| Line |
Branch |
Exec |
Source |
| 1 |
|
|
! |
| 2 |
|
|
MODULE slab_heat_transp_mod |
| 3 |
|
|
! |
| 4 |
|
|
! Slab ocean : temperature tendencies due to horizontal diffusion |
| 5 |
|
|
! and / or Ekman transport |
| 6 |
|
|
|
| 7 |
|
|
USE mod_grid_phy_lmdz, ONLY: nbp_lon, nbp_lat, klon_glo |
| 8 |
|
|
IMPLICIT NONE |
| 9 |
|
|
|
| 10 |
|
|
! Variables copied over from dyn3d dynamics: |
| 11 |
|
|
REAL,SAVE,ALLOCATABLE :: fext(:) ! Coriolis f times cell area |
| 12 |
|
|
!$OMP THREADPRIVATE(fext) |
| 13 |
|
|
REAL,SAVE,ALLOCATABLE :: beta(:) ! df/dy |
| 14 |
|
|
!$OMP THREADPRIVATE(beta) |
| 15 |
|
|
REAL,SAVE,ALLOCATABLE :: unsairez(:) ! 1/cell area |
| 16 |
|
|
!$OMP THREADPRIVATE(unsairez) |
| 17 |
|
|
REAL,SAVE,ALLOCATABLE :: unsaire(:) |
| 18 |
|
|
!$OMP THREADPRIVATE(unsaire) |
| 19 |
|
|
REAL,SAVE,ALLOCATABLE :: cu(:) ! cell longitude dim (m) |
| 20 |
|
|
!$OMP THREADPRIVATE(cu) |
| 21 |
|
|
REAL,SAVE,ALLOCATABLE :: cv(:) ! cell latitude dim (m) |
| 22 |
|
|
!$OMP THREADPRIVATE(cv) |
| 23 |
|
|
REAL,SAVE,ALLOCATABLE :: cuvsurcv(:) ! cu/cv (v points) |
| 24 |
|
|
!$OMP THREADPRIVATE(cuvsurcv) |
| 25 |
|
|
REAL,SAVE,ALLOCATABLE :: cvusurcu(:) ! cv/cu (u points) |
| 26 |
|
|
!$OMP THREADPRIVATE(cvusurcu) |
| 27 |
|
|
REAL,SAVE,ALLOCATABLE :: aire(:) ! cell area |
| 28 |
|
|
!$OMP THREADPRIVATE(aire) |
| 29 |
|
|
REAL,SAVE :: apoln ! area of north pole points |
| 30 |
|
|
!$OMP THREADPRIVATE(apoln) |
| 31 |
|
|
REAL,SAVE :: apols ! area of south pole points |
| 32 |
|
|
!$OMP THREADPRIVATE(apols) |
| 33 |
|
|
REAL,SAVE,ALLOCATABLE :: aireu(:) ! area of u cells |
| 34 |
|
|
!$OMP THREADPRIVATE(aireu) |
| 35 |
|
|
REAL,SAVE,ALLOCATABLE :: airev(:) ! area of v cells |
| 36 |
|
|
!$OMP THREADPRIVATE(airev) |
| 37 |
|
|
|
| 38 |
|
|
! Local parameters for slab transport |
| 39 |
|
|
LOGICAL,SAVE :: alpha_var ! variable coef for deep temp (1 layer) |
| 40 |
|
|
!$OMP THREADPRIVATE(alpha_var) |
| 41 |
|
|
LOGICAL,SAVE :: slab_upstream ! upstream scheme ? (1 layer) |
| 42 |
|
|
!$OMP THREADPRIVATE(slab_upstream) |
| 43 |
|
|
LOGICAL,SAVE :: slab_sverdrup ! use wind stress curl at equator |
| 44 |
|
|
!$OMP THREADPRIVATE(slab_sverdrup) |
| 45 |
|
|
LOGICAL,SAVE :: slab_tyeq ! use merid wind stress at equator |
| 46 |
|
|
!$OMP THREADPRIVATE(slab_tyeq) |
| 47 |
|
|
LOGICAL,SAVE :: ekman_zonadv ! use zonal advection by Ekman currents |
| 48 |
|
|
!$OMP THREADPRIVATE(ekman_zonadv) |
| 49 |
|
|
LOGICAL,SAVE :: ekman_zonavg ! zonally average wind stress |
| 50 |
|
|
!$OMP THREADPRIVATE(ekman_zonavg) |
| 51 |
|
|
|
| 52 |
|
|
REAL,SAVE :: alpham |
| 53 |
|
|
!$OMP THREADPRIVATE(alpham) |
| 54 |
|
|
REAL,SAVE :: gmkappa |
| 55 |
|
|
!$OMP THREADPRIVATE(gmkappa) |
| 56 |
|
|
REAL,SAVE :: gm_smax |
| 57 |
|
|
!$OMP THREADPRIVATE(gm_smax) |
| 58 |
|
|
|
| 59 |
|
|
! geometry variables : f, beta, mask... |
| 60 |
|
|
REAL,SAVE,ALLOCATABLE :: zmasqu(:) ! continent mask for zonal mass flux |
| 61 |
|
|
!$OMP THREADPRIVATE(zmasqu) |
| 62 |
|
|
REAL,SAVE,ALLOCATABLE :: zmasqv(:) ! continent mask for merid mass flux |
| 63 |
|
|
!$OMP THREADPRIVATE(zmasqv) |
| 64 |
|
|
REAL,SAVE,ALLOCATABLE :: unsfv(:) ! 1/f, v points |
| 65 |
|
|
!$OMP THREADPRIVATE(unsfv) |
| 66 |
|
|
REAL,SAVE,ALLOCATABLE :: unsbv(:) ! 1/beta |
| 67 |
|
|
!$OMP THREADPRIVATE(unsbv) |
| 68 |
|
|
REAL,SAVE,ALLOCATABLE :: unsev(:) ! 1/epsilon (drag) |
| 69 |
|
|
!$OMP THREADPRIVATE(unsev) |
| 70 |
|
|
REAL,SAVE,ALLOCATABLE :: unsfu(:) ! 1/F, u points |
| 71 |
|
|
!$OMP THREADPRIVATE(unsfu) |
| 72 |
|
|
REAL,SAVE,ALLOCATABLE :: unseu(:) |
| 73 |
|
|
!$OMP THREADPRIVATE(unseu) |
| 74 |
|
|
|
| 75 |
|
|
! Routines from dyn3d, valid on global dynamics grid only: |
| 76 |
|
|
PRIVATE :: gr_fi_dyn, gr_dyn_fi ! to go between 1D nd 2D horiz grid |
| 77 |
|
|
PRIVATE :: gr_scal_v,gr_v_scal,gr_scal_u ! change on 2D grid U,V, T points |
| 78 |
|
|
PRIVATE :: grad,diverg |
| 79 |
|
|
|
| 80 |
|
|
CONTAINS |
| 81 |
|
|
|
| 82 |
|
✗ |
SUBROUTINE ini_slab_transp_geom(ip1jm,ip1jmp1,unsairez_,fext_,unsaire_,& |
| 83 |
|
|
cu_,cuvsurcv_,cv_,cvusurcu_, & |
| 84 |
|
|
aire_,apoln_,apols_, & |
| 85 |
|
✗ |
aireu_,airev_,rlatv, rad, omeg) |
| 86 |
|
|
! number of points in lon, lat |
| 87 |
|
|
IMPLICIT NONE |
| 88 |
|
|
! Routine copies some geometry variables from the dynamical core |
| 89 |
|
|
! see global vars for meaning |
| 90 |
|
|
INTEGER,INTENT(IN) :: ip1jm |
| 91 |
|
|
INTEGER,INTENT(IN) :: ip1jmp1 |
| 92 |
|
|
REAL,INTENT(IN) :: unsairez_(ip1jm) |
| 93 |
|
|
REAL,INTENT(IN) :: fext_(ip1jm) |
| 94 |
|
|
REAL,INTENT(IN) :: unsaire_(ip1jmp1) |
| 95 |
|
|
REAL,INTENT(IN) :: cu_(ip1jmp1) |
| 96 |
|
|
REAL,INTENT(IN) :: cuvsurcv_(ip1jm) |
| 97 |
|
|
REAL,INTENT(IN) :: cv_(ip1jm) |
| 98 |
|
|
REAL,INTENT(IN) :: cvusurcu_(ip1jmp1) |
| 99 |
|
|
REAL,INTENT(IN) :: aire_(ip1jmp1) |
| 100 |
|
|
REAL,INTENT(IN) :: apoln_ |
| 101 |
|
|
REAL,INTENT(IN) :: apols_ |
| 102 |
|
|
REAL,INTENT(IN) :: aireu_(ip1jmp1) |
| 103 |
|
|
REAL,INTENT(IN) :: airev_(ip1jm) |
| 104 |
|
|
REAL,INTENT(IN) :: rlatv(nbp_lat-1) |
| 105 |
|
|
REAL,INTENT(IN) :: rad |
| 106 |
|
|
REAL,INTENT(IN) :: omeg |
| 107 |
|
|
|
| 108 |
|
|
CHARACTER (len = 20) :: modname = 'slab_heat_transp' |
| 109 |
|
|
CHARACTER (len = 80) :: abort_message |
| 110 |
|
|
|
| 111 |
|
|
! Sanity check on dimensions |
| 112 |
|
✗ |
if ((ip1jm.ne.((nbp_lon+1)*(nbp_lat-1))).or. & |
| 113 |
|
|
(ip1jmp1.ne.((nbp_lon+1)*nbp_lat))) then |
| 114 |
|
✗ |
abort_message="ini_slab_transp_geom Error: wrong array sizes" |
| 115 |
|
✗ |
CALL abort_physic(modname,abort_message,1) |
| 116 |
|
|
endif |
| 117 |
|
|
! Allocations could be done only on master process/thread... |
| 118 |
|
✗ |
allocate(unsairez(ip1jm)) |
| 119 |
|
✗ |
unsairez(:)=unsairez_(:) |
| 120 |
|
✗ |
allocate(fext(ip1jm)) |
| 121 |
|
✗ |
fext(:)=fext_(:) |
| 122 |
|
✗ |
allocate(unsaire(ip1jmp1)) |
| 123 |
|
✗ |
unsaire(:)=unsaire_(:) |
| 124 |
|
✗ |
allocate(cu(ip1jmp1)) |
| 125 |
|
✗ |
cu(:)=cu_(:) |
| 126 |
|
✗ |
allocate(cuvsurcv(ip1jm)) |
| 127 |
|
✗ |
cuvsurcv(:)=cuvsurcv_(:) |
| 128 |
|
✗ |
allocate(cv(ip1jm)) |
| 129 |
|
✗ |
cv(:)=cv_(:) |
| 130 |
|
✗ |
allocate(cvusurcu(ip1jmp1)) |
| 131 |
|
✗ |
cvusurcu(:)=cvusurcu_(:) |
| 132 |
|
✗ |
allocate(aire(ip1jmp1)) |
| 133 |
|
✗ |
aire(:)=aire_(:) |
| 134 |
|
✗ |
apoln=apoln_ |
| 135 |
|
✗ |
apols=apols_ |
| 136 |
|
✗ |
allocate(aireu(ip1jmp1)) |
| 137 |
|
✗ |
aireu(:)=aireu_(:) |
| 138 |
|
✗ |
allocate(airev(ip1jm)) |
| 139 |
|
✗ |
airev(:)=airev_(:) |
| 140 |
|
✗ |
allocate(beta(nbp_lat-1)) |
| 141 |
|
✗ |
beta(:)=2*omeg*cos(rlatv(:))/rad |
| 142 |
|
|
|
| 143 |
|
✗ |
END SUBROUTINE ini_slab_transp_geom |
| 144 |
|
|
|
| 145 |
|
✗ |
SUBROUTINE ini_slab_transp(zmasq) |
| 146 |
|
|
|
| 147 |
|
|
! USE ioipsl_getin_p_mod, only: getin_p |
| 148 |
|
|
USE IOIPSL, ONLY : getin |
| 149 |
|
|
IMPLICIT NONE |
| 150 |
|
|
|
| 151 |
|
|
REAL zmasq(klon_glo) ! ocean / continent mask, 1=continent |
| 152 |
|
✗ |
REAL zmasq_2d((nbp_lon+1)*nbp_lat) |
| 153 |
|
✗ |
REAL ff((nbp_lon+1)*(nbp_lat-1)) ! Coriolis parameter |
| 154 |
|
|
REAL eps ! epsilon friction timescale (s-1) |
| 155 |
|
|
INTEGER :: slab_ekman |
| 156 |
|
|
INTEGER i |
| 157 |
|
|
INTEGER :: iim,iip1,jjp1,ip1jm,ip1jmp1 |
| 158 |
|
|
|
| 159 |
|
|
! Some definition for grid size |
| 160 |
|
|
ip1jm=(nbp_lon+1)*(nbp_lat-1) |
| 161 |
|
|
ip1jmp1=(nbp_lon+1)*nbp_lat |
| 162 |
|
|
iim=nbp_lon |
| 163 |
|
✗ |
iip1=nbp_lon+1 |
| 164 |
|
✗ |
jjp1=nbp_lat |
| 165 |
|
|
ip1jm=(nbp_lon+1)*(nbp_lat-1) |
| 166 |
|
|
ip1jmp1=(nbp_lon+1)*nbp_lat |
| 167 |
|
|
|
| 168 |
|
|
! Options for Heat transport |
| 169 |
|
|
! Alpha variable? |
| 170 |
|
✗ |
alpha_var=.FALSE. |
| 171 |
|
✗ |
CALL getin('slab_alphav',alpha_var) |
| 172 |
|
✗ |
print *,'alpha variable',alpha_var |
| 173 |
|
|
! centered ou upstream scheme for meridional transport |
| 174 |
|
✗ |
slab_upstream=.FALSE. |
| 175 |
|
✗ |
CALL getin('slab_upstream',slab_upstream) |
| 176 |
|
✗ |
print *,'upstream slab scheme',slab_upstream |
| 177 |
|
|
! Sverdrup balance at equator ? |
| 178 |
|
✗ |
slab_sverdrup=.TRUE. |
| 179 |
|
✗ |
CALL getin('slab_sverdrup',slab_sverdrup) |
| 180 |
|
✗ |
print *,'Sverdrup balance',slab_sverdrup |
| 181 |
|
|
! Use tauy for meridional flux at equator ? |
| 182 |
|
✗ |
slab_tyeq=.TRUE. |
| 183 |
|
✗ |
CALL getin('slab_tyeq',slab_tyeq) |
| 184 |
|
✗ |
print *,'Tauy forcing at equator',slab_tyeq |
| 185 |
|
|
! Use tauy for meridional flux at equator ? |
| 186 |
|
✗ |
ekman_zonadv=.TRUE. |
| 187 |
|
✗ |
CALL getin('slab_ekman_zonadv',ekman_zonadv) |
| 188 |
|
✗ |
print *,'Use Ekman flow in zonal direction',ekman_zonadv |
| 189 |
|
|
! Use tauy for meridional flux at equator ? |
| 190 |
|
✗ |
ekman_zonavg=.FALSE. |
| 191 |
|
✗ |
CALL getin('slab_ekman_zonavg',ekman_zonavg) |
| 192 |
|
✗ |
print *,'Use zonally-averaged wind stress ?',ekman_zonavg |
| 193 |
|
|
! Value of alpha |
| 194 |
|
✗ |
alpham=2./3. |
| 195 |
|
✗ |
CALL getin('slab_alpha',alpham) |
| 196 |
|
✗ |
print *,'slab_alpha',alpham |
| 197 |
|
|
! GM k coefficient (m2/s) for 2-layers |
| 198 |
|
✗ |
gmkappa=1000. |
| 199 |
|
✗ |
CALL getin('slab_gmkappa',gmkappa) |
| 200 |
|
✗ |
print *,'slab_gmkappa',gmkappa |
| 201 |
|
|
! GM k coefficient (m2/s) for 2-layers |
| 202 |
|
✗ |
gm_smax=2e-3 |
| 203 |
|
✗ |
CALL getin('slab_gm_smax',gm_smax) |
| 204 |
|
✗ |
print *,'slab_gm_smax',gm_smax |
| 205 |
|
|
! ----------------------------------------------------------- |
| 206 |
|
|
! Define ocean / continent mask (no flux into continent cell) |
| 207 |
|
✗ |
allocate(zmasqu(ip1jmp1)) |
| 208 |
|
✗ |
allocate(zmasqv(ip1jm)) |
| 209 |
|
✗ |
zmasqu=1. |
| 210 |
|
✗ |
zmasqv=1. |
| 211 |
|
|
|
| 212 |
|
|
! convert mask to 2D grid |
| 213 |
|
✗ |
CALL gr_fi_dyn(1,iip1,jjp1,zmasq,zmasq_2d) |
| 214 |
|
|
! put flux mask to 0 at boundaries of continent cells |
| 215 |
|
✗ |
DO i=1,ip1jmp1-1 |
| 216 |
|
✗ |
IF (zmasq_2d(i).GT.1e-5 .OR. zmasq_2d(i+1).GT.1e-5) THEN |
| 217 |
|
✗ |
zmasqu(i)=0. |
| 218 |
|
|
ENDIF |
| 219 |
|
|
END DO |
| 220 |
|
✗ |
DO i=iip1,ip1jmp1,iip1 |
| 221 |
|
✗ |
zmasqu(i)=zmasqu(i-iim) |
| 222 |
|
|
END DO |
| 223 |
|
✗ |
DO i=1,ip1jm |
| 224 |
|
✗ |
IF (zmasq_2d(i).GT.1e-5 .OR. zmasq_2d(i+iip1).GT.1e-5) THEN |
| 225 |
|
✗ |
zmasqv(i)=0. |
| 226 |
|
|
END IF |
| 227 |
|
|
END DO |
| 228 |
|
|
|
| 229 |
|
|
! ----------------------------------------------------------- |
| 230 |
|
|
! Coriolis and friction for Ekman transport |
| 231 |
|
✗ |
slab_ekman=2 |
| 232 |
|
✗ |
CALL getin("slab_ekman",slab_ekman) |
| 233 |
|
✗ |
IF (slab_ekman.GT.0) THEN |
| 234 |
|
✗ |
allocate(unsfv(ip1jm)) |
| 235 |
|
✗ |
allocate(unsev(ip1jm)) |
| 236 |
|
✗ |
allocate(unsfu(ip1jmp1)) |
| 237 |
|
✗ |
allocate(unseu(ip1jmp1)) |
| 238 |
|
✗ |
allocate(unsbv(ip1jm)) |
| 239 |
|
|
|
| 240 |
|
✗ |
eps=1e-5 ! Drag |
| 241 |
|
✗ |
CALL getin('slab_eps',eps) |
| 242 |
|
✗ |
print *,'epsilon=',eps |
| 243 |
|
✗ |
ff=fext*unsairez ! Coriolis |
| 244 |
|
|
! coefs to convert tau_x, tau_y to Ekman mass fluxes |
| 245 |
|
|
! on 2D grid v points |
| 246 |
|
|
! Compute correction factor [0 1] near the equator (f<<eps) |
| 247 |
|
✗ |
IF (slab_sverdrup) THEN |
| 248 |
|
|
! New formulation, sharper near equator, when eps gives Rossby Radius |
| 249 |
|
✗ |
DO i=1,ip1jm |
| 250 |
|
✗ |
unsev(i)=exp(-ff(i)*ff(i)/eps**2) |
| 251 |
|
|
ENDDO |
| 252 |
|
|
ELSE |
| 253 |
|
✗ |
DO i=1,ip1jm |
| 254 |
|
✗ |
unsev(i)=eps**2/(ff(i)*ff(i)+eps**2) |
| 255 |
|
|
ENDDO |
| 256 |
|
|
END IF ! slab_sverdrup |
| 257 |
|
|
! 1/beta |
| 258 |
|
✗ |
DO i=1,jjp1-1 |
| 259 |
|
✗ |
unsbv((i-1)*iip1+1:i*iip1)=unsev((i-1)*iip1+1:i*iip1)/beta(i) |
| 260 |
|
|
END DO |
| 261 |
|
|
! 1/f |
| 262 |
|
✗ |
ff=SIGN(MAX(ABS(ff),eps/100.),ff) ! avoid value 0 at equator... |
| 263 |
|
✗ |
DO i=1,ip1jm |
| 264 |
|
✗ |
unsfv(i)=(1.-unsev(i))/ff(i) |
| 265 |
|
|
END DO |
| 266 |
|
|
! compute values on 2D u grid |
| 267 |
|
|
! 1/eps |
| 268 |
|
✗ |
unsev(:)=unsev(:)/eps |
| 269 |
|
✗ |
CALL gr_v_scal(1,unsfv,unsfu) |
| 270 |
|
✗ |
CALL gr_v_scal(1,unsev,unseu) |
| 271 |
|
|
END IF |
| 272 |
|
|
|
| 273 |
|
✗ |
END SUBROUTINE ini_slab_transp |
| 274 |
|
|
|
| 275 |
|
✗ |
SUBROUTINE divgrad_phy(nlevs,temp,delta) |
| 276 |
|
|
! Computes temperature tendency due to horizontal diffusion : |
| 277 |
|
|
! T Laplacian, later multiplied by diffusion coef and time-step |
| 278 |
|
|
|
| 279 |
|
|
IMPLICIT NONE |
| 280 |
|
|
|
| 281 |
|
|
INTEGER, INTENT(IN) :: nlevs ! nlevs : slab layers |
| 282 |
|
|
REAL, INTENT(IN) :: temp(klon_glo,nlevs) ! slab temperature |
| 283 |
|
|
REAL , INTENT(OUT) :: delta(klon_glo,nlevs) ! temp laplacian (heat flux div.) |
| 284 |
|
✗ |
REAL :: delta_2d((nbp_lon+1)*nbp_lat,nlevs) |
| 285 |
|
✗ |
REAL ghx((nbp_lon+1)*nbp_lat,nlevs), ghy((nbp_lon+1)*(nbp_lat-1),nlevs) |
| 286 |
|
|
INTEGER :: ll,iip1,jjp1 |
| 287 |
|
|
|
| 288 |
|
✗ |
iip1=nbp_lon+1 |
| 289 |
|
✗ |
jjp1=nbp_lat |
| 290 |
|
|
|
| 291 |
|
|
! transpose temp to 2D horiz. grid |
| 292 |
|
✗ |
CALL gr_fi_dyn(nlevs,iip1,jjp1,temp,delta_2d) |
| 293 |
|
|
! computes gradient (proportional to heat flx) |
| 294 |
|
✗ |
CALL grad(nlevs,delta_2d,ghx,ghy) |
| 295 |
|
|
! put flux to 0 at ocean / continent boundary |
| 296 |
|
✗ |
DO ll=1,nlevs |
| 297 |
|
✗ |
ghx(:,ll)=ghx(:,ll)*zmasqu |
| 298 |
|
✗ |
ghy(:,ll)=ghy(:,ll)*zmasqv |
| 299 |
|
|
END DO |
| 300 |
|
|
! flux divergence |
| 301 |
|
✗ |
CALL diverg(nlevs,ghx,ghy,delta_2d) |
| 302 |
|
|
! laplacian back to 1D grid |
| 303 |
|
✗ |
CALL gr_dyn_fi(nlevs,iip1,jjp1,delta_2d,delta) |
| 304 |
|
|
|
| 305 |
|
✗ |
RETURN |
| 306 |
|
|
END SUBROUTINE divgrad_phy |
| 307 |
|
|
|
| 308 |
|
✗ |
SUBROUTINE slab_ekman1(tx_phy,ty_phy,ts_phy,dt_phy) |
| 309 |
|
|
! 1.5 Layer Ekman transport temperature tendency |
| 310 |
|
|
! note : tendency dt later multiplied by (delta t)/(rho.H) |
| 311 |
|
|
! to convert from divergence of heat fluxes to T |
| 312 |
|
|
|
| 313 |
|
|
IMPLICIT NONE |
| 314 |
|
|
|
| 315 |
|
|
! tx, ty : wind stress (different grids) |
| 316 |
|
|
! fluxm, fluz : mass *or heat* fluxes |
| 317 |
|
|
! dt : temperature tendency |
| 318 |
|
|
INTEGER ij |
| 319 |
|
|
|
| 320 |
|
|
! ts surface temp, td deep temp (diagnosed) |
| 321 |
|
|
REAL ts_phy(klon_glo) |
| 322 |
|
✗ |
REAL ts((nbp_lon+1)*nbp_lat),td((nbp_lon+1)*nbp_lat) |
| 323 |
|
|
! zonal and meridional wind stress |
| 324 |
|
|
REAL tx_phy(klon_glo),ty_phy(klon_glo) |
| 325 |
|
✗ |
REAL tyu((nbp_lon+1)*nbp_lat),txu((nbp_lon+1)*nbp_lat) |
| 326 |
|
✗ |
REAL txv((nbp_lon+1)*(nbp_lat-1)),tyv((nbp_lon+1)*(nbp_lat-1)) |
| 327 |
|
✗ |
REAL tcurl((nbp_lon+1)*(nbp_lat-1)) |
| 328 |
|
|
! zonal and meridional Ekman mass fluxes at u, v points (2D grid) |
| 329 |
|
✗ |
REAL fluxz((nbp_lon+1)*nbp_lat),fluxm((nbp_lon+1)*(nbp_lat-1)) |
| 330 |
|
|
! vertical and absolute mass fluxes (to estimate alpha) |
| 331 |
|
✗ |
REAL fluxv((nbp_lon+1)*nbp_lat),fluxt((nbp_lon+1)*(nbp_lat-1)) |
| 332 |
|
|
! temperature tendency |
| 333 |
|
✗ |
REAL dt((nbp_lon+1)*nbp_lat),dt_phy(klon_glo) |
| 334 |
|
✗ |
REAL alpha((nbp_lon+1)*nbp_lat) ! deep temperature coef |
| 335 |
|
|
|
| 336 |
|
|
INTEGER iim,iip1,iip2,jjp1,ip1jm,ip1jmi1,ip1jmp1 |
| 337 |
|
|
|
| 338 |
|
|
! Grid definitions |
| 339 |
|
|
iim=nbp_lon |
| 340 |
|
✗ |
iip1=nbp_lon+1 |
| 341 |
|
✗ |
iip2=nbp_lon+2 |
| 342 |
|
✗ |
jjp1=nbp_lat |
| 343 |
|
|
ip1jm=(nbp_lon+1)*(nbp_lat-1) ! = iip1*jjm |
| 344 |
|
✗ |
ip1jmi1=(nbp_lon+1)*(nbp_lat-1)-(nbp_lon+1) ! = ip1jm - iip1 |
| 345 |
|
|
ip1jmp1=(nbp_lon+1)*nbp_lat ! = iip1*jjp1 |
| 346 |
|
|
|
| 347 |
|
|
! Convert taux,y to 2D scalar grid |
| 348 |
|
|
! Note: 2D grid size = iim*jjm. iip1=iim+1 |
| 349 |
|
|
! First and last points in zonal direction are the same |
| 350 |
|
|
! we use 1 index ij from 1 to (iim+1)*(jjm+1) |
| 351 |
|
|
! north and south poles |
| 352 |
|
✗ |
tx_phy(1)=0. |
| 353 |
|
✗ |
tx_phy(klon_glo)=0. |
| 354 |
|
✗ |
ty_phy(1)=0. |
| 355 |
|
✗ |
ty_phy(klon_glo)=0. |
| 356 |
|
✗ |
CALL gr_fi_dyn(1,iip1,jjp1,tx_phy,txu) |
| 357 |
|
✗ |
CALL gr_fi_dyn(1,iip1,jjp1,ty_phy,tyu) |
| 358 |
|
|
! convert to u,v grid (Arakawa C) |
| 359 |
|
|
! Multiply by f or eps to get mass flux |
| 360 |
|
|
! Meridional mass flux |
| 361 |
|
✗ |
CALL gr_scal_v(1,txu,txv) ! wind stress at v points |
| 362 |
|
✗ |
IF (slab_sverdrup) THEN ! Sverdrup bal. near equator |
| 363 |
|
✗ |
tcurl=(txu(1:ip1jm)-txu(iip2:ip1jmp1))/cv(:) |
| 364 |
|
✗ |
fluxm=-tcurl*unsbv-txv*unsfv ! in kg.s-1.m-1 (zonal distance) |
| 365 |
|
|
ELSE |
| 366 |
|
✗ |
CALL gr_scal_v(1,tyu,tyv) |
| 367 |
|
✗ |
fluxm=tyv*unsev-txv*unsfv ! in kg.s-1.m-1 (zonal distance) |
| 368 |
|
|
ENDIF |
| 369 |
|
|
! Zonal mass flux |
| 370 |
|
✗ |
CALL gr_scal_u(1,txu,txu) ! wind stress at u points |
| 371 |
|
✗ |
CALL gr_scal_u(1,tyu,tyu) |
| 372 |
|
✗ |
fluxz=tyu*unsfu+txu*unseu |
| 373 |
|
|
|
| 374 |
|
|
! Correct flux: continent mask and horiz grid size |
| 375 |
|
|
! multiply m-flux by mask and dx: flux in kg.s-1 |
| 376 |
|
✗ |
fluxm=fluxm*cv*cuvsurcv*zmasqv |
| 377 |
|
|
! multiply z-flux by mask and dy: flux in kg.s-1 |
| 378 |
|
✗ |
fluxz=fluxz*cu*cvusurcu*zmasqu |
| 379 |
|
|
|
| 380 |
|
|
! Compute vertical and absolute mass flux (for variable alpha) |
| 381 |
|
✗ |
IF (alpha_var) THEN |
| 382 |
|
✗ |
DO ij=iip2,ip1jm |
| 383 |
|
✗ |
fluxv(ij)=fluxz(ij)-fluxz(ij-1)-fluxm(ij)+fluxm(ij-iip1) |
| 384 |
|
|
fluxt(ij)=ABS(fluxz(ij))+ABS(fluxz(ij-1)) & |
| 385 |
|
✗ |
+ABS(fluxm(ij))+ABS(fluxm(ij-iip1)) |
| 386 |
|
|
ENDDO |
| 387 |
|
✗ |
DO ij=iip1,ip1jmi1,iip1 |
| 388 |
|
✗ |
fluxt(ij+1)=fluxt(ij+iip1) |
| 389 |
|
✗ |
fluxv(ij+1)=fluxv(ij+iip1) |
| 390 |
|
|
END DO |
| 391 |
|
✗ |
fluxt(1)=SUM(ABS(fluxm(1:iim))) |
| 392 |
|
✗ |
fluxt(ip1jmp1)=SUM(ABS(fluxm(ip1jm-iim:ip1jm-1))) |
| 393 |
|
✗ |
fluxv(1)=-SUM(fluxm(1:iim)) |
| 394 |
|
✗ |
fluxv(ip1jmp1)=SUM(fluxm(ip1jm-iim:ip1jm-1)) |
| 395 |
|
✗ |
fluxt=MAX(fluxt,1.e-10) |
| 396 |
|
|
ENDIF |
| 397 |
|
|
|
| 398 |
|
|
! Compute alpha coefficient. |
| 399 |
|
|
! Tdeep = Tsurf * alpha + 271.35 * (1-alpha) |
| 400 |
|
✗ |
IF (alpha_var) THEN |
| 401 |
|
|
! increase alpha (and Tdeep) in downwelling regions |
| 402 |
|
|
! and decrease in upwelling regions |
| 403 |
|
|
! to avoid "hot spots" where there is surface convergence |
| 404 |
|
✗ |
DO ij=iip2,ip1jm |
| 405 |
|
✗ |
alpha(ij)=alpham-fluxv(ij)/fluxt(ij)*(1.-alpham) |
| 406 |
|
|
ENDDO |
| 407 |
|
✗ |
alpha(1:iip1)=alpham-fluxv(1)/fluxt(1)*(1.-alpham) |
| 408 |
|
✗ |
alpha(ip1jm+1:ip1jmp1)=alpham-fluxv(ip1jmp1)/fluxt(ip1jmp1)*(1.-alpham) |
| 409 |
|
|
ELSE |
| 410 |
|
✗ |
alpha(:)=alpham |
| 411 |
|
|
! Tsurf-Tdeep ~ 10� in the Tropics |
| 412 |
|
|
ENDIF |
| 413 |
|
|
|
| 414 |
|
|
! Estimate deep temperature |
| 415 |
|
✗ |
CALL gr_fi_dyn(1,iip1,jjp1,ts_phy,ts) |
| 416 |
|
✗ |
DO ij=1,ip1jmp1 |
| 417 |
|
✗ |
td(ij)=271.35+(ts(ij)-271.35)*alpha(ij) |
| 418 |
|
✗ |
td(ij)=MIN(td(ij),ts(ij)) |
| 419 |
|
|
END DO |
| 420 |
|
|
|
| 421 |
|
|
! Meridional heat flux: multiply mass flux by (ts-td) |
| 422 |
|
|
! flux in K.kg.s-1 |
| 423 |
|
✗ |
IF (slab_upstream) THEN |
| 424 |
|
|
! upstream scheme to avoid hot spots |
| 425 |
|
✗ |
DO ij=1,ip1jm |
| 426 |
|
✗ |
IF (fluxm(ij).GE.0.) THEN |
| 427 |
|
✗ |
fluxm(ij)=fluxm(ij)*(ts(ij+iip1)-td(ij)) |
| 428 |
|
|
ELSE |
| 429 |
|
✗ |
fluxm(ij)=fluxm(ij)*(ts(ij)-td(ij+iip1)) |
| 430 |
|
|
END IF |
| 431 |
|
|
END DO |
| 432 |
|
|
ELSE |
| 433 |
|
|
! centered scheme better in mid-latitudes |
| 434 |
|
✗ |
DO ij=1,ip1jm |
| 435 |
|
✗ |
fluxm(ij)=fluxm(ij)*(ts(ij+iip1)+ts(ij)-td(ij)-td(ij+iip1))/2. |
| 436 |
|
|
END DO |
| 437 |
|
|
ENDIF |
| 438 |
|
|
|
| 439 |
|
|
! Zonal heat flux |
| 440 |
|
|
! upstream scheme |
| 441 |
|
✗ |
DO ij=iip2,ip1jm |
| 442 |
|
✗ |
fluxz(ij)=fluxz(ij)*(ts(ij)+ts(ij+1)-td(ij+1)-td(ij))/2. |
| 443 |
|
|
END DO |
| 444 |
|
✗ |
DO ij=iip1*2,ip1jmp1,iip1 |
| 445 |
|
✗ |
fluxz(ij)=fluxz(ij-iim) |
| 446 |
|
|
END DO |
| 447 |
|
|
|
| 448 |
|
|
! temperature tendency = divergence of heat fluxes |
| 449 |
|
|
! dt in K.s-1.kg.m-2 (T trend times mass/horiz surface) |
| 450 |
|
✗ |
DO ij=iip2,ip1jm |
| 451 |
|
|
dt(ij)=(fluxz(ij-1)-fluxz(ij)+fluxm(ij)-fluxm(ij-iip1)) & |
| 452 |
|
✗ |
/aire(ij) ! aire : grid area |
| 453 |
|
|
END DO |
| 454 |
|
✗ |
DO ij=iip1,ip1jmi1,iip1 |
| 455 |
|
✗ |
dt(ij+1)=dt(ij+iip1) |
| 456 |
|
|
END DO |
| 457 |
|
|
! special treatment at the Poles |
| 458 |
|
✗ |
dt(1)=SUM(fluxm(1:iim))/apoln |
| 459 |
|
✗ |
dt(ip1jmp1)=-SUM(fluxm(ip1jm-iim:ip1jm-1))/apols |
| 460 |
|
✗ |
dt(2:iip1)=dt(1) |
| 461 |
|
✗ |
dt(ip1jm+1:ip1jmp1)=dt(ip1jmp1) |
| 462 |
|
|
|
| 463 |
|
|
! tendencies back to 1D grid |
| 464 |
|
✗ |
CALL gr_dyn_fi(1,iip1,jjp1,dt,dt_phy) |
| 465 |
|
|
|
| 466 |
|
✗ |
RETURN |
| 467 |
|
|
END SUBROUTINE slab_ekman1 |
| 468 |
|
|
|
| 469 |
|
✗ |
SUBROUTINE slab_ekman2(tx_phy,ty_phy,ts_phy,dt_phy_ek,dt_phy_gm,slab_gm) |
| 470 |
|
|
! Temperature tendency for 2-layers slab ocean |
| 471 |
|
|
! note : tendency dt later multiplied by (delta time)/(rho.H) |
| 472 |
|
|
! to convert from divergence of heat fluxes to T |
| 473 |
|
|
|
| 474 |
|
|
! 11/16 : Inclusion of GM-like eddy advection |
| 475 |
|
|
|
| 476 |
|
|
IMPLICIT NONE |
| 477 |
|
|
|
| 478 |
|
|
LOGICAL,INTENT(in) :: slab_gm |
| 479 |
|
|
! Here, temperature and flux variables are on 2 layers |
| 480 |
|
|
INTEGER ij |
| 481 |
|
|
|
| 482 |
|
|
! wind stress variables |
| 483 |
|
|
REAL tx_phy(klon_glo),ty_phy(klon_glo) |
| 484 |
|
✗ |
REAL txv((nbp_lon+1)*(nbp_lat-1)), tyv((nbp_lon+1)*(nbp_lat-1)) |
| 485 |
|
✗ |
REAL tyu((nbp_lon+1)*nbp_lat),txu((nbp_lon+1)*nbp_lat) |
| 486 |
|
✗ |
REAL tcurl((nbp_lon+1)*(nbp_lat-1)) |
| 487 |
|
|
! slab temperature on 1D, 2D grid |
| 488 |
|
✗ |
REAL ts_phy(klon_glo,2), ts((nbp_lon+1)*nbp_lat,2) |
| 489 |
|
|
! Temperature gradient, v-points |
| 490 |
|
✗ |
REAL dty((nbp_lon+1)*(nbp_lat-1)),dtx((nbp_lon+1)*nbp_lat) |
| 491 |
|
|
! Vertical temperature difference, V-points |
| 492 |
|
✗ |
REAL dtz((nbp_lon+1)*(nbp_lat-1)) |
| 493 |
|
|
! zonal and meridional mass fluxes at u, v points (2D grid) |
| 494 |
|
✗ |
REAL fluxz((nbp_lon+1)*nbp_lat), fluxm((nbp_lon+1)*(nbp_lat-1)) |
| 495 |
|
|
! vertical mass flux between the 2 layers |
| 496 |
|
✗ |
REAL fluxv_ek((nbp_lon+1)*nbp_lat) |
| 497 |
|
✗ |
REAL fluxv_gm((nbp_lon+1)*nbp_lat) |
| 498 |
|
|
! zonal and meridional heat fluxes |
| 499 |
|
✗ |
REAL fluxtz((nbp_lon+1)*nbp_lat,2) |
| 500 |
|
✗ |
REAL fluxtm((nbp_lon+1)*(nbp_lat-1),2) |
| 501 |
|
|
! temperature tendency (in K.s-1.kg.m-2) |
| 502 |
|
✗ |
REAL dt_ek((nbp_lon+1)*nbp_lat,2), dt_phy_ek(klon_glo,2) |
| 503 |
|
✗ |
REAL dt_gm((nbp_lon+1)*nbp_lat,2), dt_phy_gm(klon_glo,2) |
| 504 |
|
|
! helper vars |
| 505 |
|
|
REAL zonavg, fluxv |
| 506 |
|
|
REAL, PARAMETER :: sea_den=1025. ! sea water density |
| 507 |
|
|
|
| 508 |
|
|
INTEGER iim,iip1,iip2,jjp1,ip1jm,ip1jmi1,ip1jmp1 |
| 509 |
|
|
|
| 510 |
|
|
! Grid definitions |
| 511 |
|
|
iim=nbp_lon |
| 512 |
|
✗ |
iip1=nbp_lon+1 |
| 513 |
|
✗ |
iip2=nbp_lon+2 |
| 514 |
|
✗ |
jjp1=nbp_lat |
| 515 |
|
|
ip1jm=(nbp_lon+1)*(nbp_lat-1) ! = iip1*jjm |
| 516 |
|
✗ |
ip1jmi1=(nbp_lon+1)*(nbp_lat-1)-(nbp_lon+1) ! = ip1jm - iip1 |
| 517 |
|
|
ip1jmp1=(nbp_lon+1)*nbp_lat ! = iip1*jjp1 |
| 518 |
|
|
! Convert temperature to 2D grid |
| 519 |
|
✗ |
CALL gr_fi_dyn(2,iip1,jjp1,ts_phy,ts) |
| 520 |
|
|
|
| 521 |
|
|
! ------------------------------------ |
| 522 |
|
|
! Ekman mass fluxes and Temp tendency |
| 523 |
|
|
! ------------------------------------ |
| 524 |
|
|
! Convert taux,y to 2D scalar grid |
| 525 |
|
|
! north and south poles tx,ty no meaning |
| 526 |
|
✗ |
tx_phy(1)=0. |
| 527 |
|
✗ |
tx_phy(klon_glo)=0. |
| 528 |
|
✗ |
ty_phy(1)=0. |
| 529 |
|
✗ |
ty_phy(klon_glo)=0. |
| 530 |
|
✗ |
CALL gr_fi_dyn(1,iip1,jjp1,tx_phy,txu) |
| 531 |
|
✗ |
CALL gr_fi_dyn(1,iip1,jjp1,ty_phy,tyu) |
| 532 |
|
✗ |
IF (ekman_zonavg) THEN ! use zonal average of wind stress |
| 533 |
|
✗ |
DO ij=1,jjp1-2 |
| 534 |
|
✗ |
zonavg=SUM(txu(ij*iip1+1:ij*iip1+iim))/iim |
| 535 |
|
✗ |
txu(ij*iip1+1:(ij+1)*iip1)=zonavg |
| 536 |
|
✗ |
zonavg=SUM(tyu(ij*iip1+1:ij*iip1+iim))/iim |
| 537 |
|
✗ |
tyu(ij*iip1+1:(ij+1)*iip1)=zonavg |
| 538 |
|
|
END DO |
| 539 |
|
|
END IF |
| 540 |
|
|
|
| 541 |
|
|
! Divide taux,y by f or eps, and convert to 2D u,v grids |
| 542 |
|
|
! (Arakawa C grid) |
| 543 |
|
|
! Meridional flux |
| 544 |
|
✗ |
CALL gr_scal_v(1,txu,txv) ! wind stress at v points |
| 545 |
|
✗ |
fluxm=-txv*unsfv ! in kg.s-1.m-1 (zonal distance) |
| 546 |
|
✗ |
IF (slab_sverdrup) THEN ! Sverdrup bal. near equator |
| 547 |
|
✗ |
tcurl=(txu(1:ip1jm)-txu(iip2:ip1jmp1))/cv(:) ! dtx/dy |
| 548 |
|
|
!poles curl = 0 |
| 549 |
|
✗ |
tcurl(1:iip1)=0. |
| 550 |
|
✗ |
tcurl(ip1jmi1+1:ip1jm)=0. |
| 551 |
|
✗ |
fluxm=fluxm-tcurl*unsbv |
| 552 |
|
|
ENDIF |
| 553 |
|
✗ |
IF (slab_tyeq) THEN ! meridional wind forcing at equator |
| 554 |
|
✗ |
CALL gr_scal_v(1,tyu,tyv) |
| 555 |
|
✗ |
fluxm=fluxm+tyv*unsev ! in kg.s-1.m-1 (zonal distance) |
| 556 |
|
|
ENDIF |
| 557 |
|
|
! apply continent mask, multiply by horiz grid dimension |
| 558 |
|
✗ |
fluxm=fluxm*cv*cuvsurcv*zmasqv |
| 559 |
|
|
|
| 560 |
|
|
! Zonal flux |
| 561 |
|
✗ |
IF (ekman_zonadv) THEN |
| 562 |
|
✗ |
CALL gr_scal_u(1,txu,txu) ! wind stress at u points |
| 563 |
|
✗ |
CALL gr_scal_u(1,tyu,tyu) |
| 564 |
|
✗ |
fluxz=tyu*unsfu+txu*unseu |
| 565 |
|
|
! apply continent mask, multiply by horiz grid dimension |
| 566 |
|
✗ |
fluxz=fluxz*cu*cvusurcu*zmasqu |
| 567 |
|
|
END IF |
| 568 |
|
|
|
| 569 |
|
|
! Vertical mass flux from mass budget (divergence of horiz fluxes) |
| 570 |
|
✗ |
IF (ekman_zonadv) THEN |
| 571 |
|
✗ |
DO ij=iip2,ip1jm |
| 572 |
|
✗ |
fluxv_ek(ij)=fluxz(ij)-fluxz(ij-1)-fluxm(ij)+fluxm(ij-iip1) |
| 573 |
|
|
ENDDO |
| 574 |
|
|
ELSE |
| 575 |
|
✗ |
DO ij=iip2,ip1jm |
| 576 |
|
✗ |
fluxv_ek(ij)=-fluxm(ij)+fluxm(ij-iip1) |
| 577 |
|
|
ENDDO |
| 578 |
|
|
END IF |
| 579 |
|
✗ |
DO ij=iip1,ip1jmi1,iip1 |
| 580 |
|
✗ |
fluxv_ek(ij+1)=fluxv_ek(ij+iip1) |
| 581 |
|
|
END DO |
| 582 |
|
|
! vertical mass flux at Poles |
| 583 |
|
✗ |
fluxv_ek(1)=-SUM(fluxm(1:iim)) |
| 584 |
|
✗ |
fluxv_ek(ip1jmp1)=SUM(fluxm(ip1jm-iim:ip1jm-1)) |
| 585 |
|
|
|
| 586 |
|
|
! Meridional heat fluxes |
| 587 |
|
✗ |
DO ij=1,ip1jm |
| 588 |
|
|
! centered scheme |
| 589 |
|
✗ |
fluxtm(ij,1)=fluxm(ij)*(ts(ij+iip1,1)+ts(ij,1))/2. |
| 590 |
|
✗ |
fluxtm(ij,2)=-fluxm(ij)*(ts(ij+iip1,2)+ts(ij,2))/2. |
| 591 |
|
|
END DO |
| 592 |
|
|
|
| 593 |
|
|
! Zonal heat fluxes |
| 594 |
|
|
! Schema upstream |
| 595 |
|
✗ |
IF (ekman_zonadv) THEN |
| 596 |
|
✗ |
DO ij=iip2,ip1jm |
| 597 |
|
✗ |
IF (fluxz(ij).GE.0.) THEN |
| 598 |
|
✗ |
fluxtz(ij,1)=fluxz(ij)*ts(ij,1) |
| 599 |
|
✗ |
fluxtz(ij,2)=-fluxz(ij)*ts(ij+1,2) |
| 600 |
|
|
ELSE |
| 601 |
|
✗ |
fluxtz(ij,1)=fluxz(ij)*ts(ij+1,1) |
| 602 |
|
✗ |
fluxtz(ij,2)=-fluxz(ij)*ts(ij,2) |
| 603 |
|
|
ENDIF |
| 604 |
|
|
ENDDO |
| 605 |
|
✗ |
DO ij=iip1*2,ip1jmp1,iip1 |
| 606 |
|
✗ |
fluxtz(ij,:)=fluxtz(ij-iim,:) |
| 607 |
|
|
END DO |
| 608 |
|
|
ELSE |
| 609 |
|
✗ |
fluxtz(:,:)=0. |
| 610 |
|
|
ENDIF |
| 611 |
|
|
|
| 612 |
|
|
! Temperature tendency, horizontal advection: |
| 613 |
|
✗ |
DO ij=iip2,ip1jm |
| 614 |
|
|
dt_ek(ij,:)=fluxtz(ij-1,:)-fluxtz(ij,:) & |
| 615 |
|
✗ |
+fluxtm(ij,:)-fluxtm(ij-iip1,:) |
| 616 |
|
|
END DO |
| 617 |
|
|
! Poles |
| 618 |
|
✗ |
dt_ek(1,:)=SUM(fluxtm(1:iim,:),dim=1) |
| 619 |
|
✗ |
dt_ek(ip1jmp1,:)=-SUM(fluxtm(ip1jm-iim:ip1jm-1,:),dim=1) |
| 620 |
|
|
|
| 621 |
|
|
! ------------------------------------ |
| 622 |
|
|
! GM mass fluxes and Temp tendency |
| 623 |
|
|
! ------------------------------------ |
| 624 |
|
✗ |
IF (slab_gm) THEN |
| 625 |
|
|
! Vertical Temperature difference T1-T2 on v-grid points |
| 626 |
|
✗ |
CALL gr_scal_v(1,ts(:,1)-ts(:,2),dtz) |
| 627 |
|
✗ |
dtz(:)=MAX(dtz(:),0.25) |
| 628 |
|
|
! Horizontal Temperature differences |
| 629 |
|
✗ |
CALL grad(1,(ts(:,1)+ts(:,2))/2.,dtx,dty) |
| 630 |
|
|
! Meridional flux = -k.s (s=dyT/dzT) |
| 631 |
|
|
! Continent mask, multiply by dz/dy |
| 632 |
|
✗ |
fluxm=dty/dtz*500.*cuvsurcv*zmasqv |
| 633 |
|
|
! slope limitation, multiply by kappa |
| 634 |
|
✗ |
fluxm=-gmkappa*SIGN(MIN(ABS(fluxm),gm_smax*cv*cuvsurcv),dty) |
| 635 |
|
|
! convert to kg/s |
| 636 |
|
✗ |
fluxm(:)=fluxm(:)*sea_den |
| 637 |
|
|
! Zonal flux = 0. (temporary) |
| 638 |
|
✗ |
fluxz(:)=0. |
| 639 |
|
|
! Vertical mass flux from mass budget (divergence of horiz fluxes) |
| 640 |
|
✗ |
DO ij=iip2,ip1jm |
| 641 |
|
✗ |
fluxv_gm(ij)=fluxz(ij)-fluxz(ij-1)-fluxm(ij)+fluxm(ij-iip1) |
| 642 |
|
|
ENDDO |
| 643 |
|
✗ |
DO ij=iip1,ip1jmi1,iip1 |
| 644 |
|
✗ |
fluxv_gm(ij+1)=fluxv_gm(ij+iip1) |
| 645 |
|
|
END DO |
| 646 |
|
|
! vertical mass flux at Poles |
| 647 |
|
✗ |
fluxv_gm(1)=-SUM(fluxm(1:iim)) |
| 648 |
|
✗ |
fluxv_gm(ip1jmp1)=SUM(fluxm(ip1jm-iim:ip1jm-1)) |
| 649 |
|
|
|
| 650 |
|
|
! Meridional heat fluxes |
| 651 |
|
✗ |
DO ij=1,ip1jm |
| 652 |
|
|
! centered scheme |
| 653 |
|
✗ |
fluxtm(ij,1)=fluxm(ij)*(ts(ij+iip1,1)+ts(ij,1))/2. |
| 654 |
|
✗ |
fluxtm(ij,2)=-fluxm(ij)*(ts(ij+iip1,2)+ts(ij,2))/2. |
| 655 |
|
|
END DO |
| 656 |
|
|
|
| 657 |
|
|
! Zonal heat fluxes |
| 658 |
|
|
! Schema upstream |
| 659 |
|
✗ |
DO ij=iip2,ip1jm |
| 660 |
|
✗ |
IF (fluxz(ij).GE.0.) THEN |
| 661 |
|
✗ |
fluxtz(ij,1)=fluxz(ij)*ts(ij,1) |
| 662 |
|
✗ |
fluxtz(ij,2)=-fluxz(ij)*ts(ij+1,2) |
| 663 |
|
|
ELSE |
| 664 |
|
✗ |
fluxtz(ij,1)=fluxz(ij)*ts(ij+1,1) |
| 665 |
|
✗ |
fluxtz(ij,2)=-fluxz(ij)*ts(ij,2) |
| 666 |
|
|
ENDIF |
| 667 |
|
|
ENDDO |
| 668 |
|
✗ |
DO ij=iip1*2,ip1jmp1,iip1 |
| 669 |
|
✗ |
fluxtz(ij,:)=fluxtz(ij-iim,:) |
| 670 |
|
|
END DO |
| 671 |
|
|
|
| 672 |
|
|
! Temperature tendency : |
| 673 |
|
|
! divergence of horizontal heat fluxes |
| 674 |
|
✗ |
DO ij=iip2,ip1jm |
| 675 |
|
|
dt_gm(ij,:)=fluxtz(ij-1,:)-fluxtz(ij,:) & |
| 676 |
|
✗ |
+fluxtm(ij,:)-fluxtm(ij-iip1,:) |
| 677 |
|
|
END DO |
| 678 |
|
|
! Poles |
| 679 |
|
✗ |
dt_gm(1,:)=SUM(fluxtm(1:iim,:),dim=1) |
| 680 |
|
✗ |
dt_gm(ip1jmp1,:)=-SUM(fluxtm(ip1jm-iim:ip1jm-1,:),dim=1) |
| 681 |
|
|
ELSE |
| 682 |
|
✗ |
dt_gm(:,:)=0. |
| 683 |
|
✗ |
fluxv_gm(:)=0. |
| 684 |
|
|
ENDIF ! slab_gm |
| 685 |
|
|
|
| 686 |
|
|
! ------------------------------------ |
| 687 |
|
|
! Temp tendency from vertical advection |
| 688 |
|
|
! Divide by cell area |
| 689 |
|
|
! ------------------------------------ |
| 690 |
|
|
! vertical heat flux = mass flux * T, upstream scheme |
| 691 |
|
✗ |
DO ij=iip2,ip1jm |
| 692 |
|
✗ |
fluxv=fluxv_ek(ij)+fluxv_gm(ij) ! net flux, needed for upstream scheme |
| 693 |
|
✗ |
IF (fluxv.GT.0.) THEN |
| 694 |
|
✗ |
dt_ek(ij,1)=dt_ek(ij,1)+fluxv_ek(ij)*ts(ij,2) |
| 695 |
|
✗ |
dt_ek(ij,2)=dt_ek(ij,2)-fluxv_ek(ij)*ts(ij,2) |
| 696 |
|
✗ |
dt_gm(ij,1)=dt_gm(ij,1)+fluxv_gm(ij)*ts(ij,2) |
| 697 |
|
✗ |
dt_gm(ij,2)=dt_gm(ij,2)-fluxv_gm(ij)*ts(ij,2) |
| 698 |
|
|
ELSE |
| 699 |
|
✗ |
dt_ek(ij,1)=dt_ek(ij,1)+fluxv_ek(ij)*ts(ij,1) |
| 700 |
|
✗ |
dt_ek(ij,2)=dt_ek(ij,2)-fluxv_ek(ij)*ts(ij,1) |
| 701 |
|
✗ |
dt_gm(ij,1)=dt_gm(ij,1)+fluxv_gm(ij)*ts(ij,1) |
| 702 |
|
✗ |
dt_gm(ij,2)=dt_gm(ij,2)-fluxv_gm(ij)*ts(ij,1) |
| 703 |
|
|
ENDIF |
| 704 |
|
|
! divide by cell area |
| 705 |
|
✗ |
dt_ek(ij,:)=dt_ek(ij,:)/aire(ij) |
| 706 |
|
✗ |
dt_gm(ij,:)=dt_gm(ij,:)/aire(ij) |
| 707 |
|
|
END DO |
| 708 |
|
|
! North Pole |
| 709 |
|
✗ |
fluxv=fluxv_ek(1)+fluxv_gm(1) |
| 710 |
|
✗ |
IF (fluxv.GT.0.) THEN |
| 711 |
|
✗ |
dt_ek(1,1)=dt_ek(1,1)+fluxv_ek(1)*ts(1,2) |
| 712 |
|
✗ |
dt_ek(1,2)=dt_ek(1,2)-fluxv_ek(1)*ts(1,2) |
| 713 |
|
✗ |
dt_gm(1,1)=dt_gm(1,1)+fluxv_gm(1)*ts(1,2) |
| 714 |
|
✗ |
dt_gm(1,2)=dt_gm(1,2)-fluxv_gm(1)*ts(1,2) |
| 715 |
|
|
ELSE |
| 716 |
|
✗ |
dt_ek(1,1)=dt_ek(1,1)+fluxv_ek(1)*ts(1,1) |
| 717 |
|
✗ |
dt_ek(1,2)=dt_ek(1,2)-fluxv_ek(1)*ts(1,1) |
| 718 |
|
✗ |
dt_gm(1,1)=dt_gm(1,1)+fluxv_gm(1)*ts(1,1) |
| 719 |
|
✗ |
dt_gm(1,2)=dt_gm(1,2)-fluxv_gm(1)*ts(1,1) |
| 720 |
|
|
ENDIF |
| 721 |
|
✗ |
dt_ek(1,:)=dt_ek(1,:)/apoln |
| 722 |
|
✗ |
dt_gm(1,:)=dt_gm(1,:)/apoln |
| 723 |
|
|
! South pole |
| 724 |
|
✗ |
fluxv=fluxv_ek(ip1jmp1)+fluxv_gm(ip1jmp1) |
| 725 |
|
✗ |
IF (fluxv.GT.0.) THEN |
| 726 |
|
✗ |
dt_ek(ip1jmp1,1)=dt_ek(ip1jmp1,1)+fluxv_ek(ip1jmp1)*ts(ip1jmp1,2) |
| 727 |
|
✗ |
dt_ek(ip1jmp1,2)=dt_ek(ip1jmp1,2)-fluxv_ek(ip1jmp1)*ts(ip1jmp1,2) |
| 728 |
|
✗ |
dt_gm(ip1jmp1,1)=dt_gm(ip1jmp1,1)+fluxv_gm(ip1jmp1)*ts(ip1jmp1,2) |
| 729 |
|
✗ |
dt_gm(ip1jmp1,2)=dt_gm(ip1jmp1,2)-fluxv_gm(ip1jmp1)*ts(ip1jmp1,2) |
| 730 |
|
|
ELSE |
| 731 |
|
✗ |
dt_ek(ip1jmp1,1)=dt_ek(ip1jmp1,1)+fluxv_ek(ip1jmp1)*ts(ip1jmp1,1) |
| 732 |
|
✗ |
dt_ek(ip1jmp1,2)=dt_ek(ip1jmp1,2)-fluxv_ek(ip1jmp1)*ts(ip1jmp1,1) |
| 733 |
|
✗ |
dt_gm(ip1jmp1,1)=dt_gm(ip1jmp1,1)+fluxv_gm(ip1jmp1)*ts(ip1jmp1,1) |
| 734 |
|
✗ |
dt_gm(ip1jmp1,2)=dt_gm(ip1jmp1,2)-fluxv_gm(ip1jmp1)*ts(ip1jmp1,1) |
| 735 |
|
|
ENDIF |
| 736 |
|
✗ |
dt_ek(ip1jmp1,:)=dt_ek(ip1jmp1,:)/apols |
| 737 |
|
✗ |
dt_gm(ip1jmp1,:)=dt_gm(ip1jmp1,:)/apols |
| 738 |
|
|
|
| 739 |
|
✗ |
dt_ek(2:iip1,1)=dt_ek(1,1) |
| 740 |
|
✗ |
dt_ek(2:iip1,2)=dt_ek(1,2) |
| 741 |
|
✗ |
dt_gm(2:iip1,1)=dt_gm(1,1) |
| 742 |
|
✗ |
dt_gm(2:iip1,2)=dt_gm(1,2) |
| 743 |
|
✗ |
dt_ek(ip1jm+1:ip1jmp1,1)=dt_ek(ip1jmp1,1) |
| 744 |
|
✗ |
dt_ek(ip1jm+1:ip1jmp1,2)=dt_ek(ip1jmp1,2) |
| 745 |
|
✗ |
dt_gm(ip1jm+1:ip1jmp1,1)=dt_gm(ip1jmp1,1) |
| 746 |
|
✗ |
dt_gm(ip1jm+1:ip1jmp1,2)=dt_gm(ip1jmp1,2) |
| 747 |
|
|
|
| 748 |
|
✗ |
DO ij=iip1,ip1jmi1,iip1 |
| 749 |
|
✗ |
dt_gm(ij+1,:)=dt_gm(ij+iip1,:) |
| 750 |
|
✗ |
dt_ek(ij+1,:)=dt_ek(ij+iip1,:) |
| 751 |
|
|
END DO |
| 752 |
|
|
|
| 753 |
|
|
! T tendency back to 1D grid... |
| 754 |
|
✗ |
CALL gr_dyn_fi(2,iip1,jjp1,dt_ek,dt_phy_ek) |
| 755 |
|
✗ |
CALL gr_dyn_fi(2,iip1,jjp1,dt_gm,dt_phy_gm) |
| 756 |
|
|
|
| 757 |
|
✗ |
RETURN |
| 758 |
|
|
END SUBROUTINE slab_ekman2 |
| 759 |
|
|
|
| 760 |
|
✗ |
SUBROUTINE slab_gmdiff(ts_phy,dt_phy) |
| 761 |
|
|
! Temperature tendency for 2-layers slab ocean |
| 762 |
|
|
! Due to Gent-McWilliams type eddy-induced advection |
| 763 |
|
|
|
| 764 |
|
|
IMPLICIT NONE |
| 765 |
|
|
|
| 766 |
|
|
! Here, temperature and flux variables are on 2 layers |
| 767 |
|
|
INTEGER ij |
| 768 |
|
|
! Temperature gradient, v-points |
| 769 |
|
✗ |
REAL dty((nbp_lon+1)*(nbp_lat-1)),dtx((nbp_lon+1)*nbp_lat) |
| 770 |
|
|
! Vertical temperature difference, V-points |
| 771 |
|
✗ |
REAL dtz((nbp_lon+1)*(nbp_lat-1)) |
| 772 |
|
|
! slab temperature on 1D, 2D grid |
| 773 |
|
✗ |
REAL ts_phy(klon_glo,2),ts((nbp_lon+1)*nbp_lat,2) |
| 774 |
|
|
! zonal and meridional mass fluxes at u, v points (2D grid) |
| 775 |
|
✗ |
REAL fluxz((nbp_lon+1)*nbp_lat), fluxm((nbp_lon+1)*(nbp_lat-1)) |
| 776 |
|
|
! vertical mass flux between the 2 layers |
| 777 |
|
✗ |
REAL fluxv((nbp_lon+1)*nbp_lat) |
| 778 |
|
|
! zonal and meridional heat fluxes |
| 779 |
|
✗ |
REAL fluxtz((nbp_lon+1)*nbp_lat,2) |
| 780 |
|
✗ |
REAL fluxtm((nbp_lon+1)*(nbp_lat-1),2) |
| 781 |
|
|
! temperature tendency (in K.s-1.kg.m-2) |
| 782 |
|
✗ |
REAL dt((nbp_lon+1)*nbp_lat,2), dt_phy(klon_glo,2) |
| 783 |
|
|
|
| 784 |
|
|
INTEGER iim,iip1,iip2,jjp1,ip1jm,ip1jmi1,ip1jmp1 |
| 785 |
|
|
|
| 786 |
|
|
! Grid definitions |
| 787 |
|
|
iim=nbp_lon |
| 788 |
|
✗ |
iip1=nbp_lon+1 |
| 789 |
|
✗ |
iip2=nbp_lon+2 |
| 790 |
|
✗ |
jjp1=nbp_lat |
| 791 |
|
|
ip1jm=(nbp_lon+1)*(nbp_lat-1) ! = iip1*jjm |
| 792 |
|
✗ |
ip1jmi1=(nbp_lon+1)*(nbp_lat-1)-(nbp_lon+1) ! = ip1jm - iip1 |
| 793 |
|
|
ip1jmp1=(nbp_lon+1)*nbp_lat ! = iip1*jjp1 |
| 794 |
|
|
|
| 795 |
|
|
! Convert temperature to 2D grid |
| 796 |
|
✗ |
CALL gr_fi_dyn(2,iip1,jjp1,ts_phy,ts) |
| 797 |
|
|
! Vertical Temperature difference T1-T2 on v-grid points |
| 798 |
|
✗ |
CALL gr_scal_v(1,ts(:,1)-ts(:,2),dtz) |
| 799 |
|
✗ |
dtz(:)=MAX(dtz(:),0.25) |
| 800 |
|
|
! Horizontal Temperature differences |
| 801 |
|
✗ |
CALL grad(1,(ts(:,1)+ts(:,2))/2.,dtx,dty) |
| 802 |
|
|
! Meridional flux = -k.s (s=dyT/dzT) |
| 803 |
|
|
! Continent mask, multiply by dz/dy |
| 804 |
|
✗ |
fluxm=dty/dtz*500.*cuvsurcv*zmasqv |
| 805 |
|
|
! slope limitation, multiply by kappa |
| 806 |
|
✗ |
fluxm=-gmkappa*SIGN(MIN(ABS(fluxm),gm_smax*cv*cuvsurcv),dty) |
| 807 |
|
|
! Zonal flux = 0. (temporary) |
| 808 |
|
✗ |
fluxz(:)=0. |
| 809 |
|
|
! Vertical mass flux from mass budget (divergence of horiz fluxes) |
| 810 |
|
✗ |
DO ij=iip2,ip1jm |
| 811 |
|
✗ |
fluxv(ij)=fluxz(ij)-fluxz(ij-1)-fluxm(ij)+fluxm(ij-iip1) |
| 812 |
|
|
ENDDO |
| 813 |
|
✗ |
DO ij=iip1,ip1jmi1,iip1 |
| 814 |
|
✗ |
fluxv(ij+1)=fluxv(ij+iip1) |
| 815 |
|
|
END DO |
| 816 |
|
|
! vertical mass flux at Poles |
| 817 |
|
✗ |
fluxv(1)=-SUM(fluxm(1:iim)) |
| 818 |
|
✗ |
fluxv(ip1jmp1)=SUM(fluxm(ip1jm-iim:ip1jm-1)) |
| 819 |
|
✗ |
fluxv=fluxv |
| 820 |
|
|
|
| 821 |
|
|
! Meridional heat fluxes |
| 822 |
|
✗ |
DO ij=1,ip1jm |
| 823 |
|
|
! centered scheme |
| 824 |
|
✗ |
fluxtm(ij,1)=fluxm(ij)*(ts(ij+iip1,1)+ts(ij,1))/2. |
| 825 |
|
✗ |
fluxtm(ij,2)=-fluxm(ij)*(ts(ij+iip1,2)+ts(ij,2))/2. |
| 826 |
|
|
END DO |
| 827 |
|
|
|
| 828 |
|
|
! Zonal heat fluxes |
| 829 |
|
|
! Schema upstream |
| 830 |
|
✗ |
DO ij=iip2,ip1jm |
| 831 |
|
✗ |
IF (fluxz(ij).GE.0.) THEN |
| 832 |
|
✗ |
fluxtz(ij,1)=fluxz(ij)*ts(ij,1) |
| 833 |
|
✗ |
fluxtz(ij,2)=-fluxz(ij)*ts(ij+1,2) |
| 834 |
|
|
ELSE |
| 835 |
|
✗ |
fluxtz(ij,1)=fluxz(ij)*ts(ij+1,1) |
| 836 |
|
✗ |
fluxtz(ij,2)=-fluxz(ij)*ts(ij,2) |
| 837 |
|
|
ENDIF |
| 838 |
|
|
ENDDO |
| 839 |
|
✗ |
DO ij=iip1*2,ip1jmp1,iip1 |
| 840 |
|
✗ |
fluxtz(ij,:)=fluxtz(ij-iim,:) |
| 841 |
|
|
END DO |
| 842 |
|
|
|
| 843 |
|
|
! Temperature tendency : |
| 844 |
|
✗ |
DO ij=iip2,ip1jm |
| 845 |
|
|
! divergence of horizontal heat fluxes |
| 846 |
|
|
dt(ij,:)=fluxtz(ij-1,:)-fluxtz(ij,:) & |
| 847 |
|
✗ |
+fluxtm(ij,:)-fluxtm(ij-iip1,:) |
| 848 |
|
|
! + vertical heat flux (mass flux * T, upstream scheme) |
| 849 |
|
✗ |
IF (fluxv(ij).GT.0.) THEN |
| 850 |
|
✗ |
dt(ij,1)=dt(ij,1)+fluxv(ij)*ts(ij,2) |
| 851 |
|
✗ |
dt(ij,2)=dt(ij,2)-fluxv(ij)*ts(ij,2) |
| 852 |
|
|
ELSE |
| 853 |
|
✗ |
dt(ij,1)=dt(ij,1)+fluxv(ij)*ts(ij,1) |
| 854 |
|
✗ |
dt(ij,2)=dt(ij,2)-fluxv(ij)*ts(ij,1) |
| 855 |
|
|
ENDIF |
| 856 |
|
|
! divide by cell area |
| 857 |
|
✗ |
dt(ij,:)=dt(ij,:)/aire(ij) |
| 858 |
|
|
END DO |
| 859 |
|
✗ |
DO ij=iip1,ip1jmi1,iip1 |
| 860 |
|
✗ |
dt(ij+1,:)=dt(ij+iip1,:) |
| 861 |
|
|
END DO |
| 862 |
|
|
! Poles |
| 863 |
|
✗ |
dt(1,:)=SUM(fluxtm(1:iim,:),dim=1) |
| 864 |
|
✗ |
IF (fluxv(1).GT.0.) THEN |
| 865 |
|
✗ |
dt(1,1)=dt(1,1)+fluxv(1)*ts(1,2) |
| 866 |
|
✗ |
dt(1,2)=dt(1,2)-fluxv(1)*ts(1,2) |
| 867 |
|
|
ELSE |
| 868 |
|
✗ |
dt(1,1)=dt(1,1)+fluxv(1)*ts(1,1) |
| 869 |
|
✗ |
dt(1,2)=dt(1,2)-fluxv(1)*ts(1,1) |
| 870 |
|
|
ENDIF |
| 871 |
|
✗ |
dt(1,:)=dt(1,:)/apoln |
| 872 |
|
✗ |
dt(ip1jmp1,:)=-SUM(fluxtm(ip1jm-iim:ip1jm-1,:),dim=1) |
| 873 |
|
✗ |
IF (fluxv(ip1jmp1).GT.0.) THEN |
| 874 |
|
✗ |
dt(ip1jmp1,1)=dt(ip1jmp1,1)+fluxv(ip1jmp1)*ts(ip1jmp1,2) |
| 875 |
|
✗ |
dt(ip1jmp1,2)=dt(ip1jmp1,2)-fluxv(ip1jmp1)*ts(ip1jmp1,2) |
| 876 |
|
|
ELSE |
| 877 |
|
✗ |
dt(ip1jmp1,1)=dt(ip1jmp1,1)+fluxv(ip1jmp1)*ts(ip1jmp1,1) |
| 878 |
|
✗ |
dt(ip1jmp1,2)=dt(ip1jmp1,2)-fluxv(ip1jmp1)*ts(ip1jmp1,1) |
| 879 |
|
|
ENDIF |
| 880 |
|
✗ |
dt(ip1jmp1,:)=dt(ip1jmp1,:)/apols |
| 881 |
|
✗ |
dt(2:iip1,1)=dt(1,1) |
| 882 |
|
✗ |
dt(2:iip1,2)=dt(1,2) |
| 883 |
|
✗ |
dt(ip1jm+1:ip1jmp1,1)=dt(ip1jmp1,1) |
| 884 |
|
✗ |
dt(ip1jm+1:ip1jmp1,2)=dt(ip1jmp1,2) |
| 885 |
|
|
|
| 886 |
|
|
! T tendency back to 1D grid... |
| 887 |
|
✗ |
CALL gr_dyn_fi(2,iip1,jjp1,dt,dt_phy) |
| 888 |
|
|
|
| 889 |
|
✗ |
RETURN |
| 890 |
|
|
END SUBROUTINE slab_gmdiff |
| 891 |
|
|
|
| 892 |
|
|
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! |
| 893 |
|
|
|
| 894 |
|
✗ |
SUBROUTINE gr_fi_dyn(nfield,im,jm,pfi,pdyn) |
| 895 |
|
|
! Transfer a variable from 1D "physics" grid to 2D "dynamics" grid |
| 896 |
|
|
IMPLICIT NONE |
| 897 |
|
|
|
| 898 |
|
|
INTEGER,INTENT(IN) :: im,jm,nfield |
| 899 |
|
|
REAL,INTENT(IN) :: pfi(klon_glo,nfield) ! on 1D grid |
| 900 |
|
|
REAL,INTENT(OUT) :: pdyn(im,jm,nfield) ! on 2D grid |
| 901 |
|
|
|
| 902 |
|
|
INTEGER :: i,j,ifield,ig |
| 903 |
|
|
|
| 904 |
|
✗ |
DO ifield=1,nfield |
| 905 |
|
|
! Handle poles |
| 906 |
|
✗ |
DO i=1,im |
| 907 |
|
✗ |
pdyn(i,1,ifield)=pfi(1,ifield) |
| 908 |
|
✗ |
pdyn(i,jm,ifield)=pfi(klon_glo,ifield) |
| 909 |
|
|
ENDDO |
| 910 |
|
|
! Other points |
| 911 |
|
✗ |
DO j=2,jm-1 |
| 912 |
|
✗ |
ig=2+(j-2)*(im-1) |
| 913 |
|
✗ |
CALL SCOPY(im-1,pfi(ig,ifield),1,pdyn(1,j,ifield),1) |
| 914 |
|
✗ |
pdyn(im,j,ifield)=pdyn(1,j,ifield) |
| 915 |
|
|
ENDDO |
| 916 |
|
|
ENDDO ! of DO ifield=1,nfield |
| 917 |
|
|
|
| 918 |
|
✗ |
END SUBROUTINE gr_fi_dyn |
| 919 |
|
|
|
| 920 |
|
✗ |
SUBROUTINE gr_dyn_fi(nfield,im,jm,pdyn,pfi) |
| 921 |
|
|
! Transfer a variable from 2D "dynamics" grid to 1D "physics" grid |
| 922 |
|
|
IMPLICIT NONE |
| 923 |
|
|
|
| 924 |
|
|
INTEGER,INTENT(IN) :: im,jm,nfield |
| 925 |
|
|
REAL,INTENT(IN) :: pdyn(im,jm,nfield) ! on 2D grid |
| 926 |
|
|
REAL,INTENT(OUT) :: pfi(klon_glo,nfield) ! on 1D grid |
| 927 |
|
|
|
| 928 |
|
|
INTEGER j,ifield,ig |
| 929 |
|
|
|
| 930 |
|
|
CHARACTER (len = 20) :: modname = 'slab_heat_transp' |
| 931 |
|
|
CHARACTER (len = 80) :: abort_message |
| 932 |
|
|
|
| 933 |
|
|
! Sanity check: |
| 934 |
|
✗ |
IF(klon_glo.NE.2+(jm-2)*(im-1)) THEN |
| 935 |
|
✗ |
abort_message="gr_dyn_fi error, wrong sizes" |
| 936 |
|
✗ |
CALL abort_physic(modname,abort_message,1) |
| 937 |
|
|
ENDIF |
| 938 |
|
|
|
| 939 |
|
|
! Handle poles |
| 940 |
|
✗ |
CALL SCOPY(nfield,pdyn,im*jm,pfi,klon_glo) |
| 941 |
|
✗ |
CALL SCOPY(nfield,pdyn(1,jm,1),im*jm,pfi(klon_glo,1),klon_glo) |
| 942 |
|
|
! Other points |
| 943 |
|
✗ |
DO ifield=1,nfield |
| 944 |
|
✗ |
DO j=2,jm-1 |
| 945 |
|
✗ |
ig=2+(j-2)*(im-1) |
| 946 |
|
✗ |
CALL SCOPY(im-1,pdyn(1,j,ifield),1,pfi(ig,ifield),1) |
| 947 |
|
|
ENDDO |
| 948 |
|
|
ENDDO |
| 949 |
|
|
|
| 950 |
|
✗ |
END SUBROUTINE gr_dyn_fi |
| 951 |
|
|
|
| 952 |
|
✗ |
SUBROUTINE grad(klevel,pg,pgx,pgy) |
| 953 |
|
|
! compute the covariant components pgx,pgy of the gradient of pg |
| 954 |
|
|
! pgx = d(pg)/dx * delta(x) = delta(pg) |
| 955 |
|
|
IMPLICIT NONE |
| 956 |
|
|
|
| 957 |
|
|
INTEGER,INTENT(IN) :: klevel |
| 958 |
|
|
REAL,INTENT(IN) :: pg((nbp_lon+1)*nbp_lat,klevel) |
| 959 |
|
|
REAL,INTENT(OUT) :: pgx((nbp_lon+1)*nbp_lat,klevel) |
| 960 |
|
|
REAL,INTENT(OUT) :: pgy((nbp_lon+1)*(nbp_lat-1),klevel) |
| 961 |
|
|
|
| 962 |
|
|
INTEGER :: l,ij |
| 963 |
|
|
INTEGER :: iim,iip1,ip1jm,ip1jmp1 |
| 964 |
|
|
|
| 965 |
|
|
iim=nbp_lon |
| 966 |
|
|
iip1=nbp_lon+1 |
| 967 |
|
|
ip1jm=(nbp_lon+1)*(nbp_lat-1) ! = iip1*jjm |
| 968 |
|
|
ip1jmp1=(nbp_lon+1)*nbp_lat ! = iip1*jjp1 |
| 969 |
|
|
|
| 970 |
|
✗ |
DO l=1,klevel |
| 971 |
|
✗ |
DO ij=1,ip1jmp1-1 |
| 972 |
|
✗ |
pgx(ij,l)=pg(ij+1,l)-pg(ij,l) |
| 973 |
|
|
ENDDO |
| 974 |
|
|
! correction for pgx(ip1,j,l) ... |
| 975 |
|
|
! ... pgx(iip1,j,l)=pgx(1,j,l) ... |
| 976 |
|
✗ |
DO ij=iip1,ip1jmp1,iip1 |
| 977 |
|
✗ |
pgx(ij,l)=pgx(ij-iim,l) |
| 978 |
|
|
ENDDO |
| 979 |
|
✗ |
DO ij=1,ip1jm |
| 980 |
|
✗ |
pgy(ij,l)=pg(ij,l)-pg(ij+iip1,l) |
| 981 |
|
|
ENDDO |
| 982 |
|
|
ENDDO |
| 983 |
|
|
|
| 984 |
|
✗ |
END SUBROUTINE grad |
| 985 |
|
|
|
| 986 |
|
✗ |
SUBROUTINE diverg(klevel,x,y,div) |
| 987 |
|
|
! computes the divergence of a vector field of components |
| 988 |
|
|
! x,y. x and y being covariant components |
| 989 |
|
|
IMPLICIT NONE |
| 990 |
|
|
|
| 991 |
|
|
INTEGER,INTENT(IN) :: klevel |
| 992 |
|
|
REAL,INTENT(IN) :: x((nbp_lon+1)*nbp_lat,klevel) |
| 993 |
|
|
REAL,INTENT(IN) :: y((nbp_lon+1)*(nbp_lat-1),klevel) |
| 994 |
|
|
REAL,INTENT(OUT) :: div((nbp_lon+1)*nbp_lat,klevel) |
| 995 |
|
|
|
| 996 |
|
|
INTEGER :: l,ij |
| 997 |
|
|
INTEGER :: iim,iip1,iip2,ip1jm,ip1jmp1,ip1jmi1 |
| 998 |
|
|
|
| 999 |
|
✗ |
REAL :: aiy1(nbp_lon+1),aiy2(nbp_lon+1) |
| 1000 |
|
|
REAL :: sumypn,sumyps |
| 1001 |
|
|
REAL,EXTERNAL :: SSUM |
| 1002 |
|
|
|
| 1003 |
|
✗ |
iim=nbp_lon |
| 1004 |
|
|
iip1=nbp_lon+1 |
| 1005 |
|
✗ |
iip2=nbp_lon+2 |
| 1006 |
|
|
ip1jm=(nbp_lon+1)*(nbp_lat-1) ! = iip1*jjm |
| 1007 |
|
|
ip1jmp1=(nbp_lon+1)*nbp_lat ! = iip1*jjp1 |
| 1008 |
|
✗ |
ip1jmi1=(nbp_lon+1)*(nbp_lat-1)-(nbp_lon+1) ! = ip1jm - iip1 |
| 1009 |
|
|
|
| 1010 |
|
✗ |
DO l=1,klevel |
| 1011 |
|
✗ |
DO ij=iip2,ip1jm-1 |
| 1012 |
|
|
div(ij+1,l)= & |
| 1013 |
|
|
cvusurcu(ij+1)*x(ij+1,l)-cvusurcu(ij)*x(ij,l)+ & |
| 1014 |
|
✗ |
cuvsurcv(ij-iim)*y(ij-iim,l)-cuvsurcv(ij+1)*y(ij+1,l) |
| 1015 |
|
|
ENDDO |
| 1016 |
|
|
! correction for div(1,j,l) ... |
| 1017 |
|
|
! ... div(1,j,l)= div(iip1,j,l) ... |
| 1018 |
|
✗ |
DO ij=iip2,ip1jm,iip1 |
| 1019 |
|
✗ |
div(ij,l)=div(ij+iim,l) |
| 1020 |
|
|
ENDDO |
| 1021 |
|
|
! at the poles |
| 1022 |
|
✗ |
DO ij=1,iim |
| 1023 |
|
✗ |
aiy1(ij)=cuvsurcv(ij)*y(ij,l) |
| 1024 |
|
✗ |
aiy2(ij)=cuvsurcv(ij+ip1jmi1)*y(ij+ip1jmi1,l) |
| 1025 |
|
|
ENDDO |
| 1026 |
|
✗ |
sumypn=SSUM(iim,aiy1,1)/apoln |
| 1027 |
|
✗ |
sumyps=SSUM(iim,aiy2,1)/apols |
| 1028 |
|
✗ |
DO ij=1,iip1 |
| 1029 |
|
✗ |
div(ij,l)=-sumypn |
| 1030 |
|
✗ |
div(ij+ip1jm,l)=sumyps |
| 1031 |
|
|
ENDDO |
| 1032 |
|
|
! End (poles) |
| 1033 |
|
|
ENDDO ! of DO l=1,klevel |
| 1034 |
|
|
|
| 1035 |
|
|
!!! CALL filtreg( div, jjp1, klevel, 2, 2, .TRUE., 1 ) |
| 1036 |
|
✗ |
DO l=1,klevel |
| 1037 |
|
✗ |
DO ij=iip2,ip1jm |
| 1038 |
|
✗ |
div(ij,l)=div(ij,l)*unsaire(ij) |
| 1039 |
|
|
ENDDO |
| 1040 |
|
|
ENDDO |
| 1041 |
|
|
|
| 1042 |
|
✗ |
END SUBROUTINE diverg |
| 1043 |
|
|
|
| 1044 |
|
✗ |
SUBROUTINE gr_v_scal(nx,x_v,x_scal) |
| 1045 |
|
|
! convert values from v points to scalar points on C-grid |
| 1046 |
|
|
! used to compute unsfu, unseu (u points, but depends only on latitude) |
| 1047 |
|
|
IMPLICIT NONE |
| 1048 |
|
|
|
| 1049 |
|
|
INTEGER,INTENT(IN) :: nx ! number of levels or fields |
| 1050 |
|
|
REAL,INTENT(IN) :: x_v((nbp_lon+1)*(nbp_lat-1),nx) |
| 1051 |
|
|
REAL,INTENT(OUT) :: x_scal((nbp_lon+1)*nbp_lat,nx) |
| 1052 |
|
|
|
| 1053 |
|
|
INTEGER :: l,ij |
| 1054 |
|
|
INTEGER :: iip1,iip2,ip1jm,ip1jmp1 |
| 1055 |
|
|
|
| 1056 |
|
|
iip1=nbp_lon+1 |
| 1057 |
|
✗ |
iip2=nbp_lon+2 |
| 1058 |
|
|
ip1jm=(nbp_lon+1)*(nbp_lat-1) ! = iip1*jjm |
| 1059 |
|
|
ip1jmp1=(nbp_lon+1)*nbp_lat ! = iip1*jjp1 |
| 1060 |
|
|
|
| 1061 |
|
✗ |
DO l=1,nx |
| 1062 |
|
✗ |
DO ij=iip2,ip1jm |
| 1063 |
|
|
x_scal(ij,l)= & |
| 1064 |
|
|
(airev(ij-iip1)*x_v(ij-iip1,l)+airev(ij)*x_v(ij,l)) & |
| 1065 |
|
✗ |
/(airev(ij-iip1)+airev(ij)) |
| 1066 |
|
|
ENDDO |
| 1067 |
|
✗ |
DO ij=1,iip1 |
| 1068 |
|
✗ |
x_scal(ij,l)=0. |
| 1069 |
|
|
ENDDO |
| 1070 |
|
✗ |
DO ij=ip1jm+1,ip1jmp1 |
| 1071 |
|
✗ |
x_scal(ij,l)=0. |
| 1072 |
|
|
ENDDO |
| 1073 |
|
|
ENDDO |
| 1074 |
|
|
|
| 1075 |
|
✗ |
END SUBROUTINE gr_v_scal |
| 1076 |
|
|
|
| 1077 |
|
✗ |
SUBROUTINE gr_scal_v(nx,x_scal,x_v) |
| 1078 |
|
|
! convert values from scalar points to v points on C-grid |
| 1079 |
|
|
! used to compute wind stress at V points |
| 1080 |
|
|
IMPLICIT NONE |
| 1081 |
|
|
|
| 1082 |
|
|
INTEGER,INTENT(IN) :: nx ! number of levels or fields |
| 1083 |
|
|
REAL,INTENT(OUT) :: x_v((nbp_lon+1)*(nbp_lat-1),nx) |
| 1084 |
|
|
REAL,INTENT(IN) :: x_scal((nbp_lon+1)*nbp_lat,nx) |
| 1085 |
|
|
|
| 1086 |
|
|
INTEGER :: l,ij |
| 1087 |
|
|
INTEGER :: iip1,ip1jm |
| 1088 |
|
|
|
| 1089 |
|
|
iip1=nbp_lon+1 |
| 1090 |
|
|
ip1jm=(nbp_lon+1)*(nbp_lat-1) ! = iip1*jjm |
| 1091 |
|
|
|
| 1092 |
|
✗ |
DO l=1,nx |
| 1093 |
|
✗ |
DO ij=1,ip1jm |
| 1094 |
|
|
x_v(ij,l)= & |
| 1095 |
|
|
(cu(ij)*cvusurcu(ij)*x_scal(ij,l)+ & |
| 1096 |
|
|
cu(ij+iip1)*cvusurcu(ij+iip1)*x_scal(ij+iip1,l)) & |
| 1097 |
|
✗ |
/(cu(ij)*cvusurcu(ij)+cu(ij+iip1)*cvusurcu(ij+iip1)) |
| 1098 |
|
|
ENDDO |
| 1099 |
|
|
ENDDO |
| 1100 |
|
|
|
| 1101 |
|
✗ |
END SUBROUTINE gr_scal_v |
| 1102 |
|
|
|
| 1103 |
|
✗ |
SUBROUTINE gr_scal_u(nx,x_scal,x_u) |
| 1104 |
|
|
! convert values from scalar points to U points on C-grid |
| 1105 |
|
|
! used to compute wind stress at U points |
| 1106 |
|
|
IMPLICIT NONE |
| 1107 |
|
|
|
| 1108 |
|
|
INTEGER,INTENT(IN) :: nx |
| 1109 |
|
|
REAL,INTENT(OUT) :: x_u((nbp_lon+1)*nbp_lat,nx) |
| 1110 |
|
|
REAL,INTENT(IN) :: x_scal((nbp_lon+1)*nbp_lat,nx) |
| 1111 |
|
|
|
| 1112 |
|
|
INTEGER :: l,ij |
| 1113 |
|
|
INTEGER :: iip1,jjp1,ip1jmp1 |
| 1114 |
|
|
|
| 1115 |
|
✗ |
iip1=nbp_lon+1 |
| 1116 |
|
|
jjp1=nbp_lat |
| 1117 |
|
|
ip1jmp1=(nbp_lon+1)*nbp_lat ! = iip1*jjp1 |
| 1118 |
|
|
|
| 1119 |
|
✗ |
DO l=1,nx |
| 1120 |
|
✗ |
DO ij=1,ip1jmp1-1 |
| 1121 |
|
|
x_u(ij,l)= & |
| 1122 |
|
|
(aire(ij)*x_scal(ij,l)+aire(ij+1)*x_scal(ij+1,l)) & |
| 1123 |
|
✗ |
/(aire(ij)+aire(ij+1)) |
| 1124 |
|
|
ENDDO |
| 1125 |
|
|
ENDDO |
| 1126 |
|
|
|
| 1127 |
|
✗ |
CALL SCOPY(nx*jjp1,x_u(1,1),iip1,x_u(iip1,1),iip1) |
| 1128 |
|
|
|
| 1129 |
|
✗ |
END SUBROUTINE gr_scal_u |
| 1130 |
|
|
|
| 1131 |
|
|
END MODULE slab_heat_transp_mod |
| 1132 |
|
|
|