[1] | 1 | SUBROUTINE advec_s_ups( s, var_char ) |
---|
| 2 | |
---|
| 3 | !------------------------------------------------------------------------------! |
---|
[484] | 4 | ! Current revisions: |
---|
[1] | 5 | ! ----------------- |
---|
[198] | 6 | ! |
---|
[1] | 7 | ! |
---|
| 8 | ! Former revisions: |
---|
| 9 | ! ----------------- |
---|
[3] | 10 | ! $Id: advec_s_ups.f90 484 2010-02-05 07:36:54Z suehring $ |
---|
[198] | 11 | ! |
---|
| 12 | ! 164 2008-05-15 08:46:15Z raasch |
---|
| 13 | ! Arguments removed from transpose routines |
---|
| 14 | ! |
---|
| 15 | ! February 2007 |
---|
[3] | 16 | ! RCS Log replace by Id keyword, revision history cleaned up |
---|
| 17 | ! |
---|
[1] | 18 | ! Revision 1.6 2004/04/30 08:02:43 raasch |
---|
| 19 | ! Enlarged transposition arrays introduced |
---|
| 20 | ! |
---|
| 21 | ! Revision 1.1 1999/02/05 08:44:47 raasch |
---|
| 22 | ! Initial revision |
---|
| 23 | ! |
---|
| 24 | ! |
---|
| 25 | ! Description: |
---|
| 26 | ! ------------ |
---|
| 27 | ! Upstream-Spline advection of scalar quantities (potential temperature, |
---|
| 28 | ! turbulent kinetic energy). The advection process is divided into three |
---|
| 29 | ! subsequent steps, one for each of the dimensions. The result is stored as a |
---|
| 30 | ! tendency in array tend. The computation of the cubic splines and the possible |
---|
| 31 | ! execution of the Long-filter require that all grid points of the relevant |
---|
| 32 | ! dimension are available. For model runs on more than one PE therefore both the |
---|
| 33 | ! advected and the advecting quantities are transposed accordingly. |
---|
| 34 | ! |
---|
| 35 | ! Actual arguments: |
---|
| 36 | ! s = scalar quantity to be advected (remains unchanged in this UP) |
---|
| 37 | ! var_char = character string specifying the quantity to be advected |
---|
| 38 | ! |
---|
| 39 | ! Internally used arrays: |
---|
| 40 | ! v_ad = scalar quantity to be advected, initialized = s at the beginning, |
---|
| 41 | ! also being used as temporary storage after each time step |
---|
| 42 | ! d = advecting component (u, v, or w) |
---|
| 43 | !------------------------------------------------------------------------------! |
---|
| 44 | |
---|
| 45 | USE advection |
---|
| 46 | USE arrays_3d |
---|
| 47 | USE cpulog |
---|
| 48 | USE grid_variables |
---|
| 49 | USE indices |
---|
| 50 | USE interfaces |
---|
| 51 | USE control_parameters |
---|
| 52 | |
---|
| 53 | IMPLICIT NONE |
---|
| 54 | |
---|
| 55 | CHARACTER (LEN=*) :: var_char |
---|
| 56 | |
---|
| 57 | INTEGER :: i, j, k |
---|
| 58 | REAL :: s(nzb:nzt+1,nys-1:nyn+1,nxl-1:nxr+1) |
---|
| 59 | REAL, DIMENSION(:,:,:), ALLOCATABLE :: v_ad |
---|
| 60 | |
---|
| 61 | CALL cpu_log( log_point_s(16), 'advec_s_ups', 'start' ) |
---|
| 62 | |
---|
| 63 | #if defined( __parallel ) |
---|
| 64 | |
---|
| 65 | ! |
---|
| 66 | !-- Advection of the scalar in x-direction: |
---|
| 67 | !-- Store the scalar in temporary array v_ad (component to be advected, |
---|
| 68 | !-- boundaries are not used because they disturb the transposition) |
---|
| 69 | ALLOCATE( v_ad(nzb+1:nzta,nys:nyna,nxl:nxra) ) |
---|
| 70 | v_ad = 0.0 |
---|
| 71 | v_ad(nzb+1:nzt,nys:nyn,nxl:nxr) = s(nzb+1:nzt,nys:nyn,nxl:nxr) |
---|
| 72 | |
---|
| 73 | ! |
---|
| 74 | !-- Enlarge the size of tend, used as a working array for the transpositions |
---|
| 75 | IF ( nxra > nxr .OR. nyna > nyn .OR. nza > nz ) THEN |
---|
| 76 | DEALLOCATE( tend ) |
---|
| 77 | ALLOCATE( tend(1:nza,nys:nyna,nxl:nxra) ) |
---|
| 78 | ENDIF |
---|
| 79 | |
---|
| 80 | ! |
---|
| 81 | !-- Transpose the component to be advected: z --> x |
---|
[164] | 82 | CALL transpose_zx( v_ad, tend, v_ad ) |
---|
[1] | 83 | |
---|
| 84 | #else |
---|
| 85 | |
---|
| 86 | ! |
---|
| 87 | !-- Advection of the scalar in x-direction: |
---|
| 88 | !-- Store the scalar in temporary array v_ad (component to be advected) |
---|
| 89 | ALLOCATE( v_ad(nzb:nzt+1,nys-1:nyn+1,nxl-1:nxr+1) ) |
---|
| 90 | v_ad(:,:,:) = s(:,:,:) |
---|
| 91 | |
---|
| 92 | #endif |
---|
| 93 | |
---|
| 94 | ! |
---|
| 95 | !-- Advecting component (u) must be averaged out on the scalar's grid |
---|
| 96 | DO i = nxl, nxr |
---|
| 97 | DO j = nys, nyn |
---|
| 98 | DO k = nzb+1, nzt |
---|
| 99 | d(k,j,i) = 0.5 * ( u(k,j,i) + u(k,j,i+1) ) - u_gtrans |
---|
| 100 | ENDDO |
---|
| 101 | ENDDO |
---|
| 102 | ENDDO |
---|
| 103 | |
---|
| 104 | #if defined( __parallel ) |
---|
| 105 | |
---|
| 106 | ! |
---|
| 107 | !-- Transpose the advecting componnet: z --> x |
---|
[164] | 108 | CALL transpose_zx( d, tend, d ) |
---|
[1] | 109 | |
---|
| 110 | #endif |
---|
| 111 | |
---|
| 112 | ! |
---|
| 113 | !-- Upstream-Spline advection of the scalar in x-direction |
---|
| 114 | CALL spline_x( v_ad, d, var_char ) |
---|
| 115 | |
---|
| 116 | ! |
---|
| 117 | !-- Advection of the scalar in y-direction: |
---|
| 118 | !-- advecting component (v) must be averaged out on the scalar's grid |
---|
| 119 | DO i = nxl, nxr |
---|
| 120 | DO j = nys, nyn |
---|
| 121 | DO k = nzb+1, nzt |
---|
| 122 | d(k,j,i) = 0.5 * ( v(k,j,i) + v(k,j+1,i) ) - v_gtrans |
---|
| 123 | ENDDO |
---|
| 124 | ENDDO |
---|
| 125 | ENDDO |
---|
| 126 | |
---|
| 127 | #if defined( __parallel ) |
---|
| 128 | |
---|
| 129 | ! |
---|
| 130 | !-- Transpose the advecting component: z --> y |
---|
[164] | 131 | CALL transpose_zx( d, tend, d ) |
---|
| 132 | CALL transpose_xy( d, tend, d ) |
---|
[1] | 133 | |
---|
| 134 | ! |
---|
| 135 | !-- Transpose the component to be advected: x --> y |
---|
[164] | 136 | CALL transpose_xy( v_ad, tend, v_ad ) |
---|
[1] | 137 | |
---|
| 138 | #endif |
---|
| 139 | |
---|
| 140 | ! |
---|
| 141 | !-- Upstream-Spline advection of the scalar in y-direction |
---|
| 142 | CALL spline_y( v_ad, d, var_char ) |
---|
| 143 | |
---|
| 144 | ! |
---|
| 145 | !-- Advection of the scalar in z-direction: |
---|
| 146 | !-- the advecting component (w) must be averaged out on the scalar's grid |
---|
| 147 | !-- (weighted for non-equidistant grid) |
---|
| 148 | d = 0.0 |
---|
| 149 | DO i = nxl, nxr |
---|
| 150 | DO j = nys, nyn |
---|
| 151 | DO k = nzb+1, nzt |
---|
| 152 | d(k,j,i) = ( w(k,j,i) * ( zu(k) - zw(k-1) ) + & |
---|
| 153 | w(k-1,j,i) * ( zw(k) - zu(k) ) ) * ddzw(k) |
---|
| 154 | ENDDO |
---|
| 155 | ENDDO |
---|
| 156 | ENDDO |
---|
| 157 | |
---|
| 158 | #if defined( __parallel ) |
---|
| 159 | |
---|
| 160 | ! |
---|
| 161 | !-- Transpose the component to be advected: y --> z (= y --> x + x --> z) |
---|
[164] | 162 | CALL transpose_yx( v_ad, tend, v_ad ) |
---|
| 163 | CALL transpose_xz( v_ad, tend, v_ad ) |
---|
[1] | 164 | |
---|
| 165 | ! |
---|
| 166 | !-- Resize tend to its normal size |
---|
| 167 | IF ( nxra > nxr .OR. nyna > nyn .OR. nza > nz ) THEN |
---|
| 168 | DEALLOCATE( tend ) |
---|
| 169 | ALLOCATE( tend(nzb:nzt+1,nys-1:nyn+1,nxl-1:nxr+1) ) |
---|
| 170 | ENDIF |
---|
| 171 | |
---|
| 172 | #endif |
---|
| 173 | |
---|
| 174 | ! |
---|
| 175 | !-- Upstream-Spline advection of the scalar in z-direction |
---|
| 176 | CALL spline_z( v_ad, d, dzu, spl_tri_zu, var_char ) |
---|
| 177 | |
---|
| 178 | ! |
---|
| 179 | !-- Compute the tendency term |
---|
| 180 | DO i = nxl, nxr |
---|
| 181 | DO j = nys, nyn |
---|
| 182 | DO k = nzb+1, nzt |
---|
| 183 | tend(k,j,i) = ( v_ad(k,j,i) - s(k,j,i) ) / dt_3d |
---|
| 184 | ENDDO |
---|
| 185 | ENDDO |
---|
| 186 | ENDDO |
---|
| 187 | |
---|
| 188 | DEALLOCATE( v_ad ) |
---|
| 189 | |
---|
| 190 | CALL cpu_log( log_point_s(16), 'advec_s_ups', 'stop' ) |
---|
| 191 | |
---|
| 192 | END SUBROUTINE advec_s_ups |
---|