1 |
C $Header$ |
C $Header$ |
2 |
C $Name$ |
C $Name$ |
3 |
|
|
|
CBOI |
|
|
C !TITLE: pkg/generic\_advdiff |
|
|
C !AUTHORS: adcroft@mit.edu |
|
|
C !INTRODUCTION: Generic Advection Diffusion Package |
|
|
C |
|
|
C Package "generic\_advdiff" provides a common set of routines for calculating |
|
|
C advective/diffusive fluxes for tracers (cell centered quantities on a C-grid). |
|
|
C |
|
|
C Many different advection schemes are available: the standard centered |
|
|
C second order, centered fourth order and upwind biased third order schemes |
|
|
C are known as linear methods and require some stable time-stepping method |
|
|
C such as Adams-Bashforth. Alternatives such as flux-limited schemes are |
|
|
C stable in the forward sense and are best combined with the multi-dimensional |
|
|
C method provided in gad\_advection. |
|
|
C |
|
|
C There are two high-level routines: |
|
|
C \begin{itemize} |
|
|
C \item{GAD\_CALC\_RHS} calculates all fluxes at time level "n" and is used |
|
|
C for the standard linear schemes. This must be used in conjuction with |
|
|
C Adams-Bashforth time-stepping. Diffusive and parameterized fluxes are |
|
|
C always calculated here. |
|
|
C \item{GAD\_ADVECTION} calculates just the advective fluxes using the |
|
|
C non-linear schemes and can not be used in conjuction with Adams-Bashforth |
|
|
C time-stepping. |
|
|
C \end{itemize} |
|
|
CEOI |
|
|
|
|
4 |
#include "GAD_OPTIONS.h" |
#include "GAD_OPTIONS.h" |
5 |
|
|
6 |
|
C---+----1----+----2----+----3----+----4----+----5----+----6----+----7-|--+----| |
7 |
CBOP |
CBOP |
8 |
C !ROUTINE: GAD_ADVECTION |
C !ROUTINE: GAD_ADVECTION |
9 |
|
|
10 |
C !INTERFACE: ========================================================== |
C !INTERFACE: ========================================================== |
11 |
SUBROUTINE GAD_ADVECTION(bi,bj,advectionScheme,tracerIdentity, |
SUBROUTINE GAD_ADVECTION( |
12 |
U Tracer,Gtracer, |
I implicitAdvection, advectionScheme, vertAdvecScheme, |
13 |
I myTime,myIter,myThid) |
I tracerIdentity, |
14 |
|
I uVel, vVel, wVel, tracer, |
15 |
|
O gTracer, |
16 |
|
I bi,bj, myTime,myIter,myThid) |
17 |
|
|
18 |
C !DESCRIPTION: |
C !DESCRIPTION: |
19 |
C Calculates the tendancy of a tracer due to advection. |
C Calculates the tendancy of a tracer due to advection. |
39 |
#include "SIZE.h" |
#include "SIZE.h" |
40 |
#include "EEPARAMS.h" |
#include "EEPARAMS.h" |
41 |
#include "PARAMS.h" |
#include "PARAMS.h" |
|
#include "DYNVARS.h" |
|
42 |
#include "GRID.h" |
#include "GRID.h" |
43 |
#include "GAD.h" |
#include "GAD.h" |
44 |
#ifdef ALLOW_AUTODIFF_TAMC |
#ifdef ALLOW_AUTODIFF_TAMC |
45 |
# include "tamc.h" |
# include "tamc.h" |
46 |
# include "tamc_keys.h" |
# include "tamc_keys.h" |
47 |
|
# ifdef ALLOW_PTRACERS |
48 |
|
# include "PTRACERS_SIZE.h" |
49 |
|
# endif |
50 |
#endif |
#endif |
51 |
|
#ifdef ALLOW_EXCH2 |
52 |
|
#include "W2_EXCH2_TOPOLOGY.h" |
53 |
|
#include "W2_EXCH2_PARAMS.h" |
54 |
|
#endif /* ALLOW_EXCH2 */ |
55 |
|
|
56 |
C !INPUT PARAMETERS: =================================================== |
C !INPUT PARAMETERS: =================================================== |
57 |
C bi,bj :: tile indices |
C implicitAdvection :: implicit vertical advection (later on) |
58 |
C advectionScheme :: advection scheme to use |
C advectionScheme :: advection scheme to use (Horizontal plane) |
59 |
C tracerIdentity :: identifier for the tracer (required only for OBCS) |
C vertAdvecScheme :: advection scheme to use (vertical direction) |
60 |
C Tracer :: tracer field |
C tracerIdentity :: tracer identifier (required only for OBCS) |
61 |
C myTime :: current time |
C uVel :: velocity, zonal component |
62 |
C myIter :: iteration number |
C vVel :: velocity, meridional component |
63 |
C myThid :: thread number |
C wVel :: velocity, vertical component |
64 |
INTEGER bi,bj |
C tracer :: tracer field |
65 |
INTEGER advectionScheme |
C bi,bj :: tile indices |
66 |
|
C myTime :: current time |
67 |
|
C myIter :: iteration number |
68 |
|
C myThid :: thread number |
69 |
|
LOGICAL implicitAdvection |
70 |
|
INTEGER advectionScheme, vertAdvecScheme |
71 |
INTEGER tracerIdentity |
INTEGER tracerIdentity |
72 |
_RL Tracer(1-Olx:sNx+Olx,1-Oly:sNy+Oly,Nr,nSx,nSy) |
_RL uVel (1-Olx:sNx+Olx,1-Oly:sNy+Oly,Nr,nSx,nSy) |
73 |
|
_RL vVel (1-Olx:sNx+Olx,1-Oly:sNy+Oly,Nr,nSx,nSy) |
74 |
|
_RL wVel (1-Olx:sNx+Olx,1-Oly:sNy+Oly,Nr,nSx,nSy) |
75 |
|
_RL tracer(1-Olx:sNx+Olx,1-Oly:sNy+Oly,Nr,nSx,nSy) |
76 |
|
INTEGER bi,bj |
77 |
_RL myTime |
_RL myTime |
78 |
INTEGER myIter |
INTEGER myIter |
79 |
INTEGER myThid |
INTEGER myThid |
80 |
|
|
81 |
C !OUTPUT PARAMETERS: ================================================== |
C !OUTPUT PARAMETERS: ================================================== |
82 |
C gTracer :: tendancy array |
C gTracer :: tendancy array |
83 |
_RL gTracer(1-Olx:sNx+Olx,1-Oly:sNy+Oly,Nr,nSx,nSy) |
_RL gTracer(1-Olx:sNx+Olx,1-Oly:sNy+Oly,Nr,nSx,nSy) |
84 |
|
|
85 |
C !LOCAL VARIABLES: ==================================================== |
C !LOCAL VARIABLES: ==================================================== |
86 |
C maskUp :: 2-D array for mask at W points |
C maskUp :: 2-D array for mask at W points |
87 |
C iMin,iMax,jMin,jMax :: loop range for called routines |
C iMin,iMax, :: loop range for called routines |
88 |
C i,j,k :: loop indices |
C jMin,jMax :: loop range for called routines |
89 |
C kup :: index into 2 1/2D array, toggles between 1 and 2 |
C i,j,k :: loop indices |
90 |
C kdown :: index into 2 1/2D array, toggles between 2 and 1 |
C kup :: index into 2 1/2D array, toggles between 1 and 2 |
91 |
C kp1 :: =k+1 for k<Nr, =Nr for k=Nr |
C kdown :: index into 2 1/2D array, toggles between 2 and 1 |
92 |
C xA,yA :: areas of X and Y face of tracer cells |
C kp1 :: =k+1 for k<Nr, =Nr for k=Nr |
93 |
C uTrans,vTrans,rTrans :: 2-D arrays of volume transports at U,V and W points |
C xA,yA :: areas of X and Y face of tracer cells |
94 |
C rTransKp1 :: vertical volume transport at interface k+1 |
C uTrans,vTrans :: 2-D arrays of volume transports at U,V points |
95 |
C af :: 2-D array for horizontal advective flux |
C rTrans :: 2-D arrays of volume transports at W points |
96 |
C fVerT :: 2 1/2D arrays for vertical advective flux |
C rTransKp1 :: vertical volume transport at interface k+1 |
97 |
C localTij :: 2-D array used as temporary local copy of tracer fld |
C af :: 2-D array for horizontal advective flux |
98 |
C localTijk :: 3-D array used as temporary local copy of tracer fld |
C fVerT :: 2 1/2D arrays for vertical advective flux |
99 |
C kp1Msk :: flag (0,1) to act as over-riding mask for W levels |
C localTij :: 2-D array, temporary local copy of tracer fld |
100 |
C calc_fluxes_X :: logical to indicate to calculate fluxes in X dir |
C localTijk :: 3-D array, temporary local copy of tracer fld |
101 |
C calc_fluxes_Y :: logical to indicate to calculate fluxes in Y dir |
C kp1Msk :: flag (0,1) for over-riding mask for W levels |
102 |
C nipass :: number of passes to make in multi-dimensional method |
C calc_fluxes_X :: logical to indicate to calculate fluxes in X dir |
103 |
C ipass :: number of the current pass being made |
C calc_fluxes_Y :: logical to indicate to calculate fluxes in Y dir |
104 |
|
C nipass :: number of passes in multi-dimensional method |
105 |
|
C ipass :: number of the current pass being made |
106 |
|
C myTile :: variables used to determine which cube face |
107 |
|
C nCFace :: owns a tile for cube grid runs using |
108 |
|
C :: multi-dim advection. |
109 |
_RS maskUp (1-OLx:sNx+OLx,1-OLy:sNy+OLy) |
_RS maskUp (1-OLx:sNx+OLx,1-OLy:sNy+OLy) |
110 |
INTEGER iMin,iMax,jMin,jMax |
INTEGER iMin,iMax,jMin,jMax |
111 |
INTEGER i,j,k,kup,kDown |
INTEGER i,j,k,kup,kDown |
122 |
_RL kp1Msk |
_RL kp1Msk |
123 |
LOGICAL calc_fluxes_X,calc_fluxes_Y |
LOGICAL calc_fluxes_X,calc_fluxes_Y |
124 |
INTEGER nipass,ipass |
INTEGER nipass,ipass |
125 |
|
INTEGER myTile, nCFace |
126 |
|
LOGICAL southWestCorner |
127 |
|
LOGICAL southEastCorner |
128 |
|
LOGICAL northWestCorner |
129 |
|
LOGICAL northEastCorner |
130 |
CEOP |
CEOP |
131 |
|
|
132 |
#ifdef ALLOW_AUTODIFF_TAMC |
#ifdef ALLOW_AUTODIFF_TAMC |
133 |
|
act0 = tracerIdentity - 1 |
134 |
|
max0 = maxpass |
135 |
act1 = bi - myBxLo(myThid) |
act1 = bi - myBxLo(myThid) |
136 |
max1 = myBxHi(myThid) - myBxLo(myThid) + 1 |
max1 = myBxHi(myThid) - myBxLo(myThid) + 1 |
137 |
act2 = bj - myByLo(myThid) |
act2 = bj - myByLo(myThid) |
139 |
act3 = myThid - 1 |
act3 = myThid - 1 |
140 |
max3 = nTx*nTy |
max3 = nTx*nTy |
141 |
act4 = ikey_dynamics - 1 |
act4 = ikey_dynamics - 1 |
142 |
ikey = (act1 + 1) + act2*max1 |
igadkey = (act0 + 1) |
143 |
& + act3*max1*max2 |
& + act1*max0 |
144 |
& + act4*max1*max2*max3 |
& + act2*max0*max1 |
145 |
|
& + act3*max0*max1*max2 |
146 |
|
& + act4*max0*max1*max2*max3 |
147 |
|
if (tracerIdentity.GT.maxpass) then |
148 |
|
print *, 'ph-pass gad_advection ', maxpass, tracerIdentity |
149 |
|
STOP 'maxpass seems smaller than tracerIdentity' |
150 |
|
endif |
151 |
#endif /* ALLOW_AUTODIFF_TAMC */ |
#endif /* ALLOW_AUTODIFF_TAMC */ |
152 |
|
|
153 |
C-- Set up work arrays with valid (i.e. not NaN) values |
C-- Set up work arrays with valid (i.e. not NaN) values |
176 |
C-- Start of k loop for horizontal fluxes |
C-- Start of k loop for horizontal fluxes |
177 |
DO k=1,Nr |
DO k=1,Nr |
178 |
#ifdef ALLOW_AUTODIFF_TAMC |
#ifdef ALLOW_AUTODIFF_TAMC |
179 |
kkey = (ikey-1)*Nr + k |
kkey = (igadkey-1)*Nr + k |
180 |
CADJ STORE tracer(:,:,k,bi,bj) = comlev1_bibj_k, key=kkey, byte=isbyte |
CADJ STORE tracer(:,:,k,bi,bj) = |
181 |
|
CADJ & comlev1_bibj_k_gad, key=kkey, byte=isbyte |
182 |
#endif /* ALLOW_AUTODIFF_TAMC */ |
#endif /* ALLOW_AUTODIFF_TAMC */ |
183 |
|
|
184 |
C-- Get temporary terms used by tendency routines |
C-- Get temporary terms used by tendency routines |
201 |
ENDDO |
ENDDO |
202 |
ENDDO |
ENDDO |
203 |
|
|
204 |
|
cph The following block is needed for useCubedSphereExchange only, |
205 |
|
cph but needs to be set for all cases to avoid spurious |
206 |
|
cph TAF dependencies |
207 |
|
southWestCorner = .TRUE. |
208 |
|
southEastCorner = .TRUE. |
209 |
|
northWestCorner = .TRUE. |
210 |
|
northEastCorner = .TRUE. |
211 |
|
#ifdef ALLOW_EXCH2 |
212 |
|
myTile = W2_myTileList(bi) |
213 |
|
nCFace = exch2_myFace(myTile) |
214 |
|
southWestCorner = exch2_isWedge(myTile).EQ.1 |
215 |
|
& .AND. exch2_isSedge(myTile).EQ.1 |
216 |
|
southEastCorner = exch2_isEedge(myTile).EQ.1 |
217 |
|
& .AND. exch2_isSedge(myTile).EQ.1 |
218 |
|
northEastCorner = exch2_isEedge(myTile).EQ.1 |
219 |
|
& .AND. exch2_isNedge(myTile).EQ.1 |
220 |
|
northWestCorner = exch2_isWedge(myTile).EQ.1 |
221 |
|
& .AND. exch2_isNedge(myTile).EQ.1 |
222 |
|
#else |
223 |
|
nCFace = bi |
224 |
|
#endif |
225 |
IF (useCubedSphereExchange) THEN |
IF (useCubedSphereExchange) THEN |
226 |
|
|
227 |
nipass=3 |
nipass=3 |
228 |
|
#ifdef ALLOW_AUTODIFF_TAMC |
229 |
|
if ( nipass.GT.maxcube ) |
230 |
|
& STOP 'maxcube needs to be = 3' |
231 |
|
#endif |
232 |
ELSE |
ELSE |
233 |
nipass=1 |
nipass=1 |
234 |
ENDIF |
ENDIF |
235 |
cph nipass=1 |
cph nipass=1 |
236 |
|
|
237 |
C-- Multiple passes for different directions on different tiles |
C-- Multiple passes for different directions on different tiles |
238 |
|
C-- For cube need one pass for each of red, green and blue axes. |
239 |
DO ipass=1,nipass |
DO ipass=1,nipass |
240 |
#ifdef ALLOW_AUTODIFF_TAMC |
#ifdef ALLOW_AUTODIFF_TAMC |
241 |
passkey = ipass + (k-1) *maxpass |
passkey = ipass + (k-1) *maxcube |
242 |
& + (ikey-1)*maxpass*Nr |
& + (igadkey-1)*maxcube*Nr |
243 |
IF (nipass .GT. maxpass) THEN |
IF (nipass .GT. maxpass) THEN |
244 |
STOP 'GAD_ADVECTION: nipass > maxpass. check tamc.h' |
STOP 'GAD_ADVECTION: nipass > maxcube. check tamc.h' |
245 |
ENDIF |
ENDIF |
246 |
#endif /* ALLOW_AUTODIFF_TAMC */ |
#endif /* ALLOW_AUTODIFF_TAMC */ |
247 |
|
|
248 |
IF (nipass.EQ.3) THEN |
IF (nipass.EQ.3) THEN |
249 |
calc_fluxes_X=.FALSE. |
calc_fluxes_X=.FALSE. |
250 |
calc_fluxes_Y=.FALSE. |
calc_fluxes_Y=.FALSE. |
251 |
IF (ipass.EQ.1 .AND. (bi.EQ.1 .OR. bi.EQ.2) ) THEN |
IF (ipass.EQ.1 .AND. (nCFace.EQ.1 .OR. nCFace.EQ.2) ) THEN |
252 |
calc_fluxes_X=.TRUE. |
calc_fluxes_X=.TRUE. |
253 |
ELSEIF (ipass.EQ.1 .AND. (bi.EQ.4 .OR. bi.EQ.5) ) THEN |
ELSEIF (ipass.EQ.1 .AND. (nCFace.EQ.4 .OR. nCFace.EQ.5) ) THEN |
254 |
calc_fluxes_Y=.TRUE. |
calc_fluxes_Y=.TRUE. |
255 |
ELSEIF (ipass.EQ.2 .AND. (bi.EQ.1 .OR. bi.EQ.6) ) THEN |
ELSEIF (ipass.EQ.2 .AND. (nCFace.EQ.1 .OR. nCFace.EQ.6) ) THEN |
256 |
calc_fluxes_Y=.TRUE. |
calc_fluxes_Y=.TRUE. |
257 |
ELSEIF (ipass.EQ.2 .AND. (bi.EQ.3 .OR. bi.EQ.4) ) THEN |
ELSEIF (ipass.EQ.2 .AND. (nCFace.EQ.3 .OR. nCFace.EQ.4) ) THEN |
258 |
calc_fluxes_X=.TRUE. |
calc_fluxes_X=.TRUE. |
259 |
ELSEIF (ipass.EQ.3 .AND. (bi.EQ.2 .OR. bi.EQ.3) ) THEN |
ELSEIF (ipass.EQ.3 .AND. (nCFace.EQ.2 .OR. nCFace.EQ.3) ) THEN |
260 |
calc_fluxes_Y=.TRUE. |
calc_fluxes_Y=.TRUE. |
261 |
ELSEIF (ipass.EQ.3 .AND. (bi.EQ.5 .OR. bi.EQ.6) ) THEN |
ELSEIF (ipass.EQ.3 .AND. (nCFace.EQ.5 .OR. nCFace.EQ.6) ) THEN |
262 |
calc_fluxes_X=.TRUE. |
calc_fluxes_X=.TRUE. |
263 |
ENDIF |
ENDIF |
264 |
ELSE |
ELSE |
271 |
|
|
272 |
C-- Internal exchange for calculations in X |
C-- Internal exchange for calculations in X |
273 |
IF (useCubedSphereExchange) THEN |
IF (useCubedSphereExchange) THEN |
274 |
DO j=1,Oly |
C-- For cube face corners we need to duplicate the |
275 |
DO i=1,Olx |
C-- i-1 and i+1 values into the null space as follows: |
276 |
localTij( 1-i , 1-j )=localTij( 1-j , i ) |
C |
277 |
localTij( 1-i ,sNy+j)=localTij( 1-j , sNy+1-i ) |
C |
278 |
localTij(sNx+i, 1-j )=localTij(sNx+j, i ) |
C o NW corner: copy T( 0,sNy ) into T( 0,sNy+1) e.g. |
279 |
localTij(sNx+i,sNy+j)=localTij(sNx+j, sNy+1-i ) |
C | |
280 |
|
C x T(0,sNy+1) | |
281 |
|
C /\ | |
282 |
|
C --||------------|----------- |
283 |
|
C || | |
284 |
|
C x T(0,sNy) | x T(1,sNy) |
285 |
|
C | |
286 |
|
C |
287 |
|
C o SW corner: copy T(0,1) into T(0,0) e.g. |
288 |
|
C | |
289 |
|
C x T(0,1) | x T(1,1) |
290 |
|
C || | |
291 |
|
C --||------------|----------- |
292 |
|
C \/ | |
293 |
|
C x T(0,0) | |
294 |
|
C | |
295 |
|
C |
296 |
|
C o NE corner: copy T(sNx+1,sNy ) into T(sNx+1,sNy+1) e.g. |
297 |
|
C | |
298 |
|
C | x T(sNx+1,sNy+1) |
299 |
|
C | /\ |
300 |
|
C ----------------|--||------- |
301 |
|
C | || |
302 |
|
C x T(sNx,sNy) | x T(sNx+1,sNy ) |
303 |
|
C | |
304 |
|
C o SE corner: copy T(sNx+1,1 ) into T(sNx+1,0 ) e.g. |
305 |
|
C | |
306 |
|
C x T(sNx,1) | x T(sNx+1, 1) |
307 |
|
C | || |
308 |
|
C ----------------|--||------- |
309 |
|
C | \/ |
310 |
|
C | x T(sNx+1, 0) |
311 |
|
IF ( southWestCorner ) THEN |
312 |
|
DO J=1,OLy |
313 |
|
DO I=1,OLx |
314 |
|
localTij(1-I, 1-J )= localTij(1-J ,1 ) |
315 |
|
ENDDO |
316 |
ENDDO |
ENDDO |
317 |
ENDDO |
ENDIF |
318 |
|
IF ( southEastCorner ) THEN |
319 |
|
DO J=1,OLy |
320 |
|
DO I=1,OLx |
321 |
|
localTij(sNx+I, 1-J )=localTij(sNx+J, I ) |
322 |
|
ENDDO |
323 |
|
ENDDO |
324 |
|
ENDIF |
325 |
|
IF ( northWestCorner ) THEN |
326 |
|
DO J=1,OLy |
327 |
|
DO I=1,OLx |
328 |
|
localTij( 1-I ,sNy+J)=localTij( 1-J , sNy+1-I ) |
329 |
|
ENDDO |
330 |
|
ENDDO |
331 |
|
ENDIF |
332 |
|
IF ( northEastCorner ) THEN |
333 |
|
DO J=1,OLy |
334 |
|
DO I=1,OLx |
335 |
|
localTij(sNx+I,sNy+J)=localTij(sNx+J, sNy+1-I ) |
336 |
|
ENDDO |
337 |
|
ENDDO |
338 |
|
ENDIF |
339 |
ENDIF |
ENDIF |
340 |
|
|
341 |
C- Advective flux in X |
C- Advective flux in X |
347 |
|
|
348 |
#ifdef ALLOW_AUTODIFF_TAMC |
#ifdef ALLOW_AUTODIFF_TAMC |
349 |
#ifndef DISABLE_MULTIDIM_ADVECTION |
#ifndef DISABLE_MULTIDIM_ADVECTION |
350 |
CADJ STORE localTij(:,:) = comlev1_bibj_pass, key=passkey, byte=isbyte |
CADJ STORE localTij(:,:) = |
351 |
|
CADJ & comlev1_bibj_k_gad_pass, key=passkey, byte=isbyte |
352 |
#endif |
#endif |
353 |
#endif /* ALLOW_AUTODIFF_TAMC */ |
#endif /* ALLOW_AUTODIFF_TAMC */ |
354 |
|
|
393 |
C-- Y direction |
C-- Y direction |
394 |
IF (calc_fluxes_Y) THEN |
IF (calc_fluxes_Y) THEN |
395 |
|
|
|
C-- Internal exchange for calculations in Y |
|
396 |
IF (useCubedSphereExchange) THEN |
IF (useCubedSphereExchange) THEN |
397 |
DO j=1,Oly |
C-- Internal exchange for calculations in Y |
398 |
DO i=1,Olx |
C-- For cube face corners we need to duplicate the |
399 |
localTij( 1-i , 1-j )=localTij( j , 1-i ) |
C-- j-1 and j+1 values into the null space as follows: |
400 |
localTij( 1-i ,sNy+j)=localTij( j ,sNy+i) |
C |
401 |
localTij(sNx+i, 1-j )=localTij(sNx+1-j, 1-i ) |
C o SW corner: copy T(0,1) into T(0,0) e.g. |
402 |
localTij(sNx+i,sNy+j)=localTij(sNx+1-j,sNy+i) |
C | |
403 |
|
C | x T(1,1) |
404 |
|
C | |
405 |
|
C ----------------|----------- |
406 |
|
C | |
407 |
|
C x T(0,0)<====== x T(1,0) |
408 |
|
C | |
409 |
|
C |
410 |
|
C o NW corner: copy T( 0,sNy ) into T( 0,sNy+1) e.g. |
411 |
|
C | |
412 |
|
C x T(0,sNy+1)<=== x T(1,sNy+1) |
413 |
|
C | |
414 |
|
C ----------------|----------- |
415 |
|
C | |
416 |
|
C | x T(1,sNy) |
417 |
|
C | |
418 |
|
C |
419 |
|
C o NE corner: copy T(sNx+1,sNy ) into T(sNx+1,sNy+1) e.g. |
420 |
|
C | |
421 |
|
C x T(sNx,sNy+1)=====>x T(sNx+1,sNy+1) |
422 |
|
C | |
423 |
|
C ----------------|----------- |
424 |
|
C | |
425 |
|
C x T(sNx,sNy) | |
426 |
|
C | |
427 |
|
C o SE corner: copy T(sNx+1,1 ) into T(sNx+1,0 ) e.g. |
428 |
|
C | |
429 |
|
C x T(sNx,1) | |
430 |
|
C | |
431 |
|
C ----------------|----------- |
432 |
|
C | |
433 |
|
C x T(sNx,0) =====>x T(sNx+1, 0) |
434 |
|
IF ( southWestCorner ) THEN |
435 |
|
DO J=1,Oly |
436 |
|
DO I=1,Olx |
437 |
|
localTij( 1-i , 1-j ) = localTij(j , 1-i ) |
438 |
|
ENDDO |
439 |
ENDDO |
ENDDO |
440 |
ENDDO |
ENDIF |
441 |
|
IF ( southEastCorner ) THEN |
442 |
|
DO J=1,Oly |
443 |
|
DO I=1,Olx |
444 |
|
localTij(sNx+i, 1-j ) = localTij(sNx+1-j, 1-i ) |
445 |
|
ENDDO |
446 |
|
ENDDO |
447 |
|
ENDIF |
448 |
|
IF ( northWestCorner ) THEN |
449 |
|
DO J=1,Oly |
450 |
|
DO I=1,Olx |
451 |
|
localTij( 1-i ,sNy+j) = localTij(j ,sNy+i) |
452 |
|
ENDDO |
453 |
|
ENDDO |
454 |
|
ENDIF |
455 |
|
IF ( northEastCorner ) THEN |
456 |
|
DO J=1,Oly |
457 |
|
DO I=1,Olx |
458 |
|
localTij(sNx+i,sNy+j) = localTij(sNx+1-j,sNy+i) |
459 |
|
ENDDO |
460 |
|
ENDDO |
461 |
|
ENDIF |
462 |
ENDIF |
ENDIF |
463 |
|
|
464 |
C- Advective flux in Y |
C- Advective flux in Y |
470 |
|
|
471 |
#ifdef ALLOW_AUTODIFF_TAMC |
#ifdef ALLOW_AUTODIFF_TAMC |
472 |
#ifndef DISABLE_MULTIDIM_ADVECTION |
#ifndef DISABLE_MULTIDIM_ADVECTION |
473 |
CADJ STORE localTij(:,:) = comlev1_bibj_pass, key=passkey, byte=isbyte |
CADJ STORE localTij(:,:) = |
474 |
|
CADJ & comlev1_bibj_k_gad_pass, key=passkey, byte=isbyte |
475 |
#endif |
#endif |
476 |
#endif /* ALLOW_AUTODIFF_TAMC */ |
#endif /* ALLOW_AUTODIFF_TAMC */ |
477 |
|
|
513 |
C-- End of Y direction |
C-- End of Y direction |
514 |
ENDIF |
ENDIF |
515 |
|
|
|
DO j=1-Oly,sNy+Oly |
|
|
DO i=1-Olx,sNx+Olx |
|
|
localTijk(i,j,k)=localTij(i,j) |
|
|
ENDDO |
|
|
ENDDO |
|
|
|
|
516 |
C-- End of ipass loop |
C-- End of ipass loop |
517 |
ENDDO |
ENDDO |
518 |
|
|
519 |
|
IF ( implicitAdvection ) THEN |
520 |
|
C- explicit advection is done ; store tendency in gTracer: |
521 |
|
DO j=1-Oly,sNy+Oly |
522 |
|
DO i=1-Olx,sNx+Olx |
523 |
|
gTracer(i,j,k,bi,bj)= |
524 |
|
& (localTij(i,j)-tracer(i,j,k,bi,bj))/deltaTtracer |
525 |
|
ENDDO |
526 |
|
ENDDO |
527 |
|
ELSE |
528 |
|
C- horizontal advection done; store intermediate result in 3D array: |
529 |
|
DO j=1-Oly,sNy+Oly |
530 |
|
DO i=1-Olx,sNx+Olx |
531 |
|
localTijk(i,j,k)=localTij(i,j) |
532 |
|
ENDDO |
533 |
|
ENDDO |
534 |
|
ENDIF |
535 |
|
|
536 |
C-- End of K loop for horizontal fluxes |
C-- End of K loop for horizontal fluxes |
537 |
ENDDO |
ENDDO |
538 |
|
|
539 |
|
IF ( .NOT.implicitAdvection ) THEN |
540 |
C-- Start of k loop for vertical flux |
C-- Start of k loop for vertical flux |
541 |
DO k=Nr,1,-1 |
DO k=Nr,1,-1 |
542 |
#ifdef ALLOW_AUTODIFF_TAMC |
#ifdef ALLOW_AUTODIFF_TAMC |
543 |
kkey = (ikey-1)*Nr + k |
kkey = (igadkey-1)*Nr + k |
544 |
#endif /* ALLOW_AUTODIFF_TAMC */ |
#endif /* ALLOW_AUTODIFF_TAMC */ |
|
|
|
545 |
C-- kup Cycles through 1,2 to point to w-layer above |
C-- kup Cycles through 1,2 to point to w-layer above |
546 |
C-- kDown Cycles through 2,1 to point to w-layer below |
C-- kDown Cycles through 2,1 to point to w-layer below |
547 |
kup = 1+MOD(k+1,2) |
kup = 1+MOD(k+1,2) |
548 |
kDown= 1+MOD(k,2) |
kDown= 1+MOD(k,2) |
549 |
c kp1=min(Nr,k+1) |
c kp1=min(Nr,k+1) |
550 |
kp1Msk=1. |
kp1Msk=1. |
551 |
if (k.EQ.Nr) kp1Msk=0. |
if (k.EQ.Nr) kp1Msk=0. |
|
|
|
|
#ifdef ALLOW_AUTODIFF_TAMC |
|
|
CADJ STORE localTijk(:,:,k) |
|
|
CADJ & = comlev1_bibj_k, key=kkey, byte=isbyte |
|
|
CADJ STORE rTrans(:,:) |
|
|
CADJ & = comlev1_bibj_k, key=kkey, byte=isbyte |
|
|
#endif /* ALLOW_AUTODIFF_TAMC */ |
|
552 |
|
|
553 |
C-- Compute Vertical transport |
C-- Compute Vertical transport |
554 |
C Note: wVel needs to be masked |
#ifdef ALLOW_AIM |
555 |
|
C- a hack to prevent Water-Vapor vert.transport into the stratospheric level Nr |
556 |
|
IF ( k.EQ.1 .OR. |
557 |
|
& (useAIM .AND. tracerIdentity.EQ.GAD_SALINITY .AND. k.EQ.Nr) |
558 |
|
& ) THEN |
559 |
|
#else |
560 |
|
IF ( k.EQ.1 ) THEN |
561 |
|
#endif |
562 |
|
|
|
IF (k.EQ.1) THEN |
|
563 |
C- Surface interface : |
C- Surface interface : |
564 |
|
DO j=1-Oly,sNy+Oly |
565 |
|
DO i=1-Olx,sNx+Olx |
566 |
|
rTransKp1(i,j) = kp1Msk*rTrans(i,j) |
567 |
|
rTrans(i,j) = 0. |
568 |
|
fVerT(i,j,kUp) = 0. |
569 |
|
af(i,j) = 0. |
570 |
|
ENDDO |
571 |
|
ENDDO |
572 |
|
|
573 |
DO j=1-Oly,sNy+Oly |
ELSE |
|
DO i=1-Olx,sNx+Olx |
|
|
rTransKp1(i,j) = rTrans(i,j) |
|
|
rTrans(i,j) = 0. |
|
|
fVerT(i,j,kUp) = 0. |
|
|
ENDDO |
|
|
ENDDO |
|
|
|
|
|
ELSE |
|
574 |
C- Interior interface : |
C- Interior interface : |
575 |
DO j=1-Oly,sNy+Oly |
|
576 |
DO i=1-Olx,sNx+Olx |
DO j=1-Oly,sNy+Oly |
577 |
rTransKp1(i,j) = kp1Msk*rTrans(i,j) |
DO i=1-Olx,sNx+Olx |
578 |
rTrans(i,j) = wVel(i,j,k,bi,bj)*rA(i,j,bi,bj) |
rTransKp1(i,j) = kp1Msk*rTrans(i,j) |
579 |
& *maskC(i,j,k-1,bi,bj) |
rTrans(i,j) = wVel(i,j,k,bi,bj)*rA(i,j,bi,bj) |
580 |
af(i,j) = 0. |
& *maskC(i,j,k-1,bi,bj) |
581 |
ENDDO |
af(i,j) = 0. |
582 |
ENDDO |
ENDDO |
583 |
|
ENDDO |
584 |
|
|
585 |
#ifdef ALLOW_GMREDI |
#ifdef ALLOW_GMREDI |
586 |
C-- Residual transp = Bolus transp + Eulerian transp |
C-- Residual transp = Bolus transp + Eulerian transp |
587 |
IF (useGMRedi) |
IF (useGMRedi) |
588 |
& CALL GMREDI_CALC_WFLOW( |
& CALL GMREDI_CALC_WFLOW( |
589 |
& rTrans, bi, bj, k, myThid) |
& rTrans, bi, bj, k, myThid) |
590 |
#endif /* ALLOW_GMREDI */ |
#endif /* ALLOW_GMREDI */ |
591 |
|
|
592 |
|
#ifdef ALLOW_AUTODIFF_TAMC |
593 |
|
CADJ STORE localTijk(:,:,k) |
594 |
|
CADJ & = comlev1_bibj_k_gad, key=kkey, byte=isbyte |
595 |
|
CADJ STORE rTrans(:,:) |
596 |
|
CADJ & = comlev1_bibj_k_gad, key=kkey, byte=isbyte |
597 |
|
#endif /* ALLOW_AUTODIFF_TAMC */ |
598 |
|
|
599 |
C- Compute vertical advective flux in the interior: |
C- Compute vertical advective flux in the interior: |
600 |
IF (advectionScheme.EQ.ENUM_FLUX_LIMIT) THEN |
IF (vertAdvecScheme.EQ.ENUM_FLUX_LIMIT) THEN |
601 |
CALL GAD_FLUXLIMIT_ADV_R( |
CALL GAD_FLUXLIMIT_ADV_R( |
602 |
& bi,bj,k,deltaTtracer,rTrans,wVel,localTijk,af,myThid) |
& bi,bj,k,deltaTtracer,rTrans,wVel,localTijk,af,myThid) |
603 |
ELSEIF (advectionScheme.EQ.ENUM_DST3 ) THEN |
ELSEIF (vertAdvecScheme.EQ.ENUM_DST3 ) THEN |
604 |
CALL GAD_DST3_ADV_R( |
CALL GAD_DST3_ADV_R( |
605 |
& bi,bj,k,deltaTtracer,rTrans,wVel,localTijk,af,myThid) |
& bi,bj,k,deltaTtracer,rTrans,wVel,localTijk,af,myThid) |
606 |
ELSEIF (advectionScheme.EQ.ENUM_DST3_FLUX_LIMIT ) THEN |
ELSEIF (vertAdvecScheme.EQ.ENUM_DST3_FLUX_LIMIT ) THEN |
607 |
CALL GAD_DST3FL_ADV_R( |
CALL GAD_DST3FL_ADV_R( |
608 |
& bi,bj,k,deltaTtracer,rTrans,wVel,localTijk,af,myThid) |
& bi,bj,k,deltaTtracer,rTrans,wVel,localTijk,af,myThid) |
609 |
ELSE |
ELSE |
610 |
STOP 'GAD_ADVECTION: adv. scheme incompatibale with mutli-dim' |
STOP 'GAD_ADVECTION: adv. scheme incompatibale with mutli-dim' |
611 |
ENDIF |
ENDIF |
612 |
C- add the advective flux to fVerT |
C- add the advective flux to fVerT |
613 |
DO j=1-Oly,sNy+Oly |
DO j=1-Oly,sNy+Oly |
614 |
DO i=1-Olx,sNx+Olx |
DO i=1-Olx,sNx+Olx |
615 |
fVerT(i,j,kUp) = af(i,j) |
fVerT(i,j,kUp) = af(i,j) |
616 |
ENDDO |
ENDDO |
617 |
ENDDO |
ENDDO |
618 |
|
|
619 |
C- end Surface/Interior if bloc |
C- end Surface/Interior if bloc |
620 |
ENDIF |
ENDIF |
621 |
|
|
622 |
C-- Divergence of fluxes |
#ifdef ALLOW_AUTODIFF_TAMC |
623 |
DO j=1-Oly,sNy+Oly |
CADJ STORE rTrans(:,:) |
624 |
DO i=1-Olx,sNx+Olx |
CADJ & = comlev1_bibj_k_gad, key=kkey, byte=isbyte |
625 |
localTij(i,j)=localTijk(i,j,k)-deltaTtracer* |
CADJ STORE rTranskp1(:,:) |
626 |
& _recip_hFacC(i,j,k,bi,bj)*recip_drF(k) |
CADJ & = comlev1_bibj_k_gad, key=kkey, byte=isbyte |
627 |
& *recip_rA(i,j,bi,bj) |
#endif /* ALLOW_AUTODIFF_TAMC */ |
628 |
& *( fVerT(i,j,kUp)-fVerT(i,j,kDown) |
|
629 |
& -tracer(i,j,k,bi,bj)*(rTrans(i,j)-rTransKp1(i,j)) |
C-- Divergence of vertical fluxes |
630 |
& )*rkFac |
DO j=1-Oly,sNy+Oly |
631 |
gTracer(i,j,k,bi,bj)= |
DO i=1-Olx,sNx+Olx |
632 |
& (localTij(i,j)-tracer(i,j,k,bi,bj))/deltaTtracer |
localTij(i,j)=localTijk(i,j,k)-deltaTtracer* |
633 |
ENDDO |
& _recip_hFacC(i,j,k,bi,bj)*recip_drF(k) |
634 |
ENDDO |
& *recip_rA(i,j,bi,bj) |
635 |
|
& *( fVerT(i,j,kUp)-fVerT(i,j,kDown) |
636 |
|
& -tracer(i,j,k,bi,bj)*(rTrans(i,j)-rTransKp1(i,j)) |
637 |
|
& )*rkFac |
638 |
|
gTracer(i,j,k,bi,bj)= |
639 |
|
& (localTij(i,j)-tracer(i,j,k,bi,bj))/deltaTtracer |
640 |
|
ENDDO |
641 |
|
ENDDO |
642 |
|
|
643 |
C-- End of K loop for vertical flux |
C-- End of K loop for vertical flux |
644 |
ENDDO |
ENDDO |
645 |
|
C-- end of if not.implicitAdvection block |
646 |
|
ENDIF |
647 |
|
|
648 |
RETURN |
RETURN |
649 |
END |
END |