| 1 | MODULE solsor |
|---|
| 2 | !!====================================================================== |
|---|
| 3 | !! *** MODULE solsor *** |
|---|
| 4 | !! Ocean solver : Successive Over-Relaxation solver |
|---|
| 5 | !!===================================================================== |
|---|
| 6 | |
|---|
| 7 | !!---------------------------------------------------------------------- |
|---|
| 8 | !! sol_sor : Red-Black Successive Over-Relaxation solver |
|---|
| 9 | !!---------------------------------------------------------------------- |
|---|
| 10 | !! * Modules used |
|---|
| 11 | USE oce ! ocean dynamics and tracers variables |
|---|
| 12 | USE dom_oce ! ocean space and time domain variables |
|---|
| 13 | USE zdf_oce ! ocean vertical physics variables |
|---|
| 14 | USE sol_oce ! solver variables |
|---|
| 15 | USE in_out_manager ! I/O manager |
|---|
| 16 | USE lib_mpp ! distributed memory computing |
|---|
| 17 | USE lbclnk ! ocean lateral boundary conditions (or mpp link) |
|---|
| 18 | |
|---|
| 19 | IMPLICIT NONE |
|---|
| 20 | PRIVATE |
|---|
| 21 | |
|---|
| 22 | !! * Routine accessibility |
|---|
| 23 | PUBLIC sol_sor ! ??? |
|---|
| 24 | |
|---|
| 25 | !!---------------------------------------------------------------------- |
|---|
| 26 | !! OPA 9.0 , LOCEAN-IPSL (2005) |
|---|
| 27 | !! $Header$ |
|---|
| 28 | !! This software is governed by the CeCILL licence see modipsl/doc/NEMO_CeCILL.txt |
|---|
| 29 | !!---------------------------------------------------------------------- |
|---|
| 30 | |
|---|
| 31 | CONTAINS |
|---|
| 32 | |
|---|
| 33 | SUBROUTINE sol_sor( kindic ) |
|---|
| 34 | !!---------------------------------------------------------------------- |
|---|
| 35 | !! *** ROUTINE sol_sor *** |
|---|
| 36 | !! |
|---|
| 37 | !! ** Purpose : Solve the ellipic equation for the barotropic stream |
|---|
| 38 | !! function system (lk_dynspg_rl=T) or the transport divergence |
|---|
| 39 | !! system (lk_dynspg_fsc=T) using a red-black successive-over- |
|---|
| 40 | !! relaxation method. |
|---|
| 41 | !! In the former case, the barotropic stream function trend has a |
|---|
| 42 | !! zero boundary condition along all coastlines (i.e. continent |
|---|
| 43 | !! as well as islands) while in the latter the boundary condition |
|---|
| 44 | !! specification is not required. |
|---|
| 45 | !! |
|---|
| 46 | !! ** Method : Successive-over-relaxation method using the red-black |
|---|
| 47 | !! technique. The former technique used was not compatible with |
|---|
| 48 | !! the north-fold boundary condition used in orca configurations. |
|---|
| 49 | !! |
|---|
| 50 | !! References : |
|---|
| 51 | !! Madec et al. 1988, Ocean Modelling, issue 78, 1-6. |
|---|
| 52 | !! |
|---|
| 53 | !! History : |
|---|
| 54 | !! ! 90-10 (G. Madec) Original code |
|---|
| 55 | !! ! 91-11 (G. Madec) |
|---|
| 56 | !! 7.1 ! 93-04 (G. Madec) time filter |
|---|
| 57 | !! ! 96-05 (G. Madec) merge sor and pcg formulations |
|---|
| 58 | !! ! 96-11 (A. Weaver) correction to preconditioning |
|---|
| 59 | !! 9.0 ! 03-04 (C. Deltel, G. Madec) Red-Black SOR in free form |
|---|
| 60 | !!---------------------------------------------------------------------- |
|---|
| 61 | !! * Arguments |
|---|
| 62 | INTEGER, INTENT( inout ) :: kindic ! solver indicator, < 0 if the conver- |
|---|
| 63 | ! ! gence is not reached: the model is |
|---|
| 64 | ! ! stopped in step |
|---|
| 65 | ! ! set to zero before the call of solsor |
|---|
| 66 | !! * Local declarations |
|---|
| 67 | INTEGER :: ji, jj, jn ! dummy loop indices |
|---|
| 68 | INTEGER :: ishift |
|---|
| 69 | REAL(wp) :: ztmp, zres, zres2 |
|---|
| 70 | |
|---|
| 71 | INTEGER :: ijmppodd, ijmppeven |
|---|
| 72 | !!---------------------------------------------------------------------- |
|---|
| 73 | |
|---|
| 74 | ijmppeven = MOD(nimpp+njmpp ,2) |
|---|
| 75 | ijmppodd = MOD(nimpp+njmpp+1,2) |
|---|
| 76 | ! ! ============== |
|---|
| 77 | DO jn = 1, nmax ! Iterative loop |
|---|
| 78 | ! ! ============== |
|---|
| 79 | |
|---|
| 80 | CALL lbc_lnk( gcx, c_solver_pt, 1. ) ! applied the lateral boundary conditions |
|---|
| 81 | |
|---|
| 82 | ! Residus |
|---|
| 83 | ! ------- |
|---|
| 84 | |
|---|
| 85 | ! Guess black update |
|---|
| 86 | DO jj = 2, jpjm1 |
|---|
| 87 | ishift = MOD( jj-ijmppodd, 2 ) |
|---|
| 88 | DO ji = 2+ishift, jpim1, 2 |
|---|
| 89 | ztmp = gcb(ji ,jj ) & |
|---|
| 90 | & - gcp(ji,jj,1) * gcx(ji ,jj-1) & |
|---|
| 91 | & - gcp(ji,jj,2) * gcx(ji-1,jj ) & |
|---|
| 92 | & - gcp(ji,jj,3) * gcx(ji+1,jj ) & |
|---|
| 93 | & - gcp(ji,jj,4) * gcx(ji ,jj+1) |
|---|
| 94 | ! Estimate of the residual |
|---|
| 95 | zres = ztmp - gcx(ji,jj) |
|---|
| 96 | gcr(ji,jj) = zres * gcdmat(ji,jj) * zres |
|---|
| 97 | ! Guess update |
|---|
| 98 | gcx(ji,jj) = sor * ztmp + (1-sor) * gcx(ji,jj) |
|---|
| 99 | END DO |
|---|
| 100 | END DO |
|---|
| 101 | |
|---|
| 102 | CALL lbc_lnk( gcx, c_solver_pt, 1. ) ! applied the lateral boubary conditions |
|---|
| 103 | |
|---|
| 104 | ! Guess red update |
|---|
| 105 | DO jj = 2, jpjm1 |
|---|
| 106 | ishift = MOD( jj-ijmppeven, 2 ) |
|---|
| 107 | DO ji = 2+ishift, jpim1, 2 |
|---|
| 108 | ztmp = gcb(ji ,jj ) & |
|---|
| 109 | & - gcp(ji,jj,1) * gcx(ji ,jj-1) & |
|---|
| 110 | & - gcp(ji,jj,2) * gcx(ji-1,jj ) & |
|---|
| 111 | & - gcp(ji,jj,3) * gcx(ji+1,jj ) & |
|---|
| 112 | & - gcp(ji,jj,4) * gcx(ji ,jj+1) |
|---|
| 113 | ! Estimate of the residual |
|---|
| 114 | zres = ztmp - gcx(ji,jj) |
|---|
| 115 | gcr(ji,jj) = zres * gcdmat(ji,jj) * zres |
|---|
| 116 | ! Guess update |
|---|
| 117 | gcx(ji,jj) = sor * ztmp + (1-sor) * gcx(ji,jj) |
|---|
| 118 | END DO |
|---|
| 119 | END DO |
|---|
| 120 | |
|---|
| 121 | ! test of convergence |
|---|
| 122 | IF ( jn > nmin .AND. MOD( jn-nmin, nmod ) == 0 ) then |
|---|
| 123 | |
|---|
| 124 | SELECT CASE ( nsol_arp ) |
|---|
| 125 | CASE ( 0 ) ! absolute precision (maximum value of the residual) |
|---|
| 126 | zres2 = MAXVAL( gcr(2:jpim1,2:jpjm1) ) |
|---|
| 127 | IF( lk_mpp ) CALL mpp_max( zres2 ) ! max over the global domain |
|---|
| 128 | ! test of convergence |
|---|
| 129 | IF( zres2 < resmax .OR. jn == nmax ) THEN |
|---|
| 130 | res = SQRT( zres2 ) |
|---|
| 131 | niter = jn |
|---|
| 132 | ncut = 999 |
|---|
| 133 | ENDIF |
|---|
| 134 | CASE ( 1 ) ! relative precision |
|---|
| 135 | rnorme = SUM( gcr(2:jpim1,2:jpjm1) ) |
|---|
| 136 | IF( lk_mpp ) CALL mpp_sum( rnorme ) ! sum over the global domain |
|---|
| 137 | ! test of convergence |
|---|
| 138 | IF( rnorme < epsr .OR. jn == nmax ) THEN |
|---|
| 139 | res = SQRT( rnorme ) |
|---|
| 140 | niter = jn |
|---|
| 141 | ncut = 999 |
|---|
| 142 | ENDIF |
|---|
| 143 | END SELECT |
|---|
| 144 | |
|---|
| 145 | !**** |
|---|
| 146 | ! IF(lwp)WRITE(numsol,9300) jn, res, sqrt( epsr ) / eps |
|---|
| 147 | 9300 FORMAT(' niter :',i4,' res :',e20.10,' b :',e20.10) |
|---|
| 148 | !**** |
|---|
| 149 | |
|---|
| 150 | ENDIF |
|---|
| 151 | ! indicator of non-convergence or explosion |
|---|
| 152 | IF( jn == nmax .OR. SQRT(epsr)/eps > 1.e+20 ) kindic = -2 |
|---|
| 153 | IF( ncut == 999 ) GOTO 999 |
|---|
| 154 | |
|---|
| 155 | ! ! ===================== |
|---|
| 156 | END DO ! END of iterative loop |
|---|
| 157 | ! ! ===================== |
|---|
| 158 | |
|---|
| 159 | 999 CONTINUE |
|---|
| 160 | |
|---|
| 161 | |
|---|
| 162 | ! Output in gcx |
|---|
| 163 | ! ------------- |
|---|
| 164 | |
|---|
| 165 | CALL lbc_lnk( gcx, c_solver_pt, 1. ) ! boundary conditions |
|---|
| 166 | |
|---|
| 167 | |
|---|
| 168 | END SUBROUTINE sol_sor |
|---|
| 169 | |
|---|
| 170 | !!===================================================================== |
|---|
| 171 | END MODULE solsor |
|---|
