| 1 |
C $Header: /u/gcmpack/MITgcm/pkg/opps/opps_calc.F,v 1.11 2012/06/27 22:34:07 jmc Exp $ |
| 2 |
C $Name: $ |
| 3 |
|
| 4 |
#include "OPPS_OPTIONS.h" |
| 5 |
#undef OPPS_ORGCODE |
| 6 |
|
| 7 |
C-- File opps_calc.F: |
| 8 |
C-- Contents |
| 9 |
C-- o OPPS_CALC |
| 10 |
C-- o STATE1 |
| 11 |
C-- o NLOPPS |
| 12 |
|
| 13 |
C---+----1----+----2----+----3----+----4----+----5----+----6----+----7-|--+----| |
| 14 |
CBOP |
| 15 |
C !ROUTINE: OPPS_CALC |
| 16 |
|
| 17 |
C !INTERFACE: ====================================================== |
| 18 |
SUBROUTINE OPPS_CALC( |
| 19 |
U tracerEnv, |
| 20 |
I wVel, |
| 21 |
I kMax, nTracer, nTracerInuse, |
| 22 |
I I, J, bi, bj, myTime, myIter, myThid ) |
| 23 |
|
| 24 |
C !DESCRIPTION: \bv |
| 25 |
C *=====================================================================* |
| 26 |
C | SUBROUTINE OPPS_CALC | |
| 27 |
C | o Compute all OPPS fields defined in OPPS.h | |
| 28 |
C *=====================================================================* |
| 29 |
C | This subroutine is based on the routine 3dconvection.F | |
| 30 |
C | by E. Skyllingstad (?) | |
| 31 |
C | plenty of modifications to make it work: | |
| 32 |
C | - removed many unused parameters and variables | |
| 33 |
C | - turned everything (back) into 1D code | |
| 34 |
C | - pass variables, that are originally in common blocks: | |
| 35 |
C | maxDepth | |
| 36 |
C | - pass vertical velocity, set in OPPS_INTERFACE | |
| 37 |
C | - do not use convadj for now (whatever that is) | |
| 38 |
C | - changed two .LT. 0 to .LE. 0 statements (because of possible | |
| 39 |
C | division) | |
| 40 |
C | - replaced statement function state1 by call to a real function | |
| 41 |
C | - removed range check, actually moved it up to OPPS_INTERFACE | |
| 42 |
C | - avoid division by zero: if (Wd.EQ.0) dt = ...1/Wd | |
| 43 |
C | - cleaned-up debugging | |
| 44 |
C | - replaced local dz and GridThickness by global drF | |
| 45 |
C | - replaced 1/dz by 1*recip_drF | |
| 46 |
C | - replaced 9.81 with gravity (=9.81) | |
| 47 |
C | - added a lot of comments that relate code to equation in paper | |
| 48 |
C | (Paluszkiewicz+Romea, 1997, Dynamics of Atmospheres and Oceans, | |
| 49 |
C | 26, pp. 95-130) | |
| 50 |
C | - included passive tracer support. This is the main change and may | |
| 51 |
C | not improve the readability of the code because of the joint | |
| 52 |
C | treatment of active (theta, salt) and passive tracers. The array | |
| 53 |
C | tracerEnv(Nr,2+PTRACERS_num) contains | |
| 54 |
C | theta = tracerEnv(:,1), | |
| 55 |
C | salt = tracerEnv(:,2), and | |
| 56 |
C | ptracers = tracerEnv(:,3:PTRACERS_num+2). | |
| 57 |
C | All related array names have been changed accordingly, so that | |
| 58 |
C | instead of Sd(Nr) and Td(Nr) (plume salinity and temperature), we | |
| 59 |
C | have Pd(Nr,nTracer) (tracer in plume), with Sd(:) = Pd(:,2), | |
| 60 |
C | Td(:) = Pd(:,1), etc. | |
| 61 |
C | o TODO: | |
| 62 |
C | clean up the logic of the vertical loops and get rid off the | |
| 63 |
C | GOTO statements | |
| 64 |
C *=====================================================================* |
| 65 |
C \ev |
| 66 |
|
| 67 |
IMPLICIT NONE |
| 68 |
|
| 69 |
C !USES: ============================================================ |
| 70 |
#include "SIZE.h" |
| 71 |
#include "EEPARAMS.h" |
| 72 |
#include "PARAMS.h" |
| 73 |
#include "OPPS.h" |
| 74 |
#include "GRID.h" |
| 75 |
|
| 76 |
C !INPUT PARAMETERS: =================================================== |
| 77 |
C Routine arguments |
| 78 |
C bi, bj :: array indices on which to apply calculations |
| 79 |
C myTime :: Current time in simulation |
| 80 |
|
| 81 |
INTEGER I, J, bi, bj, KMax, nTracer, nTracerInUse |
| 82 |
INTEGER myThid, myIter |
| 83 |
_RL myTime |
| 84 |
_RL tracerEnv(Nr,nTracer),wVel(Nr) |
| 85 |
|
| 86 |
#ifdef ALLOW_OPPS |
| 87 |
C !FUNCTIONS: ========================================================== |
| 88 |
c EXTERNAL DIFFERENT_MULTIPLE |
| 89 |
c LOGICAL DIFFERENT_MULTIPLE |
| 90 |
_RL STATE1 |
| 91 |
|
| 92 |
C !LOCAL VARIABLES: ==================================================== |
| 93 |
C Local constants |
| 94 |
C msgBuf :: Informational/error message buffer |
| 95 |
CHARACTER*(MAX_LEN_MBUF) msgBuf |
| 96 |
INTEGER K, K2, K2m1, K2p1, ktr |
| 97 |
INTEGER ntime,nn,kmx,ic |
| 98 |
INTEGER maxDepth |
| 99 |
|
| 100 |
_RL wsqr,oldflux,newflux,entrainrate |
| 101 |
_RL pmix |
| 102 |
_RL D1, D2 |
| 103 |
_RL dz1,dz2 |
| 104 |
_RL radius,StartingFlux |
| 105 |
_RL dtts,dt |
| 106 |
C Arrays |
| 107 |
_RL Paa(Nr,nTracer) |
| 108 |
_RL wda(Nr), mda(Nr), pda(Nr,nTracer) |
| 109 |
C |
| 110 |
C Pd, Wd :: tracers, vertical velocity in plume |
| 111 |
C Md :: plume mass flux (?) |
| 112 |
C Ad :: fractional area covered by plume |
| 113 |
C Dd :: density in plume |
| 114 |
C De :: density of environment |
| 115 |
C PlumeEntrainment :: |
| 116 |
_RL Ad(Nr),Wd(Nr),Dd(Nr),Md(Nr) |
| 117 |
_RL De(Nr) |
| 118 |
_RL PlumeEntrainment(Nr) |
| 119 |
_RL Pd(Nr,nTracer) |
| 120 |
CEOP |
| 121 |
|
| 122 |
C-- Check to see if should convect now |
| 123 |
C IF ( DIFFERENT_MULTIPLE(cAdjFreq,myTime,deltaTClock) ) THEN |
| 124 |
IF ( .true. ) THEN |
| 125 |
C local initialization |
| 126 |
|
| 127 |
C Copy some arrays |
| 128 |
dtts = dTtracerLev(1) |
| 129 |
|
| 130 |
C start k-loop |
| 131 |
|
| 132 |
DO k=1,KMax-1 |
| 133 |
|
| 134 |
C initialize the plume T,S,density, and w velocity |
| 135 |
|
| 136 |
DO ktr=1,nTracerInUse |
| 137 |
Pd(k,ktr) = tracerEnv(k,ktr) |
| 138 |
ENDDO |
| 139 |
Dd(k)=STATE1(Pd(k,2),Pd(k,1),i,j,k,bi,bj,myThid) |
| 140 |
De(k)=Dd(k) |
| 141 |
CML print *, 'ml-opps:', i,j,k,tracerEnv(k,2),tracerEnv(k,1), |
| 142 |
CML & Dd(k),Pd(k,1),Pd(k,2) |
| 143 |
CML compute vertical velocity at cell centers from GCM velocity |
| 144 |
Wd(k)= - .5*(wVel(K)+wVel(K+1)) |
| 145 |
CML( |
| 146 |
CML avoid division by zero |
| 147 |
CML IF (Wd(K) .EQ. 0.D0) Wd(K) = 2.23e-16 |
| 148 |
CML) |
| 149 |
|
| 150 |
C guess at initial top grid cell vertical velocity |
| 151 |
|
| 152 |
CML Wd(k) = 0.03 |
| 153 |
|
| 154 |
C these estimates of initial plume velocity based on plume size and |
| 155 |
C top grid cell water mass |
| 156 |
|
| 157 |
c Wd(k) = 0.5*drF(k)/(dtts*FRACTIONAL_AREA) |
| 158 |
c Wd(k) = 0.5*drF(k)/dtts |
| 159 |
|
| 160 |
wsqr=Wd(k)*Wd(k) |
| 161 |
PlumeEntrainment(k) = 0.0 |
| 162 |
|
| 163 |
#ifdef ALLOW_OPPS_DEBUG |
| 164 |
IF ( OPPSdebugLevel.GE.debLevB ) THEN |
| 165 |
WRITE(msgBuf,'(A,I3)') |
| 166 |
& 'S/R OPPS_CALC: doing old lowerparcel', k |
| 167 |
CALL PRINT_MESSAGE( msgBuf, standardMessageUnit, |
| 168 |
& SQUEEZE_RIGHT , 1) |
| 169 |
ENDIF |
| 170 |
#endif /* ALLOW_OPPS_DEBUG */ |
| 171 |
radius=PlumeRadius |
| 172 |
StartingFlux=radius*radius*Wd(k)*Dd(k) |
| 173 |
oldflux=StartingFlux |
| 174 |
|
| 175 |
dz2=DrF(k) |
| 176 |
DO k2=k,KMax-1 |
| 177 |
D1=STATE1( Pd(k2,2), Pd(k2,1),i,j,k2+1,bi,bj,myThid) |
| 178 |
D2=STATE1( tracerEnv(k2+1,2), tracerEnv(k2+1,1), |
| 179 |
& i,j,k2+1,bi,bj,myThid) |
| 180 |
De(k2+1)=D2 |
| 181 |
|
| 182 |
C To start downward, parcel has to initially be heavier than environment |
| 183 |
C but after it has started moving, we continue plume until plume tke or |
| 184 |
C flux goes negative |
| 185 |
|
| 186 |
CML & _hFacC(i,j,k-1,bi,bj) |
| 187 |
CML & *_hFacC(i,j,k,bi,bj) .GT. 0. |
| 188 |
CML & .AND. |
| 189 |
IF (D2-D1 .LT. STABILITY_THRESHOLD.or.k2.ne.k) THEN |
| 190 |
dz1=dz2 |
| 191 |
dz2=DrF(k2+1) |
| 192 |
|
| 193 |
C find mass flux according to eq.(3) from paper by vertical integration |
| 194 |
|
| 195 |
newflux=oldflux+e2*radius*Wd(k2)*Dd(k2)* |
| 196 |
& .5*(dz1+dz2) |
| 197 |
CML print *, 'ml-opps:', i,j,k,oldflux,newflux,e2,radius, |
| 198 |
CML & Wd(k2),Dd(k2),Pd(k2,1),Pd(k2,2),dz1,dz2 |
| 199 |
|
| 200 |
PlumeEntrainment(k2+1) = newflux/StartingFlux |
| 201 |
|
| 202 |
IF(newflux.LE.0.0) then |
| 203 |
#ifdef ALLOW_OPPS_DEBUG |
| 204 |
IF ( OPPSdebugLevel.GE.debLevA ) THEN |
| 205 |
WRITE(msgBuf,'(A,I3)') |
| 206 |
& 'S/R OPPS_CALC: Plume entrained to zero at level ', k2 |
| 207 |
CALL PRINT_MESSAGE( msgBuf, standardMessageUnit, |
| 208 |
& SQUEEZE_RIGHT , 1) |
| 209 |
ENDIF |
| 210 |
#endif /* ALLOW_OPPS_DEBUG */ |
| 211 |
maxdepth = k2 |
| 212 |
if(maxdepth.eq.k) goto 1000 |
| 213 |
goto 1 |
| 214 |
endif |
| 215 |
|
| 216 |
C entrainment rate is basically a scaled mass flux dM/M |
| 217 |
|
| 218 |
entrainrate = (newflux - oldflux)/newflux |
| 219 |
oldflux = newflux |
| 220 |
|
| 221 |
C mix var(s) are the average environmental values over the two grid levels |
| 222 |
|
| 223 |
DO ktr=1,nTracerInUse |
| 224 |
pmix=(dz1*tracerEnv(k2,ktr)+dz2*tracerEnv(k2+1,ktr)) |
| 225 |
& /(dz1+dz2) |
| 226 |
Pd(k2+1,ktr)=Pd(k2,ktr) |
| 227 |
& - entrainrate*(pmix - Pd(k2,ktr)) |
| 228 |
ENDDO |
| 229 |
|
| 230 |
C compute the density at this level for the buoyancy term in the |
| 231 |
C vertical k.e. equation |
| 232 |
|
| 233 |
Dd(k2+1)=STATE1(Pd(k2+1,2),Pd(k2+1,1),i,j,k2+1,bi,bj,myThid) |
| 234 |
|
| 235 |
C next, solve for the vertical velocity k.e. using combined eq. (4) |
| 236 |
C and eq (5) from the paper |
| 237 |
|
| 238 |
#ifdef ALLOW_OPPS_DEBUG |
| 239 |
IF ( OPPSdebugLevel.GE.debLevA ) THEN |
| 240 |
WRITE(msgBuf,'(A,3E12.4,I3)') |
| 241 |
& 'S/R OPPS_CALC: Dd,De,entr,k ',Dd(k2),De(k2),entrainrate,k2 |
| 242 |
CALL PRINT_MESSAGE( msgBuf, standardMessageUnit, |
| 243 |
& SQUEEZE_RIGHT , 1) |
| 244 |
ENDIF |
| 245 |
#endif /* ALLOW_OPPS_DEBUG */ |
| 246 |
CML insert Eq. (4) into Eq. (5) to get something like this for wp^2 |
| 247 |
wsqr = wsqr - wsqr*abs(entrainrate)+ gravity* |
| 248 |
& (dz1*(Dd(k2)-De(k2))/De(k2) |
| 249 |
& +dz2*(Dd(k2+1)-De(k2+1))/De(k2+1)) |
| 250 |
|
| 251 |
C if negative k.e. then plume has reached max depth, get out of loop |
| 252 |
|
| 253 |
IF(wsqr.LE.0.0)then |
| 254 |
maxdepth = k2 |
| 255 |
#ifdef ALLOW_OPPS_DEBUG |
| 256 |
IF ( OPPSdebugLevel.GE.debLevA ) THEN |
| 257 |
WRITE(msgBuf,'(A,I3)') |
| 258 |
& 'S/R OPPS_CALC: Plume velocity went to zero at level ', k2 |
| 259 |
CALL PRINT_MESSAGE( msgBuf, standardMessageUnit, |
| 260 |
& SQUEEZE_RIGHT , 1) |
| 261 |
WRITE(msgBuf,'(A,4A14)') |
| 262 |
& 'S/R OPPS_CALC: ', 'wsqr', 'entrainrate', |
| 263 |
& '(Dd-De)/De up', '(Dd-De)/De do' |
| 264 |
CALL PRINT_MESSAGE( msgBuf, standardMessageUnit, |
| 265 |
& SQUEEZE_RIGHT , 1) |
| 266 |
WRITE(msgBuf,'(A,4E14.6)') |
| 267 |
& 'S/R OPPS_CALC: ', wsqr, entrainrate, |
| 268 |
& (Dd(k2)-De(k2))/De(k2), (Dd(k2+1)-De(k2+1))/De(k2+1) |
| 269 |
CALL PRINT_MESSAGE( msgBuf, standardMessageUnit, |
| 270 |
& SQUEEZE_RIGHT , 1) |
| 271 |
ENDIF |
| 272 |
#endif /* ALLOW_OPPS_DEBUG */ |
| 273 |
if(maxdepth.eq.k) goto 1000 |
| 274 |
goto 1 |
| 275 |
endif |
| 276 |
Wd(k2+1)=sqrt(wsqr) |
| 277 |
|
| 278 |
C compute a new radius based on the new mass flux at this grid level |
| 279 |
C from Eq. (4) |
| 280 |
|
| 281 |
radius=sqrt(newflux/(Wd(k2)*Dd(k2))) |
| 282 |
ELSE |
| 283 |
maxdepth=k2 |
| 284 |
if(maxdepth.eq.k) goto 1000 |
| 285 |
GOTO 1 |
| 286 |
ENDIF |
| 287 |
ENDDO |
| 288 |
|
| 289 |
C plume has reached the bottom |
| 290 |
|
| 291 |
MaxDepth=kMax |
| 292 |
|
| 293 |
1 CONTINUE |
| 294 |
|
| 295 |
Ad(k)=FRACTIONAL_AREA |
| 296 |
IC=0 |
| 297 |
|
| 298 |
C start iteration on fractional area, not used in OGCM implementation |
| 299 |
|
| 300 |
DO IC=1,Max_ABE_Iterations |
| 301 |
|
| 302 |
C next compute the mass flux beteen each grid box using the entrainment |
| 303 |
|
| 304 |
Md(k)=Wd(k)*Ad(k) |
| 305 |
|
| 306 |
DO k2=k+1,maxDepth |
| 307 |
Md(k2)=Md(k)*PlumeEntrainment(k2) |
| 308 |
#ifdef ALLOW_OPPS_DEBUG |
| 309 |
IF ( OPPSdebugLevel.GE.debLevA ) THEN |
| 310 |
WRITE(msgBuf,'(A,2E12.4,I3)') |
| 311 |
& 'S/R OPPS_CALC: Md, Wd, and k are ',Md(k2),Wd(k2),k2 |
| 312 |
CALL PRINT_MESSAGE( msgBuf, standardMessageUnit, |
| 313 |
& SQUEEZE_RIGHT , 1) |
| 314 |
ENDIF |
| 315 |
#endif /* ALLOW_OPPS_DEBUG */ |
| 316 |
ENDDO |
| 317 |
|
| 318 |
C Now move on to calculate new temperature using flux from |
| 319 |
C Td, Sd, Wd, ta, sa, and we. Values for these variables are at |
| 320 |
C center of grid cell, use weighted average to get boundary values |
| 321 |
|
| 322 |
C use a timestep limited by the GCM model timestep and the maximum plume |
| 323 |
C velocity (CFL criteria) |
| 324 |
|
| 325 |
C calculate the weighted wd, td, and sd |
| 326 |
dt = dtts |
| 327 |
do k2=k,maxDepth-1 |
| 328 |
IF ( Wd(K2) .NE. 0. _d 0 ) dt = min(dt,drF(k2)/Wd(k2)) |
| 329 |
|
| 330 |
C time integration will be integer number of steps to get one |
| 331 |
C gcm time step |
| 332 |
|
| 333 |
ntime = nint(0.5*int(dtts/dt)) |
| 334 |
if(ntime.eq.0) then |
| 335 |
ntime = 1 |
| 336 |
endif |
| 337 |
|
| 338 |
C make sure area weighted vertical velocities match; in other words |
| 339 |
C make sure mass in equals mass out at the intersection of each grid |
| 340 |
C cell. Eq. (20) |
| 341 |
|
| 342 |
mda(k2) = (md(k2)*drF(k2)+md(k2+1)*drF(k2+1))/ |
| 343 |
* (drF(k2)+drF(k2+1)) |
| 344 |
|
| 345 |
wda(k2) = (wd(k2)*drF(k2)+wd(k2+1)*drF(k2+1))/ |
| 346 |
* (drF(k2)+drF(k2+1)) |
| 347 |
|
| 348 |
DO ktr = 1, nTracerInUse |
| 349 |
Pda(k2,ktr) = Pd(k2,ktr) |
| 350 |
Paa(k2,ktr) = tracerEnv(k2+1,ktr) |
| 351 |
ENDDO |
| 352 |
|
| 353 |
enddo |
| 354 |
dt = min(dt,dtts) |
| 355 |
#ifdef ALLOW_OPPS_DEBUG |
| 356 |
IF ( OPPSdebugLevel.GE.debLevA ) THEN |
| 357 |
WRITE(msgBuf,'(A,F14.4)') |
| 358 |
& 'S/R OPPS_CALC: time step = ', dt |
| 359 |
CALL PRINT_MESSAGE( msgBuf, standardMessageUnit, |
| 360 |
& SQUEEZE_RIGHT , 1) |
| 361 |
ENDIF |
| 362 |
#endif /* ALLOW_OPPS_DEBUG */ |
| 363 |
DO ktr=1,nTracerInUse |
| 364 |
Pda(maxdepth,ktr) = Pd(maxdepth,ktr) |
| 365 |
ENDDO |
| 366 |
|
| 367 |
kmx = maxdepth-1 |
| 368 |
do nn=1,ntime |
| 369 |
|
| 370 |
C top point |
| 371 |
|
| 372 |
DO ktr = 1,nTracerInUse |
| 373 |
tracerEnv(k,ktr) = tracerEnv(k,ktr)- |
| 374 |
& (mda(k)*(Pda(k,ktr)-Paa(k,ktr)))*dt*recip_drF(k) |
| 375 |
ENDDO |
| 376 |
|
| 377 |
C now do inner points if there are any |
| 378 |
|
| 379 |
CML if(Maxdepth-k.gt.1) then |
| 380 |
CML This if statement is superfluous |
| 381 |
CML IF ( k .LT. Maxdepth-1 ) THEN |
| 382 |
CML DO k2=k+1,Maxdepth-1 |
| 383 |
CML mda(maxDepth) = 0. |
| 384 |
DO k2=k+1,kmx |
| 385 |
k2m1 = max(k,k2-1) |
| 386 |
k2p1 = max(k2+1,maxDepth) |
| 387 |
|
| 388 |
DO ktr = 1,nTracerInUse |
| 389 |
tracerEnv(k2,ktr) = tracerEnv(k2,ktr) + |
| 390 |
& (mda(k2m1)*(Pda(k2m1,ktr)-Paa(k2m1,ktr)) |
| 391 |
& -mda(k2) *(Pda(k2,ktr) -Paa(k2,ktr)) ) |
| 392 |
& *dt*recip_drF(k2) |
| 393 |
ENDDO |
| 394 |
ENDDO |
| 395 |
CML This if statement is superfluous |
| 396 |
CML ENDIF |
| 397 |
|
| 398 |
C bottom point |
| 399 |
|
| 400 |
DO ktr=1,nTracerInUse |
| 401 |
tracerEnv(kmx+1,ktr) = tracerEnv(kmx+1,ktr)+ |
| 402 |
& mda(kmx)*(Pda(kmx,ktr)-Paa(kmx,ktr))*dt*recip_drF(kmx+1) |
| 403 |
ENDDO |
| 404 |
|
| 405 |
C set the environmental temp and salinity to equal new fields |
| 406 |
|
| 407 |
DO ktr=1,nTracerInUse |
| 408 |
DO k2=1,kmx |
| 409 |
paa(k2,ktr) = tracerEnv(k2+1,ktr) |
| 410 |
ENDDO |
| 411 |
ENDDO |
| 412 |
|
| 413 |
C end loop on number of time integration steps |
| 414 |
|
| 415 |
enddo |
| 416 |
ENDDO |
| 417 |
|
| 418 |
C count convection event in this grid cell |
| 419 |
|
| 420 |
OPPSconvectCount(I,J,K,bi,bj) = |
| 421 |
& OPPSconvectCount(I,J,K,bi,bj) + 1. _d 0 |
| 422 |
|
| 423 |
C jump here if k = maxdepth or if level not unstable, go to next |
| 424 |
C profile point |
| 425 |
|
| 426 |
1000 continue |
| 427 |
|
| 428 |
C end of k-loop |
| 429 |
|
| 430 |
ENDDO |
| 431 |
|
| 432 |
C-- End IF (DIFFERENT_MULTIPLE) |
| 433 |
ENDIF |
| 434 |
|
| 435 |
RETURN |
| 436 |
END |
| 437 |
|
| 438 |
C---+----1----+----2----+----3----+----4----+----5----+----6----+----7-|--+----| |
| 439 |
|
| 440 |
_RL FUNCTION STATE1( sLoc, tLoc, i, j, kRef, bi, bj, myThid ) |
| 441 |
|
| 442 |
C !DESCRIPTION: \bv |
| 443 |
C *===============================================================* |
| 444 |
C | o SUBROUTINE STATE1 |
| 445 |
C | Calculates rho(S,T,p) |
| 446 |
C | It is absolutely necessary to compute |
| 447 |
C | the full rho and not sigma=rho-rhoConst, because |
| 448 |
C | density is used as a scale factor for fluxes and velocities |
| 449 |
C *===============================================================* |
| 450 |
C \ev |
| 451 |
|
| 452 |
C !USES: |
| 453 |
IMPLICIT NONE |
| 454 |
C == Global variables == |
| 455 |
#include "SIZE.h" |
| 456 |
#include "EEPARAMS.h" |
| 457 |
#include "PARAMS.h" |
| 458 |
#include "GRID.h" |
| 459 |
#include "DYNVARS.h" |
| 460 |
|
| 461 |
C !INPUT/OUTPUT PARAMETERS: |
| 462 |
C == Routine arguments == |
| 463 |
INTEGER i, j, kRef, bi, bj, myThid |
| 464 |
_RL tLoc, sLoc |
| 465 |
|
| 466 |
C !LOCAL VARIABLES: |
| 467 |
C == Local variables == |
| 468 |
_RL rhoLoc |
| 469 |
_RL pLoc |
| 470 |
|
| 471 |
CMLC estimate pressure from depth at cell centers |
| 472 |
CML mtoSI = gravity*rhoConst |
| 473 |
CML pLoc = ABS(rC(kRef))*mtoSI |
| 474 |
|
| 475 |
IF ( usingZCoords ) THEN |
| 476 |
C in Z coordinates the pressure is rho0 * (hydrostatic) Potential |
| 477 |
IF ( useDynP_inEos_Zc ) THEN |
| 478 |
C---------- |
| 479 |
C NOTE: For now, totPhiHyd only contains the Potential anomaly |
| 480 |
C since PhiRef is not available for Atmos and has not (yet) |
| 481 |
C been added in S/R DIAGS_PHI_HYD |
| 482 |
C---------- |
| 483 |
pLoc = rhoConst*( totPhiHyd(i,j,kRef,bi,bj) |
| 484 |
& -rC(kRef)*gravity |
| 485 |
& )*maskC(i,j,kRef,bi,bj) |
| 486 |
ELSE |
| 487 |
pLoc = -rhoConst*rC(kRef)*gravity*maskC(i,j,kRef,bi,bj) |
| 488 |
ENDIF |
| 489 |
ELSEIF ( usingPCoords ) THEN |
| 490 |
C in P coordinates the pressure is just the coordinate of |
| 491 |
C the tracer point |
| 492 |
pLoc = rC(kRef)* maskC(i,j,kRef,bi,bj) |
| 493 |
ENDIF |
| 494 |
|
| 495 |
CALL FIND_RHO_SCALAR( tLoc, sLoc, pLoc, rhoLoc, myThid ) |
| 496 |
STATE1 = rhoLoc |
| 497 |
|
| 498 |
#endif /* ALLOW_OPPS */ |
| 499 |
RETURN |
| 500 |
END |
| 501 |
|
| 502 |
#ifdef OPPS_ORGCODE |
| 503 |
C---+----1----+----2----+----3----+----4----+----5----+----6----+----7-|--+----| |
| 504 |
|
| 505 |
C Listed below is the subroutine for use in parallel 3-d circulation code. |
| 506 |
C It has been used in the parallel semtner-chervin code and is now being used |
| 507 |
C In the POP code. The subroutine is called nlopps (long story to explain why). |
| 508 |
|
| 509 |
C I've attached the version of lopps that we've been using in the simulations. |
| 510 |
C There is one common block that is different from the standard model commons |
| 511 |
C (countc) and it is not needed if the array convadj is not used. The routine |
| 512 |
C does need "kmp" which is why the boundc common is included. For massively |
| 513 |
C parallel codes (like POP) we think this will work well when converted from a |
| 514 |
C "slab" (i=is,ie) to a column, which just means removing the "do i=is,ie" loop. |
| 515 |
C There are differences between this |
| 516 |
C code and the 1-d code and the earlier scheme implemented in 3-d models. These |
| 517 |
C differences are described below. |
| 518 |
|
| 519 |
subroutine nlopps(j,is,ie,ta,sa,gcmdz) |
| 520 |
|
| 521 |
parameter (imt = 361 , jmt = 301 , km = 30 ) |
| 522 |
|
| 523 |
C Nlopps: E. Skyllingstad and T. Paluszkiewicz |
| 524 |
|
| 525 |
C Version: December 11, 1996 |
| 526 |
|
| 527 |
C Nlopps: This version of lopps is significantly different from |
| 528 |
C the original code developed by R. Romea and T. Paluskiewicz. The |
| 529 |
C code uses a flux constraint to control the change in T and S at |
| 530 |
C each grid level. First, a plume profile of T,S, and W are |
| 531 |
C determined using the standard plume model, but with a detraining |
| 532 |
C mass instead of entraining. Thus, the T and S plume |
| 533 |
C characteristics still change, but the plume contracts in size |
| 534 |
C rather than expanding ala classical entraining plumes. This |
| 535 |
C is heuristically more in line with large eddy simulation results. |
| 536 |
C At each grid level, the convergence of plume velocity determines |
| 537 |
C the flux of T and S, which is conserved by using an upstream |
| 538 |
C advection. The vertical velocity is balanced so that the area |
| 539 |
C weighted upward velocity equals the area weighted downdraft |
| 540 |
C velocity, ensuring mass conservation. The present implementation |
| 541 |
C adjusts the plume for a time period equal to the time for 1/2 of |
| 542 |
C the mass of the fastest moving level to move downward. As a |
| 543 |
C consequence, the model does not completely adjust the profile at |
| 544 |
C each model time step, but provides a smooth adjustment over time. |
| 545 |
|
| 546 |
c include "params.h" |
| 547 |
c include "plume_fast_inc.h" |
| 548 |
c include "plume_fast.h" |
| 549 |
c #include "loppsd.h" |
| 550 |
|
| 551 |
real ta(imt,km),sa(imt,km),gcmdz(km),dz(km) |
| 552 |
real pdensity,wsqr,oldflux,newflux,entrainrate,adtemp |
| 553 |
REAL Del,D,dza1,dza2,kd,kd1,Smix,Thmix,PlumeS,PlumeT,PlumeD |
| 554 |
|
| 555 |
INTEGER i,j,k |
| 556 |
Clfh |
| 557 |
integer is,ie,k2 |
| 558 |
Clfh |
| 559 |
REAL D1,D2,state1,Density |
| 560 |
REAL dz1,dz2 |
| 561 |
REAL radius,StartingFlux |
| 562 |
real ttemp(km),stemp(km),taa(km),saa(km) |
| 563 |
real wda(km),tda(km),sda(km),mda(km) |
| 564 |
real dtts,dt,sumo,sumn |
| 565 |
integer ntime,nn,kmx,ic |
| 566 |
|
| 567 |
LOGICAL debug,done |
| 568 |
INTEGER MAX_ABE_ITERATIONS |
| 569 |
PARAMETER(MAX_ABE_ITERATIONS=1) |
| 570 |
REAL PlumeRadius |
| 571 |
REAL STABILITY_THRESHOLD |
| 572 |
REAL FRACTIONAL_AREA |
| 573 |
REAL MAX_FRACTIONAL_AREA |
| 574 |
REAL VERTICAL_VELOCITY |
| 575 |
REAL ENTRAINMENT_RATE |
| 576 |
REAL e2 |
| 577 |
PARAMETER ( PlumeRadius = 100.D0 ) |
| 578 |
PARAMETER ( STABILITY_THRESHOLD = -1.E-4 ) |
| 579 |
PARAMETER ( FRACTIONAL_AREA = .1E0 ) |
| 580 |
PARAMETER ( MAX_FRACTIONAL_AREA = .8E0 ) |
| 581 |
PARAMETER ( VERTICAL_VELOCITY = .02E0 ) |
| 582 |
PARAMETER ( ENTRAINMENT_RATE = -.05E0 ) |
| 583 |
PARAMETER ( e2 = 2.E0*ENTRAINMENT_RATE ) |
| 584 |
! Arrays. |
| 585 |
REAL Ad(km),Sd(km),Td(km),Wd(km),Dd(km),Md(km) |
| 586 |
REAL Se(km),Te(km),We(km),De(km) |
| 587 |
REAL PlumeEntrainment(km) |
| 588 |
REAL GridThickness(km) |
| 589 |
|
| 590 |
C input kmp through a common block |
| 591 |
|
| 592 |
common / boundc / wsx(imt,jmt),wsy(imt,jmt),hfs(imt,jmt), |
| 593 |
1 ple(imt,jmt),kmp(imt,jmt),kmq(imt,jmt) |
| 594 |
cwmseas |
| 595 |
& ,wsx1(imt,jmt),wsy1(imt,jmt) |
| 596 |
1 ,wsx2(imt,jmt),wsy2(imt,jmt) |
| 597 |
|
| 598 |
C input the variables through a common |
| 599 |
|
| 600 |
logical problem |
| 601 |
common /countc/ convadj(imt,jmt,km),ics,depth(km),problem |
| 602 |
|
| 603 |
C-----may want to setup an option to get this only on first call |
| 604 |
C otherwise it is repetive |
| 605 |
C griddz is initialize by call to setupgrid |
| 606 |
|
| 607 |
dtts = 2400 |
| 608 |
|
| 609 |
do k=1,km |
| 610 |
dz(k) = 0.01*gcmdz(k) |
| 611 |
enddo |
| 612 |
|
| 613 |
do k=1,km |
| 614 |
GridThickness(k) = dz(k) |
| 615 |
enddo |
| 616 |
|
| 617 |
C modified to loop over slab |
| 618 |
|
| 619 |
DO i=is,ie |
| 620 |
|
| 621 |
numgridpoints=kmp(i,j) |
| 622 |
|
| 623 |
C go to next column if only 1 grid point or on land |
| 624 |
|
| 625 |
if(numgridpoints.le.1) goto 1100 |
| 626 |
|
| 627 |
C loop over depth |
| 628 |
|
| 629 |
debug = .false. |
| 630 |
|
| 631 |
C first save copy of initial profile |
| 632 |
|
| 633 |
DO k=1,NumGridPoints |
| 634 |
stemp(k)=sa(i,k) |
| 635 |
ttemp(k)=ta(i,k) |
| 636 |
|
| 637 |
C do a check of t and s range, if out of bounds set flag |
| 638 |
|
| 639 |
if(problem) then |
| 640 |
write(0,*)"Code in trouble before this nlopps call" |
| 641 |
return |
| 642 |
endif |
| 643 |
|
| 644 |
if(sa(i,k).gt.40..or.ta(i,k).lt.-4.0) then |
| 645 |
problem = .true. |
| 646 |
write(0,*)"t out of range at j ",j |
| 647 |
debug = .true. |
| 648 |
return |
| 649 |
endif |
| 650 |
ENDDO |
| 651 |
|
| 652 |
if(debug) then |
| 653 |
write(*,*)"T and S Profile at ",i,j |
| 654 |
write(*,*)(ta(i,k),sa(i,k),k=1,NumGridPoints) |
| 655 |
endif |
| 656 |
|
| 657 |
DO k=1,NumGridPoints-1 |
| 658 |
|
| 659 |
C initialize the plume T,S,density, and w velocity |
| 660 |
|
| 661 |
Sd(k)=stemp(k) |
| 662 |
Td(k)=ttemp(k) |
| 663 |
Dd(k)=state1(stemp(k),ttemp(k),k) |
| 664 |
De(k)=Dd(k) |
| 665 |
c Wd(k)=VERTICAL_VELOCITY |
| 666 |
|
| 667 |
C guess at initial top grid cell vertical velocity |
| 668 |
|
| 669 |
Wd(k) = 0.03 |
| 670 |
|
| 671 |
C these estimates of initial plume velocity based on plume size and |
| 672 |
C top grid cell water mass |
| 673 |
|
| 674 |
c Wd(k) = 0.5*dz(k)/(dtts*FRACTIONAL_AREA) |
| 675 |
c Wd(k) = 0.5*dz(k)/dtts |
| 676 |
|
| 677 |
wsqr=Wd(k)*Wd(k) |
| 678 |
PlumeEntrainment(k) = 0.0 |
| 679 |
|
| 680 |
if(debug) write(0,*) 'Doing old lowerparcel' |
| 681 |
radius=PlumeRadius |
| 682 |
StartingFlux=radius*radius*Wd(k)*Dd(k) |
| 683 |
oldflux=StartingFlux |
| 684 |
|
| 685 |
dz2=GridThickness(k) |
| 686 |
DO k2=k,NumGridPoints-1 |
| 687 |
D1=state1(Sd(k2),Td(k2),k2+1) |
| 688 |
D2=state1(stemp(k2+1),ttemp(k2+1),k2+1) |
| 689 |
De(k2+1)=D2 |
| 690 |
|
| 691 |
C To start downward, parcel has to initially be heavier than environment |
| 692 |
C but after it has started moving, we continue plume until plume tke or |
| 693 |
C flux goes negative |
| 694 |
|
| 695 |
IF (D2-D1 .LT. STABILITY_THRESHOLD.or.k2.ne.k) THEN |
| 696 |
dz1=dz2 |
| 697 |
dz2=GridThickness(k2+1) |
| 698 |
|
| 699 |
C define mass flux according to eq. 4 from paper |
| 700 |
|
| 701 |
newflux=oldflux+e2*radius*Wd(k2)*Dd(k2)*0.50* |
| 702 |
. (dz1+dz2) |
| 703 |
|
| 704 |
PlumeEntrainment(k2+1) = newflux/StartingFlux |
| 705 |
|
| 706 |
IF(newflux.LT.0.0) then |
| 707 |
if(debug) then |
| 708 |
write(0,*)"Plume entrained to zero at ",k2 |
| 709 |
endif |
| 710 |
maxdepth = k2 |
| 711 |
if(maxdepth.eq.k) goto 1000 |
| 712 |
goto 1 |
| 713 |
endif |
| 714 |
|
| 715 |
C entrainment rate is basically a scaled mass flux dM/M |
| 716 |
|
| 717 |
entrainrate = (newflux - oldflux)/newflux |
| 718 |
oldflux = newflux |
| 719 |
|
| 720 |
C mix var(s) are the average environmental values over the two grid levels |
| 721 |
|
| 722 |
smix=(dz1*stemp(k2)+dz2*stemp(k2+1))/(dz1+dz2) |
| 723 |
thmix=(dz1*ttemp(k2)+dz2*ttemp(k2+1))/(dz1+dz2) |
| 724 |
|
| 725 |
C first compute the new salinity and temperature for this level |
| 726 |
C using equations 3.6 and 3.7 from the paper |
| 727 |
|
| 728 |
sd(k2+1)=sd(k2) - entrainrate*(smix - sd(k2)) |
| 729 |
td(k2+1)=td(k2) - entrainrate*(thmix - td(k2)) |
| 730 |
|
| 731 |
C compute the density at this level for the buoyancy term in the |
| 732 |
C vertical k.e. equation |
| 733 |
|
| 734 |
Dd(k2+1)=state1(Sd(k2+1),Td(k2+1),k2+1) |
| 735 |
|
| 736 |
C next, solve for the vertical velocity k.e. using combined eq. 4 |
| 737 |
C and eq 5 from the paper |
| 738 |
|
| 739 |
if(debug) then |
| 740 |
write(0,*)"Dd,De,entr,k ",Dd(k2),De(k2),entrainrate,k2 |
| 741 |
endif |
| 742 |
|
| 743 |
wsqr = wsqr - wsqr*abs(entrainrate)+ 9.81* |
| 744 |
. (dz1*(Dd(k2)-De(k2))/De(k2) |
| 745 |
. +dz2*(Dd(k2+1)-De(k2+1))/De(k2+1)) |
| 746 |
|
| 747 |
C if negative k.e. then plume has reached max depth, get out of loop |
| 748 |
|
| 749 |
IF(wsqr.LT.0.0)then |
| 750 |
maxdepth = k2 |
| 751 |
if(debug) then |
| 752 |
write(0,*)"Plume velocity went to zero at ",k2 |
| 753 |
endif |
| 754 |
if(maxdepth.eq.k) goto 1000 |
| 755 |
goto 1 |
| 756 |
endif |
| 757 |
Wd(k2+1)=sqrt(wsqr) |
| 758 |
|
| 759 |
C compute a new radius based on the new mass flux at this grid level |
| 760 |
|
| 761 |
radius=sqrt(newflux/(Wd(k2)*Dd(k2))) |
| 762 |
ELSE |
| 763 |
maxdepth=k2 |
| 764 |
if(maxdepth.eq.k) goto 1000 |
| 765 |
GOTO 1 |
| 766 |
ENDIF |
| 767 |
ENDDO |
| 768 |
|
| 769 |
C plume has reached the bottom |
| 770 |
|
| 771 |
MaxDepth=NumGridPoints |
| 772 |
|
| 773 |
1 continue |
| 774 |
|
| 775 |
Ad(k)=FRACTIONAL_AREA |
| 776 |
IC=0 |
| 777 |
|
| 778 |
C start iteration on fractional area, not used in OGCM implementation |
| 779 |
|
| 780 |
DO IC=1,Max_ABE_Iterations |
| 781 |
|
| 782 |
C next compute the mass flux beteen each grid box using the entrainment |
| 783 |
|
| 784 |
92 continue |
| 785 |
Md(k)=Wd(k)*Ad(k) |
| 786 |
|
| 787 |
DO k2=k+1,MaxDepth |
| 788 |
Md(k2)=Md(k)*PlumeEntrainment(k2) |
| 789 |
if(debug) then |
| 790 |
write(0,*)"Md, Wd, and k are ",Md(k2),Wd(k2),k2 |
| 791 |
endif |
| 792 |
ENDDO |
| 793 |
|
| 794 |
C Now move on to calculate new temperature using flux from |
| 795 |
C Td, Sd, Wd, ta, sa, and we. Values for these variables are at |
| 796 |
C center of grid cell, use weighted average to get boundary values |
| 797 |
|
| 798 |
C use a timestep limited by the GCM model timestep and the maximum plume |
| 799 |
C velocity (CFL criteria) |
| 800 |
|
| 801 |
C calculate the weighted wd, td, and sd |
| 802 |
|
| 803 |
dt = dtts |
| 804 |
do k2=k,maxdepth-1 |
| 805 |
dt = min(dt,dz(k2)/wd(k2)) |
| 806 |
|
| 807 |
C time integration will be integer number of steps to get one |
| 808 |
C gcm time step |
| 809 |
|
| 810 |
ntime = nint(0.5*int(dtts/dt)) |
| 811 |
if(ntime.eq.0) then |
| 812 |
ntime = 1 |
| 813 |
endif |
| 814 |
|
| 815 |
C make sure area weighted vertical velocities match; in other words |
| 816 |
C make sure mass in equals mass out at the intersection of each grid |
| 817 |
C cell. |
| 818 |
|
| 819 |
mda(k2) = (md(k2)*dz(k2)+md(k2+1)*dz(k2+1))/ |
| 820 |
* (dz(k2)+dz(k2+1)) |
| 821 |
|
| 822 |
wda(k2) = (wd(k2)*dz(k2)+wd(k2+1)*dz(k2+1))/ |
| 823 |
* (dz(k2)+dz(k2+1)) |
| 824 |
|
| 825 |
tda(k2) = td(k2) |
| 826 |
sda(k2) = sd(k2) |
| 827 |
|
| 828 |
taa(k2) = ttemp(k2+1) |
| 829 |
saa(k2) = stemp(k2+1) |
| 830 |
|
| 831 |
enddo |
| 832 |
dt = min(dt,dtts) |
| 833 |
if(debug) then |
| 834 |
write(0,*)"Time step is ", dt |
| 835 |
endif |
| 836 |
tda(maxdepth) = td(maxdepth) |
| 837 |
sda(maxdepth) = sd(maxdepth) |
| 838 |
|
| 839 |
C do top and bottom points first |
| 840 |
|
| 841 |
kmx = maxdepth-1 |
| 842 |
do nn=1,ntime |
| 843 |
|
| 844 |
ttemp(k) = ttemp(k)- |
| 845 |
* (mda(k)*(tda(k)-taa(k)))*dt*recip_drF(k) |
| 846 |
|
| 847 |
stemp(k) = stemp(k)- |
| 848 |
* (mda(k)*(sda(k)-saa(k)))*dt*recip_drF(k) |
| 849 |
|
| 850 |
C now do inner points if there are any |
| 851 |
|
| 852 |
if(Maxdepth-k.gt.1) then |
| 853 |
do k2=k+1,Maxdepth-1 |
| 854 |
|
| 855 |
ttemp(k2) = ttemp(k2) + |
| 856 |
* (mda(k2-1)*(tda(k2-1)-taa(k2-1))- |
| 857 |
* mda(k2)*(tda(k2)-taa(k2))) |
| 858 |
* *dt*recip_drF(k2) |
| 859 |
|
| 860 |
stemp(k2) = stemp(k2) + |
| 861 |
* (mda(k2-1)*(sda(k2-1)-saa(k2-1))- |
| 862 |
* mda(k2)*(sda(k2)-saa(k2))) |
| 863 |
* *dt*recip_drF(k2) |
| 864 |
|
| 865 |
enddo |
| 866 |
endif |
| 867 |
ttemp(kmx+1) = ttemp(kmx+1)+ |
| 868 |
* (mda(kmx)*(tda(kmx)-taa(kmx)))* |
| 869 |
* dt*recip_drF(kmx+1) |
| 870 |
|
| 871 |
stemp(kmx+1) = stemp(kmx+1)+ |
| 872 |
* (mda(kmx)*(sda(kmx)-saa(kmx)))* |
| 873 |
* dt*recip_drF(kmx+1) |
| 874 |
|
| 875 |
C set the environmental temp and salinity to equal new fields |
| 876 |
|
| 877 |
do k2=1,maxdepth-1 |
| 878 |
taa(k2) = ttemp(k2+1) |
| 879 |
saa(k2) = stemp(k2+1) |
| 880 |
enddo |
| 881 |
|
| 882 |
C end loop on number of time integration steps |
| 883 |
|
| 884 |
enddo |
| 885 |
ENDDO |
| 886 |
999 continue |
| 887 |
|
| 888 |
C assume that it converged, so update the ta and sa with new fields |
| 889 |
|
| 890 |
c if(i.gt.180.and.j.gt.200.and.i.lt.240) then |
| 891 |
c write(*,*)"Converged ",i,j,k,maxdepth,ttemp(k+1),ta(i,k+1) |
| 892 |
c endif |
| 893 |
do k2=k,maxdepth |
| 894 |
convadj(i,j,k2) = convadj(i,j,k2) + (ttemp(k2)- |
| 895 |
* ta(i,k2)) |
| 896 |
sa(i,k2) = stemp(k2) |
| 897 |
ta(i,k2) = ttemp(k2) |
| 898 |
c if(i.gt.180.and.j.gt.200.and.i.lt.240) then |
| 899 |
c write(*,*)"convadj ",convadj(i,j,k2) |
| 900 |
c endif |
| 901 |
|
| 902 |
C see if nlopps messed up |
| 903 |
|
| 904 |
if(sa(i,k).gt.40..or.ta(i,k).lt.-4.0) then |
| 905 |
problem = .true. |
| 906 |
write(0,*)"t out of range at j after adjust",j |
| 907 |
debug = .true. |
| 908 |
endif |
| 909 |
|
| 910 |
enddo |
| 911 |
|
| 912 |
C jump here if k = maxdepth or if level not unstable, go to next |
| 913 |
C profile point |
| 914 |
|
| 915 |
1000 continue |
| 916 |
|
| 917 |
C end loop on k, move on to next possible plume |
| 918 |
|
| 919 |
ENDDO |
| 920 |
1100 continue |
| 921 |
|
| 922 |
C i loop |
| 923 |
|
| 924 |
ENDDO |
| 925 |
END |
| 926 |
|
| 927 |
#endif /* OPPS_ORGCODE */ |
Parent Directory
| 
Revision Log
| 
Revision Graph