) |
| 76 |
REAL vGradyOp (nx ,ny ) |
| 77 |
REAL vGradzOpK (nx ,ny ) |
| 78 |
REAL vGradzOpKP1(nx ,ny ) |
| 79 |
REAL zaK (0:nx,0:ny) |
| 80 |
REAL zaKP1 (0:nx,0:ny) |
| 81 |
REAL rVk (nx ,ny ) |
| 82 |
C /--------------------------------------------------------------\ |
| 83 |
C |State variables | |
| 84 |
C | o u, v, w - X,Y,Z velocity slices. | |
| 85 |
C | o ph - Hydrostatic pressure. | |
| 86 |
C \--------------------------------------------------------------/ |
| 87 |
REAL uK (nx+1,0:ny) |
| 88 |
REAL uMskK (nx+1,0:ny) |
| 89 |
REAL uKNM1 (nx+1,0:ny) |
| 90 |
REAL wK (0:nx,0:ny) |
| 91 |
REAL wkP1 (0:nx,0:ny) |
| 92 |
REAL phK (nx ,ny ) |
| 93 |
REAL vK (0:nx,ny+1) |
| 94 |
REAL vMskK (nx+1,0:ny) |
| 95 |
REAL vKNM1 (0:nx,ny+1) |
| 96 |
REAL vKP1 (nx ,ny ) |
| 97 |
REAL vKM1 (nx ,ny ) |
| 98 |
C /--------------------------------------------------------------\ |
| 99 |
C |Work space arrays | |
| 100 |
C \--------------------------------------------------------------/ |
| 101 |
REAL tmp (0:nx+1,0:ny+1) |
| 102 |
REAL tmp2 (0:nx+1,0:ny+1) |
| 103 |
REAL stmp |
| 104 |
C /--------------------------------------------------------------\ |
| 105 |
C |Arrays for accumulating fluxes | |
| 106 |
C | o fX, fY, fZ - Flux in X, Y and Z. | |
| 107 |
C \--------------------------------------------------------------/ |
| 108 |
REAL fX (0:nx+1,ny) |
| 109 |
REAL fY (nx,0:ny+1) |
| 110 |
REAL fZK (nx,ny) |
| 111 |
REAL fZKP1 (nx,ny) |
| 112 |
C Loop counters: |
| 113 |
INTEGER I, J |
| 114 |
C Flags: |
| 115 |
LOGICAL TOP_LAYER |
| 116 |
LOGICAL BOTTOM_LAYER |
| 117 |
C === Set up controlling flags ======================================== |
| 118 |
TOP_LAYER = K .EQ. 1 |
| 119 |
BOTTOM_LAYER = K .EQ. nz |
| 120 |
C === Set up the slices =============================================== |
| 121 |
C /---------------------------------------------------------------\ |
| 122 |
C | xaK <- xa(K) | |
| 123 |
C | yaK <- ya(K) | |
| 124 |
C | za[K,KP1] <- za[(K),(K+1)] | |
| 125 |
C \---------------------------------------------------------------/ |
| 126 |
xaK (1:nx,1:ny) = xa (_I3(K,:,:)) |
| 127 |
xaK (nx+1,1:ny) = xaK (1,1:ny ) |
| 128 |
xaK (:,0 ) = xaK (:,ny ) |
| 129 |
yaK (1:nx,1:ny) = ya (_I3(K,:,:)) |
| 130 |
yaK (0,1:ny ) = yaK (nx,1:ny ) |
| 131 |
yaK (:,ny+1 ) = yaK (:,1 ) |
| 132 |
zaK (1:nx,1:ny) = za (_I3(K,:,:)) |
| 133 |
zaK (0,1:ny ) = zaK (nx,1:ny ) |
| 134 |
zaK (:,0 ) = zaK (:,ny ) |
| 135 |
IF ( .NOT. BOTTOM_LAYER ) THEN |
| 136 |
zaKP1(1:nx,1:ny) = za (_I3(K+1,:,:)) |
| 137 |
zaKP1(0,1:ny ) = zaKP1(nx,1:ny ) |
| 138 |
zaKP1(:,0 ) = zaKP1(:,ny ) |
| 139 |
ENDIF |
| 140 |
C /---------------------------------------------------------------\ |
| 141 |
C | uGradXOp <- 1/DXu*Bx{xa} | |
| 142 |
C | uGradYOp <- 1/DYu*Bx{ya} | |
| 143 |
C | uGradZOp[K,KP1] <- 1/DZu*Bz{ZA[K,K+1]} | |
| 144 |
C \---------------------------------------------------------------/ |
| 145 |
DO J = 1, ny |
| 146 |
DO I = 1, nx |
| 147 |
uGradxOp(I,J) = 0.5*( xaK(I,J)+xaK(I+1,J) )*uDdxOp(_I3(K,I ,J)) |
| 148 |
ENDDO |
| 149 |
ENDDO |
| 150 |
DO J = 1, ny |
| 151 |
DO I = 2, nx |
| 152 |
uGradyOp(I,J) = 0.5*( yaK(I,J)+yaK(I-1,J) )*uDdyOp(_I3(K,I,J)) |
| 153 |
uGradzOpK(I,J) = |
| 154 |
& 0.5*(ZA(_I3(K,I,J))+ZA(_I3(K,I-1,J)))*uDdzOp(_I3(K,I,J)) |
| 155 |
ENDDO |
| 156 |
uGradyOp(1,J) = 0.5*( yaK(1,J)+yaK(nx,J) )*uDdyOp(_I3(K,1,J)) |
| 157 |
uGradzOpK(1,J) = |
| 158 |
& 0.5*(ZA(_I3(K,1,J))+ZA(_I3(K,nx,J)))*uDdzOp(_I3(K,1,J)) |
| 159 |
ENDDO |
| 160 |
IF ( .NOT. BOTTOM_LAYER ) THEN |
| 161 |
DO J = 1, ny |
| 162 |
DO I = 2, nx |
| 163 |
uGradzOpKP1(I,J) = |
| 164 |
& 0.5*(ZA(_I3(K+1,I,J))+ZA(_I3(K+1,I-1,J)))*uDdzOp(_I3(K+1,I,J)) |
| 165 |
ENDDO |
| 166 |
uGradzOpKP1(1,J) = |
| 167 |
& 0.5*( ZA(_I3(K+1,1,J))+ZA(_I3(K+1,nx,J)) )*uDdzOp(_I3(K+1,1,J)) |
| 168 |
ENDDO |
| 169 |
ENDIF |
| 170 |
C /---------------------------------------------------------------\ |
| 171 |
C | vGradXOp <- 1/DXu*By{xa} | |
| 172 |
C | vGradYOp <- 1/DYu*By{ya} | |
| 173 |
C | vGradZOp[K,KP1] <- 1/DZu*Bz{ZA[(K),(K+1)]} | |
| 174 |
C \---------------------------------------------------------------/ |
| 175 |
DO J = 2, ny |
| 176 |
DO I = 1, nx |
| 177 |
vGradxOp(I,J) = 0.5*( xaK(I,J)+xaK(I,J-1) )*vDdxOp(_I3(K,I,J)) |
| 178 |
vGradzOpK(I,J)= |
| 179 |
& 0.5*(ZA(_I3(K,I,J))+ZA(_I3(K,I,J-1)))*vDdzOp(_I3(K,I,J)) |
| 180 |
ENDDO |
| 181 |
ENDDO |
| 182 |
DO I = 1, nx |
| 183 |
vGradxOp (I,1) = 0.5*( xaK(I,1)+xaK(I,ny) )*vDdxOp(_I3(K,I,1 )) |
| 184 |
vGradzOpK(I,1) = 0.5*(ZA(_I3(K,I,1))+ZA(_I3(K,I,ny)))*vDdzOp(_I3(K,I,1)) |
| 185 |
ENDDO |
| 186 |
IF ( .NOT. BOTTOM_LAYER ) THEN |
| 187 |
DO J = 2, ny |
| 188 |
DO I = 1, nx |
| 189 |
vGradzOpKP1(I,J) = |
| 190 |
& 0.5*(ZA(_I3(K+1,I,J))+ZA(_I3(K+1,I,J-1)))*vDdzOp(_I3(K+1,I,J)) |
| 191 |
ENDDO |
| 192 |
ENDDO |
| 193 |
DO I = 1, nx |
| 194 |
vGradzOpKP1(I,1) = |
| 195 |
& 0.5*(ZA(_I3(K+1,I,1))+ZA(_I3(K+1,I,ny)))*vDdzOp(_I3(K+1,I,1)) |
| 196 |
ENDDO |
| 197 |
ENDIF |
| 198 |
DO J = 1, ny-1 |
| 199 |
DO I = 1, nx |
| 200 |
vGradyOp(I,J) = 0.5*( yaK(I,J)+yaK(I,J+1) )*vDdyOp(_I3(K,I,J)) |
| 201 |
ENDDO |
| 202 |
ENDDO |
| 203 |
DO I = 1, nx |
| 204 |
vGradyOp(I,ny) = 0.5*( yaK(I,ny)+yaK(I,1) )*vDdyOp(_I3(K,I,ny)) |
| 205 |
ENDDO |
| 206 |
C /---------------------------------------------------------------\ |
| 207 |
C | w[K,KP1] <- W[(K),(K+1)] | |
| 208 |
C \---------------------------------------------------------------/ |
| 209 |
wKP1 = 0. |
| 210 |
wK(1:nx,1:ny) = W(_I3(K,:,:)) |
| 211 |
wK(0,1:ny) = wK(nx,1:ny) |
| 212 |
wK(:,0) = wK(:,ny) |
| 213 |
IF ( .NOT. BOTTOM_LAYER ) wKP1(1:nx,1:ny) = W(_I3(K+1,:,:)) |
| 214 |
wKP1(0,1:ny) = wKP1(nx,1:ny) |
| 215 |
wKP1(:,0) = wKP1(:,ny) |
| 216 |
C /---------------------------------------------------------------\ |
| 217 |
C | uK <- U(K) | |
| 218 |
C \---------------------------------------------------------------/ |
| 219 |
uK(1:nx,1:ny) = U(_I3(K,:,:)) |
| 220 |
uK(nx+1,1:ny) = uK(1,1:ny) |
| 221 |
uK(:,0) = uK(:,ny) |
| 222 |
uKNM1(1:nx,1:ny) = UNM1(_I3(K,:,:)) |
| 223 |
uKNM1(nx+1,1:ny) = uKNM1(1,1:ny) |
| 224 |
uKNM1(:,0) = uKNM1(:,ny) |
| 225 |
uMskK(1:nx,1:ny) = UMASK(_I3(K,:,:)) |
| 226 |
uMskK(nx+1,1:ny) = uMskK(1,1:ny) |
| 227 |
uMskK(:,0) = uMskK(:,ny) |
| 228 |
C /---------------------------------------------------------------\ |
| 229 |
C | v[KM1,K,KP1] = V[(K-1),(K),(K+1)] | |
| 230 |
C \---------------------------------------------------------------/ |
| 231 |
vKP1 = 0. |
| 232 |
vKM1 = 0. |
| 233 |
vK(1:nx,1:ny) = V(_I3(K,:,:)) |
| 234 |
vK(0,1:ny) = vK(nx,1:ny) |
| 235 |
vK(:,ny+1) = vK(:,1) |
| 236 |
vKNM1(1:nx,1:ny) = VNM1(_I3(K,:,:)) |
| 237 |
vKNM1(0,1:ny) = vKNM1(nx,1:ny) |
| 238 |
vKNM1(:,ny+1) = vKNM1(:,1) |
| 239 |
vMskK(1:nx,1:ny) = VMASK(_I3(K,:,:)) |
| 240 |
vMskK(0,1:ny) = vMskK(nx,1:ny) |
| 241 |
vMskK(:,ny+1) = vMskK(:,1) |
| 242 |
IF ( .NOT. TOP_LAYER ) vKM1(1:nx,1:ny) = V(_I3(K-1,:,:)) |
| 243 |
IF ( .NOT. BOTTOM_LAYER ) vKP1(1:nx,1:ny) = V(_I3(K+1,:,:)) |
| 244 |
C /---------------------------------------------------------------\ |
| 245 |
C | Gu Section | |
| 246 |
C |===============================================================| |
| 247 |
C | In this section Gu = GuDiffusion | |
| 248 |
C | + GuAdvection | |
| 249 |
C | + GuMetric | |
| 250 |
C | + GuXZCoriolis | |
| 251 |
C | + GuForcing | |
| 252 |
C \---------------------------------------------------------------/ |
| 253 |
C /---------------------------------------------------------------\ |
| 254 |
C | Initialise local terms | |
| 255 |
C |===============================================================| |
| 256 |
C | rVk <- 1/Vu | |
| 257 |
C \---------------------------------------------------------------/ |
| 258 |
rVk = rUvol(_I3(K,:,:)) |
| 259 |
fZK = 0. |
| 260 |
fZKP1 = 0. |
| 261 |
C /---------------------------------------------------------------\ |
| 262 |
C | XY GuDiffusion + XY GuAdvection | |
| 263 |
C |===============================================================| |
| 264 |
C | Note: XY GuDiffusion = Laplacian + Biharmonic in XY-plane. | |
| 265 |
C | o Laplacian ( -a2MomXY.del^2(u) ) | |
| 266 |
C | o Biharmonic ( +a4MomXY.del^4(u) ) | |
| 267 |
C \---------------------------------------------------------------/ |
| 268 |
DO J = 1, ny |
| 269 |
DO I = 1, nx |
| 270 |
C /-------------------------------------------------------------\ |
| 271 |
C | fX <- 1/DXu*XAu*Ddx{U} | |
| 272 |
C | fY <- 1/DYu*YAu*Ddy{U} | |
| 273 |
C \-------------------------------------------------------------/ |
| 274 |
fX(I,J) = uGradxOp(I,J)*(uK(I+1,J)-uK(I,J)) |
| 275 |
fY(I,J) = uGradyOp(I,J)*(uK(I,J)-uK(I,J-1)) |
| 276 |
ENDDO |
| 277 |
ENDDO |
| 278 |
fX(0 ,1:ny) = fX(nx,1:ny) |
| 279 |
fY(1:nx,ny+1) = fY(1:nx,1) |
| 280 |
C WRITE(0,*) ' CALC_G fY ' |
| 281 |
C CALL PLOT_FIELD( fY(1:nx,1:ny), nX, nY ) |
| 282 |
DO J=1,ny |
| 283 |
DO I=1,nx |
| 284 |
C /-------------------------------------------------------------\ |
| 285 |
C | tmp <- 1/Vu*(Dx{fX}+Dy{fY}) | |
| 286 |
C \-------------------------------------------------------------/ |
| 287 |
tmp(I,J)=rVk(I,J)*fX(I,J)-rVk(I,J)*fX(I-1,J) |
| 288 |
& +rVk(I,J)*fY(I,J+1)-rVk(I,J)*fY(I,J) |
| 289 |
ENDDO |
| 290 |
ENDDO |
| 291 |
tmp(nx+1,1:ny) = tmp(1,1:ny) |
| 292 |
tmp(:,0) = tmp(:,ny) |
| 293 |
DO J = 1,ny |
| 294 |
DO I = 1,nx |
| 295 |
C /-------------------------------------------------------------\ |
| 296 |
C | Diffusive and advective flux at eastern face | |
| 297 |
C |=============================================================| |
| 298 |
C | fX <- -k2*fX + ( Laplacian,<Line 1> ) | |
| 299 |
C | k4/DXu*XAu*Ddx{tmp} + ( Biharmonic,<Line 2> ) | |
| 300 |
C | Bx{XA*U}*Bx{U} ( Advection,<Line 3-4> ) | |
| 301 |
C \-------------------------------------------------------------/ |
| 302 |
fX(I,J) = 0. |
| 303 |
_LPM(& -_DB1*a2MomXY*fX(I,J) ) |
| 304 |
_BHM(& +a4MomXY*uGradxOp(I,J)*(tmp(I+1,J)-tmp(I,J)) ) |
| 305 |
_ADM(& +0.25D0*(uK(I,J)*XAK(I,J)+uK(I+1,J)*XAK(I+1,J))* ) |
| 306 |
_ADM(& (uK(I,J)+uK(I+1,J)) ) |
| 307 |
C /-------------------------------------------------------------\ |
| 308 |
C | Diffusive and advective flux at southern face | |
| 309 |
C |=============================================================| |
| 310 |
C | fY <- -k2*fY + ( Laplacian,<Line 1> ) | |
| 311 |
C | k4/DYu*YAu*Ddy{tmp} + ( Biharmonic,<Line 2> ) | |
| 312 |
C | Bx{YA*V}*By{U} ( Advection,<Line 3-4> ) | |
| 313 |
C \-------------------------------------------------------------/ |
| 314 |
fY(I,J) = 0. |
| 315 |
_LPM(& -_DB1*a2MomXY*fY(I,J) ) |
| 316 |
_BHM(& +a4MomXY*uGradyOp(I,J)*(tmp(I,J)-tmp(I,J-1)) ) |
| 317 |
_ADM(& +0.25D0*(uK(I,J)+uK(I,J-1))* ) |
| 318 |
_ADM(& (vK(I,J)*YAK(I,J)+vK(I-1,J)*YAK(I-1,J)) ) |
| 319 |
_ADM(& *uMskK(I,J)*uMskK(I,J-1) ) |
| 320 |
C Note!!!! The last line above "*uMskK(I,J)*uMskK(I,J-1)" is the |
| 321 |
C Note!!!! boundary condition in the standard v010 CM5 code. The |
| 322 |
C Note!!!! FV paper used a different bc in which the last line is |
| 323 |
C Note!!!! omitted. |
| 324 |
ENDDO |
| 325 |
ENDDO |
| 326 |
|
| 327 |
C /---------------------------------------------------------------\ |
| 328 |
C | Z GuDiffusion + Z GuAdvection | |
| 329 |
C |===============================================================| |
| 330 |
C \---------------------------------------------------------------/ |
| 331 |
IF ( .NOT. TOP_LAYER ) THEN |
| 332 |
DO J = 1, ny |
| 333 |
DO I = 1, nx |
| 334 |
C /------------------------------------------------------------\ |
| 335 |
C | Diffusive and advective flux at upper face | |
| 336 |
C |============================================================| |
| 337 |
C | fZK <- -k2Z/DZu*ZAu*Ddz{U} + ( Laplacian,<Line 1> ) | |
| 338 |
C | Bx{ZA*-w}*Bz{U} ( Advection,<Line 2-3> ) | |
| 339 |
C \------------------------------------------------------------/ |
| 340 |
fZK(I,J) = 0. |
| 341 |
_LPM(& -_DB1*a2MomZ*uGradzOpK(I,J)*( U(_I3(K-1,I,J))-uK(I,J) ) ) |
| 342 |
_ADM(& +0.25D0*(uK(I,J)+U(_I3(K-1,I,J)))* ) |
| 343 |
_ADM(& (-wK(I,J)*ZAK(I,J)-wK(I-1,J)*ZAK(I-1,J)) ) |
| 344 |
_ADM(& *uMskK(I,J)*UMASK(_I3(K-1,I,J)) ) |
| 345 |
C Note!!!! The last line above "*uMskK(I,J)*UMASK(_I3(K-1,I,J))" is |
| 346 |
C Note!!!! boundary condition in the standard v010 CM5 code. The |
| 347 |
C Note!!!! FV paper used a different bc in which the last line is |
| 348 |
C Note!!!! omitted. |
| 349 |
ENDDO |
| 350 |
ENDDO |
| 351 |
ELSE |
| 352 |
DO J = 1, ny |
| 353 |
DO I = 1, nx |
| 354 |
C /------------------------------------------------------------\ |
| 355 |
C | Advective flux at top layer | |
| 356 |
C |============================================================| |
| 357 |
C | o Only active when using implcit free surface. | |
| 358 |
C \------------------------------------------------------------/ |
| 359 |
fZK(I,J) = 0. |
| 360 |
_ADM(& +freeSurfFac*0.25D0*(uK(I,J)+uK(I,J) )* ) |
| 361 |
_ADM(& (-wK(I,J)*ZAK(I,J)-wK(I-1,J)*ZAK(I-1,J)) ) |
| 362 |
ENDDO |
| 363 |
ENDDO |
| 364 |
ENDIF |
| 365 |
IF ( .NOT. BOTTOM_LAYER ) THEN |
| 366 |
DO J = 1, ny |
| 367 |
DO I = 1, nx |
| 368 |
C /------------------------------------------------------------\ |
| 369 |
C | Diffusive and advective flux at lower face | |
| 370 |
C |============================================================| |
| 371 |
C | fZKP1 <- -k2Z/DZu*ZAu*Ddz{U} + ( Laplacian,<Line 1> ) | |
| 372 |
C | Bx{ZA*-w}*Bz{U} ( Advection,<Line 2-3> ) | |
| 373 |
C \------------------------------------------------------------/ |
| 374 |
fZKP1(I,J) = 0. |
| 375 |
_LPM(& -_DB1*a2MomZ*uGradzOpKP1(I,J)*(uK(I,J)-U(_I3(K+1,I,J))) ) |
| 376 |
_ADM(& +0.25D0*(uK(I,J)+U(_I3(K+1,I,J)))* ) |
| 377 |
_ADM(& (-wKP1(I,J)*ZAKP1(I,J)-wKP1(I-1,J)*ZAKP1(I-1,J)) ) |
| 378 |
_ADM(& *uMskK(I,J)*UMASK(_I3(K+1,I,J)) ) |
| 379 |
C Note!!!! The last line above "*uMskK(I,J)*UMASK(_I3(K-1,I,J))" is |
| 380 |
C Note!!!! boundary condition in the standard v010 CM5 code. The |
| 381 |
C Note!!!! FV paper used a different bc in which the last line is |
| 382 |
C Note!!!! omitted. |
| 383 |
ENDDO |
| 384 |
ENDDO |
| 385 |
ELSE |
| 386 |
DO J = 1, ny |
| 387 |
DO I = 1, nx |
| 388 |
fZKP1(I,J) = 0. |
| 389 |
_LPM(& -_DB1*a2MomZ*rDzAtP(K)*2.D0*uK(I,J) ) |
| 390 |
_LPM(& *0.5*( ZAk(I,J)+ZAk(I-1,J) ) ) |
| 391 |
ENDDO |
| 392 |
ENDDO |
| 393 |
ENDIF |
| 394 |
C /---------------------------------------------------------------\ |
| 395 |
C | gU <- GuDiffusion + GuAdvection = -div(fX,fY,fZ) | |
| 396 |
C \---------------------------------------------------------------/ |
| 397 |
fX(0,1:ny) = fX(nx,1:ny) |
| 398 |
fY(:,ny+1) = fY(:,1) |
| 399 |
DO J = 1, ny |
| 400 |
DO I = 1, nx |
| 401 |
C /-------------------------------------------------------------\ |
| 402 |
C | o Minus divergence of the lateral fluxes | |
| 403 |
C | gU <- -1/Vu*(Ddx{fX}+Ddy{fY}) | |
| 404 |
C \-------------------------------------------------------------/ |
| 405 |
gU(_I3(K,I,J))=-fX(I,J)*rVk(I,J)+fX(I-1,J)*rVk(I,J) |
| 406 |
& -fY(I,J+1)*rVk(I,J)+fY(I,J)*rVk(I,J) |
| 407 |
ENDDO |
| 408 |
ENDDO |
| 409 |
DO J = 1, ny |
| 410 |
DO I = 1, nx |
| 411 |
C /------------------------------------------------------------\ |
| 412 |
C | o Minus flux at upper face | |
| 413 |
C | gU <- gU-1/Vu*fZK | |
| 414 |
C \------------------------------------------------------------/ |
| 415 |
gU(_I3(K,I,J))=gU(_I3(K,I,J))-fZK(I,J)*rVk(I,J) |
| 416 |
ENDDO |
| 417 |
ENDDO |
| 418 |
CcnhDebugStarts |
| 419 |
C Changed to make noslip bottom boundary. |
| 420 |
Cdbg IF ( .NOT. BOTTOM_LAYER ) THEN |
| 421 |
CcnhDebugEnds |
| 422 |
DO J = 1, ny |
| 423 |
DO I = 1, nx |
| 424 |
C /------------------------------------------------------------\ |
| 425 |
C | o Plus flux at lower face | |
| 426 |
C | gU <- gU+1/Vu*fZP1 | |
| 427 |
C \------------------------------------------------------------/ |
| 428 |
gU(_I3(K,I,J))=gU(_I3(K,I,J))+fZKP1(I,J)*rVk(I,J) |
| 429 |
ENDDO |
| 430 |
ENDDO |
| 431 |
CcnhDebugStarts |
| 432 |
C Changed to make noslip bottom boundary. |
| 433 |
Cdbg ENDIF |
| 434 |
CcnhDebugEnds |
| 435 |
|
| 436 |
C /---------------------------------------------------------------\ |
| 437 |
C | gU <- gU + GuMetric + GuXZCoriolis | |
| 438 |
C \---------------------------------------------------------------/ |
| 439 |
DO J = 1, ny |
| 440 |
DO I = 1, nx |
| 441 |
# ifdef _SPHERICAL_POLAR_METRIC_TERMS |
| 442 |
C /-------------------------------------------------------------\ |
| 443 |
C | o Spherical polar metric terms for U | |
| 444 |
C |=============================================================| |
| 445 |
C | gU <- gU - U/aEarth*BzBx{-w/G/RONIL} - U*tan(lat)*ByBx{V} | |
| 446 |
C \-------------------------------------------------------------/ |
| 447 |
gU(_I3(K,I,J)) = gU(_I3(K,I,J)) |
| 448 |
_SPM(& -uMask(_I3(K,I,J))*uK(I,J)/rZero* ) |
| 449 |
_SPM(&0.25D0*(wK(I-1,J)+wK(I,J)+wKP1(I-1,J)+wKP1(I,J))/(-G*RONIL)) |
| 450 |
_SPM(& +uMask(_I3(K,I,J))*uK(I,J)*uTphiOP(_I3(K,I,J))/rZero ) |
| 451 |
_SPM(& *0.25D0*(vK(I,J)+vK(I-1,J)+vK(I,J+1)+vK(I-1,J+1)) ) |
| 452 |
# endif |
| 453 |
C /-------------------------------------------------------------\ |
| 454 |
C | o XZ-plane coriolis terms for U | |
| 455 |
C |=============================================================| |
| 456 |
C | gU <- gU - 2*omega*cos(lat)*BzBx{-w/G/RONIL} | |
| 457 |
C \-------------------------------------------------------------/ |
| 458 |
gU(_I3(K,I,J)) = gU(_I3(K,I,J)) |
| 459 |
_XZC(& -fCorW(_I3(K,I,J))/ ) |
| 460 |
_XZC(&G/(-RONIL)*0.25D0*(wK(I-1,J)+wK(I,J)+wKP1(I-1,J)+wKP1(I,J))) |
| 461 |
ENDDO |
| 462 |
ENDDO |
| 463 |
|
| 464 |
#ifdef _MOMENTUM_FORCING |
| 465 |
C /---------------------------------------------------------------\ |
| 466 |
C | gU <- gU + GuForcing | |
| 467 |
C \---------------------------------------------------------------/ |
| 468 |
IF ( TOP_LAYER ) THEN |
| 469 |
C /--------------------------------------------------------------\ |
| 470 |
C | o Add a drag term to the top-layer of U | |
| 471 |
C | u* is the velocity of the disk relative to the surface. | |
| 472 |
C | lambda is the inverse of the time scale with which the | |
| 473 |
C | flow equilibrates withthe rotating disk. | |
| 474 |
C | to get lambda we use the following | |
| 475 |
C | du/dt = stress/rho/dz | |
| 476 |
C | stress = rho.Kz.DU/DZ | |
| 477 |
C | DZ = (2Kz/f)^1/2 (Ekman layer thiickness ) | |
| 478 |
C | DU = (U-u*) | |
| 479 |
C | F=1e-4, Kz=1e-3, dz of 25m gives an equilibration time | |
| 480 |
C | scale of 1.3 days. | |
| 481 |
C | gU(K=1) <- gU(K=1) - lambda(U-u*) | |
| 482 |
C \--------------------------------------------------------------/ |
| 483 |
gU(_I3(K,:,:)) = gU(_I3(K,:,:)) |
| 484 |
& + fu |
| 485 |
ENDIF |
| 486 |
#endif |
| 487 |
|
| 488 |
C ********************************************************************* |
| 489 |
C ****Gv SECTION******************************************************* |
| 490 |
C ********************************************************************* |
| 491 |
C In this section Gv = GvDiffusion |
| 492 |
C + GvAdvection |
| 493 |
C + GvMetric |
| 494 |
C + GvHorizontalCoriolis |
| 495 |
C + GvForcing |
| 496 |
C === Initialise local operators ====================================== |
| 497 |
C rVk: 1/VVolume |
| 498 |
fZK = 0. |
| 499 |
fZKP1 = 0. |
| 500 |
rVk = rVvol(_I3(K,:,:)) |
| 501 |
C === Lateral diffusive flux of V ===================================== |
| 502 |
C o Laplacian ( -a2MomXY.del^2(v) ) |
| 503 |
C o Biharmonic ( +a4MomXY.del^4(v) ) |
| 504 |
DO J = 1, ny |
| 505 |
DO I = 1, nx |
| 506 |
C fX: VAreaX*dx[V] |
| 507 |
C fY: VAreaY*dy[V] |
| 508 |
fX(I,J) = vGradxOp(I,J)*(vK(I,J)-vK(I-1,J)) |
| 509 |
fY(I,J) = vGradyOp(I,J)*(vK(I,J+1)-vK(I,J)) |
| 510 |
ENDDO |
| 511 |
ENDDO |
| 512 |
fX(nx+1,1:ny) = fX(1 ,1:ny) |
| 513 |
fY(1:nx,0 ) = fY(1:nx,ny) |
| 514 |
DO J=1,ny |
| 515 |
DO I=1,nx |
| 516 |
C o Del^2 V ( for use in biharmonic mixing ). |
| 517 |
C tmp: div(fX,fY) = 1/VVolume*(dx[fX]+dy[fy]) |
| 518 |
tmp(I,J)= fX(I+1,J)*rVk(I,J)-fX(I,J)*rVk(I,J) |
| 519 |
& +fY(I,J)*rVk(I,J)-fY(I,J-1)*rVk(I,J) |
| 520 |
ENDDO |
| 521 |
ENDDO |
| 522 |
tmp(0,1:ny) = tmp(nx,1:ny) |
| 523 |
tmp(:,ny+1) = tmp(:,1) |
| 524 |
DO J = 1,ny |
| 525 |
DO I = 1,nx |
| 526 |
C o Laplacian<Line 1> + Biharmonic<Line 2> + Advection<Line 3,4>. |
| 527 |
C fX: -K2xyMomentum*fX + K4xyMomentum*VAreaX*dx[tmp] + By[Ax*U]*Bx[V] |
| 528 |
C fY: -K2xyMomentum*fY + K4xyMomentum*VAreaY*dy[tmp] + By[Ay*V]*By[V] |
| 529 |
fX(I,J) = 0. |
| 530 |
_LPM(& -_DB1*a2MomXY*fX(I,J) ) |
| 531 |
_BHM(& +a4MomXY*vGradxOp(I,J)*(tmp(I,J)-tmp(I-1,J)) ) |
| 532 |
_ADM(& +0.25D0*(uK(I,J-1)*XAK(I,J-1)+uK(I,J)*XAK(I,J))* ) |
| 533 |
_ADM(& (vK(I-1,J)+vK(I,J)) ) |
| 534 |
_ADM(& *vMskK(I,J)*vMskK(I-1,J) ) |
| 535 |
C Note!!!! The last line above "*vMskK(I,J)*vMskK(I-1,J)" is |
| 536 |
C Note!!!! boundary condition in the standard v010 CM5 code. The |
| 537 |
C Note!!!! FV paper used a different bc in which the last line is |
| 538 |
C Note!!!! omitted. |
| 539 |
fY(I,J) = 0. |
| 540 |
_LPM(& -_DB1*a2MomXY*fY(I,J) ) |
| 541 |
_BHM(& +a4MomXY*vGradyOp(I,J)*(tmp(I,J+1)-tmp(I,J)) ) |
| 542 |
_ADM(& +0.25D0*(vK(I,J+1)*YAK(I,J+1)+vK(I,J)*YAK(I,J))* ) |
| 543 |
_ADM(& (vK(I,J+1)+vK(I,J)) ) |
| 544 |
C Note!!!! The last line above "*vMskK(I,J)*vMskK(I-1,J)" is |
| 545 |
ENDDO |
| 546 |
ENDDO |
| 547 |
C === Vertical diffusive flux of V ==================================== |
| 548 |
C === Vertical advective flux of V ==================================== |
| 549 |
C o Laplacian Diffusion<Line 1> + Advection<Line 2,3> |
| 550 |
C fZK: -KzMomentum*VAreaZ*dz[V] + Bz[V]*By[Az*W] |
| 551 |
IF ( .NOT. TOP_LAYER ) THEN |
| 552 |
c WRITE(0,*) ' fZK subsurface layer ' |
| 553 |
DO J = 1, ny |
| 554 |
DO I = 1, nx |
| 555 |
fZK(I,J) = 0. |
| 556 |
_LPM(& -_DB1*a2MomZ*vGradzOpK(I,J)*(V(_I3(K-1,I,J))-vK(I,J) ) ) |
| 557 |
_ADM(& +0.25D0*(vK(I,J)+V(_I3(K-1,I,J)))* ) |
| 558 |
_ADM(& (-wK(I,J)*ZAK(I,J)-wK(I,J-1)*ZAK(I,J-1)) ) |
| 559 |
_ADM(& *vMskK(I,J)*vMask(_I3(K-1,I,J))*_wdvdz ) |
| 560 |
C Note!!!! The last line above "*vMskK(I,J)*vMask(_I3(K-1,I,J))" is |
| 561 |
C Note!!!! boundary condition in the standard v010 CM5 code. The |
| 562 |
C Note!!!! FV paper used a different bc in which the last line is |
| 563 |
C Note!!!! omitted. |
| 564 |
ENDDO |
| 565 |
ENDDO |
| 566 |
ELSE |
| 567 |
c WRITE(0,*) ' fZK top layer ' |
| 568 |
DO J = 1, ny |
| 569 |
DO I = 1, nx |
| 570 |
fZK(I,J) = 0.D0 |
| 571 |
_ADM(& +freeSurfFac*0.25D0*(vK(I,J)+vK(I,J)) ) |
| 572 |
_ADM(& *(-wK(I,J)*ZAK(I,J)-wK(I,J-1)*ZAK(I,J-1))*_wdvdz ) |
| 573 |
ENDDO |
| 574 |
ENDDO |
| 575 |
ENDIF |
| 576 |
IF ( .NOT. BOTTOM_LAYER ) THEN |
| 577 |
DO J = 1, ny |
| 578 |
DO I = 1, nx |
| 579 |
fZKP1(I,J) = 0. |
| 580 |
_LPM(& -_DB1*a2MomZ*vGradzOpKP1(I,J)*( vK(I,J)-V(_I3(K+1,I,J))) ) |
| 581 |
_ADM(& +0.25D0*(vK(I,J)+V(_I3(K+1,I,J)))* ) |
| 582 |
_ADM(& (-wKP1(I,J)*ZAKP1(I,J)-wKP1(I,J-1)*ZAKP1(I,J-1)) ) |
| 583 |
_ADM(& *vMskK(I,J)*vMask(_I3(K+1,I,J))*_wdvdz ) |
| 584 |
C Note!!!! The last line above "*vMskK(I,J)*vMask(_I3(K+1,I,J))" is |
| 585 |
C Note!!!! boundary condition in the standard v010 CM5 code. The |
| 586 |
C Note!!!! FV paper used a different bc in which the last line is |
| 587 |
C Note!!!! omitted. |
| 588 |
ENDDO |
| 589 |
ENDDO |
| 590 |
ELSE |
| 591 |
DO J = 1, ny |
| 592 |
DO I = 1, nx |
| 593 |
fZKP1(I,J) = 0. |
| 594 |
_LPM(& -_DB1*a2MomZ*rDzAtP(K)*2.*vK(I,J) ) |
| 595 |
_LPM(& *0.5*( ZAk(I,J)+ZAk(I,J-1) ) ) |
| 596 |
|
| 597 |
ENDDO |
| 598 |
ENDDO |
| 599 |
ENDIF |
| 600 |
c WRITE(0,*) 'K, MAXVAL(fZK) = ', K, MAXVAL(fZK) |
| 601 |
c WRITE(0,*) 'K, MAXVAL(fZKP1) = ', K, MAXVAL(fZKP1) |
| 602 |
C === Combine advective and diffusive fluxes of V ===================== |
| 603 |
C o V tendency is minus divergence of diffusive and advective flux. |
| 604 |
C gV: -div(fX,fY,fZ) = -1/VVolume*(dx[fX]+dy[fY]+fZK-fZKP1) |
| 605 |
fX(nx+1,1:ny) = fX(1,1:ny) |
| 606 |
fY(:,0) = fY(:,ny) |
| 607 |
DO I = 1, nx |
| 608 |
DO J = 1, ny |
| 609 |
gV(_I3(K,I,J))=-fX(I+1,J)*rVK(I,J)+fX(I,J)*rVK(I,J) |
| 610 |
& -fY(I,J)*rVK(I,J)+fY(I,J-1)*rVK(I,J) |
| 611 |
ENDDO |
| 612 |
ENDDO |
| 613 |
DO J = 1, ny |
| 614 |
DO I = 1, nx |
| 615 |
gV(_I3(K,I,J))=gV(_I3(K,I,J))-fZK(I,J)*rVK(I,J) |
| 616 |
ENDDO |
| 617 |
ENDDO |
| 618 |
CcnhDebugStarts |
| 619 |
C Changed to provide noslip bottom boundary |
| 620 |
Cdbg IF ( .NOT. BOTTOM_LAYER ) THEN |
| 621 |
CcnhDebugEnds |
| 622 |
DO J = 1, ny |
| 623 |
DO I = 1, nx |
| 624 |
gV(_I3(K,I,J))=gV(_I3(K,I,J))+fZKP1(I,J)*rVK(I,J) |
| 625 |
ENDDO |
| 626 |
ENDDO |
| 627 |
CcnhDebugStarts |
| 628 |
Cdbg ENDIF |
| 629 |
CcnhDebugEnds |
| 630 |
C === Metric terms due to curvilinear coordinates ===================== |
| 631 |
DO J = 1, ny |
| 632 |
DO I = 1, nx |
| 633 |
C o Spherical polar metric terms. |
| 634 |
C gV: gV - V/radius*ByBz[w]+tan(latitude)*BxBy[u] |
| 635 |
gV(_I3(K,I,J)) = gV(_I3(K,I,J)) |
| 636 |
_SPM(& -vMask(_I3(K,I,J))*vK(I,J)/rZero* ) |
| 637 |
_SPM(& 0.25D0*(wK(I,J-1)+wK(I,J)+wKP1(I,J-1)+wKP1(I,J))/(-G*RONIL) ) |
| 638 |
_SPM(& -vMask(_I3(K,I,J))*vTphiOP(_I3(K,I,J))/rZero*0.25D0*0.25D0 ) |
| 639 |
_SPM(& *(uK(I,J)+uK(I+1,J)+uK(I,J-1)+uK(I+1,J-1)) ) |
| 640 |
_SPM(& *(uK(I,J)+uK(I+1,J)+uK(I,J-1)+uK(I+1,J-1)) ) |
| 641 |
ENDDO |
| 642 |
ENDDO |
| 643 |
|
| 644 |
#ifdef _MOMENTUM_FORCING |
| 645 |
IF ( TOP_LAYER ) THEN |
| 646 |
gV(_I3(K,:,:)) = gV(_I3(K,:,:)) |
| 647 |
& + fv |
| 648 |
ENDIF |
| 649 |
#endif |
| 650 |
|
| 651 |
C Hydrostatic pressure term |
| 652 |
DO J=1,Ny |
| 653 |
DO I=1,Nx |
| 654 |
tmp(I,J)=G*PH(_I3(K,I,J)) |
| 655 |
ENDDO |
| 656 |
ENDDO |
| 657 |
DO J=1,Ny |
| 658 |
tmp(0,J)=tmp(Nx,J) |
| 659 |
ENDDO |
| 660 |
DO I=1,Nx |
| 661 |
tmp(I,0)=tmp(I,Ny) |
| 662 |
ENDDO |
| 663 |
DO J = 1, ny |
| 664 |
DO I = 1, nx |
| 665 |
gU(_I3(K,I,J)) = gU(_I3(K,I,J)) -pDdxOp(_I3(K,I,J))*(tmp(I,J)-tmp(I-1,J)) |
| 666 |
gV(_I3(K,I,J)) = gV(_I3(K,I,J)) -pDdyOp(_I3(K,I,J))*(tmp(I,J)-tmp(I,J-1)) |
| 667 |
ENDDO |
| 668 |
ENDDO |
| 669 |
|
| 670 |
|
| 671 |
#ifdef _XY_CORIOLIS |
| 672 |
C /---------------------------------------------------------------\ |
| 673 |
C | Cd Scheme | |
| 674 |
C |===============================================================| |
| 675 |
C | o Gu = Gu + GuXYCoriolis | |
| 676 |
C | o Gv = Gv + GvXYCoriolis | |
| 677 |
C \---------------------------------------------------------------/ |
| 678 |
C /---------------------------------------------------------------\ |
| 679 |
C | tmp2 <- P(n) = P extrapolated to time level n. | |
| 680 |
C \---------------------------------------------------------------/ |
| 681 |
tmp2(1:nx,1:ny) = (1.5+abEps)*(pS+pH(_I3(K,:,:)) ) |
| 682 |
& -(0.5+abEps)*(psNM1+phNM1(_I3(K,:,:)) ) |
| 683 |
tmp2(1:nx,1:ny) = (1.5+abEps)*(pS ) |
| 684 |
& -(0.5+abEps)*(psNM1 ) |
| 685 |
tmp2(0,1:ny) = tmp2(nx,1:ny) |
| 686 |
DO J = 1,ny |
| 687 |
DO I = 1,nx |
| 688 |
C /-------------------------------------------------------------\ |
| 689 |
C | tmp <- -G*Ddx{P(n)} + gU | |
| 690 |
C \-------------------------------------------------------------/ |
| 691 |
tmp(I,J) = |
| 692 |
& -G*pDdxOp(_I3(K,I,J))*(tmp2(I,J)-tmp2(I-1,J))+ |
| 693 |
& gU(_I3(K,I,J))*uMask(_I3(K,I,J)) |
| 694 |
ENDDO |
| 695 |
ENDDO |
| 696 |
tmp(nx+1,1:ny) = tmp(1,1:ny) |
| 697 |
tmp(:,0 ) = tmp(:,ny ) |
| 698 |
DO J = 1,ny |
| 699 |
DO I = 1,nx |
| 700 |
C /-------------------------------------------------------------\ |
| 701 |
C | uAja <- uAja + DT*BxBy{-d/dx P(n) + gU}+f*V | |
| 702 |
C \-------------------------------------------------------------/ |
| 703 |
uAja(_I3(K,I,J)) = uAja(_I3(K,I,J)) |
| 704 |
& +1.*DelT*(0.*tmp(I,J)+ |
| 705 |
& 1.*0.25D0*(tmp(I,J)+tmp(I+1,J)+tmp(I,J-1)+tmp(I+1,J-1)) |
| 706 |
& +fCorV(_I3(K,I,J))*V(_I3(K,I,J)) ) |
| 707 |
C uAja: uAja - lambda(uAja-BxBy[U(n+0.5)] ( note rAja = (1-lambda) ) |
| 708 |
uAja(_I3(K,I,J)) = rAja*uAja(_I3(K,I,J)) |
| 709 |
& +(1.-rAja)*( (1.5+abEps)*0.25D0*(uK(I,J)+uK(I+1,J)+uK(I,J-1)+uK(I+1,J-1)) |
| 710 |
& -(0.5+abEps)*0.25D0*(uKNM1(I,J)+uKNM1(I+1,J)+uKNM1(I,J-1)+uKNM1(I+1,J-1)) ) |
| 711 |
ENDDO |
| 712 |
ENDDO |
| 713 |
CcnhDebugStarts |
| 714 |
C WRITE(0,*) ' CG2D MAXVAL(uAja) ', MAXVAL(uAja) |
| 715 |
CcnhDebugEnds |
| 716 |
C tmp2: P(n) = P extrapolated to time level n. |
| 717 |
tmp2(1:nx,1:ny) = (1.5+abEps)*(pS+pH(_I3(K,:,:)) ) |
| 718 |
& -(0.5+abEps)*(psNM1+phNM1(_I3(K,:,:)) ) |
| 719 |
tmp2(1:nx,1:ny) = (1.5+abEps)*(pS ) |
| 720 |
& -(0.5+abEps)*(psNM1 ) |
| 721 |
tmp2(1:nx,0) = tmp2(1:nx,ny) |
| 722 |
DO J = 1,ny |
| 723 |
DO I = 1,nx |
| 724 |
C tmp: -d/dy P(n) + gV |
| 725 |
tmp(I,J) = -G*pDdyOp(_I3(K,I,J))*(tmp2(I,J)-tmp2(I,J-1))+ |
| 726 |
& gV(_I3(K,I,J))*vMask(_I3(K,I,J)) |
| 727 |
ENDDO |
| 728 |
ENDDO |
| 729 |
tmp(0,1:ny) = tmp(nx,1:ny) |
| 730 |
tmp(:,ny+1) = tmp(: ,1 ) |
| 731 |
DO J = 1,ny |
| 732 |
DO I = 1,nx |
| 733 |
C vAja: vAja + DT*(BxBy[-d/dy P(n) + gV]-f*U) |
| 734 |
vAja(_I3(K,I,J)) = vAja(_I3(K,I,J)) |
| 735 |
& +DelT*(0.25D0*(tmp(I-1,J)+tmp(I,J)+tmp(I-1,J+1)+tmp(I,J+1)) |
| 736 |
& -fCorU(_I3(K,I,J))*U(_I3(K,I,J)) ) |
| 737 |
C vAja: vAja - lambda(vAja-BxBy[V(n+0.5)] ( note rAja = (1-lambda) ) |
| 738 |
vAja(_I3(K,I,J)) = rAja*vAja(_I3(K,I,J)) |
| 739 |
& +(1.-rAja)*( (1.5+abEps)*0.25D0*(vK(I,J)+vK(I-1,J)+vK(I,J+1)+vK(I-1,J+1)) |
| 740 |
& -(0.5+abEps)*0.25D0*(vKNM1(I,J)+vKNM1(I-1,J)+vKNM1(I,J+1)+vKNM1(I-1,J+1)) ) |
| 741 |
ENDDO |
| 742 |
ENDDO |
| 743 |
#endif |
| 744 |
|
| 745 |
C Add coriolis term from Cd grid into Gu and Gv. |
| 746 |
C Note: AB2 does not apply to Cd variables. |
| 747 |
gU(_I3(K,:,:)) = gU(_I3(K,:,:))*uMask(_I3(K,:,:)) |
| 748 |
tmp(1:nx,1:ny) = gUNM1(_I3(K,:,:)) |
| 749 |
gUNM1(_I3(K,:,:)) = gU(_I3(K,:,:)) |
| 750 |
gU(_I3(K,:,:)) = |
| 751 |
& (1.5+abEps)*gU(_I3(K,:,:))-(0.5+abEps)*tmp(1:nx,1:ny) |
| 752 |
& +fCorU(_I3(K,:,:))*vAja(_I3(K,:,:))*uMask(_I3(K,:,:)) |
| 753 |
IF ( BOTTOM_LAYER ) THEN |
| 754 |
_D(( ' S/R G_CALC: MAXVAL(GU)', MAXVAL(GU) )) |
| 755 |
_D(( ' MAXVAL(GU) @', MAXLOC(GU) )) |
| 756 |
ENDIF |
| 757 |
|
| 758 |
gV(_I3(K,:,:)) = gV(_I3(K,:,:))*vMask(_I3(K,:,:)) |
| 759 |
tmp(1:nx,1:ny) = gVNM1(_I3(K,:,:)) |
| 760 |
gVNM1(_I3(K,:,:)) = gV(_I3(K,:,:)) |
| 761 |
gV(_I3(K,:,:)) = |
| 762 |
& (1.5+abEps)*gV(_I3(K,:,:))-(0.5+abEps)*tmp(1:nx,1:ny) |
| 763 |
& -fCorV(_I3(K,:,:))*uAja(_I3(K,:,:))*vMask(_I3(K,:,:)) |
| 764 |
_D(( ' S/R G_CALC: MAXVAL(GV)', MAXVAL(GV) )) |
| 765 |
C ********************************************************************* |
| 766 |
C *** FILTER SECTION ************************************************** |
| 767 |
C ********************************************************************* |
| 768 |
C Apply filter to control energy build up at high-latitudes. |
| 769 |
|
| 770 |
C ********************************************************************* |
| 771 |
C *** RHS SECTION ***************************************************** |
| 772 |
C ********************************************************************* |
| 773 |
C Form right-hand side of elliptic eqaution |
| 774 |
C RHS2d: dx[Ax*Gu]+dy[Ay*Gv] |
| 775 |
C +1/delt*(dx[Ax*U]+dy[Ay*V]) |
| 776 |
tmp(1:nx,1:ny) = |
| 777 |
& gU(_I3(K,:,:))/RONIL/G |
| 778 |
& +rDelt*Uk(1:nx,1:ny)/RONIL/G |
| 779 |
tmp(nx+1,1:ny) = tmp(1,1:ny) |
| 780 |
DO J=1,ny |
| 781 |
DO I=1,nx |
| 782 |
RHS2d(I,J)=RHS2d(I,J)+XAk(I+1,J)*tmp(I+1,J)-XAk(I,J)*tmp(I,J) |
| 783 |
ENDDO |
| 784 |
ENDDO |
| 785 |
tmp(1:nx,1:ny) = |
| 786 |
& gV(_I3(K,:,:))/RONIL/G |
| 787 |
& +rDelt*Vk(1:nx,1:ny)/RONIL/G |
| 788 |
tmp(1:nx,ny+1) = tmp(1:nx,1) |
| 789 |
DO J=1,ny |
| 790 |
DO I=1,nx |
| 791 |
RHS2d(I,J)=RHS2d(I,J)+YAk(I,J+1)*tmp(I,J+1)-YAk(I,J)*tmp(I,J) |
| 792 |
ENDDO |
| 793 |
ENDDO |
| 794 |
CcnhDebugStarts |
| 795 |
IF ( BOTTOM_LAYER ) THEN |
| 796 |
C DO I=1,Nx |
| 797 |
C WRITE(0,*) ' CG2D RHS2d(I,30) ', rhs2d(I,30), I |
| 798 |
C ENDDO |
| 799 |
C DO J=1,Ny |
| 800 |
C WRITE(0,*) ' CG2D MAXVAL(RHS2d) ', |
| 801 |
C & MAXVAL(RHS2d(1:Nx,J)), J |
| 802 |
C WRITE(0,*) ' CG2D MINVAL(RHS2d) ', |
| 803 |
C & MINVAL(RHS2d(1:Nx,J)), J |
| 804 |
C & MAXVAL(ABS(fCorV(_I3(:,I,:))*uAja(_I3(:,I,:))*vMask(_I3(:,I,:)))) |
| 805 |
C ENDDO |
| 806 |
C CALL PLOT_FIELD( RHS2d(1:nx,1:ny), nX, nY ) |
| 807 |
C STOP |
| 808 |
ENDIF |
| 809 |
CcnhDebugEnds |
| 810 |
DO J=1,ny |
| 811 |
DO I=1,nx |
| 812 |
divH(I,J)= divH(I,J)+ |
| 813 |
& XAk(I+1,J)*uK(I+1,J)-XAk(I,J)*uK(I,J) |
| 814 |
& +YAk(I,J+1)*vK(I,J+1)-YAk(I,J)*vK(I,J) |
| 815 |
ENDDO |
| 816 |
ENDDO |
| 817 |
|
| 818 |
RETURN |
| 819 |
END |
Parent Directory
| 
Revision Log
| 
Revision Graph