| 1 |
! |
| 2 |
! $Header: /home/cvsroot/LMDZ4/libf/filtrez/inifilr.F,v 1.1.1.1 2004/05/19 12:53:09 lmdzadmin Exp $ |
| 3 |
! |
| 4 |
SUBROUTINE inifilr |
| 5 |
c |
| 6 |
c ... H. Upadhyaya, O.Sharma ... |
| 7 |
c |
| 8 |
use dimens_m |
| 9 |
use paramet_m |
| 10 |
use logic |
| 11 |
use comgeom |
| 12 |
use serre |
| 13 |
use parafilt |
| 14 |
use coefils |
| 15 |
IMPLICIT NONE |
| 16 |
c |
| 17 |
c version 3 ..... |
| 18 |
|
| 19 |
c Correction le 28/10/97 P. Le Van . |
| 20 |
c ------------------------------------------------------------------- |
| 21 |
c ------------------------------------------------------------------- |
| 22 |
|
| 23 |
REAL dlonu(iim),dlatu(jjm) |
| 24 |
REAL rlamda( iim ), eignvl( iim ) |
| 25 |
c |
| 26 |
|
| 27 |
REAL lamdamax,pi,cof |
| 28 |
INTEGER i,j,modemax,imx,k,kf,ii |
| 29 |
REAL dymin,dxmin,colat0 |
| 30 |
REAL eignft(iim,iim), coff |
| 31 |
REAL matriceun,matriceus,matricevn,matricevs,matrinvn,matrinvs |
| 32 |
COMMON/matrfil/matriceun(iim,iim,nfilun),matriceus(iim,iim,nfilus) |
| 33 |
, , matricevn(iim,iim,nfilvn),matricevs(iim,iim,nfilvs) |
| 34 |
, , matrinvn(iim,iim,nfilun),matrinvs (iim,iim,nfilus) |
| 35 |
EXTERNAL inifgn |
| 36 |
c |
| 37 |
c ------------------------------------------------------------ |
| 38 |
c This routine computes the eigenfunctions of the laplacien |
| 39 |
c on the stretched grid, and the filtering coefficients |
| 40 |
c |
| 41 |
c We designate: |
| 42 |
c eignfn eigenfunctions of the discrete laplacien |
| 43 |
c eigenvl eigenvalues |
| 44 |
c jfiltn indexof the last scalar line filtered in NH |
| 45 |
c jfilts index of the first line filtered in SH |
| 46 |
c modfrst index of the mode from where modes are filtered |
| 47 |
c modemax maximum number of modes ( im ) |
| 48 |
c coefil filtering coefficients ( lamda_max*cos(rlat)/lamda ) |
| 49 |
c sdd SQRT( dx ) |
| 50 |
c |
| 51 |
c the modes are filtered from modfrst to modemax |
| 52 |
c |
| 53 |
c----------------------------------------------------------- |
| 54 |
c |
| 55 |
|
| 56 |
pi = 2. * ASIN( 1. ) |
| 57 |
|
| 58 |
DO i = 1,iim |
| 59 |
dlonu(i) = xprimu( i ) |
| 60 |
ENDDO |
| 61 |
c |
| 62 |
CALL inifgn(eignvl) |
| 63 |
c |
| 64 |
print *,' EIGNVL ' |
| 65 |
PRINT 250,eignvl |
| 66 |
250 FORMAT( 1x,5e13.6) |
| 67 |
c |
| 68 |
c compute eigenvalues and eigenfunctions |
| 69 |
c |
| 70 |
c |
| 71 |
c................................................................. |
| 72 |
c |
| 73 |
c compute the filtering coefficients for scalar lines and |
| 74 |
c meridional wind v-lines |
| 75 |
c |
| 76 |
c we filter all those latitude lines where coefil < 1 |
| 77 |
c NO FILTERING AT POLES |
| 78 |
c |
| 79 |
c colat0 is to be used when alpha (stretching coefficient) |
| 80 |
c is set equal to zero for the regular grid case |
| 81 |
c |
| 82 |
c ....... Calcul de colat0 ......... |
| 83 |
c ..... colat0 = minimum de ( 0.5, min dy/ min dx ) ... |
| 84 |
c |
| 85 |
c |
| 86 |
DO 45 j = 1,jjm |
| 87 |
dlatu( j ) = rlatu( j ) - rlatu( j+1 ) |
| 88 |
45 CONTINUE |
| 89 |
c |
| 90 |
dxmin = dlonu(1) |
| 91 |
DO i = 2, iim |
| 92 |
dxmin = MIN( dxmin,dlonu(i) ) |
| 93 |
ENDDO |
| 94 |
dymin = dlatu(1) |
| 95 |
DO j = 2, jjm |
| 96 |
dymin = MIN( dymin,dlatu(j) ) |
| 97 |
ENDDO |
| 98 |
c |
| 99 |
c |
| 100 |
colat0 = MIN( 0.5, dymin/dxmin ) |
| 101 |
c |
| 102 |
IF( .NOT.fxyhypb.AND.ysinus ) THEN |
| 103 |
colat0 = 0.6 |
| 104 |
c ...... a revoir pour ysinus ! ....... |
| 105 |
alphax = 0. |
| 106 |
ENDIF |
| 107 |
c |
| 108 |
PRINT 50, colat0,alphax |
| 109 |
50 FORMAT(/15x,' Inifilr colat0 alphax ',2e16.7) |
| 110 |
c |
| 111 |
IF(alphax.EQ.1. ) THEN |
| 112 |
PRINT *,' Inifilr alphax doit etre < a 1. Corriger ' |
| 113 |
STOP 1 |
| 114 |
ENDIF |
| 115 |
c |
| 116 |
lamdamax = iim / ( pi * colat0 * ( 1. - alphax ) ) |
| 117 |
|
| 118 |
cc ... Correction le 28/10/97 ( P.Le Van ) .. |
| 119 |
c |
| 120 |
DO 71 i = 2,iim |
| 121 |
rlamda( i ) = lamdamax/ SQRT( ABS( eignvl(i) ) ) |
| 122 |
71 CONTINUE |
| 123 |
c |
| 124 |
|
| 125 |
DO 72 j = 1,jjm |
| 126 |
DO 73 i = 1,iim |
| 127 |
coefilu( i,j ) = 0.0 |
| 128 |
coefilv( i,j ) = 0.0 |
| 129 |
coefilu2( i,j ) = 0.0 |
| 130 |
coefilv2( i,j ) = 0.0 |
| 131 |
73 CONTINUE |
| 132 |
72 CONTINUE |
| 133 |
|
| 134 |
c |
| 135 |
c ... Determination de jfiltnu,jfiltnv,jfiltsu,jfiltsv .... |
| 136 |
c ......................................................... |
| 137 |
c |
| 138 |
modemax = iim |
| 139 |
|
| 140 |
cccc imx = modemax - 4 * (modemax/iim) |
| 141 |
|
| 142 |
imx = iim |
| 143 |
c |
| 144 |
PRINT *,' TRUNCATION AT ',imx |
| 145 |
c |
| 146 |
DO 75 j = 2, jjm/2+1 |
| 147 |
cof = COS( rlatu(j) )/ colat0 |
| 148 |
IF ( cof .LT. 1. ) THEN |
| 149 |
IF( rlamda(imx) * COS(rlatu(j) ).LT.1. ) jfiltnu= j |
| 150 |
ENDIF |
| 151 |
|
| 152 |
cof = COS( rlatu(jjp1-j+1) )/ colat0 |
| 153 |
IF ( cof .LT. 1. ) THEN |
| 154 |
IF( rlamda(imx) * COS(rlatu(jjp1-j+1) ).LT.1. ) |
| 155 |
$ jfiltsu= jjp1-j+1 |
| 156 |
ENDIF |
| 157 |
75 CONTINUE |
| 158 |
c |
| 159 |
DO 76 j = 1, jjm/2 |
| 160 |
cof = COS( rlatv(j) )/ colat0 |
| 161 |
IF ( cof .LT. 1. ) THEN |
| 162 |
IF( rlamda(imx) * COS(rlatv(j) ).LT.1. ) jfiltnv= j |
| 163 |
ENDIF |
| 164 |
|
| 165 |
cof = COS( rlatv(jjm-j+1) )/ colat0 |
| 166 |
IF ( cof .LT. 1. ) THEN |
| 167 |
IF( rlamda(imx) * COS(rlatv(jjm-j+1) ).LT.1. ) |
| 168 |
$ jfiltsv= jjm-j+1 |
| 169 |
ENDIF |
| 170 |
76 CONTINUE |
| 171 |
c |
| 172 |
|
| 173 |
if ( jfiltnu.LE.0 ) jfiltnu=1 |
| 174 |
IF( jfiltnu.GT. jjm/2 +1 ) THEN |
| 175 |
PRINT *,' jfiltnu en dehors des valeurs acceptables ' ,jfiltnu |
| 176 |
STOP 1 |
| 177 |
ENDIF |
| 178 |
|
| 179 |
IF( jfiltsu.LE.0) jfiltsu=1 |
| 180 |
IF( jfiltsu.GT. jjm +1 ) THEN |
| 181 |
PRINT *,' jfiltsu en dehors des valeurs acceptables ' ,jfiltsu |
| 182 |
STOP 1 |
| 183 |
ENDIF |
| 184 |
|
| 185 |
IF( jfiltnv.LE.0) jfiltnv=1 |
| 186 |
IF( jfiltnv.GT. jjm/2 ) THEN |
| 187 |
PRINT *,' jfiltnv en dehors des valeurs acceptables ' ,jfiltnv |
| 188 |
STOP 1 |
| 189 |
ENDIF |
| 190 |
|
| 191 |
IF( jfiltsv.LE.0) jfiltsv=1 |
| 192 |
IF( jfiltsv.GT. jjm ) THEN |
| 193 |
PRINT *,' jfiltsv en dehors des valeurs acceptables ' ,jfiltsv |
| 194 |
STOP 1 |
| 195 |
ENDIF |
| 196 |
|
| 197 |
PRINT *,' jfiltnv jfiltsv jfiltnu jfiltsu ' , |
| 198 |
* jfiltnv,jfiltsv,jfiltnu,jfiltsu |
| 199 |
|
| 200 |
c |
| 201 |
c ... Determination de coefilu,coefilv,n=modfrstu,modfrstv .... |
| 202 |
c................................................................ |
| 203 |
c |
| 204 |
c |
| 205 |
DO 77 j = 1,jjm |
| 206 |
modfrstu( j ) = iim |
| 207 |
modfrstv( j ) = iim |
| 208 |
77 CONTINUE |
| 209 |
c |
| 210 |
DO 84 j = 2,jfiltnu |
| 211 |
DO 81 k = 2,modemax |
| 212 |
cof = rlamda(k) * COS( rlatu(j) ) |
| 213 |
IF ( cof .LT. 1. ) GOTO 82 |
| 214 |
81 CONTINUE |
| 215 |
GOTO 84 |
| 216 |
82 modfrstu( j ) = k |
| 217 |
c |
| 218 |
kf = modfrstu( j ) |
| 219 |
DO 83 k = kf , modemax |
| 220 |
cof = rlamda(k) * COS( rlatu(j) ) |
| 221 |
coefilu(k,j) = cof - 1. |
| 222 |
coefilu2(k,j) = cof*cof - 1. |
| 223 |
83 CONTINUE |
| 224 |
84 CONTINUE |
| 225 |
c |
| 226 |
c |
| 227 |
DO 89 j = 1,jfiltnv |
| 228 |
c |
| 229 |
DO 86 k = 2,modemax |
| 230 |
cof = rlamda(k) * COS( rlatv(j) ) |
| 231 |
IF ( cof .LT. 1. ) GOTO 87 |
| 232 |
86 CONTINUE |
| 233 |
GOTO 89 |
| 234 |
87 modfrstv( j ) = k |
| 235 |
c |
| 236 |
kf = modfrstv( j ) |
| 237 |
DO 88 k = kf , modemax |
| 238 |
cof = rlamda(k) * COS( rlatv(j) ) |
| 239 |
coefilv(k,j) = cof - 1. |
| 240 |
coefilv2(k,j) = cof*cof - 1. |
| 241 |
88 CONTINUE |
| 242 |
c |
| 243 |
89 CONTINUE |
| 244 |
c |
| 245 |
DO 94 j = jfiltsu,jjm |
| 246 |
DO 91 k = 2,modemax |
| 247 |
cof = rlamda(k) * COS( rlatu(j) ) |
| 248 |
IF ( cof .LT. 1. ) GOTO 92 |
| 249 |
91 CONTINUE |
| 250 |
GOTO 94 |
| 251 |
92 modfrstu( j ) = k |
| 252 |
c |
| 253 |
kf = modfrstu( j ) |
| 254 |
DO 93 k = kf , modemax |
| 255 |
cof = rlamda(k) * COS( rlatu(j) ) |
| 256 |
coefilu(k,j) = cof - 1. |
| 257 |
coefilu2(k,j) = cof*cof - 1. |
| 258 |
93 CONTINUE |
| 259 |
94 CONTINUE |
| 260 |
c |
| 261 |
DO 99 j = jfiltsv,jjm |
| 262 |
DO 96 k = 2,modemax |
| 263 |
cof = rlamda(k) * COS( rlatv(j) ) |
| 264 |
IF ( cof .LT. 1. ) GOTO 97 |
| 265 |
96 CONTINUE |
| 266 |
GOTO 99 |
| 267 |
97 modfrstv( j ) = k |
| 268 |
c |
| 269 |
kf = modfrstv( j ) |
| 270 |
DO 98 k = kf , modemax |
| 271 |
cof = rlamda(k) * COS( rlatv(j) ) |
| 272 |
coefilv(k,j) = cof - 1. |
| 273 |
coefilv2(k,j) = cof*cof - 1. |
| 274 |
98 CONTINUE |
| 275 |
99 CONTINUE |
| 276 |
c |
| 277 |
|
| 278 |
IF(jfiltnv.GE.jjm/2 .OR. jfiltnu.GE.jjm/2)THEN |
| 279 |
|
| 280 |
IF(jfiltnv.EQ.jfiltsv)jfiltsv=1+jfiltnv |
| 281 |
IF(jfiltnu.EQ.jfiltsu)jfiltsu=1+jfiltnu |
| 282 |
|
| 283 |
PRINT *,'jfiltnv jfiltsv jfiltnu jfiltsu' , |
| 284 |
* jfiltnv,jfiltsv,jfiltnu,jfiltsu |
| 285 |
ENDIF |
| 286 |
|
| 287 |
PRINT *,' Modes premiers v ' |
| 288 |
PRINT 334,modfrstv |
| 289 |
PRINT *,' Modes premiers u ' |
| 290 |
PRINT 334,modfrstu |
| 291 |
|
| 292 |
|
| 293 |
IF( nfilun.LT. jfiltnu ) THEN |
| 294 |
PRINT *,' le parametre nfilun utilise pour la matrice ', |
| 295 |
* ' matriceun est trop petit ! ' |
| 296 |
PRINT *,'Le changer dans parafilt.h et le mettre a ',jfiltnu |
| 297 |
PRINT *,' Pour information, nfilun,nfilus,nfilvn,nfilvs ' |
| 298 |
* ,'doivent etre egaux successivement a ',jfiltnu,jjm-jfiltsu+1 |
| 299 |
* ,jfiltnv,jjm-jfiltsv+1 |
| 300 |
STOP 1 |
| 301 |
ENDIF |
| 302 |
IF( nfilun.GT. jfiltnu+ 2 ) THEN |
| 303 |
PRINT *,' le parametre nfilun utilise pour la matrice ', |
| 304 |
*' matriceun est trop grand ! Gachis de memoire ! ' |
| 305 |
PRINT *,'Le changer dans parafilt.h et le mettre a ',jfiltnu |
| 306 |
PRINT *,' Pour information, nfilun,nfilus,nfilvn,nfilvs ' |
| 307 |
* ,'doivent etre egaux successivement a ',jfiltnu,jjm-jfiltsu+1 |
| 308 |
* ,jfiltnv,jjm-jfiltsv+1 |
| 309 |
c STOP 1 |
| 310 |
ENDIF |
| 311 |
IF( nfilus.LT. jjm - jfiltsu +1 ) THEN |
| 312 |
PRINT *,' le parametre nfilus utilise pour la matrice ', |
| 313 |
* ' matriceus est trop petit ! ' |
| 314 |
PRINT *,' Le changer dans parafilt.h et le mettre a ', |
| 315 |
* jjm - jfiltsu + 1 |
| 316 |
PRINT *,' Pour information , nfilun,nfilus,nfilvn,nfilvs ' |
| 317 |
* ,'doivent etre egaux successivement a ',jfiltnu,jjm-jfiltsu+1 |
| 318 |
* ,jfiltnv,jjm-jfiltsv+1 |
| 319 |
STOP 1 |
| 320 |
ENDIF |
| 321 |
IF( nfilus.GT. jjm - jfiltsu + 3 ) THEN |
| 322 |
PRINT *,' le parametre nfilus utilise pour la matrice ', |
| 323 |
* ' matriceus est trop grand ! ' |
| 324 |
PRINT *,' Le changer dans parafilt.h et le mettre a ' , |
| 325 |
* jjm - jfiltsu + 1 |
| 326 |
PRINT *,' Pour information , nfilun,nfilus,nfilvn,nfilvs ' |
| 327 |
* ,'doivent etre egaux successivement a ',jfiltnu,jjm-jfiltsu+1 |
| 328 |
* ,jfiltnv,jjm-jfiltsv+1 |
| 329 |
c STOP 1 |
| 330 |
ENDIF |
| 331 |
IF( nfilvn.LT. jfiltnv ) THEN |
| 332 |
PRINT *,' le parametre nfilvn utilise pour la matrice ', |
| 333 |
* ' matricevn est trop petit ! ' |
| 334 |
PRINT *,'Le changer dans parafilt.h et le mettre a ',jfiltnv |
| 335 |
PRINT *,' Pour information , nfilun,nfilus,nfilvn,nfilvs ' |
| 336 |
* ,'doivent etre egaux successivement a ',jfiltnu,jjm-jfiltsu+1 |
| 337 |
* ,jfiltnv,jjm-jfiltsv+1 |
| 338 |
STOP 1 |
| 339 |
ENDIF |
| 340 |
IF( nfilvn.GT. jfiltnv+ 2 ) THEN |
| 341 |
PRINT *,' le parametre nfilvn utilise pour la matrice ', |
| 342 |
*' matricevn est trop grand ! Gachis de memoire ! ' |
| 343 |
PRINT *,'Le changer dans parafilt.h et le mettre a ',jfiltnv |
| 344 |
PRINT *,' Pour information , nfilun,nfilus,nfilvn,nfilvs ' |
| 345 |
* ,'doivent etre egaux successivement a ',jfiltnu,jjm-jfiltsu+1 |
| 346 |
* ,jfiltnv,jjm-jfiltsv+1 |
| 347 |
c STOP 1 |
| 348 |
ENDIF |
| 349 |
IF( nfilvs.LT. jjm - jfiltsv +1 ) THEN |
| 350 |
PRINT *,' le parametre nfilvs utilise pour la matrice ', |
| 351 |
* ' matricevs est trop petit ! Le changer dans parafilt.h ' |
| 352 |
PRINT *,' Le changer dans parafilt.h et le mettre a ' |
| 353 |
* , jjm - jfiltsv + 1 |
| 354 |
PRINT *,' Pour information , nfilun,nfilus,nfilvn,nfilvs ' |
| 355 |
* ,'doivent etre egaux successivement a ',jfiltnu,jjm-jfiltsu+1 |
| 356 |
* ,jfiltnv,jjm-jfiltsv+1 |
| 357 |
STOP 1 |
| 358 |
ENDIF |
| 359 |
IF( nfilvs.GT. jjm - jfiltsv + 3 ) THEN |
| 360 |
PRINT *,' le parametre nfilvs utilise pour la matrice ', |
| 361 |
* ' matricevs est trop grand ! Gachis de memoire ! ' |
| 362 |
PRINT *,' Le changer dans parafilt.h et le mettre a ' |
| 363 |
* , jjm - jfiltsv + 1 |
| 364 |
PRINT *,' Pour information , nfilun,nfilus,nfilvn,nfilvs ' |
| 365 |
* ,'doivent etre egaux successivement a ',jfiltnu,jjm-jfiltsu+1 |
| 366 |
* ,jfiltnv,jjm-jfiltsv+1 |
| 367 |
c STOP 1 |
| 368 |
ENDIF |
| 369 |
|
| 370 |
c |
| 371 |
c ................................................................... |
| 372 |
c |
| 373 |
c ... Calcul de la matrice filtre 'matriceu' pour les champs situes |
| 374 |
c sur la grille scalaire ........ |
| 375 |
c ................................................................... |
| 376 |
c |
| 377 |
DO j = 2, jfiltnu |
| 378 |
|
| 379 |
DO i=1,iim |
| 380 |
coff = coefilu(i,j) |
| 381 |
IF( i.LT.modfrstu(j) ) coff = 0. |
| 382 |
DO k=1,iim |
| 383 |
eignft(i,k) = eignfnv(k,i) * coff |
| 384 |
ENDDO |
| 385 |
ENDDO |
| 386 |
DO k = 1, iim |
| 387 |
DO i = 1, iim |
| 388 |
matriceun(i,k,j) = 0.0 |
| 389 |
DO ii = 1, iim |
| 390 |
matriceun(i,k,j) = matriceun(i,k,j) |
| 391 |
. + eignfnv(i,ii)*eignft(ii,k) |
| 392 |
ENDDO |
| 393 |
ENDDO |
| 394 |
ENDDO |
| 395 |
|
| 396 |
ENDDO |
| 397 |
|
| 398 |
DO j = jfiltsu, jjm |
| 399 |
|
| 400 |
DO i=1,iim |
| 401 |
coff = coefilu(i,j) |
| 402 |
IF( i.LT.modfrstu(j) ) coff = 0. |
| 403 |
DO k=1,iim |
| 404 |
eignft(i,k) = eignfnv(k,i) * coff |
| 405 |
ENDDO |
| 406 |
ENDDO |
| 407 |
DO k = 1, iim |
| 408 |
DO i = 1, iim |
| 409 |
matriceus(i,k,j-jfiltsu+1) = 0.0 |
| 410 |
DO ii = 1, iim |
| 411 |
matriceus(i,k,j-jfiltsu+1) = matriceus(i,k,j-jfiltsu+1) |
| 412 |
. + eignfnv(i,ii)*eignft(ii,k) |
| 413 |
ENDDO |
| 414 |
ENDDO |
| 415 |
ENDDO |
| 416 |
|
| 417 |
ENDDO |
| 418 |
|
| 419 |
c ................................................................... |
| 420 |
c |
| 421 |
c ... Calcul de la matrice filtre 'matricev' pour les champs situes |
| 422 |
c sur la grille de V ou de Z ........ |
| 423 |
c ................................................................... |
| 424 |
c |
| 425 |
DO j = 1, jfiltnv |
| 426 |
|
| 427 |
DO i = 1, iim |
| 428 |
coff = coefilv(i,j) |
| 429 |
IF( i.LT.modfrstv(j) ) coff = 0. |
| 430 |
DO k = 1, iim |
| 431 |
eignft(i,k) = eignfnu(k,i) * coff |
| 432 |
ENDDO |
| 433 |
ENDDO |
| 434 |
DO k = 1, iim |
| 435 |
DO i = 1, iim |
| 436 |
matricevn(i,k,j) = 0.0 |
| 437 |
DO ii = 1, iim |
| 438 |
matricevn(i,k,j) = matricevn(i,k,j) |
| 439 |
. + eignfnu(i,ii)*eignft(ii,k) |
| 440 |
ENDDO |
| 441 |
ENDDO |
| 442 |
ENDDO |
| 443 |
|
| 444 |
ENDDO |
| 445 |
|
| 446 |
DO j = jfiltsv, jjm |
| 447 |
|
| 448 |
DO i = 1, iim |
| 449 |
coff = coefilv(i,j) |
| 450 |
IF( i.LT.modfrstv(j) ) coff = 0. |
| 451 |
DO k = 1, iim |
| 452 |
eignft(i,k) = eignfnu(k,i) * coff |
| 453 |
ENDDO |
| 454 |
ENDDO |
| 455 |
DO k = 1, iim |
| 456 |
DO i = 1, iim |
| 457 |
matricevs(i,k,j-jfiltsv+1) = 0.0 |
| 458 |
DO ii = 1, iim |
| 459 |
matricevs(i,k,j-jfiltsv+1) = matricevs(i,k,j-jfiltsv+1) |
| 460 |
. + eignfnu(i,ii)*eignft(ii,k) |
| 461 |
ENDDO |
| 462 |
ENDDO |
| 463 |
ENDDO |
| 464 |
|
| 465 |
ENDDO |
| 466 |
|
| 467 |
c ................................................................... |
| 468 |
c |
| 469 |
c ... Calcul de la matrice filtre 'matrinv' pour les champs situes |
| 470 |
c sur la grille scalaire , pour le filtre inverse ........ |
| 471 |
c ................................................................... |
| 472 |
c |
| 473 |
DO j = 2, jfiltnu |
| 474 |
|
| 475 |
DO i = 1,iim |
| 476 |
coff = coefilu(i,j)/ ( 1. + coefilu(i,j) ) |
| 477 |
IF( i.LT.modfrstu(j) ) coff = 0. |
| 478 |
DO k=1,iim |
| 479 |
eignft(i,k) = eignfnv(k,i) * coff |
| 480 |
ENDDO |
| 481 |
ENDDO |
| 482 |
DO k = 1, iim |
| 483 |
DO i = 1, iim |
| 484 |
matrinvn(i,k,j) = 0.0 |
| 485 |
DO ii = 1, iim |
| 486 |
matrinvn(i,k,j) = matrinvn(i,k,j) |
| 487 |
. + eignfnv(i,ii)*eignft(ii,k) |
| 488 |
ENDDO |
| 489 |
ENDDO |
| 490 |
ENDDO |
| 491 |
|
| 492 |
ENDDO |
| 493 |
|
| 494 |
DO j = jfiltsu, jjm |
| 495 |
|
| 496 |
DO i = 1,iim |
| 497 |
coff = coefilu(i,j) / ( 1. + coefilu(i,j) ) |
| 498 |
IF( i.LT.modfrstu(j) ) coff = 0. |
| 499 |
DO k=1,iim |
| 500 |
eignft(i,k) = eignfnv(k,i) * coff |
| 501 |
ENDDO |
| 502 |
ENDDO |
| 503 |
DO k = 1, iim |
| 504 |
DO i = 1, iim |
| 505 |
matrinvs(i,k,j-jfiltsu+1) = 0.0 |
| 506 |
DO ii = 1, iim |
| 507 |
matrinvs(i,k,j-jfiltsu+1) = matrinvs(i,k,j-jfiltsu+1) |
| 508 |
. + eignfnv(i,ii)*eignft(ii,k) |
| 509 |
ENDDO |
| 510 |
ENDDO |
| 511 |
ENDDO |
| 512 |
|
| 513 |
ENDDO |
| 514 |
|
| 515 |
c ................................................................... |
| 516 |
|
| 517 |
c |
| 518 |
334 FORMAT(1x,24i3) |
| 519 |
755 FORMAT(1x,6f10.3,i3) |
| 520 |
|
| 521 |
RETURN |
| 522 |
END |
Parent Directory
| 
Revision Log