Changeset 12586
- Timestamp:
- 2020-03-23T13:14:40+01:00 (5 years ago)
- Location:
- NEMO/branches/2020/dev_r12558_HPC-08_epico_Extra_Halo/src
- Files:
-
- 1 added
- 21 edited
-
OCE/DOM/dom_oce.F90 (modified) (3 diffs)
-
OCE/LBC/halo_mng.F90 (added)
-
OCE/LBC/lbc_lnk_multi_generic.h90 (modified) (3 diffs)
-
OCE/LBC/lbc_nfd_nogather_generic.h90 (modified) (10 diffs)
-
OCE/LBC/lbcnfd.F90 (modified) (1 diff)
-
OCE/LBC/mpp_lnk_generic.h90 (modified) (14 diffs)
-
OCE/LBC/mpp_nfd_generic.h90 (modified) (13 diffs)
-
OCE/LBC/mppini.F90 (modified) (3 diffs)
-
OCE/SBC/sbcrnf.F90 (modified) (1 diff)
-
OCE/TRA/traadv.F90 (modified) (2 diffs)
-
OCE/TRA/traadv_mus.F90 (modified) (15 diffs)
-
OCE/ZDF/zdfdrg.F90 (modified) (1 diff)
-
OCE/ZDF/zdfphy.F90 (modified) (1 diff)
-
OCE/do_loop_substitute.h90 (modified) (3 diffs)
-
OCE/nemogcm.F90 (modified) (4 diffs)
-
OCE/oce.F90 (modified) (1 diff)
-
OCE/par_oce.F90 (modified) (2 diffs)
-
OFF/nemogcm.F90 (modified) (3 diffs)
-
SAO/nemogcm.F90 (modified) (3 diffs)
-
SAS/nemogcm.F90 (modified) (4 diffs)
-
TOP/TRP/trcadv.F90 (modified) (2 diffs)
-
TOP/trc.F90 (modified) (1 diff)
Legend:
- Unmodified
- Added
- Removed
-
NEMO/branches/2020/dev_r12558_HPC-08_epico_Extra_Halo/src/OCE/DOM/dom_oce.F90
r12489 r12586 114 114 REAL(wp), PUBLIC, ALLOCATABLE, SAVE, TARGET, DIMENSION(:,:) :: e1f , e2f , r1_e1f, r1_e2f !: horizontal scale factors at f-point [m] 115 115 ! 116 REAL(wp), PUBLIC, ALLOCATABLE, SAVE , DIMENSION(:,:) :: e1e2t , r1_e1e2t !: associated metrics at t-point 117 REAL(wp), PUBLIC, ALLOCATABLE, SAVE , DIMENSION(:,:) :: e1e2u , r1_e1e2u , e2_e1u !: associated metrics at u-point 118 REAL(wp), PUBLIC, ALLOCATABLE, SAVE , DIMENSION(:,:) :: e1e2v , r1_e1e2v , e1_e2v !: associated metrics at v-point 116 REAL(wp), PUBLIC, POINTER, SAVE , DIMENSION(:,:) :: r1_e1e2t, r1_e1e2u, r1_e1e2v !: associated metrics at t-point 117 REAL(wp), PUBLIC, ALLOCATABLE, SAVE , DIMENSION(:,:) :: e1e2t !: associated metrics at t-point 118 REAL(wp), PUBLIC, ALLOCATABLE, SAVE , DIMENSION(:,:) :: e1e2u , e2_e1u !: associated metrics at u-point 119 REAL(wp), PUBLIC, ALLOCATABLE, SAVE , DIMENSION(:,:) :: e1e2v , e1_e2v !: associated metrics at v-point 119 120 REAL(wp), PUBLIC, ALLOCATABLE, SAVE , DIMENSION(:,:) :: e1e2f , r1_e1e2f !: associated metrics at f-point 120 121 ! … … 136 137 REAL(wp), PUBLIC, ALLOCATABLE, SAVE, DIMENSION(:,:,:) :: e3vw_0 !: vw-vert. scale factor [m] 137 138 ! ! time-dependent scale factors 138 REAL(wp), PUBLIC, ALLOCATABLE, SAVE, DIMENSION(:,:,:,:) :: e3t, e3u, e3v, e3w, e3uw, e3vw !: vert. scale factor [m] 139 REAL(wp), PUBLIC, POINTER, SAVE, DIMENSION(:,:,:,:) :: e3t, e3u, e3v, e3w !: vert. scale factor [m] 140 REAL(wp), PUBLIC, ALLOCATABLE, SAVE, DIMENSION(:,:,:,:) :: e3uw, e3vw !: vert. scale factor [m] 139 141 REAL(wp), PUBLIC, ALLOCATABLE, SAVE, DIMENSION(:,:,:) :: e3f !: F-point vert. scale factor [m] 140 142 … … 176 178 REAL(wp), PUBLIC, ALLOCATABLE, SAVE, DIMENSION(:,:) :: tmask_h !: internal domain T-point mask (Figure 8.5 NEMO book) 177 179 178 INTEGER , PUBLIC, ALLOCATABLE, SAVE, DIMENSION(:,:) :: mikt, miku, mikv, mikf !: top first wet T-, U-, V-, F-level (ISF) 180 INTEGER , PUBLIC, ALLOCATABLE, SAVE, DIMENSION(:,:) :: miku, mikv, mikf !: top first wet U-, V-, F-level (ISF) 181 REAL(wp), PUBLIC, POINTER, SAVE, DIMENSION(:,:) :: mikt !: top first wet T-level (ISF) 179 182 180 183 REAL(wp), PUBLIC, ALLOCATABLE, SAVE, DIMENSION(:,:) :: ssmask, ssumask, ssvmask !: surface mask at T-,U-, V- and F-pts 181 REAL(wp), PUBLIC, ALLOCATABLE, SAVE, DIMENSION(:,:,:), TARGET :: tmask, umask, vmask, fmask !: land/ocean mask at T-, U-, V- and F-pts 182 REAL(wp), PUBLIC, ALLOCATABLE, SAVE, DIMENSION(:,:,:), TARGET :: wmask, wumask, wvmask !: land/ocean mask at WT-, WU- and WV-pts 184 REAL(wp), PUBLIC, ALLOCATABLE, SAVE, DIMENSION(:,:,:), TARGET :: fmask !: land/ocean mask at F-pts 185 REAL(wp), PUBLIC, POINTER, SAVE, DIMENSION(:,:,:) :: tmask, umask, vmask, wmask !: land/ocean mask at T-, U-, V-pts 186 REAL(wp), PUBLIC, ALLOCATABLE, SAVE, DIMENSION(:,:,:), TARGET :: wumask, wvmask !: land/ocean mask at WT-, WU- and WV-pts 183 187 184 188 REAL(wp), PUBLIC, ALLOCATABLE, SAVE, DIMENSION(:) :: tpol, fpol !: north fold mask (jperio= 3 or 4) -
NEMO/branches/2020/dev_r12558_HPC-08_epico_Extra_Halo/src/OCE/LBC/lbc_lnk_multi_generic.h90
r11536 r12586 19 19 & , pt5, cdna5, psgn5, pt6 , cdna6 , psgn6 , pt7 , cdna7 , psgn7 , pt8, cdna8, psgn8 & 20 20 & , pt9, cdna9, psgn9, pt10, cdna10, psgn10, pt11, cdna11, psgn11 & 21 & , kfillmode, pfillval, lsend, lrecv , ihlcom)21 & , kfillmode, pfillval, lsend, lrecv ) 22 22 !!--------------------------------------------------------------------- 23 23 CHARACTER(len=*) , INTENT(in ) :: cdname ! name of the calling subroutine … … 31 31 REAL(wp) , OPTIONAL , INTENT(in ) :: pfillval ! background value (used at closed boundaries) 32 32 LOGICAL, DIMENSION(4), OPTIONAL , INTENT(in ) :: lsend, lrecv ! indicate how communications are to be carried out 33 INTEGER , OPTIONAL , INTENT(in ) :: ihlcom ! number of ranks and rows to be communicated34 33 !! 35 34 INTEGER :: kfld ! number of elements that will be attributed … … 56 55 IF( PRESENT(psgn11) ) CALL ROUTINE_LOAD( pt11, cdna11, psgn11, ptab_ptr, cdna_ptr, psgn_ptr, kfld ) 57 56 ! 58 CALL lbc_lnk_ptr ( cdname, ptab_ptr, cdna_ptr, psgn_ptr, kfld, kfillmode, pfillval, lsend, lrecv , ihlcom)57 CALL lbc_lnk_ptr ( cdname, ptab_ptr, cdna_ptr, psgn_ptr, kfld, kfillmode, pfillval, lsend, lrecv ) 59 58 ! 60 59 END SUBROUTINE ROUTINE_MULTI -
NEMO/branches/2020/dev_r12558_HPC-08_epico_Extra_Halo/src/OCE/LBC/lbc_nfd_nogather_generic.h90
r11536 r12586 44 44 # define L_SIZE(ptab) SIZE(ptab,4) 45 45 # endif 46 # define ARRAY2_IN(i,j,k,l,f) ptab2(i,j,k,l)47 46 # define J_SIZE(ptab2) SIZE(ptab2,2) 47 # define ARRAY2_IN(i,j,k,l,f) ptab2(i,j,k,l) 48 48 # define ARRAY_TYPE(i,j,k,l,f) REAL(wp),INTENT(inout)::ARRAY_IN(i,j,k,l,f) 49 49 # define ARRAY2_TYPE(i,j,k,l,f) REAL(wp),INTENT(inout)::ARRAY2_IN(i,j,k,l,f) … … 57 57 !! 58 58 !!---------------------------------------------------------------------- 59 ARRAY_TYPE( :,:,:,:,:) ! array or pointer of arrays on which the boundary condition is applied60 ARRAY2_TYPE( :,:,:,:,:) ! array or pointer of arrays on which the boundary condition is applied59 ARRAY_TYPE(1-nn_hls+1:,1-nn_hls+1:,:,:,:) 60 ARRAY2_TYPE(1-nn_hls+1:,:,:,:,:) 61 61 CHARACTER(len=1) , INTENT(in ) :: NAT_IN(:) ! nature of array grid-points 62 62 REAL(wp) , INTENT(in ) :: SGN_IN(:) ! sign used across the north fold boundary 63 63 INTEGER, OPTIONAL, INTENT(in ) :: kfld ! number of pt3d arrays 64 64 ! 65 INTEGER :: ji, jj, jk, 66 INTEGER :: ipi, ipj, ipk, ipl, ipf 67 INTEGER :: ijt, iju, ij pj, ijpjp1, ijta, ijua, jia, startloop, endloop65 INTEGER :: ji, jj, jk, jn, ii, jl, jh, jf ! dummy loop indices 66 INTEGER :: ipi, ipj, ipk, ipl, ipf, ijj ! dimension of the input array 67 INTEGER :: ijt, iju, ijta, ijua, jia, startloop, endloop 68 68 LOGICAL :: l_fast_exchanges 69 69 !!---------------------------------------------------------------------- … … 75 75 ! Security check for further developments 76 76 IF ( ipf > 1 ) CALL ctl_stop( 'STOP', 'lbc_nfd_nogather: multiple fields not allowed. Revise implementation...' ) 77 !78 ijpj = 1 ! index of first modified line79 ijpjp1 = 2 ! index + 180 81 77 ! 2nd dimension determines exchange speed 82 78 IF (ipj == 1 ) THEN … … 95 91 ! 96 92 CASE ( 'T' , 'W' ) ! T-, W-point 97 IF ( nimpp /= 1 ) THEN ; startloop =193 IF ( nimpp - nn_hls+1 /= 1 ) THEN ; startloop = 1 - nn_hls + 1 98 94 ELSE ; startloop = 2 99 95 ENDIF 100 96 ! 101 97 DO jl = 1, ipl; DO jk = 1, ipk 102 DO ji = startloop, nlci 103 ijt = jpiglo - ji - nimpp - nfiimpp(isendto(1),jpnj) + 4 104 ARRAY_IN(ji,nlcj,jk,jl,jf) = SGN_IN(jf) * ARRAY2_IN(ijt,ijpj,jk,jl,jf) 105 END DO 106 END DO; END DO 107 IF( nimpp == 1 ) THEN 108 DO jl = 1, ipl; DO jk = 1, ipk 109 ARRAY_IN(1,nlcj,jk,jl,jf) = SGN_IN(jf) * ARRAY_IN(3,nlcj-2,jk,jl,jf) 110 END DO; END DO 111 ENDIF 112 ! 113 IF ( .NOT. l_fast_exchanges ) THEN 114 IF( nimpp >= jpiglo/2+1 ) THEN 115 startloop = 1 116 ELSEIF( nimpp+nlci-1 >= jpiglo/2+1 .AND. nimpp < jpiglo/2+1 ) THEN 117 startloop = jpiglo/2+1 - nimpp + 1 98 DO jj = 1, nn_hls 99 ijj = nlcj -jj +1 100 DO ji = startloop, nlci 101 ijt = jpiglo - ji - nimpp - nfiimpp(isendto(1),jpnj) + 4 102 ARRAY_IN(ji,ijj,jk,jl,jf) = SGN_IN(jf) * ARRAY2_IN(ijt,jj,jk,jl,jf) 103 END DO 104 END DO 105 END DO; END DO 106 IF( nimpp - nn_hls+1 == 1 ) THEN 107 DO jl = 1, ipl; DO jk = 1, ipk 108 DO jj = 1, nn_hls 109 ijj = nlcj -jj +1 110 DO ii = 1, nn_hls 111 ARRAY_IN(1-ii+1,ijj,jk,jl,jf) = SGN_IN(jf) * ARRAY_IN(ii+2,nlcj-2*nn_hls+jj-1,jk,jl,jf) 112 END DO 113 END DO 114 END DO; END DO 115 ENDIF 116 ! 117 IF ( .NOT. l_fast_exchanges ) THEN 118 IF( nimpp - nn_hls +1 >= jpiglo/2+1 ) THEN 119 startloop = 1 - nn_hls +1 120 ELSEIF( nimpp+nlci-1 >= jpiglo/2+1 .AND. nimpp - nn_hls +1 < jpiglo/2+1 ) THEN 121 startloop = jpiglo/2+1 - nimpp + nn_hls 118 122 ELSE 119 123 startloop = nlci + 1 … … 126 130 ijta = jpiglo - jia + 2 127 131 IF( ijta >= startloop+nimpp-1 .AND. ijta < jia ) THEN 128 ARRAY_IN(ji,nlcj- 1,jk,jl,jf) = SGN_IN(jf) * ARRAY_IN(ijta-nimpp+1,nlcj-1,jk,jl,jf)132 ARRAY_IN(ji,nlcj-nn_hls,jk,jl,jf) = SGN_IN(jf) * ARRAY_IN(ijta-nimpp+1,nlcj-nn_hls,jk,jl,jf) 129 133 ELSE 130 ARRAY_IN(ji,nlcj- 1,jk,jl,jf) = SGN_IN(jf) * ARRAY2_IN(ijt,ijpjp1,jk,jl,jf)134 ARRAY_IN(ji,nlcj-nn_hls,jk,jl,jf) = SGN_IN(jf) * ARRAY2_IN(ijt,nn_hls+1,jk,jl,jf) 131 135 ENDIF 132 136 END DO … … 134 138 ENDIF 135 139 ENDIF 136 140 CASE ( 'U' ) ! U-point 141 IF( nimpp + nlci - nn_hls /= jpiglo ) THEN 142 endloop = nlci 143 ELSE 144 endloop = nlci - nn_hls 145 ENDIF 146 DO jl = 1, ipl; DO jk = 1, ipk 147 DO jj = 1, nn_hls 148 ijj = nlcj -jj +1 149 DO ji = 1-nn_hls+1, endloop 150 iju = jpiglo - ji - nimpp - nfiimpp(isendto(1),jpnj) + 3 151 ARRAY_IN(ji,ijj,jk,jl,jf) = SGN_IN(jf) * ARRAY2_IN(iju,jj,jk,jl,jf) 152 END DO 153 END DO 154 END DO; END DO 155 IF (nimpp - nn_hls+1 .eq. 1) THEN 156 DO jj = 1, nn_hls 157 ijj = nlcj -jj +1 158 DO ii = 1, nn_hls 159 ARRAY_IN(2-ii,ijj,:,:,jf) = SGN_IN(jf) * ARRAY_IN(ii+1,nlcj-2*nn_hls+jj-1,:,:,jf) 160 END DO 161 END DO 162 ENDIF 163 IF((nimpp + nlci - nn_hls) .eq. jpiglo) THEN 164 DO jj = 1, nn_hls 165 ijj = nlcj -jj +1 166 DO ii = 1, nn_hls 167 ARRAY_IN(nlci-ii+1,ijj,:,:,jf) = SGN_IN(jf) * ARRAY_IN(nlci-2*nn_hls+ii,nlcj-2*nn_hls+jj-1,:,:,jf) 168 END DO 169 END DO 170 ENDIF 171 ! 172 IF ( .NOT. l_fast_exchanges ) THEN 173 IF( nimpp + nlci - nn_hls /= jpiglo ) THEN 174 endloop = nlci 175 ELSE 176 endloop = nlci - nn_hls 177 ENDIF 178 IF( nimpp - nn_hls+1 >= jpiglo/2 ) THEN 179 startloop = 1- nn_hls + 1 180 ELSEIF( ( nimpp+nlci-1 >= jpiglo/2 ) .AND. ( nimpp - nn_hls+1 < jpiglo/2 ) ) THEN 181 startloop = jpiglo/2 - nimpp + nn_hls 182 ELSE 183 startloop = endloop + 1 184 ENDIF 185 IF( startloop <= endloop ) THEN 186 DO jl = 1, ipl; DO jk = 1, ipk 187 DO ji = startloop, endloop 188 iju = jpiglo - ji - nimpp - nfiimpp(isendto(1),jpnj) + 3 189 jia = ji + nimpp - 1 190 ijua = jpiglo - jia + 1 191 IF( ijua >= startloop+nimpp-1 .AND. ijua < jia ) THEN 192 ARRAY_IN(ji,nlcj-nn_hls,jk,jl,jf) = SGN_IN(jf) * ARRAY_IN(ijua-nimpp+1,nlcj-nn_hls,jk,jl,jf) 193 ELSE 194 ARRAY_IN(ji,nlcj-nn_hls,jk,jl,jf) = SGN_IN(jf) * ARRAY2_IN(iju,nn_hls+1,jk,jl,jf) 195 ENDIF 196 END DO 197 END DO; END DO 198 ENDIF 199 ENDIF 200 ! 201 CASE ( 'V' ) ! V-point 202 IF( nimpp - nn_hls+1 /= 1 ) THEN 203 startloop = 1 - nn_hls + 1 204 ELSE 205 startloop = 2 206 ENDIF 207 IF ( .NOT. l_fast_exchanges ) THEN 208 DO jl = 1, ipl; DO jk = 1, ipk 209 DO jj = 2, nn_hls+1 210 ijj = nlcj -jj +1 211 DO ji = startloop, nlci 212 ijt=jpiglo - ji - nimpp - nfiimpp(isendto(1),jpnj) + 4 213 ARRAY_IN(ji,ijj,jk,jl,jf) = SGN_IN(jf) * ARRAY2_IN(ijt,jj,jk,jl,jf) 214 END DO 215 END DO 216 END DO; END DO 217 ENDIF 218 DO jl = 1, ipl; DO jk = 1, ipk 219 DO ji = startloop, nlci 220 ijt=jpiglo - ji - nimpp - nfiimpp(isendto(1),jpnj) + 4 221 ARRAY_IN(ji,nlcj,jk,jl,jf) = SGN_IN(jf) * ARRAY2_IN(ijt,1,jk,jl,jf) 222 END DO 223 END DO; END DO 224 IF (nimpp - nn_hls+1.eq. 1) THEN 225 DO jj = 1, nn_hls 226 ijj = nlcj-jj+1 227 DO ii = 1, nn_hls 228 ARRAY_IN(1-ii+1,ijj,:,:,jf) = SGN_IN(jf) * ARRAY_IN(ii+2,nlcj-2*nn_hls+jj-1,:,:,jf) 229 END DO 230 END DO 231 ENDIF 232 CASE ( 'F' ) ! F-point 233 IF( nimpp + nlci - nn_hls /= jpiglo ) THEN 234 endloop = nlci 235 ELSE 236 endloop = nlci - nn_hls 237 ENDIF 238 IF ( .NOT. l_fast_exchanges ) THEN 239 DO jl = 1, ipl; DO jk = 1, ipk 240 DO jj = 2, nn_hls+1 241 ijj = nlcj -jj +1 242 DO ji = 1 - nn_hls +1, endloop 243 iju = jpiglo - ji - nimpp - nfiimpp(isendto(1),jpnj) + 3 244 ARRAY_IN(ji,ijj,jk,jl,jf) = SGN_IN(jf) * ARRAY2_IN(iju,jj,jk,jl,jf) 245 END DO 246 END DO 247 END DO; END DO 248 ENDIF 249 DO jl = 1, ipl; DO jk = 1, ipk 250 DO ji = 1- nn_hls +1, endloop 251 iju = jpiglo - ji - nimpp - nfiimpp(isendto(1),jpnj) + 3 252 ARRAY_IN(ji,nlcj,jk,jl,jf) = SGN_IN(jf) * ARRAY2_IN(iju,1,jk,jl,jf) 253 END DO 254 END DO; END DO 255 IF (nimpp - nn_hls+1.eq. 1) THEN 256 DO ii = 1, nn_hls 257 ARRAY_IN(2-ii,nlcj,:,:,jf) = SGN_IN(jf) * ARRAY_IN(ii+1,nlcj-2*nn_hls-1,:,:,jf) 258 END DO 259 IF ( .NOT. l_fast_exchanges ) THEN 260 DO jj = 1, nn_hls 261 ijj = nlcj -jj 262 DO ii = 1, nn_hls 263 ARRAY_IN(2-ii,ijj,:,:,jf) = SGN_IN(jf) * ARRAY_IN(ii+1,nlcj-2*nn_hls+jj-1,:,:,jf) 264 END DO 265 END DO 266 ENDIF 267 ENDIF 268 IF((nimpp + nlci - nn_hls ) .eq. jpiglo) THEN 269 DO ii = 1, nn_hls 270 ARRAY_IN(nlci-ii+1,nlcj,:,:,jf) = SGN_IN(jf) * ARRAY_IN(nlci-2*nn_hls+ii,nlcj-2*nn_hls-1,:,:,jf) 271 END DO 272 IF ( .NOT. l_fast_exchanges ) THEN 273 DO jj = 1, nn_hls 274 ijj = nlcj -jj 275 DO ii = 1, nn_hls 276 ARRAY_IN(nlci-ii+1,ijj,:,:,jf) = SGN_IN(jf) * ARRAY_IN(nlci-2*nn_hls+ii,nlcj-2*nn_hls+jj-1,:,:,jf) 277 END DO 278 END DO 279 ENDIF 280 ENDIF 281 ! 282 END SELECT 283 ! 284 CASE ( 5, 6 ) ! * North fold F-point pivot 285 ! 286 WRITE(*,*) 'extrahalo not handled in this case', __FILE__, __LINE__ 287 SELECT CASE ( NAT_IN(jf) ) 288 CASE ( 'T' , 'W' ) ! T-, W-point 289 DO jl = 1, ipl; DO jk = 1, ipk 290 DO jj = 1, nn_hls 291 ijj = nlcj -jj+1 292 DO ji = 1, nlci 293 ijt = jpiglo - ji - nimpp - nfiimpp(isendto(1),jpnj) + 3 294 ARRAY_IN(ji,ijj,jk,jl,jf) = SGN_IN(jf) * ARRAY2_IN(ijt,jj,jk,jl,jf) 295 END DO 296 END DO 297 END DO; END DO 298 ! 137 299 CASE ( 'U' ) ! U-point 138 300 IF( nimpp + nlci - 1 /= jpiglo ) THEN … … 143 305 DO jl = 1, ipl; DO jk = 1, ipk 144 306 DO ji = 1, endloop 145 iju = jpiglo - ji - nimpp - nfiimpp(isendto(1),jpnj) + 3 146 ARRAY_IN(ji,nlcj,jk,jl,jf) = SGN_IN(jf) * ARRAY2_IN(iju,ijpj,jk,jl,jf) 147 END DO 148 END DO; END DO 149 IF (nimpp .eq. 1) THEN 150 ARRAY_IN(1,nlcj,:,:,jf) = SGN_IN(jf) * ARRAY_IN(2,nlcj-2,:,:,jf) 151 ENDIF 307 iju = jpiglo - ji - nimpp - nfiimpp(isendto(1),jpnj) + 2 308 ARRAY_IN(ji,nlcj,jk,jl,jf) = SGN_IN(jf) * ARRAY2_IN(iju,ijpj-1,jk,jl,jf) 309 END DO 310 END DO; END DO 152 311 IF((nimpp + nlci - 1) .eq. jpiglo) THEN 153 ARRAY_IN(nlci,nlcj,:,:,jf) = SGN_IN(jf) * ARRAY_IN(nlci-1,nlcj-2,:,:,jf) 154 ENDIF 155 ! 156 IF ( .NOT. l_fast_exchanges ) THEN 157 IF( nimpp + nlci - 1 /= jpiglo ) THEN 158 endloop = nlci 159 ELSE 160 endloop = nlci - 1 161 ENDIF 162 IF( nimpp >= jpiglo/2 ) THEN 163 startloop = 1 164 ELSEIF( ( nimpp+nlci-1 >= jpiglo/2 ) .AND. ( nimpp < jpiglo/2 ) ) THEN 165 startloop = jpiglo/2 - nimpp + 1 166 ELSE 167 startloop = endloop + 1 168 ENDIF 169 IF( startloop <= endloop ) THEN 170 DO jl = 1, ipl; DO jk = 1, ipk 171 DO ji = startloop, endloop 172 iju = jpiglo - ji - nimpp - nfiimpp(isendto(1),jpnj) + 3 173 jia = ji + nimpp - 1 174 ijua = jpiglo - jia + 1 175 IF( ijua >= startloop+nimpp-1 .AND. ijua < jia ) THEN 176 ARRAY_IN(ji,nlcj-1,jk,jl,jf) = SGN_IN(jf) * ARRAY_IN(ijua-nimpp+1,nlcj-1,jk,jl,jf) 177 ELSE 178 ARRAY_IN(ji,nlcj-1,jk,jl,jf) = SGN_IN(jf) * ARRAY2_IN(iju,ijpjp1,jk,jl,jf) 179 ENDIF 180 END DO 181 END DO; END DO 182 ENDIF 312 DO jl = 1, ipl; DO jk = 1, ipk 313 ARRAY_IN(nlci,nlcj,jk,jl,jf) = SGN_IN(jf) * ARRAY_IN(nlci-2,nlcj-1,jk,jl,jf) 314 END DO; END DO 183 315 ENDIF 184 316 ! 185 317 CASE ( 'V' ) ! V-point 186 IF( nimpp /= 1 ) THEN187 startloop = 1188 ELSE189 startloop = 2190 ENDIF191 IF ( .NOT. l_fast_exchanges ) THEN192 DO jl = 1, ipl; DO jk = 1, ipk193 DO ji = startloop, nlci194 ijt=jpiglo - ji - nimpp - nfiimpp(isendto(1),jpnj) + 4195 ARRAY_IN(ji,nlcj-1,jk,jl,jf) = SGN_IN(jf) * ARRAY2_IN(ijt,ijpjp1,jk,jl,jf)196 END DO197 END DO; END DO198 ENDIF199 DO jl = 1, ipl; DO jk = 1, ipk200 DO ji = startloop, nlci201 ijt=jpiglo - ji - nimpp - nfiimpp(isendto(1),jpnj) + 4202 ARRAY_IN(ji,nlcj,jk,jl,jf) = SGN_IN(jf) * ARRAY2_IN(ijt,ijpj,jk,jl,jf)203 END DO204 END DO; END DO205 IF (nimpp .eq. 1) THEN206 ARRAY_IN(1,nlcj,:,:,jf) = SGN_IN(jf) * ARRAY_IN(3,nlcj-3,:,:,jf)207 ENDIF208 CASE ( 'F' ) ! F-point209 IF( nimpp + nlci - 1 /= jpiglo ) THEN210 endloop = nlci211 ELSE212 endloop = nlci - 1213 ENDIF214 IF ( .NOT. l_fast_exchanges ) THEN215 DO jl = 1, ipl; DO jk = 1, ipk216 DO ji = 1, endloop217 iju = jpiglo - ji - nimpp - nfiimpp(isendto(1),jpnj) + 3218 ARRAY_IN(ji,nlcj-1,jk,jl,jf) = SGN_IN(jf) * ARRAY2_IN(iju,ijpjp1,jk,jl,jf)219 END DO220 END DO; END DO221 ENDIF222 DO jl = 1, ipl; DO jk = 1, ipk223 DO ji = 1, endloop224 iju = jpiglo - ji - nimpp - nfiimpp(isendto(1),jpnj) + 3225 ARRAY_IN(ji,nlcj,jk,jl,jf) = SGN_IN(jf) * ARRAY2_IN(iju,ijpj,jk,jl,jf)226 END DO227 END DO; END DO228 IF (nimpp .eq. 1) THEN229 ARRAY_IN(1,nlcj,:,:,jf) = SGN_IN(jf) * ARRAY_IN(2,nlcj-3,:,:,jf)230 IF ( .NOT. l_fast_exchanges ) &231 ARRAY_IN(1,nlcj-1,:,:,jf) = SGN_IN(jf) * ARRAY_IN(2,nlcj-2,:,:,jf)232 ENDIF233 IF((nimpp + nlci - 1) .eq. jpiglo) THEN234 ARRAY_IN(nlci,nlcj,:,:,jf) = SGN_IN(jf) * ARRAY_IN(nlci-1,nlcj-3,:,:,jf)235 IF ( .NOT. l_fast_exchanges ) &236 ARRAY_IN(nlci,nlcj-1,:,:,jf) = SGN_IN(jf) * ARRAY_IN(nlci-1,nlcj-2,:,:,jf)237 ENDIF238 !239 END SELECT240 !241 CASE ( 5, 6 ) ! * North fold F-point pivot242 !243 SELECT CASE ( NAT_IN(jf) )244 CASE ( 'T' , 'W' ) ! T-, W-point245 318 DO jl = 1, ipl; DO jk = 1, ipk 246 319 DO ji = 1, nlci 247 320 ijt = jpiglo - ji - nimpp - nfiimpp(isendto(1),jpnj) + 3 248 ARRAY_IN(ji,nlcj,jk,jl,jf) = SGN_IN(jf) * ARRAY2_IN(ijt,ijpj,jk,jl,jf) 249 END DO 250 END DO; END DO 251 ! 252 CASE ( 'U' ) ! U-point 253 IF( nimpp + nlci - 1 /= jpiglo ) THEN 254 endloop = nlci 255 ELSE 256 endloop = nlci - 1 257 ENDIF 258 DO jl = 1, ipl; DO jk = 1, ipk 259 DO ji = 1, endloop 260 iju = jpiglo - ji - nimpp - nfiimpp(isendto(1),jpnj) + 2 261 ARRAY_IN(ji,nlcj,jk,jl,jf) = SGN_IN(jf) * ARRAY2_IN(iju,ijpj,jk,jl,jf) 262 END DO 263 END DO; END DO 264 IF((nimpp + nlci - 1) .eq. jpiglo) THEN 265 DO jl = 1, ipl; DO jk = 1, ipk 266 ARRAY_IN(nlci,nlcj,jk,jl,jf) = SGN_IN(jf) * ARRAY_IN(nlci-2,nlcj-1,jk,jl,jf) 267 END DO; END DO 268 ENDIF 269 ! 270 CASE ( 'V' ) ! V-point 271 DO jl = 1, ipl; DO jk = 1, ipk 272 DO ji = 1, nlci 273 ijt = jpiglo - ji - nimpp - nfiimpp(isendto(1),jpnj) + 3 274 ARRAY_IN(ji,nlcj,jk,jl,jf) = SGN_IN(jf) * ARRAY2_IN(ijt,ijpj,jk,jl,jf) 321 ARRAY_IN(ji,nlcj,jk,jl,jf) = SGN_IN(jf) * ARRAY2_IN(ijt,ijpj-1,jk,jl,jf) 275 322 END DO 276 323 END DO; END DO … … 288 335 DO ji = startloop, nlci 289 336 ijt = jpiglo - ji - nimpp - nfiimpp(isendto(1),jpnj) + 3 290 ARRAY_IN(ji,nlcj-1,jk,jl,jf) = SGN_IN(jf) * ARRAY2_IN(ijt,ijpj p1,jk,jl,jf)337 ARRAY_IN(ji,nlcj-1,jk,jl,jf) = SGN_IN(jf) * ARRAY2_IN(ijt,ijpj,jk,jl,jf) 291 338 END DO 292 339 END DO; END DO … … 303 350 DO ji = 1, endloop 304 351 iju = jpiglo - ji - nimpp - nfiimpp(isendto(1),jpnj) + 2 305 ARRAY_IN(ji,nlcj ,jk,jl,jf) = SGN_IN(jf) * ARRAY2_IN(iju,ijpj ,jk,jl,jf)352 ARRAY_IN(ji,nlcj ,jk,jl,jf) = SGN_IN(jf) * ARRAY2_IN(iju,ijpj-1,jk,jl,jf) 306 353 END DO 307 354 END DO; END DO … … 329 376 DO ji = startloop, endloop 330 377 iju = jpiglo - ji - nimpp - nfiimpp(isendto(1),jpnj) + 2 331 ARRAY_IN(ji,nlcj-1,jk,jl,jf) = SGN_IN(jf) * ARRAY2_IN(iju,ijpj p1,jk,jl,jf)378 ARRAY_IN(ji,nlcj-1,jk,jl,jf) = SGN_IN(jf) * ARRAY2_IN(iju,ijpj,jk,jl,jf) 332 379 END DO 333 380 END DO; END DO -
NEMO/branches/2020/dev_r12558_HPC-08_epico_Extra_Halo/src/OCE/LBC/lbcnfd.F90
r11536 r12586 53 53 54 54 INTEGER, PUBLIC, PARAMETER :: jpmaxngh = 3 !: 55 INTEGER, PUBLIC :: nsndto , nfsloop, nfeloop!:55 INTEGER, PUBLIC :: nsndto !: 56 56 INTEGER, PUBLIC, DIMENSION (jpmaxngh) :: isendto !: processes to which communicate 57 INTEGER, PUBLIC :: ijpj 57 58 58 59 !!---------------------------------------------------------------------- -
NEMO/branches/2020/dev_r12558_HPC-08_epico_Extra_Halo/src/OCE/LBC/mpp_lnk_generic.h90
r11536 r12586 46 46 47 47 #if defined MULTI 48 SUBROUTINE ROUTINE_LNK( cdname, ptab, cd_nat, psgn, kfld, kfillmode, pfillval, lsend, lrecv , ihlcom)48 SUBROUTINE ROUTINE_LNK( cdname, ptab, cd_nat, psgn, kfld, kfillmode, pfillval, lsend, lrecv ) 49 49 INTEGER , INTENT(in ) :: kfld ! number of pt3d arrays 50 50 #else 51 SUBROUTINE ROUTINE_LNK( cdname, ptab, cd_nat, psgn , kfillmode, pfillval, lsend, lrecv , ihlcom)51 SUBROUTINE ROUTINE_LNK( cdname, ptab, cd_nat, psgn , kfillmode, pfillval, lsend, lrecv ) 52 52 #endif 53 ARRAY_TYPE( :,:,:,:,:) ! array or pointer of arrays on which the boundary condition is applied53 ARRAY_TYPE(1-nn_hls+1:,1-nn_hls+1:,:,:,:) ! array or pointer of arrays on which the boundary condition is applied 54 54 CHARACTER(len=*) , INTENT(in ) :: cdname ! name of the calling subroutine 55 55 CHARACTER(len=1) , INTENT(in ) :: NAT_IN(:) ! nature of array grid-points … … 58 58 REAL(wp), OPTIONAL, INTENT(in ) :: pfillval ! background value (used at closed boundaries) 59 59 LOGICAL, DIMENSION(4),OPTIONAL, INTENT(in ) :: lsend, lrecv ! communication with other 4 proc 60 INTEGER ,OPTIONAL, INTENT(in ) :: ihlcom ! number of ranks and rows to be communicated61 60 ! 62 61 INTEGER :: ji, jj, jk, jl, jf ! dummy loop indices … … 66 65 INTEGER :: ierr 67 66 INTEGER :: ifill_we, ifill_ea, ifill_so, ifill_no 68 INTEGER :: ihl ! number of ranks and rows to be communicated69 67 REAL(wp) :: zland 70 68 INTEGER , DIMENSION(MPI_STATUS_SIZE) :: istat ! for mpi_isend … … 83 81 ipl = L_SIZE(ptab) ! 4th - 84 82 ipf = F_SIZE(ptab) ! 5th - use in "multi" case (array of pointers) 85 !86 IF( PRESENT(ihlcom) ) THEN ; ihl = ihlcom87 ELSE ; ihl = 188 END IF89 83 ! 90 84 IF( narea == 1 .AND. numcom == -1 ) CALL mpp_report( cdname, ipk, ipl, ipf, ld_lbc = .TRUE. ) … … 148 142 ! -------------------------------------------------- ! 149 143 ! 144 150 145 ! Must exchange the whole column (from 1 to jpj) to get the corners if we have no south/north neighbourg 151 isize = ihl * jpj* ipk * ipl * ipf146 isize = nn_hls * ( jpj + nn_hls - 1 ) * ipk * ipl * ipf 152 147 ! 153 148 ! Allocate local temporary arrays to be sent/received. Fill arrays to be sent 154 IF( llsend_we ) ALLOCATE( zsnd_we( ihl,jpj,ipk,ipl,ipf) )155 IF( llsend_ea ) ALLOCATE( zsnd_ea( ihl,jpj,ipk,ipl,ipf) )156 IF( llrecv_we ) ALLOCATE( zrcv_we( ihl,jpj,ipk,ipl,ipf) )157 IF( llrecv_ea ) ALLOCATE( zrcv_ea( ihl,jpj,ipk,ipl,ipf) )149 IF( llsend_we ) ALLOCATE( zsnd_we(nn_hls,1-nn_hls+1:jpj,ipk,ipl,ipf) ) 150 IF( llsend_ea ) ALLOCATE( zsnd_ea(nn_hls,1-nn_hls+1:jpj,ipk,ipl,ipf) ) 151 IF( llrecv_we ) ALLOCATE( zrcv_we(nn_hls,1-nn_hls+1:jpj,ipk,ipl,ipf) ) 152 IF( llrecv_ea ) ALLOCATE( zrcv_ea(nn_hls,1-nn_hls+1:jpj,ipk,ipl,ipf) ) 158 153 ! 159 154 IF( llsend_we ) THEN ! copy western side of the inner mpi domain in local temporary array to be sent by MPI 160 ishift = ihl161 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1 , jpj ; DO ji = 1, ihl162 zsnd_we(ji,jj,jk,jl,jf) = ARRAY_IN(ishift+ji,jj,jk,jl,jf) ! ihl + 1 -> 2*ihl163 END DO ; END DO ; END DO ; END DO ; END DO 164 ENDIF 165 ! 166 IF( llsend_ea) THEN ! copy eastern side of the inner mpi domain in local temporary array to be sent by MPI167 ishift = jpi - 2 * ihl168 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1 , jpj ; DO ji = 1, ihl169 zsnd_ea(ji,jj,jk,jl,jf) = ARRAY_IN(ishift+ji,jj,jk,jl,jf) ! jpi - 2* ihl + 1 -> jpi - ihl155 ishift = 1 156 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1-nn_hls+1, jpj ; DO ji = 1, nn_hls 157 zsnd_we(ji,jj,jk,jl,jf) = ARRAY_IN(ishift+ji,jj,jk,jl,jf) ! nn_hls + 1 -> 2*nn_hls 158 END DO ; END DO ; END DO ; END DO ; END DO 159 ENDIF 160 ! 161 IF( llsend_ea ) THEN ! copy eastern side of the inner mpi domain in local temporary array to be sent by MPI 162 ishift = jpi - 2 * nn_hls 163 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1-nn_hls+1, jpj ; DO ji = 1, nn_hls 164 zsnd_ea(ji,jj,jk,jl,jf) = ARRAY_IN(ishift+ji,jj,jk,jl,jf) ! jpi - 2*nn_hls + 1 -> jpi - nn_hls 170 165 END DO ; END DO ; END DO ; END DO ; END DO 171 166 ENDIF … … 174 169 ! 175 170 ! non-blocking send of the western/eastern side using local temporary arrays 176 IF( llsend_we ) CALL mppsend( 1, zsnd_we(1,1 ,1,1,1), isize, nowe, ireq_we )177 IF( llsend_ea ) CALL mppsend( 2, zsnd_ea(1,1 ,1,1,1), isize, noea, ireq_ea )171 IF( llsend_we ) CALL mppsend( 1, zsnd_we(1,1-nn_hls+1,1,1,1), isize, nowe, ireq_we ) 172 IF( llsend_ea ) CALL mppsend( 2, zsnd_ea(1,1-nn_hls+1,1,1,1), isize, noea, ireq_ea ) 178 173 ! blocking receive of the western/eastern halo in local temporary arrays 179 IF( llrecv_we ) CALL mpprecv( 2, zrcv_we(1,1 ,1,1,1), isize, nowe )180 IF( llrecv_ea ) CALL mpprecv( 1, zrcv_ea(1,1 ,1,1,1), isize, noea )174 IF( llrecv_we ) CALL mpprecv( 2, zrcv_we(1,1-nn_hls+1,1,1,1), isize, nowe ) 175 IF( llrecv_ea ) CALL mpprecv( 1, zrcv_ea(1,1-nn_hls+1,1,1,1), isize, noea ) 181 176 ! 182 177 IF( ln_timing ) CALL tic_tac(.FALSE.) … … 189 184 ! 2.1 fill weastern halo 190 185 ! ---------------------- 191 ! ishift = 0 ! fill halo from ji = 1 to ihl186 ! ishift = 0 ! fill halo from ji = 1 to nn_hls 192 187 SELECT CASE ( ifill_we ) 193 188 CASE ( jpfillnothing ) ! no filling 194 189 CASE ( jpfillmpi ) ! use data received by MPI 195 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1 , jpj ; DO ji = 1, ihl196 ARRAY_IN(ji ,jj,jk,jl,jf) = zrcv_we(ji,jj,jk,jl,jf) ! 1 -> ihl190 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1-nn_hls+1, jpj ; DO ji = 1, nn_hls 191 ARRAY_IN(ji-nn_hls+1,jj,jk,jl,jf) = zrcv_we(ji,jj,jk,jl,jf) ! 1 -> nn_hls 197 192 END DO; END DO ; END DO ; END DO ; END DO 198 193 CASE ( jpfillperio ) ! use east-weast periodicity 199 ishift2 = jpi - 2 * ihl200 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1 , jpj ; DO ji = 1, ihl201 ARRAY_IN(ji ,jj,jk,jl,jf) = ARRAY_IN(ishift2+ji,jj,jk,jl,jf)194 ishift2 = jpi - 2 * nn_hls 195 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1-nn_hls+1, jpj ; DO ji = 1, nn_hls 196 ARRAY_IN(ji-nn_hls+1,jj,jk,jl,jf) = ARRAY_IN(ishift2+ji,jj,jk,jl,jf) 202 197 END DO; END DO ; END DO ; END DO ; END DO 203 198 CASE ( jpfillcopy ) ! filling with inner domain values 204 199 DO jf = 1, ipf ! number of arrays to be treated 205 200 IF( .NOT. NAT_IN(jf) == 'F' ) THEN ! do nothing for F point 206 DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1 , jpj ; DO ji = 1, ihl207 ARRAY_IN(ji ,jj,jk,jl,jf) = ARRAY_IN(ihl+1,jj,jk,jl,jf)201 DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1-nn_hls+1, jpj ; DO ji = 1, nn_hls 202 ARRAY_IN(ji-nn_hls+1,jj,jk,jl,jf) = ARRAY_IN(1+ji,jj,jk,jl,jf) 208 203 END DO ; END DO ; END DO ; END DO 209 204 ENDIF … … 212 207 DO jf = 1, ipf ! number of arrays to be treated 213 208 IF( .NOT. NAT_IN(jf) == 'F' ) THEN ! do nothing for F point 214 DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1 , jpj ; DO ji = 1, ihl215 ARRAY_IN(ji ,jj,jk,jl,jf) = zland209 DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1-nn_hls+1, jpj ; DO ji = 1, nn_hls 210 ARRAY_IN(ji-nn_hls+1,jj,jk,jl,jf) = zland 216 211 END DO; END DO ; END DO ; END DO 217 212 ENDIF … … 221 216 ! 2.2 fill eastern halo 222 217 ! --------------------- 223 ishift = jpi - ihl ! fill halo from ji = jpi-ihl+1 to jpi218 ishift = jpi - nn_hls ! fill halo from ji = jpi-nn_hls+1 to jpi 224 219 SELECT CASE ( ifill_ea ) 225 220 CASE ( jpfillnothing ) ! no filling 226 221 CASE ( jpfillmpi ) ! use data received by MPI 227 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1 , jpj ; DO ji = 1, ihl228 ARRAY_IN(ishift+ji,jj,jk,jl,jf) = zrcv_ea(ji,jj,jk,jl,jf) ! jpi - ihl+ 1 -> jpi222 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1-nn_hls+1, jpj ; DO ji = 1, nn_hls 223 ARRAY_IN(ishift+ji,jj,jk,jl,jf) = zrcv_ea(ji,jj,jk,jl,jf) ! jpi - nn_hls + 1 -> jpi 229 224 END DO ; END DO ; END DO ; END DO ; END DO 230 225 CASE ( jpfillperio ) ! use east-weast periodicity 231 ishift2 = ihl232 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1 , jpj ; DO ji = 1, ihl226 ishift2 = 1 227 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1-nn_hls+1, jpj ; DO ji = 1, nn_hls 233 228 ARRAY_IN(ishift+ji,jj,jk,jl,jf) = ARRAY_IN(ishift2+ji,jj,jk,jl,jf) 234 229 END DO ; END DO ; END DO ; END DO ; END DO 235 230 CASE ( jpfillcopy ) ! filling with inner domain values 236 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1, jpj ; DO ji = 1, ihl 237 ARRAY_IN(ishift+ji,jj,jk,jl,jf) = ARRAY_IN(ishift,jj,jk,jl,jf) 231 ishift2 = jpi - 2*nn_hls 232 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1-nn_hls+1, jpj ; DO ji = 1, nn_hls 233 ARRAY_IN(ishift+ji,jj,jk,jl,jf) = ARRAY_IN(ishift2+ji,jj,jk,jl,jf) 238 234 END DO ; END DO ; END DO ; END DO ; END DO 239 235 CASE ( jpfillcst ) ! filling with constant value 240 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1 , jpj ; DO ji = 1, ihl236 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1-nn_hls+1, jpj ; DO ji = 1, nn_hls 241 237 ARRAY_IN(ishift+ji,jj,jk,jl,jf) = zland 242 238 END DO; END DO ; END DO ; END DO ; END DO … … 264 260 ! ---------------------------------------------------- ! 265 261 ! 266 IF( llsend_so ) ALLOCATE( zsnd_so( jpi,ihl,ipk,ipl,ipf) )267 IF( llsend_no ) ALLOCATE( zsnd_no( jpi,ihl,ipk,ipl,ipf) )268 IF( llrecv_so ) ALLOCATE( zrcv_so( jpi,ihl,ipk,ipl,ipf) )269 IF( llrecv_no ) ALLOCATE( zrcv_no( jpi,ihl,ipk,ipl,ipf) )270 ! 271 isize = jpi * ihl* ipk * ipl * ipf262 IF( llsend_so ) ALLOCATE( zsnd_so(1-nn_hls+1:jpi,nn_hls,ipk,ipl,ipf) ) 263 IF( llsend_no ) ALLOCATE( zsnd_no(1-nn_hls+1:jpi,nn_hls,ipk,ipl,ipf) ) 264 IF( llrecv_so ) ALLOCATE( zrcv_so(1-nn_hls+1:jpi,nn_hls,ipk,ipl,ipf) ) 265 IF( llrecv_no ) ALLOCATE( zrcv_no(1-nn_hls+1:jpi,nn_hls,ipk,ipl,ipf) ) 266 ! 267 isize = ( jpi + nn_hls - 1 ) * nn_hls * ipk * ipl * ipf 272 268 273 269 ! allocate local temporary arrays to be sent/received. Fill arrays to be sent 274 270 IF( llsend_so ) THEN ! copy sourhern side of the inner mpi domain in local temporary array to be sent by MPI 275 ishift = ihl276 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1, ihl ; DO ji =1, jpi277 zsnd_so(ji,jj,jk,jl,jf) = ARRAY_IN(ji,ishift+jj,jk,jl,jf) ! ihl+1 -> 2*ihl271 ishift = 1 272 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1, nn_hls ; DO ji = 1-nn_hls+1, jpi 273 zsnd_so(ji,jj,jk,jl,jf) = ARRAY_IN(ji,ishift+jj,jk,jl,jf) ! nn_hls+1 -> 2*nn_hls 278 274 END DO ; END DO ; END DO ; END DO ; END DO 279 275 ENDIF 280 276 ! 281 277 IF( llsend_no ) THEN ! copy eastern side of the inner mpi domain in local temporary array to be sent by MPI 282 ishift = jpj - 2 * ihl283 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1, ihl ; DO ji =1, jpi284 zsnd_no(ji,jj,jk,jl,jf) = ARRAY_IN(ji,ishift+jj,jk,jl,jf) ! jpj-2* ihl+1 -> jpj-ihl278 ishift = jpj - 2 * nn_hls 279 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1, nn_hls ; DO ji = 1-nn_hls+1, jpi 280 zsnd_no(ji,jj,jk,jl,jf) = ARRAY_IN(ji,ishift+jj,jk,jl,jf) ! jpj-2*nn_hls+1 -> jpj-nn_hls 285 281 END DO ; END DO ; END DO ; END DO ; END DO 286 282 ENDIF … … 289 285 ! 290 286 ! non-blocking send of the southern/northern side 291 IF( llsend_so ) CALL mppsend( 3, zsnd_so(1 ,1,1,1,1), isize, noso, ireq_so )292 IF( llsend_no ) CALL mppsend( 4, zsnd_no(1 ,1,1,1,1), isize, nono, ireq_no )287 IF( llsend_so ) CALL mppsend( 3, zsnd_so(1-nn_hls+1,1,1,1,1), isize, noso, ireq_so ) 288 IF( llsend_no ) CALL mppsend( 4, zsnd_no(1-nn_hls+1,1,1,1,1), isize, nono, ireq_no ) 293 289 ! blocking receive of the southern/northern halo 294 IF( llrecv_so ) CALL mpprecv( 4, zrcv_so(1 ,1,1,1,1), isize, noso )295 IF( llrecv_no ) CALL mpprecv( 3, zrcv_no(1 ,1,1,1,1), isize, nono )290 IF( llrecv_so ) CALL mpprecv( 4, zrcv_so(1-nn_hls+1,1,1,1,1), isize, noso ) 291 IF( llrecv_no ) CALL mpprecv( 3, zrcv_no(1-nn_hls+1,1,1,1,1), isize, nono ) 296 292 ! 297 293 IF( ln_timing ) CALL tic_tac(.FALSE.) … … 303 299 ! 5.1 fill southern halo 304 300 ! ---------------------- 305 ! ishift = 0 ! fill halo from jj = 1 to ihl301 ! ishift = 0 ! fill halo from jj = 1 to nn_hls 306 302 SELECT CASE ( ifill_so ) 307 303 CASE ( jpfillnothing ) ! no filling 308 304 CASE ( jpfillmpi ) ! use data received by MPI 309 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1, ihl ; DO ji =1, jpi310 ARRAY_IN(ji,jj ,jk,jl,jf) = zrcv_so(ji,jj,jk,jl,jf) ! 1 -> ihl305 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1, nn_hls ; DO ji = 1-nn_hls+1, jpi 306 ARRAY_IN(ji,jj-nn_hls+1,jk,jl,jf) = zrcv_so(ji,jj,jk,jl,jf) ! 1 -> nn_hls 311 307 END DO; END DO ; END DO ; END DO ; END DO 312 308 CASE ( jpfillperio ) ! use north-south periodicity 313 ishift2 = jpj - 2 * ihl314 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1, ihl ; DO ji =1, jpi315 ARRAY_IN(ji,jj ,jk,jl,jf) = ARRAY_IN(ji,ishift2+jj,jk,jl,jf)309 ishift2 = jpj - 2 * nn_hls 310 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1, nn_hls ; DO ji = 1-nn_hls+1, jpi 311 ARRAY_IN(ji,jj-nn_hls+1,jk,jl,jf) = ARRAY_IN(ji,ishift2+jj,jk,jl,jf) 316 312 END DO; END DO ; END DO ; END DO ; END DO 317 313 CASE ( jpfillcopy ) ! filling with inner domain values 318 314 DO jf = 1, ipf ! number of arrays to be treated 319 315 IF( .NOT. NAT_IN(jf) == 'F' ) THEN ! do nothing for F point 320 DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1, ihl ; DO ji =1, jpi321 ARRAY_IN(ji,jj ,jk,jl,jf) = ARRAY_IN(ji,ihl+1,jk,jl,jf)316 DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1, nn_hls ; DO ji = 1-nn_hls+1, jpi 317 ARRAY_IN(ji,jj-nn_hls+1,jk,jl,jf) = ARRAY_IN(ji,1+jj,jk,jl,jf) 322 318 END DO ; END DO ; END DO ; END DO 323 319 ENDIF … … 326 322 DO jf = 1, ipf ! number of arrays to be treated 327 323 IF( .NOT. NAT_IN(jf) == 'F' ) THEN ! do nothing for F point 328 DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1, ihl ; DO ji =1, jpi329 ARRAY_IN(ji,jj ,jk,jl,jf) = zland324 DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1, nn_hls ; DO ji = 1-nn_hls+1, jpi 325 ARRAY_IN(ji,jj-nn_hls+1,jk,jl,jf) = zland 330 326 END DO; END DO ; END DO ; END DO 331 327 ENDIF … … 335 331 ! 5.2 fill northern halo 336 332 ! ---------------------- 337 ishift = jpj - ihl ! fill halo from jj = jpj-ihl+1 to jpj333 ishift = jpj - nn_hls ! fill halo from jj = jpj-nn_hls+1 to jpj 338 334 SELECT CASE ( ifill_no ) 339 335 CASE ( jpfillnothing ) ! no filling 340 336 CASE ( jpfillmpi ) ! use data received by MPI 341 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1, ihl ; DO ji =1, jpi342 ARRAY_IN(ji,ishift+jj,jk,jl,jf) = zrcv_no(ji,jj,jk,jl,jf) ! jpj- ihl+1 -> jpj337 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1, nn_hls ; DO ji = 1-nn_hls+1, jpi 338 ARRAY_IN(ji,ishift+jj,jk,jl,jf) = zrcv_no(ji,jj,jk,jl,jf) ! jpj-nn_hls+1 -> jpj 343 339 END DO ; END DO ; END DO ; END DO ; END DO 344 340 CASE ( jpfillperio ) ! use north-south periodicity 345 ishift2 = ihl346 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1, ihl ; DO ji =1, jpi341 ishift2 = 1 342 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1, nn_hls ; DO ji = 1-nn_hls+1, jpi 347 343 ARRAY_IN(ji,ishift+jj,jk,jl,jf) = ARRAY_IN(ji,ishift2+jj,jk,jl,jf) 348 344 END DO; END DO ; END DO ; END DO ; END DO 349 345 CASE ( jpfillcopy ) ! filling with inner domain values 350 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1, ihl ; DO ji = 1, jpi 351 ARRAY_IN(ji,ishift+jj,jk,jl,jf) = ARRAY_IN(ji,ishift,jk,jl,jf) 346 ishift2 = jpj - 2*nn_hls 347 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1, nn_hls ; DO ji = 1-nn_hls+1, jpi 348 ARRAY_IN(ji,ishift+jj,jk,jl,jf) = ARRAY_IN(ji,ishift2+jj,jk,jl,jf) 352 349 END DO; END DO ; END DO ; END DO ; END DO 353 350 CASE ( jpfillcst ) ! filling with constant value 354 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1, ihl ; DO ji =1, jpi351 DO jf = 1, ipf ; DO jl = 1, ipl ; DO jk = 1, ipk ; DO jj = 1, nn_hls ; DO ji = 1-nn_hls+1, jpi 355 352 ARRAY_IN(ji,ishift+jj,jk,jl,jf) = zland 356 353 END DO; END DO ; END DO ; END DO ; END DO -
NEMO/branches/2020/dev_r12558_HPC-08_epico_Extra_Halo/src/OCE/LBC/mpp_nfd_generic.h90
r11536 r12586 48 48 SUBROUTINE ROUTINE_NFD( ptab, cd_nat, psgn, kfld ) 49 49 !!---------------------------------------------------------------------- 50 ARRAY_TYPE( :,:,:,:,:) ! array or pointer of arrays on which the boundary condition is applied50 ARRAY_TYPE(1-nn_hls+1:,1-nn_hls+1:,:,:,:) ! array or pointer of arrays on which the boundary condition is applied 51 51 CHARACTER(len=1) , INTENT(in ) :: NAT_IN(:) ! nature of array grid-points 52 52 REAL(wp) , INTENT(in ) :: SGN_IN(:) ! sign used across the north fold boundary … … 54 54 ! 55 55 INTEGER :: ji, jj, jk, jl, jh, jf, jr ! dummy loop indices 56 INTEGER :: ipi, ip j, ipk, ipl, ipf ! dimension of the input array56 INTEGER :: ipi, ipk, ipl, ipf ! dimension of the input array 57 57 INTEGER :: imigr, iihom, ijhom ! local integers 58 58 INTEGER :: ierr, ibuffsize, ilci, ildi, ilei, iilb … … 80 80 ALLOCATE(ipj_s(ipf)) 81 81 82 i pj = 2! Max 2nd dimension of message transfers (last two j-line only)83 ipj_s(:) = 1 ! Real 2nd dimension of message transfers (depending on perf requirement)82 ijpj = 2 + nn_hls - 1 ! Max 2nd dimension of message transfers (last two j-line only) 83 ipj_s(:) = 1 + nn_hls - 1 ! Real 2nd dimension of message transfers (depending on perf requirement) 84 84 ! by default, only one line is exchanged 85 85 86 ALLOCATE( jj_s(ipf, 2) )86 ALLOCATE( jj_s(ipf,ijpj) ) 87 87 88 88 ! re-define number of exchanged lines : … … 98 98 IF ( l_full_nf_update .OR. & ! if coupling fields 99 99 ( ncom_stp == nit000 .AND. .NOT. ln_rstart ) ) & ! at first time step, if not restart 100 ipj_s(:) = 2 100 ipj_s(:) = 2 + nn_hls - 1 101 101 102 102 ! Index of modifying lines in input … … 110 110 ! 111 111 CASE ( 'T' , 'W' ,'U' ) ! T-, U-, W-point 112 jj_s(jf,1) = nlcj - 2 ; jj_s(jf,2) = nlcj - 1 112 DO ji = 1, nn_hls+1 113 jj_s(jf,ji) = nlcj - 2*nn_hls +ji -1 114 ENDDO 113 115 CASE ( 'V' , 'F' ) ! V-, F-point 114 jj_s(jf,1) = nlcj - 3 ; jj_s(jf,2) = nlcj - 2 116 DO ji = 1, nn_hls+1 117 jj_s(jf,ji) = nlcj - 2*nn_hls +ji - 2 118 ENDDO 115 119 END SELECT 116 120 ! … … 119 123 ! 120 124 CASE ( 'T' , 'W' ,'U' ) ! T-, U-, W-point 121 jj_s(jf,1) = nlcj - 1 122 ipj_s(jf) = 1 ! need only one line anyway 125 DO ji = 1, nn_hls 126 jj_s(jf,ji) = nlcj - 2*nn_hls + ji 127 ENDDO 128 ipj_s(jf) = nn_hls ! need only one line anyway 123 129 CASE ( 'V' , 'F' ) ! V-, F-point 124 jj_s(jf,1) = nlcj - 2 ; jj_s(jf,2) = nlcj - 1 130 DO ji = 1, nn_hls+1 131 jj_s(jf,ji) = nlcj - 2*nn_hls +ji -1 132 ENDDO 125 133 END SELECT 126 134 ! … … 131 139 ipf_j = sum (ipj_s(:)) ! Total number of lines to be exchanged 132 140 ! 133 ALLOCATE( znorthloc( jpimax,ipf_j,ipk,ipl,1) )141 ALLOCATE( znorthloc(1-nn_hls+1:jpimax,ipf_j,ipk,ipl,1) ) 134 142 ! 135 143 js = 0 … … 139 147 DO jl = 1, ipl 140 148 DO jk = 1, ipk 141 znorthloc(1 :jpi,js,jk,jl,1) = ARRAY_IN(1:jpi,jj_s(jf,jj),jk,jl,jf)149 znorthloc(1-nn_hls+1:jpi,js,jk,jl,1) = ARRAY_IN(1-nn_hls+1:jpi,jj_s(jf,jj),jk,jl,jf) 142 150 END DO 143 151 END DO … … 145 153 END DO 146 154 ! 147 ibuffsize = jpimax* ipf_j * ipk * ipl148 ! 149 ALLOCATE( zfoldwk( jpimax,ipf_j,ipk,ipl,1) )150 ALLOCATE( ztabr( jpimax*jpmaxngh,ipj,ipk,ipl,ipf) )155 ibuffsize = (jpimax + nn_hls -1) * ipf_j * ipk * ipl 156 ! 157 ALLOCATE( zfoldwk(1-nn_hls+1:jpimax,ipf_j,ipk,ipl,1) ) 158 ALLOCATE( ztabr(1-nn_hls+1:(jpi+nn_hls-1)*jpmaxngh-nn_hls+1,ijpj,ipk,ipl,ipf) ) 151 159 ! when some processors of the north fold are suppressed, 152 160 ! values of ztab* arrays corresponding to these suppressed domain won't be defined … … 218 226 ELSE !== allgather exchanges ==! 219 227 ! 220 i pj = 4 ! 2nd dimension of message transfers (last j-lines)221 ! 222 ALLOCATE( znorthloc(jpimax,i pj,ipk,ipl,ipf) )223 ! 224 DO jf = 1, ipf ! put in znorthloc the last i pj j-lines of ptab228 ijpj = 4 ! 2nd dimension of message transfers (last j-lines) 229 ! 230 ALLOCATE( znorthloc(jpimax,ijpj,ipk,ipl,ipf) ) 231 ! 232 DO jf = 1, ipf ! put in znorthloc the last ijpj j-lines of ptab 225 233 DO jl = 1, ipl 226 234 DO jk = 1, ipk 227 DO jj = nlcj - i pj +1, nlcj228 ij = jj - nlcj + i pj235 DO jj = nlcj - ijpj +1, nlcj 236 ij = jj - nlcj + ijpj 229 237 znorthloc(1:jpi,ij,jk,jl,jf) = ARRAY_IN(1:jpi,jj,jk,jl,jf) 230 238 END DO … … 233 241 END DO 234 242 ! 235 ibuffsize = jpimax * i pj * ipk * ipl * ipf236 ! 237 ALLOCATE( ztab (jpiglo,i pj,ipk,ipl,ipf ) )238 ALLOCATE( znorthgloio(jpimax,i pj,ipk,ipl,ipf,jpni) )243 ibuffsize = jpimax * ijpj * ipk * ipl * ipf 244 ! 245 ALLOCATE( ztab (jpiglo,ijpj,ipk,ipl,ipf ) ) 246 ALLOCATE( znorthgloio(jpimax,ijpj,ipk,ipl,ipf,jpni) ) 239 247 ! 240 248 ! when some processors of the north fold are suppressed, … … 263 271 DO jl = 1, ipl 264 272 DO jk = 1, ipk 265 DO jj = 1, i pj273 DO jj = 1, ijpj 266 274 DO ji = ildi, ilei 267 275 ztab(ji+iilb-1,jj,jk,jl,jf) = znorthgloio(ji,jj,jk,jl,jf,jr) … … 279 287 DO jl = 1, ipl 280 288 DO jk = 1, ipk 281 DO jj = nlcj-i pj+1, nlcj ! Scatter back to ARRAY_IN282 ij = jj - nlcj + i pj289 DO jj = nlcj-ijpj+1, nlcj ! Scatter back to ARRAY_IN 290 ij = jj - nlcj + ijpj 283 291 DO ji= 1, nlci 284 292 ARRAY_IN(ji,jj,jk,jl,jf) = ztab(ji+nimpp-1,ij,jk,jl,jf) -
NEMO/branches/2020/dev_r12558_HPC-08_epico_Extra_Halo/src/OCE/LBC/mppini.F90
r12377 r12586 25 25 USE bdy_oce ! open BounDarY 26 26 ! 27 USE lbcnfd , ONLY : isendto, nsndto , nfsloop, nfeloop! Setup of north fold exchanges27 USE lbcnfd , ONLY : isendto, nsndto ! Setup of north fold exchanges 28 28 USE lib_mpp ! distribued memory computing library 29 29 USE iom ! nemo I/O library … … 699 699 WRITE(inum,*) 700 700 WRITE(inum,*) 'north fold exchanges with explicit point-to-point messaging :' 701 WRITE(inum,*) 'nfsloop : ', nfsloop702 WRITE(inum,*) 'nfeloop : ', nfeloop703 701 WRITE(inum,*) 'nsndto : ', nsndto 704 702 WRITE(inum,*) 'isendto : ', isendto … … 1261 1259 ! 1262 1260 END DO 1263 nfsloop = 11264 nfeloop = nlci1265 DO jn = 2,jpni-11266 IF( nfipproc(jn,jpnj) == (narea - 1) ) THEN1267 IF( nfipproc(jn-1,jpnj) == -1 ) nfsloop = nldi1268 IF( nfipproc(jn+1,jpnj) == -1 ) nfeloop = nlei1269 ENDIF1270 END DO1271 1261 ! 1272 1262 ENDIF -
NEMO/branches/2020/dev_r12558_HPC-08_epico_Extra_Halo/src/OCE/SBC/sbcrnf.F90
r12489 r12586 59 59 INTEGER , PUBLIC :: nkrnf = 0 !: nb of levels over which Kz is increased at river mouths 60 60 61 REAL(wp), PUBLIC, ALLOCATABLE, SAVE, DIMENSION(:,:) :: rnfmsk !: river mouth mask (hori.)61 REAL(wp), PUBLIC, POINTER, SAVE, DIMENSION(:,:) :: rnfmsk !: river mouth mask (hori.) 62 62 REAL(wp), PUBLIC, ALLOCATABLE, SAVE, DIMENSION(:) :: rnfmsk_z !: river mouth mask (vert.) 63 63 REAL(wp), PUBLIC, ALLOCATABLE, SAVE, DIMENSION(:,:) :: h_rnf !: depth of runoff in m -
NEMO/branches/2020/dev_r12558_HPC-08_epico_Extra_Halo/src/OCE/TRA/traadv.F90
r12489 r12586 83 83 INTEGER , INTENT(in) :: kt ! ocean time-step index 84 84 INTEGER , INTENT(in) :: Kbb, Kmm, Krhs ! time level indices 85 REAL(wp), DIMENSION(jpi,jpj,jpk,jpts,jpt), INTENT(inout) :: pts ! active tracers and RHS of tracer equation85 REAL(wp), POINTER, DIMENSION(:,:,:,:,:) , INTENT(inout) :: pts ! active tracers and RHS of tracer equation 86 86 ! 87 87 INTEGER :: jk ! dummy loop index 88 REAL(wp), DIMENSION(jpi,jpj,jpk) :: zuu, zvv, zww ! 3D workspace88 REAL(wp), POINTER, DIMENSION(:,:,:) :: zuu, zvv, zww ! 3D workspace 89 89 REAL(wp), DIMENSION(:,:,:), ALLOCATABLE :: ztrdt, ztrds 90 90 !!---------------------------------------------------------------------- 91 91 ! 92 92 IF( ln_timing ) CALL timing_start('tra_adv') 93 ALLOCATE( zuu(jpi,jpj,jpk), zvv(jpi,jpj,jpk), zww(jpi,jpj,jpk) ) 93 94 ! 94 95 ! !== effective transport ==! … … 167 168 & tab3d_2=pts(:,:,:,jp_sal,Krhs), clinfo2= ' Sa: ', mask2=tmask, clinfo3='tra' ) 168 169 ! 170 DEALLOCATE( zuu, zvv, zww ) 169 171 IF( ln_timing ) CALL timing_stop( 'tra_adv' ) 170 172 ! -
NEMO/branches/2020/dev_r12558_HPC-08_epico_Extra_Halo/src/OCE/TRA/traadv_mus.F90
r12377 r12586 31 31 USE lbclnk ! ocean lateral boundary condition (or mpp link) 32 32 USE lib_fortran ! Fortran utilities (allows no signed zero when 'key_nosignedzero' defined) 33 USE halo_mng 33 34 34 35 IMPLICIT NONE … … 37 38 PUBLIC tra_adv_mus ! routine called by traadv.F90 38 39 39 REAL(wp), ALLOCATABLE, SAVE, DIMENSION(:,:) :: upsmsk !: mixed upstream/centered scheme near some straits40 REAL(wp), POINTER, SAVE, DIMENSION(:,:) :: upsmsk !: mixed upstream/centered scheme near some straits 40 41 ! ! and in closed seas (orca 2 and 1 configurations) 41 42 REAL(wp), ALLOCATABLE, SAVE, DIMENSION(:,:,:) :: xind !: mixed upstream/centered index … … 44 45 LOGICAL :: l_ptr ! flag to compute poleward transport 45 46 LOGICAL :: l_hst ! flag to compute heat/salt transport 47 48 INTEGER :: jphls = 2 46 49 47 50 !! * Substitutions … … 80 83 LOGICAL , INTENT(in ) :: ld_msc_ups ! use upstream scheme within muscl 81 84 REAL(wp) , INTENT(in ) :: p2dt ! tracer time-step 82 REAL(wp), DIMENSION(jpi,jpj,jpk ), INTENT(in ) :: pU, pV, pW ! 3 ocean volume flux components83 REAL(wp), DIMENSION(jpi,jpj,jpk,kjpt,jpt), INTENT(inout) :: pt ! tracers and RHS of tracer equation85 REAL(wp), POINTER, DIMENSION(:,:,: ) , INTENT(in ) :: pU, pV, pW ! 3 ocean volume flux components 86 REAL(wp), POINTER, DIMENSION(:,:,:,:,:) , INTENT(inout) :: pt ! tracers and RHS of tracer equation 84 87 ! 85 88 INTEGER :: ji, jj, jk, jn ! dummy loop indices … … 87 90 REAL(wp) :: zu, z0u, zzwx, zw , zalpha ! local scalars 88 91 REAL(wp) :: zv, z0v, zzwy, z0w ! - - 89 REAL(wp), DIMENSION(jpi,jpj,jpk) :: zwx, zslpx ! 3D workspace90 REAL(wp), DIMENSION(jpi,jpj,jpk) :: zwy, zslpy ! - -92 REAL(wp), ALLOCATABLE, DIMENSION(:,:,:) :: zwx, zslpx ! 3D workspace 93 REAL(wp), ALLOCATABLE, DIMENSION(:,:,:) :: zwy, zslpy ! - - 91 94 !!---------------------------------------------------------------------- 92 95 ! 96 CALL halo_mng_set(jphls) 97 98 ALLOCATE(zwx(jplbi:jpi,jplbj:jpj,jpk)) 99 ALLOCATE(zwy(jplbi:jpi,jplbj:jpj,jpk)) 100 ALLOCATE(zslpx(jplbi:jpi,jplbj:jpj,jpk)) 101 ALLOCATE(zslpy(jplbi:jpi,jplbj:jpj,jpk)) 102 103 CALL halo_mng_resize(r1_e1e2t,'T', 1._wp) 104 CALL halo_mng_resize(r1_e1e2u,'U', 1._wp) 105 CALL halo_mng_resize(r1_e1e2v,'V', 1._wp) 106 CALL halo_mng_resize(tmask,'T', 1._wp) 107 CALL halo_mng_resize(wmask, 'W', 1._wp) 108 CALL halo_mng_resize(umask, 'U', 1._wp) 109 CALL halo_mng_resize(vmask, 'V', 1._wp) 110 CALL halo_mng_resize(pt, 'T', 1._wp, kjpt=kjpt, fjpt=Kbb ) 111 CALL halo_mng_resize(pt, 'T', 1._wp, kjpt=kjpt, fjpt=Krhs ) 112 CALL halo_mng_resize(e3t,'T', 1._wp, fillval=1._wp, fjpt=Kmm) 113 CALL halo_mng_resize(e3u, 'U', 1._wp, fillval=1._wp, fjpt=Kmm ) 114 CALL halo_mng_resize(e3v, 'V', 1._wp, fillval=1._wp, fjpt=Kmm) 115 CALL halo_mng_resize(e3w, 'W', 1._wp, fillval=1._wp, fjpt=Kmm) 116 CALL halo_mng_resize(pU, 'U', 1._wp) 117 CALL halo_mng_resize(pV, 'V', 1._wp) 118 CALL halo_mng_resize(pW, 'W', 1._wp) 119 ! 120 IF( ln_isfcav ) CALL halo_mng_resize(mikt, 'T', 1._wp) 121 IF( ld_msc_ups) CALL halo_mng_resize(rnfmsk, 'T', 1._wp) 122 IF( ld_msc_ups) CALL halo_mng_resize(upsmsk, 'T', 1._wp) 123 93 124 IF( kt == kit000 ) THEN 94 125 IF(lwp) WRITE(numout,*) … … 100 131 ! Upstream / MUSCL scheme indicator 101 132 ! 102 ALLOCATE( xind(jp i,jpj,jpk), STAT=ierr )133 ALLOCATE( xind(jplbi:jpi,jplbj:jpj,jpk), STAT=ierr ) 103 134 xind(:,:,:) = 1._wp ! set equal to 1 where up-stream is not needed 104 135 ! 105 136 IF( ld_msc_ups ) THEN ! define the upstream indicator (if asked) 106 ALLOCATE( upsmsk(jp i,jpj), STAT=ierr )137 ALLOCATE( upsmsk(jplbi:jpi,jplbj:jpj), STAT=ierr ) 107 138 upsmsk(:,:) = 0._wp ! not upstream by default 108 139 ! … … 115 146 ! 116 147 ENDIF 117 ! 148 118 149 l_trd = .FALSE. 119 150 l_hst = .FALSE. … … 131 162 zwx(:,:,jpk) = 0._wp ! bottom values 132 163 zwy(:,:,jpk) = 0._wp 133 DO_3D_ 10_10( 1, jpkm1 )164 DO_3D_20_20( 1, jpkm1 ) 134 165 zwx(ji,jj,jk) = umask(ji,jj,jk) * ( pt(ji+1,jj,jk,jn,Kbb) - pt(ji,jj,jk,jn,Kbb) ) 135 166 zwy(ji,jj,jk) = vmask(ji,jj,jk) * ( pt(ji,jj+1,jk,jn,Kbb) - pt(ji,jj,jk,jn,Kbb) ) 136 167 END_3D 137 168 ! lateral boundary conditions (changed sign) 138 CALL lbc_lnk_multi( 'traadv_mus', zwx, 'U', -1. , zwy, 'V', -1. ) 169 CALL lbc_lnk( 'traadv_mus', zwx, 'U', -1. ) ! lateral boundary conditions (changed sign) 170 CALL lbc_lnk( 'traadv_mus', zwy, 'V', -1. ) ! lateral boundary conditions (changed sign) 139 171 ! !-- Slopes of tracer 140 172 zslpx(:,:,jpk) = 0._wp ! bottom values 141 173 zslpy(:,:,jpk) = 0._wp 142 DO_3D_ 01_01( 1, jpkm1 )174 DO_3D_31_31( 1, jpkm1 ) 143 175 zslpx(ji,jj,jk) = ( zwx(ji,jj,jk) + zwx(ji-1,jj ,jk) ) & 144 176 & * ( 0.25 + SIGN( 0.25, zwx(ji,jj,jk) * zwx(ji-1,jj ,jk) ) ) … … 147 179 END_3D 148 180 ! 149 DO_3D_ 01_01( 1, jpkm1 )181 DO_3D_31_31( 1, jpkm1 ) 150 182 zslpx(ji,jj,jk) = SIGN( 1., zslpx(ji,jj,jk) ) * MIN( ABS( zslpx(ji ,jj,jk) ), & 151 183 & 2.*ABS( zwx (ji-1,jj,jk) ), & … … 156 188 END_3D 157 189 ! 158 DO_3D_ 00_00( 1, jpkm1 )190 DO_3D_30_30( 1, jpkm1 ) 159 191 ! MUSCL fluxes 160 192 z0u = SIGN( 0.5, pU(ji,jj,jk) ) … … 172 204 zwy(ji,jj,jk) = pV(ji,jj,jk) * ( zalpha * zzwx + (1.-zalpha) * zzwy ) 173 205 END_3D 174 CALL lbc_lnk_multi( 'traadv_mus', zwx, 'U', -1. , zwy, 'V', -1. ) ! lateral boundary conditions (changed sign) 175 ! 176 DO_3D_00_00( 1, jpkm1 ) 206 CALL lbc_lnk( 'traadv_mus', zwx, 'U', -1. ) ! lateral boundary conditions (changed sign) 207 CALL lbc_lnk( 'traadv_mus', zwy, 'V', -1. ) ! lateral boundary conditions (changed sign) 208 ! 209 DO_3D_30_30( 1, jpkm1 ) 177 210 pt(ji,jj,jk,jn,Krhs) = pt(ji,jj,jk,jn,Krhs) - ( zwx(ji,jj,jk) - zwx(ji-1,jj ,jk ) & 178 211 & + zwy(ji,jj,jk) - zwy(ji ,jj-1,jk ) ) & … … 199 232 ! !-- Slopes of tracer 200 233 zslpx(:,:,1) = 0._wp ! surface values 201 DO_3D_ 11_11( 2, jpkm1 )234 DO_3D_21_21( 2, jpkm1 ) 202 235 zslpx(ji,jj,jk) = ( zwx(ji,jj,jk) + zwx(ji,jj,jk+1) ) & 203 236 & * ( 0.25 + SIGN( 0.25, zwx(ji,jj,jk) * zwx(ji,jj,jk+1) ) ) 204 237 END_3D 205 DO_3D_ 11_11( 2, jpkm1 )238 DO_3D_21_21( 2, jpkm1 ) 206 239 zslpx(ji,jj,jk) = SIGN( 1., zslpx(ji,jj,jk) ) * MIN( ABS( zslpx(ji,jj,jk ) ), & 207 240 & 2.*ABS( zwx (ji,jj,jk+1) ), & 208 241 & 2.*ABS( zwx (ji,jj,jk ) ) ) 209 242 END_3D 210 DO_3D_ 00_00( 1, jpk-2 )243 DO_3D_30_30( 1, jpk-2 ) 211 244 z0w = SIGN( 0.5, pW(ji,jj,jk+1) ) 212 245 zalpha = 0.5 + z0w … … 218 251 IF( ln_linssh ) THEN ! top values, linear free surface only 219 252 IF( ln_isfcav ) THEN ! ice-shelf cavities (top of the ocean) 220 DO_2D_ 11_11253 DO_2D_21_21 221 254 zwx(ji,jj, mikt(ji,jj) ) = pW(ji,jj,mikt(ji,jj)) * pt(ji,jj,mikt(ji,jj),jn,Kbb) 222 255 END_2D … … 226 259 ENDIF 227 260 ! 228 DO_3D_ 00_00( 1, jpkm1 )261 DO_3D_30_30( 1, jpkm1 ) 229 262 pt(ji,jj,jk,jn,Krhs) = pt(ji,jj,jk,jn,Krhs) - ( zwx(ji,jj,jk) - zwx(ji,jj,jk+1) ) * r1_e1e2t(ji,jj) / e3t(ji,jj,jk,Kmm) 230 263 END_3D … … 234 267 END DO ! end of tracer loop 235 268 ! 269 DEALLOCATE(zwx,zwy) 270 DEALLOCATE(zslpx,zslpy) 271 272 CALL halo_mng_set(1) 273 274 CALL halo_mng_resize(r1_e1e2t,'T', 1._wp) 275 CALL halo_mng_resize(r1_e1e2u,'U', 1._wp) 276 CALL halo_mng_resize(r1_e1e2v,'V', 1._wp) 277 CALL halo_mng_resize(pt, 'T', 1._wp, kjpt=kjpt, fjpt=Kbb ) 278 CALL halo_mng_resize(pt, 'T', 1._wp, kjpt=kjpt, fjpt=Krhs ) 279 CALL halo_mng_resize(tmask,'T', 1._wp) 280 CALL halo_mng_resize(wmask, 'W', 1._wp) 281 CALL halo_mng_resize(umask, 'U', 1._wp) 282 CALL halo_mng_resize(vmask, 'V', 1._wp) 283 CALL halo_mng_resize(e3t,'T', 1._wp, fillval=1._wp, fjpt=Kmm) 284 CALL halo_mng_resize(e3u, 'U', 1._wp, fillval=1._wp, fjpt=Kmm) 285 CALL halo_mng_resize(e3v, 'V', 1._wp, fillval=1._wp, fjpt=Kmm) 286 CALL halo_mng_resize(e3w, 'W', 1._wp, fillval=1._wp, fjpt=Kmm) 287 CALL halo_mng_resize(pU, 'U', 1._wp) 288 CALL halo_mng_resize(pV, 'V', 1._wp) 289 CALL halo_mng_resize(pW, 'W', 1._wp) 290 291 IF( ln_isfcav ) CALL halo_mng_resize(mikt, 'T', 1._wp) 292 IF( ld_msc_ups) CALL halo_mng_resize(rnfmsk, 'T', 1._wp) 293 IF( ld_msc_ups) CALL halo_mng_resize(upsmsk, 'T', 1._wp) 294 236 295 END SUBROUTINE tra_adv_mus 237 296 -
NEMO/branches/2020/dev_r12558_HPC-08_epico_Extra_Halo/src/OCE/ZDF/zdfdrg.F90
r12489 r12586 268 268 IF( ln_isfcav ) THEN ! Ocean cavities: top friction setting 269 269 ALLOCATE( rCd0_top(jpi,jpj), rCdU_top(jpi,jpj) ) 270 CALL drg_init( 'TOP ' , mikt, & ! <== in270 CALL drg_init( 'TOP ' , INT(mikt) , & ! <== in 271 271 & r_Cdmin_top, r_Cdmax_top, r_z0_top, r_ke0_top, rCd0_top, rCdU_top ) ! ==> out 272 272 ENDIF -
NEMO/branches/2020/dev_r12558_HPC-08_epico_Extra_Halo/src/OCE/ZDF/zdfphy.F90
r12377 r12586 247 247 & rCdU_bot ) ! ==>> out : bottom drag [m/s] 248 248 IF( ln_isfcav ) THEN !* top drag (ocean cavities) 249 CALL zdf_drg( kt, Kmm, mikt, r_Cdmin_top, r_Cdmax_top, & ! <<== in249 CALL zdf_drg( kt, Kmm, INT(mikt) , r_Cdmin_top, r_Cdmax_top, & ! <<== in 250 250 & r_z0_top, r_ke0_top, rCd0_top, & 251 251 & rCdU_top ) ! ==>> out : bottom drag [m/s] -
NEMO/branches/2020/dev_r12558_HPC-08_epico_Extra_Halo/src/OCE/do_loop_substitute.h90
r12377 r12586 59 59 #define __kIsm1_ 1 60 60 #define __kJsm1_ 1 61 # 62 #define __kIsmh_ jplbi 63 #define __kJsmh_ jplbj 64 #define __kIsmhp1_ jplbi+1 65 #define __kJsmhp1_ jplbj+1 61 66 62 #define __kIe_ jpi m163 #define __kJe_ jpj m167 #define __kIe_ jpi-1 68 #define __kJe_ jpj-1 64 69 #define __kIep1_ jpi 65 70 #define __kJep1_ jpj … … 78 83 #define DO_2D_10_10 DO jj = __kJsm1_, __kJe_ ; DO ji = __kIsm1_, __kIe_ 79 84 #define DO_2D_10_11 DO jj = __kJsm1_, __kJe_ ; DO ji = __kIsm1_, __kIep1_ 85 # 86 #define DO_2D_20_20 DO jj = __kJsmh_, __kJe_ ; DO ji = __kIsmh_, __kIe_ 87 #define DO_2D_21_21 DO jj = __kJsmh_, __kJep1_ ; DO ji = __kIsmh_, __kIep1_ 88 #define DO_2D_31_31 DO jj = __kJsmhp1_, __kJep1_ ; DO ji = __kIsmhp1_, __kIep1_ 89 #define DO_2D_30_30 DO jj = __kJsmhp1_, __kJe_ ; DO ji = __kIsmhp1_, __kIe_ 80 90 81 91 #define DO_2D_11_00 DO jj = __kJsm1_, __kJep1_ ; DO ji = __kIs_, __kIe_ … … 92 102 #define DO_3D_10_10(ks,ke) DO jk = ks, ke ; DO_2D_10_10 93 103 #define DO_3D_10_11(ks,ke) DO jk = ks, ke ; DO_2D_10_11 104 # 105 #define DO_3D_20_20(ks,ke) DO jk = ks, ke ; DO_2D_20_20 106 #define DO_3D_21_21(ks,ke) DO jk = ks, ke ; DO_2D_21_21 107 #define DO_3D_31_31(ks,ke) DO jk = ks, ke ; DO_2D_31_31 108 #define DO_3D_30_30(ks,ke) DO jk = ks, ke ; DO_2D_30_30 94 109 95 110 #define DO_3D_11_11(ks,ke) DO jk = ks, ke ; DO_2D_11_11 111 #define DO_3D_21_21(ks,ke) DO jk = ks, ke ; DO_2D_21_21 96 112 97 113 #define DO_3DS_00_00(ks,ke,ki) DO jk = ks, ke, ki ; DO_2D_00_00 -
NEMO/branches/2020/dev_r12558_HPC-08_epico_Extra_Halo/src/OCE/nemogcm.F90
r12489 r12586 86 86 USE lib_mpp ! distributed memory computing 87 87 USE mppini ! shared/distributed memory setting (mpp_init routine) 88 USE lbcnfd , ONLY : isendto, nsndto , nfsloop, nfeloop! Setup of north fold exchanges88 USE lbcnfd , ONLY : isendto, nsndto ! Setup of north fold exchanges 89 89 USE lib_fortran ! Fortran utilities (allows no signed zero when 'key_nosignedzero' defined) 90 90 #if defined key_iomput … … 94 94 USE agrif_all_update ! Master Agrif update 95 95 #endif 96 USE halo_mng 96 97 97 98 IMPLICIT NONE … … 276 277 ! 277 278 cxios_context = 'nemo' 279 nn_hls = 1 278 280 ! 279 281 ! !-------------------------------------------------! … … 402 404 CALL mpp_init 403 405 406 CALL halo_mng_init() 404 407 ! Now we know the dimensions of the grid and numout has been set: we can allocate arrays 405 408 CALL nemo_alloc() -
NEMO/branches/2020/dev_r12558_HPC-08_epico_Extra_Halo/src/OCE/oce.F90
r12489 r12586 24 24 REAL(wp), PUBLIC, ALLOCATABLE, SAVE, DIMENSION(:,:,:) :: wi !: vertical vel. (adaptive-implicit) [m/s] 25 25 REAL(wp), PUBLIC, ALLOCATABLE, SAVE, DIMENSION(:,:,:) :: hdiv !: horizontal divergence [s-1] 26 REAL(wp), PUBLIC, ALLOCATABLE, SAVE, DIMENSION(:,:,:,:,:) :: ts !: 4D T-S fields [Celsius,psu]26 REAL(wp), PUBLIC, POINTER, SAVE, DIMENSION(:,:,:,:,:) :: ts !: 4D T-S fields [Celsius,psu] 27 27 REAL(wp), PUBLIC, ALLOCATABLE, SAVE, DIMENSION(:,:,:,:) :: rab_b, rab_n !: thermal/haline expansion coef. [Celsius-1,psu-1] 28 28 REAL(wp), PUBLIC, ALLOCATABLE, SAVE, DIMENSION(:,:,:) :: rn2b , rn2 !: brunt-vaisala frequency**2 [s-2] -
NEMO/branches/2020/dev_r12558_HPC-08_epico_Extra_Halo/src/OCE/par_oce.F90
r12377 r12586 61 61 INTEGER, PUBLIC :: jpimax! = ( jpiglo-2*nn_hls + (jpni-1) ) / jpni + 2*nn_hls !: maximum jpi 62 62 INTEGER, PUBLIC :: jpjmax! = ( jpjglo-2*nn_hls + (jpnj-1) ) / jpnj + 2*nn_hls !: maximum jpj 63 INTEGER, PUBLIC :: jplbi 64 INTEGER, PUBLIC :: jplbj 63 65 64 66 !!--------------------------------------------------------------------- … … 78 80 INTEGER, PUBLIC, PARAMETER :: jpr2di = 0 !: number of columns for extra outer halo 79 81 INTEGER, PUBLIC, PARAMETER :: jpr2dj = 0 !: number of rows for extra outer halo 80 INTEGER, PUBLIC , PARAMETER :: nn_hls = 1!: halo width (applies to both rows and columns)82 INTEGER, PUBLIC :: nn_hls !: halo width (applies to both rows and columns) 81 83 82 84 !!---------------------------------------------------------------------- -
NEMO/branches/2020/dev_r12558_HPC-08_epico_Extra_Halo/src/OFF/nemogcm.F90
r12377 r12586 59 59 USE timing ! Timing 60 60 USE lib_fortran ! Fortran utilities (allows no signed zero when 'key_nosignedzero' defined) 61 USE lbcnfd , ONLY : isendto, nsndto , nfsloop, nfeloop! Setup of north fold exchanges61 USE lbcnfd , ONLY : isendto, nsndto ! Setup of north fold exchanges 62 62 USE step, ONLY : Nbb, Nnn, Naa, Nrhs ! time level indices 63 USE halo_mng 63 64 64 65 IMPLICIT NONE … … 182 183 ! 183 184 cxios_context = 'nemo' 185 nn_hls = 1 184 186 ! 185 187 ! !-------------------------------------------------! … … 296 298 CALL mpp_init 297 299 300 CALL halo_mng_init() 298 301 ! Now we know the dimensions of the grid and numout has been set: we can allocate arrays 299 302 CALL nemo_alloc() -
NEMO/branches/2020/dev_r12558_HPC-08_epico_Extra_Halo/src/SAO/nemogcm.F90
r12377 r12586 31 31 USE lib_mpp ! distributed memory computing 32 32 USE mppini ! shared/distributed memory setting (mpp_init routine) 33 USE lbcnfd , ONLY : isendto, nsndto , nfsloop, nfeloop! Setup of north fold exchanges33 USE lbcnfd , ONLY : isendto, nsndto ! Setup of north fold exchanges 34 34 USE lib_fortran ! Fortran utilities (allows no signed zero when 'key_nosignedzero' defined) 35 35 #if defined key_iomput 36 36 USE xios ! xIOserver 37 37 #endif 38 USE halo_mng 38 39 39 40 IMPLICIT NONE … … 98 99 ! 99 100 cxios_context = 'nemo' 101 nn_hls = 1 100 102 ! 101 103 ! !-------------------------------------------------! … … 224 226 CALL mpp_init 225 227 228 CALL halo_mng_init() 226 229 ! Now we know the dimensions of the grid and numout has been set: we can allocate arrays 227 230 CALL nemo_alloc() -
NEMO/branches/2020/dev_r12558_HPC-08_epico_Extra_Halo/src/SAS/nemogcm.F90
r12489 r12586 37 37 USE lib_mpp ! distributed memory computing 38 38 USE mppini ! shared/distributed memory setting (mpp_init routine) 39 USE lbcnfd , ONLY : isendto, nsndto , nfsloop, nfeloop! Setup of north fold exchanges39 USE lbcnfd , ONLY : isendto, nsndto ! Setup of north fold exchanges 40 40 USE lib_fortran ! Fortran utilities (allows no signed zero when 'key_nosignedzero' defined) 41 41 #if defined key_iomput … … 45 45 USE agrif_ice_update ! ice update 46 46 #endif 47 USE halo_mng 47 48 48 49 IMPLICIT NONE … … 207 208 ELSE ; cxios_context = 'nemo' 208 209 ENDIF 210 nn_hls = 1 209 211 ! 210 212 ! !-------------------------------------------------! … … 345 347 CALL mpp_init 346 348 349 CALL halo_mng_init() 347 350 ! Now we know the dimensions of the grid and numout has been set: we can allocate arrays 348 351 CALL nemo_alloc() -
NEMO/branches/2020/dev_r12558_HPC-08_epico_Extra_Halo/src/TOP/TRP/trcadv.F90
r12489 r12586 76 76 INTEGER , INTENT(in) :: kt ! ocean time-step index 77 77 INTEGER , INTENT(in) :: Kbb, Kmm, Krhs ! time level indices 78 REAL(wp), DIMENSION(jpi,jpj,jpk,jptra,jpt), INTENT(inout) :: ptr ! passive tracers and RHS of tracer equation78 REAL(wp), POINTER, DIMENSION(:,:,:,:,:) , INTENT(inout) :: ptr ! passive tracers and RHS of tracer equation 79 79 ! 80 80 INTEGER :: jk ! dummy loop index 81 81 CHARACTER (len=22) :: charout 82 REAL(wp), DIMENSION(jpi,jpj,jpk) :: zuu, zvv, zww ! effective velocity82 REAL(wp), POINTER, DIMENSION(:,:,:) :: zuu, zvv, zww ! effective velocity 83 83 !!---------------------------------------------------------------------- 84 84 ! 85 85 IF( ln_timing ) CALL timing_start('trc_adv') 86 ALLOCATE( zuu(jpi,jpj,jpk), zvv(jpi,jpj,jpk), zww(jpi,jpj,jpk) ) 86 87 ! 87 88 ! !== effective transport ==! … … 143 144 IF( ln_timing ) CALL timing_stop('trc_adv') 144 145 ! 146 DEALLOCATE( zuu, zvv, zww ) 145 147 END SUBROUTINE trc_adv 146 148 -
NEMO/branches/2020/dev_r12558_HPC-08_epico_Extra_Halo/src/TOP/trc.F90
r12489 r12586 33 33 REAL(wp), PUBLIC :: areatot !: total volume 34 34 REAL(wp), PUBLIC, ALLOCATABLE, SAVE, DIMENSION(:,:,: ) :: cvol !: volume correction -degrad option- 35 REAL(wp), PUBLIC, ALLOCATABLE, SAVE, DIMENSION(:,:,:,:,:) :: tr !: tracer concentration35 REAL(wp), PUBLIC, POINTER, SAVE, DIMENSION(:,:,:,:,:) :: tr !: tracer concentration 36 36 REAL(wp), PUBLIC, ALLOCATABLE, SAVE, DIMENSION(:,:,: ) :: sbc_trc_b !: Before sbc fluxes for tracers 37 37 REAL(wp), PUBLIC, ALLOCATABLE, SAVE, DIMENSION(:,:,: ) :: sbc_trc !: Now sbc fluxes for tracers
Note: See TracChangeset
for help on using the changeset viewer.
