1 |
C $Header: /u/gcmpack/MITgcm/pkg/generic_advdiff/gad_advection.F,v 1.15 2003/06/27 01:57:28 heimbach Exp $ |
2 |
C $Name: $ |
3 |
|
4 |
CBOI |
5 |
C !TITLE: pkg/generic\_advdiff |
6 |
C !AUTHORS: adcroft@mit.edu |
7 |
C !INTRODUCTION: Generic Advection Diffusion Package |
8 |
C |
9 |
C Package "generic\_advdiff" provides a common set of routines for calculating |
10 |
C advective/diffusive fluxes for tracers (cell centered quantities on a C-grid). |
11 |
C |
12 |
C Many different advection schemes are available: the standard centered |
13 |
C second order, centered fourth order and upwind biased third order schemes |
14 |
C are known as linear methods and require some stable time-stepping method |
15 |
C such as Adams-Bashforth. Alternatives such as flux-limited schemes are |
16 |
C stable in the forward sense and are best combined with the multi-dimensional |
17 |
C method provided in gad\_advection. |
18 |
C |
19 |
C There are two high-level routines: |
20 |
C \begin{itemize} |
21 |
C \item{GAD\_CALC\_RHS} calculates all fluxes at time level "n" and is used |
22 |
C for the standard linear schemes. This must be used in conjuction with |
23 |
C Adams-Bashforth time-stepping. Diffusive and parameterized fluxes are |
24 |
C always calculated here. |
25 |
C \item{GAD\_ADVECTION} calculates just the advective fluxes using the |
26 |
C non-linear schemes and can not be used in conjuction with Adams-Bashforth |
27 |
C time-stepping. |
28 |
C \end{itemize} |
29 |
CEOI |
30 |
|
31 |
#include "PACKAGES_CONFIG.h" |
32 |
#include "GAD_OPTIONS.h" |
33 |
#ifdef ALLOW_AUTODIFF |
34 |
# include "CPP_OPTIONS.h" |
35 |
#endif |
36 |
|
37 |
CBOP |
38 |
C !ROUTINE: GAD_ADVECTION |
39 |
|
40 |
C !INTERFACE: ========================================================== |
41 |
SUBROUTINE GAD_ADVECTION(bi,bj,advectionScheme,tracerIdentity, |
42 |
U Tracer,Gtracer, |
43 |
I myTime,myIter,myThid) |
44 |
|
45 |
C !DESCRIPTION: |
46 |
C Calculates the tendancy of a tracer due to advection. |
47 |
C It uses the multi-dimensional method given in \ref{sect:multiDimAdvection} |
48 |
C and can only be used for the non-linear advection schemes such as the |
49 |
C direct-space-time method and flux-limiters. |
50 |
C |
51 |
C The algorithm is as follows: |
52 |
C \begin{itemize} |
53 |
C \item{$\theta^{(n+1/3)} = \theta^{(n)} |
54 |
C - \Delta t \partial_x (u\theta^{(n)}) + \theta^{(n)} \partial_x u$} |
55 |
C \item{$\theta^{(n+2/3)} = \theta^{(n+1/3)} |
56 |
C - \Delta t \partial_y (v\theta^{(n+1/3)}) + \theta^{(n)} \partial_y v$} |
57 |
C \item{$\theta^{(n+3/3)} = \theta^{(n+2/3)} |
58 |
C - \Delta t \partial_r (w\theta^{(n+2/3)}) + \theta^{(n)} \partial_r w$} |
59 |
C \item{$G_\theta = ( \theta^{(n+3/3)} - \theta^{(n)} )/\Delta t$} |
60 |
C \end{itemize} |
61 |
C |
62 |
C The tendancy (output) is over-written by this routine. |
63 |
|
64 |
C !USES: =============================================================== |
65 |
IMPLICIT NONE |
66 |
#include "SIZE.h" |
67 |
#include "EEPARAMS.h" |
68 |
#include "PARAMS.h" |
69 |
#include "DYNVARS.h" |
70 |
#include "GRID.h" |
71 |
#include "GAD.h" |
72 |
#ifdef ALLOW_AUTODIFF_TAMC |
73 |
# include "tamc.h" |
74 |
# include "tamc_keys.h" |
75 |
#endif |
76 |
|
77 |
C !INPUT PARAMETERS: =================================================== |
78 |
C bi,bj :: tile indices |
79 |
C advectionScheme :: advection scheme to use |
80 |
C tracerIdentity :: identifier for the tracer (required only for OBCS) |
81 |
C Tracer :: tracer field |
82 |
C myTime :: current time |
83 |
C myIter :: iteration number |
84 |
C myThid :: thread number |
85 |
INTEGER bi,bj |
86 |
INTEGER advectionScheme |
87 |
INTEGER tracerIdentity |
88 |
_RL Tracer(1-Olx:sNx+Olx,1-Oly:sNy+Oly,Nr,nSx,nSy) |
89 |
_RL myTime |
90 |
INTEGER myIter |
91 |
INTEGER myThid |
92 |
|
93 |
C !OUTPUT PARAMETERS: ================================================== |
94 |
C gTracer :: tendancy array |
95 |
_RL gTracer(1-Olx:sNx+Olx,1-Oly:sNy+Oly,Nr,nSx,nSy) |
96 |
|
97 |
C !LOCAL VARIABLES: ==================================================== |
98 |
C maskUp :: 2-D array for mask at W points |
99 |
C iMin,iMax,jMin,jMax :: loop range for called routines |
100 |
C i,j,k :: loop indices |
101 |
C kup :: index into 2 1/2D array, toggles between 1 and 2 |
102 |
C kdown :: index into 2 1/2D array, toggles between 2 and 1 |
103 |
C kp1 :: =k+1 for k<Nr, =Nr for k=Nr |
104 |
C xA,yA :: areas of X and Y face of tracer cells |
105 |
C uTrans,vTrans,rTrans :: 2-D arrays of volume transports at U,V and W points |
106 |
C rTransKp1 :: vertical volume transport at interface k+1 |
107 |
C af :: 2-D array for horizontal advective flux |
108 |
C fVerT :: 2 1/2D arrays for vertical advective flux |
109 |
C localTij :: 2-D array used as temporary local copy of tracer fld |
110 |
C localTijk :: 3-D array used as temporary local copy of tracer fld |
111 |
C kp1Msk :: flag (0,1) to act as over-riding mask for W levels |
112 |
C calc_fluxes_X :: logical to indicate to calculate fluxes in X dir |
113 |
C calc_fluxes_Y :: logical to indicate to calculate fluxes in Y dir |
114 |
C nipass :: number of passes to make in multi-dimensional method |
115 |
C ipass :: number of the current pass being made |
116 |
_RS maskUp (1-OLx:sNx+OLx,1-OLy:sNy+OLy) |
117 |
INTEGER iMin,iMax,jMin,jMax |
118 |
INTEGER i,j,k,kup,kDown |
119 |
_RS xA (1-OLx:sNx+OLx,1-OLy:sNy+OLy) |
120 |
_RS yA (1-OLx:sNx+OLx,1-OLy:sNy+OLy) |
121 |
_RL uTrans (1-OLx:sNx+OLx,1-OLy:sNy+OLy) |
122 |
_RL vTrans (1-OLx:sNx+OLx,1-OLy:sNy+OLy) |
123 |
_RL rTrans (1-OLx:sNx+OLx,1-OLy:sNy+OLy) |
124 |
_RL rTransKp1(1-OLx:sNx+OLx,1-OLy:sNy+OLy) |
125 |
_RL af (1-OLx:sNx+OLx,1-OLy:sNy+OLy) |
126 |
_RL fVerT (1-OLx:sNx+OLx,1-OLy:sNy+OLy,2) |
127 |
_RL localTij(1-OLx:sNx+OLx,1-OLy:sNy+OLy) |
128 |
_RL localTijk(1-OLx:sNx+OLx,1-OLy:sNy+OLy,Nr) |
129 |
_RL kp1Msk |
130 |
LOGICAL calc_fluxes_X,calc_fluxes_Y |
131 |
INTEGER nipass,ipass |
132 |
CEOP |
133 |
|
134 |
#ifdef ALLOW_AUTODIFF_TAMC |
135 |
act0 = tracerIdentity - 1 |
136 |
max0 = maxpass |
137 |
act1 = bi - myBxLo(myThid) |
138 |
max1 = myBxHi(myThid) - myBxLo(myThid) + 1 |
139 |
act2 = bj - myByLo(myThid) |
140 |
max2 = myByHi(myThid) - myByLo(myThid) + 1 |
141 |
act3 = myThid - 1 |
142 |
max3 = nTx*nTy |
143 |
act4 = ikey_dynamics - 1 |
144 |
igadkey = (act0 + 1) |
145 |
& + act1*max0 |
146 |
& + act2*max0*max1 |
147 |
& + act3*max0*max1*max2 |
148 |
& + act4*max0*max1*max2*max3 |
149 |
if (tracerIdentity.GT.maxpass) then |
150 |
print *, 'ph-pass gad_advection ', maxpass, tracerIdentity |
151 |
STOP 'maxpass seems smaller than tracerIdentity' |
152 |
endif |
153 |
#endif /* ALLOW_AUTODIFF_TAMC */ |
154 |
|
155 |
C-- Set up work arrays with valid (i.e. not NaN) values |
156 |
C These inital values do not alter the numerical results. They |
157 |
C just ensure that all memory references are to valid floating |
158 |
C point numbers. This prevents spurious hardware signals due to |
159 |
C uninitialised but inert locations. |
160 |
DO j=1-OLy,sNy+OLy |
161 |
DO i=1-OLx,sNx+OLx |
162 |
xA(i,j) = 0. _d 0 |
163 |
yA(i,j) = 0. _d 0 |
164 |
uTrans(i,j) = 0. _d 0 |
165 |
vTrans(i,j) = 0. _d 0 |
166 |
rTrans(i,j) = 0. _d 0 |
167 |
fVerT(i,j,1) = 0. _d 0 |
168 |
fVerT(i,j,2) = 0. _d 0 |
169 |
rTransKp1(i,j)= 0. _d 0 |
170 |
ENDDO |
171 |
ENDDO |
172 |
|
173 |
iMin = 1-OLx |
174 |
iMax = sNx+OLx |
175 |
jMin = 1-OLy |
176 |
jMax = sNy+OLy |
177 |
|
178 |
C-- Start of k loop for horizontal fluxes |
179 |
DO k=1,Nr |
180 |
#ifdef ALLOW_AUTODIFF_TAMC |
181 |
kkey = (igadkey-1)*Nr + k |
182 |
CADJ STORE tracer(:,:,k,bi,bj) = |
183 |
CADJ & comlev1_bibj_k_gad, key=kkey, byte=isbyte |
184 |
#endif /* ALLOW_AUTODIFF_TAMC */ |
185 |
|
186 |
C-- Get temporary terms used by tendency routines |
187 |
CALL CALC_COMMON_FACTORS ( |
188 |
I bi,bj,iMin,iMax,jMin,jMax,k, |
189 |
O xA,yA,uTrans,vTrans,rTrans,maskUp, |
190 |
I myThid) |
191 |
|
192 |
#ifdef ALLOW_GMREDI |
193 |
C-- Residual transp = Bolus transp + Eulerian transp |
194 |
IF (useGMRedi) |
195 |
& CALL GMREDI_CALC_UVFLOW( |
196 |
& uTrans, vTrans, bi, bj, k, myThid) |
197 |
#endif /* ALLOW_GMREDI */ |
198 |
|
199 |
C-- Make local copy of tracer array |
200 |
DO j=1-OLy,sNy+OLy |
201 |
DO i=1-OLx,sNx+OLx |
202 |
localTij(i,j)=tracer(i,j,k,bi,bj) |
203 |
ENDDO |
204 |
ENDDO |
205 |
|
206 |
IF (useCubedSphereExchange) THEN |
207 |
nipass=3 |
208 |
#ifdef ALLOW_AUTODIFF_TAMC |
209 |
if ( nipass.GT.maxcube ) |
210 |
& STOP 'maxcube needs to be = 3' |
211 |
#endif |
212 |
ELSE |
213 |
nipass=1 |
214 |
ENDIF |
215 |
cph nipass=1 |
216 |
|
217 |
C-- Multiple passes for different directions on different tiles |
218 |
DO ipass=1,nipass |
219 |
#ifdef ALLOW_AUTODIFF_TAMC |
220 |
passkey = ipass + (k-1) *maxcube |
221 |
& + (igadkey-1)*maxcube*Nr |
222 |
IF (nipass .GT. maxpass) THEN |
223 |
STOP 'GAD_ADVECTION: nipass > maxcube. check tamc.h' |
224 |
ENDIF |
225 |
#endif /* ALLOW_AUTODIFF_TAMC */ |
226 |
|
227 |
IF (nipass.EQ.3) THEN |
228 |
calc_fluxes_X=.FALSE. |
229 |
calc_fluxes_Y=.FALSE. |
230 |
IF (ipass.EQ.1 .AND. (bi.EQ.1 .OR. bi.EQ.2) ) THEN |
231 |
calc_fluxes_X=.TRUE. |
232 |
ELSEIF (ipass.EQ.1 .AND. (bi.EQ.4 .OR. bi.EQ.5) ) THEN |
233 |
calc_fluxes_Y=.TRUE. |
234 |
ELSEIF (ipass.EQ.2 .AND. (bi.EQ.1 .OR. bi.EQ.6) ) THEN |
235 |
calc_fluxes_Y=.TRUE. |
236 |
ELSEIF (ipass.EQ.2 .AND. (bi.EQ.3 .OR. bi.EQ.4) ) THEN |
237 |
calc_fluxes_X=.TRUE. |
238 |
ELSEIF (ipass.EQ.3 .AND. (bi.EQ.2 .OR. bi.EQ.3) ) THEN |
239 |
calc_fluxes_Y=.TRUE. |
240 |
ELSEIF (ipass.EQ.3 .AND. (bi.EQ.5 .OR. bi.EQ.6) ) THEN |
241 |
calc_fluxes_X=.TRUE. |
242 |
ENDIF |
243 |
ELSE |
244 |
calc_fluxes_X=.TRUE. |
245 |
calc_fluxes_Y=.TRUE. |
246 |
ENDIF |
247 |
|
248 |
C-- X direction |
249 |
IF (calc_fluxes_X) THEN |
250 |
|
251 |
C-- Internal exchange for calculations in X |
252 |
IF (useCubedSphereExchange) THEN |
253 |
DO j=1,Oly |
254 |
DO i=1,Olx |
255 |
localTij( 1-i , 1-j )=localTij( 1-j , i ) |
256 |
localTij( 1-i ,sNy+j)=localTij( 1-j , sNy+1-i ) |
257 |
localTij(sNx+i, 1-j )=localTij(sNx+j, i ) |
258 |
localTij(sNx+i,sNy+j)=localTij(sNx+j, sNy+1-i ) |
259 |
ENDDO |
260 |
ENDDO |
261 |
ENDIF |
262 |
|
263 |
C- Advective flux in X |
264 |
DO j=1-Oly,sNy+Oly |
265 |
DO i=1-Olx,sNx+Olx |
266 |
af(i,j) = 0. |
267 |
ENDDO |
268 |
ENDDO |
269 |
|
270 |
#ifdef ALLOW_AUTODIFF_TAMC |
271 |
#ifndef DISABLE_MULTIDIM_ADVECTION |
272 |
CADJ STORE localTij(:,:) = |
273 |
CADJ & comlev1_bibj_k_gad_pass, key=passkey, byte=isbyte |
274 |
#endif |
275 |
#endif /* ALLOW_AUTODIFF_TAMC */ |
276 |
|
277 |
IF (advectionScheme.EQ.ENUM_FLUX_LIMIT) THEN |
278 |
CALL GAD_FLUXLIMIT_ADV_X( |
279 |
& bi,bj,k,deltaTtracer,uTrans,uVel,localTij,af,myThid) |
280 |
ELSEIF (advectionScheme.EQ.ENUM_DST3 ) THEN |
281 |
CALL GAD_DST3_ADV_X( |
282 |
& bi,bj,k,deltaTtracer,uTrans,uVel,localTij,af,myThid) |
283 |
ELSEIF (advectionScheme.EQ.ENUM_DST3_FLUX_LIMIT ) THEN |
284 |
CALL GAD_DST3FL_ADV_X( |
285 |
& bi,bj,k,deltaTtracer,uTrans,uVel,localTij,af,myThid) |
286 |
ELSE |
287 |
STOP 'GAD_ADVECTION: adv. scheme incompatibale with multi-dim' |
288 |
ENDIF |
289 |
|
290 |
DO j=1-Oly,sNy+Oly |
291 |
DO i=1-Olx,sNx+Olx-1 |
292 |
localTij(i,j)=localTij(i,j)-deltaTtracer* |
293 |
& _recip_hFacC(i,j,k,bi,bj)*recip_drF(k) |
294 |
& *recip_rA(i,j,bi,bj) |
295 |
& *( af(i+1,j)-af(i,j) |
296 |
& -tracer(i,j,k,bi,bj)*(uTrans(i+1,j)-uTrans(i,j)) |
297 |
& ) |
298 |
ENDDO |
299 |
ENDDO |
300 |
|
301 |
#ifdef ALLOW_OBCS |
302 |
C-- Apply open boundary conditions |
303 |
IF (useOBCS) THEN |
304 |
IF (tracerIdentity.EQ.GAD_TEMPERATURE) THEN |
305 |
CALL OBCS_APPLY_TLOC( bi, bj, k, localTij, myThid ) |
306 |
ELSEIF (tracerIdentity.EQ.GAD_SALINITY) THEN |
307 |
CALL OBCS_APPLY_SLOC( bi, bj, k, localTij, myThid ) |
308 |
END IF |
309 |
END IF |
310 |
#endif /* ALLOW_OBCS */ |
311 |
|
312 |
C-- End of X direction |
313 |
ENDIF |
314 |
|
315 |
C-- Y direction |
316 |
IF (calc_fluxes_Y) THEN |
317 |
|
318 |
C-- Internal exchange for calculations in Y |
319 |
IF (useCubedSphereExchange) THEN |
320 |
DO j=1,Oly |
321 |
DO i=1,Olx |
322 |
localTij( 1-i , 1-j )=localTij( j , 1-i ) |
323 |
localTij( 1-i ,sNy+j)=localTij( j ,sNy+i) |
324 |
localTij(sNx+i, 1-j )=localTij(sNx+1-j, 1-i ) |
325 |
localTij(sNx+i,sNy+j)=localTij(sNx+1-j,sNy+i) |
326 |
ENDDO |
327 |
ENDDO |
328 |
ENDIF |
329 |
|
330 |
C- Advective flux in Y |
331 |
DO j=1-Oly,sNy+Oly |
332 |
DO i=1-Olx,sNx+Olx |
333 |
af(i,j) = 0. |
334 |
ENDDO |
335 |
ENDDO |
336 |
|
337 |
#ifdef ALLOW_AUTODIFF_TAMC |
338 |
#ifndef DISABLE_MULTIDIM_ADVECTION |
339 |
CADJ STORE localTij(:,:) = |
340 |
CADJ & comlev1_bibj_k_gad_pass, key=passkey, byte=isbyte |
341 |
#endif |
342 |
#endif /* ALLOW_AUTODIFF_TAMC */ |
343 |
|
344 |
IF (advectionScheme.EQ.ENUM_FLUX_LIMIT) THEN |
345 |
CALL GAD_FLUXLIMIT_ADV_Y( |
346 |
& bi,bj,k,deltaTtracer,vTrans,vVel,localTij,af,myThid) |
347 |
ELSEIF (advectionScheme.EQ.ENUM_DST3 ) THEN |
348 |
CALL GAD_DST3_ADV_Y( |
349 |
& bi,bj,k,deltaTtracer,vTrans,vVel,localTij,af,myThid) |
350 |
ELSEIF (advectionScheme.EQ.ENUM_DST3_FLUX_LIMIT ) THEN |
351 |
CALL GAD_DST3FL_ADV_Y( |
352 |
& bi,bj,k,deltaTtracer,vTrans,vVel,localTij,af,myThid) |
353 |
ELSE |
354 |
STOP 'GAD_ADVECTION: adv. scheme incompatibale with mutli-dim' |
355 |
ENDIF |
356 |
|
357 |
DO j=1-Oly,sNy+Oly-1 |
358 |
DO i=1-Olx,sNx+Olx |
359 |
localTij(i,j)=localTij(i,j)-deltaTtracer* |
360 |
& _recip_hFacC(i,j,k,bi,bj)*recip_drF(k) |
361 |
& *recip_rA(i,j,bi,bj) |
362 |
& *( af(i,j+1)-af(i,j) |
363 |
& -tracer(i,j,k,bi,bj)*(vTrans(i,j+1)-vTrans(i,j)) |
364 |
& ) |
365 |
ENDDO |
366 |
ENDDO |
367 |
|
368 |
#ifdef ALLOW_OBCS |
369 |
C-- Apply open boundary conditions |
370 |
IF (useOBCS) THEN |
371 |
IF (tracerIdentity.EQ.GAD_TEMPERATURE) THEN |
372 |
CALL OBCS_APPLY_TLOC( bi, bj, k, localTij, myThid ) |
373 |
ELSEIF (tracerIdentity.EQ.GAD_SALINITY) THEN |
374 |
CALL OBCS_APPLY_SLOC( bi, bj, k, localTij, myThid ) |
375 |
END IF |
376 |
END IF |
377 |
#endif /* ALLOW_OBCS */ |
378 |
|
379 |
C-- End of Y direction |
380 |
ENDIF |
381 |
|
382 |
DO j=1-Oly,sNy+Oly |
383 |
DO i=1-Olx,sNx+Olx |
384 |
localTijk(i,j,k)=localTij(i,j) |
385 |
ENDDO |
386 |
ENDDO |
387 |
|
388 |
C-- End of ipass loop |
389 |
ENDDO |
390 |
|
391 |
C-- End of K loop for horizontal fluxes |
392 |
ENDDO |
393 |
|
394 |
C-- Start of k loop for vertical flux |
395 |
DO k=Nr,1,-1 |
396 |
#ifdef ALLOW_AUTODIFF_TAMC |
397 |
kkey = (igadkey-1)*Nr + k |
398 |
#endif /* ALLOW_AUTODIFF_TAMC */ |
399 |
|
400 |
C-- kup Cycles through 1,2 to point to w-layer above |
401 |
C-- kDown Cycles through 2,1 to point to w-layer below |
402 |
kup = 1+MOD(k+1,2) |
403 |
kDown= 1+MOD(k,2) |
404 |
c kp1=min(Nr,k+1) |
405 |
kp1Msk=1. |
406 |
if (k.EQ.Nr) kp1Msk=0. |
407 |
|
408 |
C-- Compute Vertical transport |
409 |
C Note: wVel needs to be masked |
410 |
|
411 |
IF (k.EQ.1) THEN |
412 |
C- Surface interface : |
413 |
|
414 |
DO j=1-Oly,sNy+Oly |
415 |
DO i=1-Olx,sNx+Olx |
416 |
rTransKp1(i,j) = rTrans(i,j) |
417 |
rTrans(i,j) = 0. |
418 |
fVerT(i,j,kUp) = 0. |
419 |
af(i,j) = 0. |
420 |
ENDDO |
421 |
ENDDO |
422 |
|
423 |
ELSE |
424 |
C- Interior interface : |
425 |
DO j=1-Oly,sNy+Oly |
426 |
DO i=1-Olx,sNx+Olx |
427 |
rTransKp1(i,j) = kp1Msk*rTrans(i,j) |
428 |
rTrans(i,j) = wVel(i,j,k,bi,bj)*rA(i,j,bi,bj) |
429 |
& *maskC(i,j,k-1,bi,bj) |
430 |
af(i,j) = 0. |
431 |
ENDDO |
432 |
ENDDO |
433 |
|
434 |
#ifdef ALLOW_GMREDI |
435 |
C-- Residual transp = Bolus transp + Eulerian transp |
436 |
IF (useGMRedi) |
437 |
& CALL GMREDI_CALC_WFLOW( |
438 |
& rTrans, bi, bj, k, myThid) |
439 |
#endif /* ALLOW_GMREDI */ |
440 |
|
441 |
#ifdef ALLOW_AUTODIFF_TAMC |
442 |
CADJ STORE localTijk(:,:,k) |
443 |
CADJ & = comlev1_bibj_k_gad, key=kkey, byte=isbyte |
444 |
CADJ STORE rTrans(:,:) |
445 |
CADJ & = comlev1_bibj_k_gad, key=kkey, byte=isbyte |
446 |
#endif /* ALLOW_AUTODIFF_TAMC */ |
447 |
|
448 |
C- Compute vertical advective flux in the interior: |
449 |
IF (advectionScheme.EQ.ENUM_FLUX_LIMIT) THEN |
450 |
CALL GAD_FLUXLIMIT_ADV_R( |
451 |
& bi,bj,k,deltaTtracer,rTrans,wVel,localTijk,af,myThid) |
452 |
ELSEIF (advectionScheme.EQ.ENUM_DST3 ) THEN |
453 |
CALL GAD_DST3_ADV_R( |
454 |
& bi,bj,k,deltaTtracer,rTrans,wVel,localTijk,af,myThid) |
455 |
ELSEIF (advectionScheme.EQ.ENUM_DST3_FLUX_LIMIT ) THEN |
456 |
CALL GAD_DST3FL_ADV_R( |
457 |
& bi,bj,k,deltaTtracer,rTrans,wVel,localTijk,af,myThid) |
458 |
ELSE |
459 |
STOP 'GAD_ADVECTION: adv. scheme incompatibale with mutli-dim' |
460 |
ENDIF |
461 |
C- add the advective flux to fVerT |
462 |
DO j=1-Oly,sNy+Oly |
463 |
DO i=1-Olx,sNx+Olx |
464 |
fVerT(i,j,kUp) = af(i,j) |
465 |
ENDDO |
466 |
ENDDO |
467 |
|
468 |
C- end Surface/Interior if bloc |
469 |
ENDIF |
470 |
|
471 |
#ifdef ALLOW_AUTODIFF_TAMC |
472 |
CADJ STORE rTrans(:,:) |
473 |
CADJ & = comlev1_bibj_k_gad, key=kkey, byte=isbyte |
474 |
CADJ STORE rTranskp1(:,:) |
475 |
CADJ & = comlev1_bibj_k_gad, key=kkey, byte=isbyte |
476 |
#endif /* ALLOW_AUTODIFF_TAMC */ |
477 |
|
478 |
C-- Divergence of fluxes |
479 |
DO j=1-Oly,sNy+Oly |
480 |
DO i=1-Olx,sNx+Olx |
481 |
localTij(i,j)=localTijk(i,j,k)-deltaTtracer* |
482 |
& _recip_hFacC(i,j,k,bi,bj)*recip_drF(k) |
483 |
& *recip_rA(i,j,bi,bj) |
484 |
& *( fVerT(i,j,kUp)-fVerT(i,j,kDown) |
485 |
& -tracer(i,j,k,bi,bj)*(rTrans(i,j)-rTransKp1(i,j)) |
486 |
& )*rkFac |
487 |
gTracer(i,j,k,bi,bj)= |
488 |
& (localTij(i,j)-tracer(i,j,k,bi,bj))/deltaTtracer |
489 |
ENDDO |
490 |
ENDDO |
491 |
|
492 |
C-- End of K loop for vertical flux |
493 |
ENDDO |
494 |
|
495 |
RETURN |
496 |
END |