| 1 |
C $Header: /u/gcmpack/MITgcm/model/src/ini_cylinder_grid.F,v 1.5 2009/01/27 15:35:27 jmc Exp $ |
| 2 |
C $Name: $ |
| 3 |
|
| 4 |
c#include "PACKAGES_CONFIG.h" |
| 5 |
#include "CPP_OPTIONS.h" |
| 6 |
|
| 7 |
CBOP |
| 8 |
C !ROUTINE: INI_CYLINDER_GRID |
| 9 |
C !INTERFACE: |
| 10 |
SUBROUTINE INI_CYLINDER_GRID( myThid ) |
| 11 |
|
| 12 |
C !DESCRIPTION: \bv |
| 13 |
C *==========================================================* |
| 14 |
C | SUBROUTINE INI_CYLINDER_GRID |
| 15 |
C | o Initialise model coordinate system arrays |
| 16 |
C *==========================================================* |
| 17 |
C | These arrays are used throughout the code in evaluating |
| 18 |
C | gradients, integrals and spatial avarages. This routine |
| 19 |
C | is called separately by each thread and initialise only |
| 20 |
C | the region of the domain it is "responsible" for. |
| 21 |
C | Under the spherical polar grid mode primitive distances |
| 22 |
C | in X is in degrees and Y in meters. |
| 23 |
C | Distance in Z are in m or Pa |
| 24 |
C | depending on the vertical gridding mode. |
| 25 |
C *==========================================================* |
| 26 |
C \ev |
| 27 |
|
| 28 |
C !USES: |
| 29 |
IMPLICIT NONE |
| 30 |
C === Global variables === |
| 31 |
#include "SIZE.h" |
| 32 |
#include "EEPARAMS.h" |
| 33 |
#include "PARAMS.h" |
| 34 |
#include "GRID.h" |
| 35 |
c#ifdef ALLOW_EXCH2 |
| 36 |
c#include "W2_EXCH2_SIZE.h" |
| 37 |
c#include "W2_EXCH2_TOPOLOGY.h" |
| 38 |
c#include "W2_EXCH2_PARAMS.h" |
| 39 |
c#endif /* ALLOW_EXCH2 */ |
| 40 |
|
| 41 |
C !INPUT/OUTPUT PARAMETERS: |
| 42 |
C == Routine arguments == |
| 43 |
C myThid :: my Thread Id number |
| 44 |
INTEGER myThid |
| 45 |
|
| 46 |
C !LOCAL VARIABLES: |
| 47 |
C == Local variables == |
| 48 |
C xG0,yG0 :: coordinate of South-West tile-corner |
| 49 |
C iG, jG :: Global coordinate index. Usually used to hold |
| 50 |
C :: the south-west global coordinate of a tile. |
| 51 |
C bi,bj :: tile indices |
| 52 |
C i, j :: loop counters |
| 53 |
INTEGER iG, jG |
| 54 |
INTEGER bi, bj |
| 55 |
INTEGER i, j |
| 56 |
_RL dtheta, thisRad, xG0, yG0 |
| 57 |
|
| 58 |
C "Long" real for temporary coordinate calculation |
| 59 |
C NOTICE the extended range of indices!! |
| 60 |
_RL xGloc(1-Olx:sNx+Olx+1,1-Oly:sNy+Oly+1) |
| 61 |
_RL yGloc(1-Olx:sNx+Olx+1,1-Oly:sNy+Oly+1) |
| 62 |
|
| 63 |
C The functions iGl, jGl return the "global" index with valid values beyond |
| 64 |
C halo regions |
| 65 |
C cnh wrote: |
| 66 |
C > I dont understand why we would ever have to multiply the |
| 67 |
C > overlap by the total domain size e.g |
| 68 |
C > OLx*Nx, OLy*Ny. |
| 69 |
C > Can anybody explain? Lines are in ini_spherical_polar_grid.F. |
| 70 |
C > Surprised the code works if its wrong, so I am puzzled. |
| 71 |
C jmc replied: |
| 72 |
C Yes, I can explain this since I put this modification to work |
| 73 |
C with small domain (where Oly > Ny, as for instance, zonal-average |
| 74 |
C case): |
| 75 |
C This has no effect on the acuracy of the evaluation of iGl(I,bi) |
| 76 |
C and jGl(j,bj) since we take mod(a+OLx*Nx,Nx) and mod(b+OLy*Ny,Ny). |
| 77 |
C But in case a or b is negative, then the FORTRAN function "mod" |
| 78 |
C does not return the matematical value of the "modulus" function, |
| 79 |
C and this is not good for your purpose. |
| 80 |
C This is why I add +OLx*Nx and +OLy*Ny to be sure that the 1rst |
| 81 |
C argument of the mod function is positive. |
| 82 |
INTEGER iGl,jGl |
| 83 |
iGl(i,bi) = 1+MOD(myXGlobalLo-1+(bi-1)*sNx+i+Olx*Nx-1,Nx) |
| 84 |
jGl(j,bj) = 1+MOD(myYGlobalLo-1+(bj-1)*sNy+j+Oly*Ny-1,Ny) |
| 85 |
c#ifdef ALLOW_EXCH2 |
| 86 |
c INTEGER tN |
| 87 |
c#endif /* ALLOW_EXCH2 */ |
| 88 |
CEOP |
| 89 |
|
| 90 |
C For each tile ... |
| 91 |
DO bj = myByLo(myThid), myByHi(myThid) |
| 92 |
DO bi = myBxLo(myThid), myBxHi(myThid) |
| 93 |
|
| 94 |
C-- "Global" index (place holder) |
| 95 |
jG = myYGlobalLo + (bj-1)*sNy |
| 96 |
iG = myXGlobalLo + (bi-1)*sNx |
| 97 |
c#ifdef ALLOW_EXCH2 |
| 98 |
c IF ( W2_useE2ioLayOut ) THEN |
| 99 |
cC- note: does not work for non-uniform delX or delY |
| 100 |
c tN = W2_myTileList(bi,bj) |
| 101 |
c iG = exch2_txGlobalo(tN) |
| 102 |
c jG = exch2_tyGlobalo(tN) |
| 103 |
c ENDIF |
| 104 |
c#endif /* ALLOW_EXCH2 */ |
| 105 |
|
| 106 |
C-- First find coordinate of tile corner (meaning outer corner of halo) |
| 107 |
xG0 = xgOrigin |
| 108 |
C Find the X-coordinate of the outer grid-line of the "real" tile |
| 109 |
DO i=1, iG-1 |
| 110 |
xG0 = xG0 + delX(i) |
| 111 |
ENDDO |
| 112 |
C Back-step to the outer grid-line of the "halo" region |
| 113 |
DO i=1, Olx |
| 114 |
xG0 = xG0 - delX( 1+mod(Olx*Nx-1+iG-i,Nx) ) |
| 115 |
ENDDO |
| 116 |
C Find the Y-coordinate of the outer grid-line of the "real" tile |
| 117 |
yG0 = ygOrigin |
| 118 |
DO j=1, jG-1 |
| 119 |
yG0 = yG0 + delY(j) |
| 120 |
ENDDO |
| 121 |
C Back-step to the outer grid-line of the "halo" region |
| 122 |
DO j=1, Oly |
| 123 |
yG0 = yG0 - delY( 1+mod(Oly*Ny-1+jG-j,Ny) ) |
| 124 |
ENDDO |
| 125 |
|
| 126 |
C-- Calculate coordinates of cell corners for N+1 grid-lines |
| 127 |
DO j=1-Oly,sNy+Oly +1 |
| 128 |
xGloc(1-Olx,j) = xG0 |
| 129 |
DO i=1-Olx,sNx+Olx |
| 130 |
xGloc(i+1,j) = xGloc(i,j) + delX( iGl(i,bi) ) |
| 131 |
ENDDO |
| 132 |
ENDDO |
| 133 |
DO i=1-Olx,sNx+Olx +1 |
| 134 |
yGloc(i,1-Oly) = yG0 |
| 135 |
DO j=1-Oly,sNy+Oly |
| 136 |
yGloc(i,j+1) = yGloc(i,j) + delY( jGl(j,bj) ) |
| 137 |
ENDDO |
| 138 |
ENDDO |
| 139 |
|
| 140 |
C-- Make a permanent copy of [xGloc,yGloc] in [xG,yG] |
| 141 |
DO j=1-Oly,sNy+Oly |
| 142 |
DO i=1-Olx,sNx+Olx |
| 143 |
xG(i,j,bi,bj) = xGloc(i,j) |
| 144 |
yG(i,j,bi,bj) = yGloc(i,j) |
| 145 |
ENDDO |
| 146 |
ENDDO |
| 147 |
|
| 148 |
C-- Calculate [xC,yC], coordinates of cell centers |
| 149 |
DO j=1-Oly,sNy+Oly |
| 150 |
DO i=1-Olx,sNx+Olx |
| 151 |
C by averaging |
| 152 |
xC(i,j,bi,bj) = 0.25 _d 0*( |
| 153 |
& xGloc(i,j)+xGloc(i+1,j)+xGloc(i,j+1)+xGloc(i+1,j+1) ) |
| 154 |
yC(i,j,bi,bj) = 0.25 _d 0*( |
| 155 |
& yGloc(i,j)+yGloc(i+1,j)+yGloc(i,j+1)+yGloc(i+1,j+1) ) |
| 156 |
ENDDO |
| 157 |
ENDDO |
| 158 |
|
| 159 |
C-- Calculate [dxF,dyF], lengths between cell faces (through center) |
| 160 |
DO j=1-Oly,sNy+Oly |
| 161 |
DO i=1-Olx,sNx+Olx |
| 162 |
thisRad = yC(i,j,bi,bj) |
| 163 |
dtheta = delX( iGl(i,bi) ) |
| 164 |
dxF(i,j,bi,bj) = thisRad*dtheta*deg2rad |
| 165 |
dyF(i,j,bi,bj) = delY( jGl(j,bj) ) |
| 166 |
ENDDO |
| 167 |
ENDDO |
| 168 |
|
| 169 |
C-- Calculate [dxG,dyG], lengths along cell boundaries |
| 170 |
DO j=1-Oly,sNy+Oly |
| 171 |
DO i=1-Olx,sNx+Olx |
| 172 |
thisRad = 0.5 _d 0*(yGloc(i,j)+yGloc(i+1,j)) |
| 173 |
dtheta = delX( iGl(i,bi) ) |
| 174 |
dxG(i,j,bi,bj) = thisRad*dtheta*deg2rad |
| 175 |
dyG(i,j,bi,bj) = delY( jGl(j,bj) ) |
| 176 |
ENDDO |
| 177 |
ENDDO |
| 178 |
|
| 179 |
C-- The following arrays are not defined in some parts of the halo |
| 180 |
C region. We set them to zero here for safety. If they are ever |
| 181 |
C referred to, especially in the denominator then it is a mistake! |
| 182 |
DO j=1-Oly,sNy+Oly |
| 183 |
DO i=1-Olx,sNx+Olx |
| 184 |
dxC(i,j,bi,bj) = 0. |
| 185 |
dyC(i,j,bi,bj) = 0. |
| 186 |
dxV(i,j,bi,bj) = 0. |
| 187 |
dyU(i,j,bi,bj) = 0. |
| 188 |
rAw(i,j,bi,bj) = 0. |
| 189 |
rAs(i,j,bi,bj) = 0. |
| 190 |
ENDDO |
| 191 |
ENDDO |
| 192 |
|
| 193 |
C-- Calculate [dxC], zonal length between cell centers |
| 194 |
DO j=1-Oly,sNy+Oly |
| 195 |
DO i=1-Olx+1,sNx+Olx ! NOTE range |
| 196 |
C by averaging |
| 197 |
dxC(i,j,bi,bj) = 0.5 _d 0*(dxF(i,j,bi,bj)+dxF(i-1,j,bi,bj)) |
| 198 |
ENDDO |
| 199 |
ENDDO |
| 200 |
|
| 201 |
C-- Calculate [dyC], meridional length between cell centers |
| 202 |
DO j=1-Oly+1,sNy+Oly ! NOTE range |
| 203 |
DO i=1-Olx,sNx+Olx |
| 204 |
C by averaging |
| 205 |
dyC(i,j,bi,bj) = 0.5 _d 0*(dyF(i,j,bi,bj)+dyF(i,j-1,bi,bj)) |
| 206 |
ENDDO |
| 207 |
ENDDO |
| 208 |
|
| 209 |
C-- Calculate [dxV,dyU], length between velocity points (through corners) |
| 210 |
DO j=1-Oly+1,sNy+Oly ! NOTE range |
| 211 |
DO i=1-Olx+1,sNx+Olx ! NOTE range |
| 212 |
C by averaging (method I) |
| 213 |
dxV(i,j,bi,bj) = 0.5 _d 0*(dxG(i,j,bi,bj)+dxG(i-1,j,bi,bj)) |
| 214 |
dyU(i,j,bi,bj) = 0.5 _d 0*(dyG(i,j,bi,bj)+dyG(i,j-1,bi,bj)) |
| 215 |
ENDDO |
| 216 |
ENDDO |
| 217 |
|
| 218 |
C-- Calculate vertical face area |
| 219 |
DO j=1-Oly,sNy+Oly |
| 220 |
DO i=1-Olx,sNx+Olx |
| 221 |
C- All r(dr)(dtheta) |
| 222 |
rA (i,j,bi,bj) = dxF(i,j,bi,bj)*dyF(i,j,bi,bj) |
| 223 |
rAw(i,j,bi,bj) = dxC(i,j,bi,bj)*dyG(i,j,bi,bj) |
| 224 |
rAs(i,j,bi,bj) = dxG(i,j,bi,bj)*dyC(i,j,bi,bj) |
| 225 |
rAz(i,j,bi,bj) = dxV(i,j,bi,bj)*dyU(i,j,bi,bj) |
| 226 |
C-- Set trigonometric terms & grid orientation: |
| 227 |
tanPhiAtU(i,j,bi,bj) = 0. |
| 228 |
tanPhiAtV(i,j,bi,bj) = 0. |
| 229 |
angleCosC(i,j,bi,bj) = 1. |
| 230 |
angleSinC(i,j,bi,bj) = 0. |
| 231 |
ENDDO |
| 232 |
ENDDO |
| 233 |
|
| 234 |
C-- Cosine(lat) scaling |
| 235 |
DO j=1-OLy,sNy+OLy |
| 236 |
cosFacU(j,bi,bj)=1. |
| 237 |
cosFacV(j,bi,bj)=1. |
| 238 |
sqcosFacU(j,bi,bj)=1. |
| 239 |
sqcosFacV(j,bi,bj)=1. |
| 240 |
ENDDO |
| 241 |
|
| 242 |
ENDDO ! bi |
| 243 |
ENDDO ! bj |
| 244 |
|
| 245 |
C-- Set default (=whole domain) for where relaxation to climatology applies |
| 246 |
IF ( latBandClimRelax.EQ.UNSET_RL ) THEN |
| 247 |
_BEGIN_MASTER(myThid) |
| 248 |
latBandClimRelax = 0. |
| 249 |
DO j=1,Ny |
| 250 |
latBandClimRelax = latBandClimRelax + delY(j) |
| 251 |
ENDDO |
| 252 |
latBandClimRelax = latBandClimRelax*3. _d 0 |
| 253 |
_END_MASTER(myThid) |
| 254 |
ENDIF |
| 255 |
|
| 256 |
RETURN |
| 257 |
END |
Parent Directory
| 
Revision Log
| 
Revision Graph