| 1 | !> @file timestep.f90 |
|---|
| 2 | !------------------------------------------------------------------------------! |
|---|
| 3 | ! This file is part of the PALM model system. |
|---|
| 4 | ! |
|---|
| 5 | ! PALM is free software: you can redistribute it and/or modify it under the |
|---|
| 6 | ! terms of the GNU General Public License as published by the Free Software |
|---|
| 7 | ! Foundation, either version 3 of the License, or (at your option) any later |
|---|
| 8 | ! version. |
|---|
| 9 | ! |
|---|
| 10 | ! PALM is distributed in the hope that it will be useful, but WITHOUT ANY |
|---|
| 11 | ! WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR |
|---|
| 12 | ! A PARTICULAR PURPOSE. See the GNU General Public License for more details. |
|---|
| 13 | ! |
|---|
| 14 | ! You should have received a copy of the GNU General Public License along with |
|---|
| 15 | ! PALM. If not, see <http://www.gnu.org/licenses/>. |
|---|
| 16 | ! |
|---|
| 17 | ! Copyright 1997-2020 Leibniz Universitaet Hannover |
|---|
| 18 | !------------------------------------------------------------------------------! |
|---|
| 19 | ! |
|---|
| 20 | ! Current revisions: |
|---|
| 21 | ! ------------------ |
|---|
| 22 | ! |
|---|
| 23 | ! |
|---|
| 24 | ! Former revisions: |
|---|
| 25 | ! ----------------- |
|---|
| 26 | ! $Id: timestep.f90 4444 2020-03-05 15:59:50Z raasch $ |
|---|
| 27 | ! bugfix: cpp-directives for serial mode added |
|---|
| 28 | ! |
|---|
| 29 | ! 4360 2020-01-07 11:25:50Z suehring |
|---|
| 30 | ! Added missing OpenMP directives |
|---|
| 31 | ! |
|---|
| 32 | ! 4233 2019-09-20 09:55:54Z knoop |
|---|
| 33 | ! OpenACC data update host removed |
|---|
| 34 | ! |
|---|
| 35 | ! 4182 2019-08-22 15:20:23Z scharf |
|---|
| 36 | ! Corrected "Former revisions" section |
|---|
| 37 | ! |
|---|
| 38 | ! 4101 2019-07-17 15:14:26Z gronemeier |
|---|
| 39 | ! - consider 2*Km within diffusion criterion as Km is considered twice within |
|---|
| 40 | ! the diffusion of e, |
|---|
| 41 | ! - in RANS mode, instead of considering each wind component individually use |
|---|
| 42 | ! the wind speed of 3d wind vector in CFL criterion |
|---|
| 43 | ! - do not limit the increase of dt based on its previous value in RANS mode |
|---|
| 44 | ! |
|---|
| 45 | ! 3658 2019-01-07 20:28:54Z knoop |
|---|
| 46 | ! OpenACC port for SPEC |
|---|
| 47 | ! |
|---|
| 48 | ! Revision 1.1 1997/08/11 06:26:19 raasch |
|---|
| 49 | ! Initial revision |
|---|
| 50 | ! |
|---|
| 51 | ! |
|---|
| 52 | ! Description: |
|---|
| 53 | ! ------------ |
|---|
| 54 | !> Compute the time step under consideration of the FCL and diffusion criterion. |
|---|
| 55 | !------------------------------------------------------------------------------! |
|---|
| 56 | SUBROUTINE timestep |
|---|
| 57 | |
|---|
| 58 | |
|---|
| 59 | USE arrays_3d, & |
|---|
| 60 | ONLY: dzu, dzw, kh, km, u, u_stokes_zu, v, v_stokes_zu, w |
|---|
| 61 | |
|---|
| 62 | USE control_parameters, & |
|---|
| 63 | ONLY: cfl_factor, dt_3d, dt_fixed, dt_max, galilei_transformation, & |
|---|
| 64 | message_string, rans_mode, stop_dt, timestep_reason, u_gtrans, & |
|---|
| 65 | use_ug_for_galilei_tr, v_gtrans |
|---|
| 66 | |
|---|
| 67 | #if defined( __parallel ) |
|---|
| 68 | USE control_parameters, & |
|---|
| 69 | ONLY: coupling_mode, terminate_coupled, terminate_coupled_remote |
|---|
| 70 | #endif |
|---|
| 71 | |
|---|
| 72 | USE cpulog, & |
|---|
| 73 | ONLY: cpu_log, log_point |
|---|
| 74 | |
|---|
| 75 | USE grid_variables, & |
|---|
| 76 | ONLY: dx, dx2, dy, dy2 |
|---|
| 77 | |
|---|
| 78 | USE indices, & |
|---|
| 79 | ONLY: nxl, nxlg, nxr, nxrg, nyn, nyng, nys, nysg, nzb, nzt |
|---|
| 80 | |
|---|
| 81 | USE interfaces |
|---|
| 82 | |
|---|
| 83 | USE kinds |
|---|
| 84 | |
|---|
| 85 | USE bulk_cloud_model_mod, & |
|---|
| 86 | ONLY: dt_precipitation |
|---|
| 87 | |
|---|
| 88 | USE pegrid |
|---|
| 89 | |
|---|
| 90 | USE pmc_interface, & |
|---|
| 91 | ONLY: nested_run |
|---|
| 92 | |
|---|
| 93 | USE statistics, & |
|---|
| 94 | ONLY: flow_statistics_called, hom, u_max, u_max_ijk, v_max, v_max_ijk,& |
|---|
| 95 | w_max, w_max_ijk |
|---|
| 96 | |
|---|
| 97 | #if defined( __parallel ) |
|---|
| 98 | USE vertical_nesting_mod, & |
|---|
| 99 | ONLY: vnested, vnest_timestep_sync |
|---|
| 100 | #endif |
|---|
| 101 | |
|---|
| 102 | IMPLICIT NONE |
|---|
| 103 | |
|---|
| 104 | INTEGER(iwp) :: i !< |
|---|
| 105 | INTEGER(iwp) :: j !< |
|---|
| 106 | INTEGER(iwp) :: k !< |
|---|
| 107 | INTEGER(iwp) :: km_max_ijk(3) = -1 !< index values (i,j,k) of location where km_max occurs |
|---|
| 108 | INTEGER(iwp) :: kh_max_ijk(3) = -1 !< index values (i,j,k) of location where kh_max occurs |
|---|
| 109 | |
|---|
| 110 | LOGICAL :: stop_dt_local !< local switch for controlling the time stepping |
|---|
| 111 | |
|---|
| 112 | REAL(wp) :: div !< |
|---|
| 113 | REAL(wp) :: dt_diff !< |
|---|
| 114 | REAL(wp) :: dt_diff_l !< |
|---|
| 115 | REAL(wp) :: dt_u !< |
|---|
| 116 | REAL(wp) :: dt_u_l !< |
|---|
| 117 | REAL(wp) :: dt_v !< |
|---|
| 118 | REAL(wp) :: dt_v_l !< |
|---|
| 119 | REAL(wp) :: dt_w !< |
|---|
| 120 | REAL(wp) :: dt_w_l !< |
|---|
| 121 | REAL(wp) :: km_max !< maximum of Km in entire domain |
|---|
| 122 | REAL(wp) :: kh_max !< maximum of Kh in entire domain |
|---|
| 123 | REAL(wp) :: u_gtrans_l !< |
|---|
| 124 | REAL(wp) :: v_gtrans_l !< |
|---|
| 125 | |
|---|
| 126 | REAL(wp), DIMENSION(2) :: uv_gtrans_l !< |
|---|
| 127 | #if defined( __parallel ) |
|---|
| 128 | REAL(wp), DIMENSION(2) :: uv_gtrans !< |
|---|
| 129 | REAL(wp), DIMENSION(3) :: reduce !< |
|---|
| 130 | REAL(wp), DIMENSION(3) :: reduce_l !< |
|---|
| 131 | #endif |
|---|
| 132 | REAL(wp), DIMENSION(nzb+1:nzt) :: dxyz2_min !< |
|---|
| 133 | !$ACC DECLARE CREATE(dxyz2_min) |
|---|
| 134 | |
|---|
| 135 | |
|---|
| 136 | CALL cpu_log( log_point(12), 'calculate_timestep', 'start' ) |
|---|
| 137 | |
|---|
| 138 | ! |
|---|
| 139 | !-- In case of Galilei-transform not using the geostrophic wind as translation |
|---|
| 140 | !-- velocity, compute the volume-averaged horizontal velocity components, which |
|---|
| 141 | !-- will then be subtracted from the horizontal wind for the time step and |
|---|
| 142 | !-- horizontal advection routines. |
|---|
| 143 | IF ( galilei_transformation .AND. .NOT. use_ug_for_galilei_tr ) THEN |
|---|
| 144 | IF ( flow_statistics_called ) THEN |
|---|
| 145 | ! |
|---|
| 146 | !-- Horizontal averages already existent, just need to average them |
|---|
| 147 | !-- vertically. |
|---|
| 148 | u_gtrans = 0.0_wp |
|---|
| 149 | v_gtrans = 0.0_wp |
|---|
| 150 | DO k = nzb+1, nzt |
|---|
| 151 | u_gtrans = u_gtrans + hom(k,1,1,0) |
|---|
| 152 | v_gtrans = v_gtrans + hom(k,1,2,0) |
|---|
| 153 | ENDDO |
|---|
| 154 | u_gtrans = u_gtrans / REAL( nzt - nzb, KIND=wp ) |
|---|
| 155 | v_gtrans = v_gtrans / REAL( nzt - nzb, KIND=wp ) |
|---|
| 156 | ELSE |
|---|
| 157 | ! |
|---|
| 158 | !-- Averaging over the entire model domain. |
|---|
| 159 | u_gtrans_l = 0.0_wp |
|---|
| 160 | v_gtrans_l = 0.0_wp |
|---|
| 161 | DO i = nxl, nxr |
|---|
| 162 | DO j = nys, nyn |
|---|
| 163 | DO k = nzb+1, nzt |
|---|
| 164 | u_gtrans_l = u_gtrans_l + u(k,j,i) |
|---|
| 165 | v_gtrans_l = v_gtrans_l + v(k,j,i) |
|---|
| 166 | ENDDO |
|---|
| 167 | ENDDO |
|---|
| 168 | ENDDO |
|---|
| 169 | uv_gtrans_l(1) = u_gtrans_l / & |
|---|
| 170 | REAL( (nxr-nxl+1)*(nyn-nys+1)*(nzt-nzb), KIND=wp ) |
|---|
| 171 | uv_gtrans_l(2) = v_gtrans_l / & |
|---|
| 172 | REAL( (nxr-nxl+1)*(nyn-nys+1)*(nzt-nzb), KIND=wp ) |
|---|
| 173 | #if defined( __parallel ) |
|---|
| 174 | IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr ) |
|---|
| 175 | CALL MPI_ALLREDUCE( uv_gtrans_l, uv_gtrans, 2, MPI_REAL, MPI_SUM, & |
|---|
| 176 | comm2d, ierr ) |
|---|
| 177 | u_gtrans = uv_gtrans(1) / REAL( numprocs, KIND=wp ) |
|---|
| 178 | v_gtrans = uv_gtrans(2) / REAL( numprocs, KIND=wp ) |
|---|
| 179 | #else |
|---|
| 180 | u_gtrans = uv_gtrans_l(1) |
|---|
| 181 | v_gtrans = uv_gtrans_l(2) |
|---|
| 182 | #endif |
|---|
| 183 | ENDIF |
|---|
| 184 | ENDIF |
|---|
| 185 | |
|---|
| 186 | ! |
|---|
| 187 | !-- Determine the maxima of the velocity components, including their |
|---|
| 188 | !-- grid index positions. |
|---|
| 189 | CALL global_min_max( nzb, nzt+1, nysg, nyng, nxlg, nxrg, u, 'abs', 0.0_wp, & |
|---|
| 190 | u_max, u_max_ijk ) |
|---|
| 191 | CALL global_min_max( nzb, nzt+1, nysg, nyng, nxlg, nxrg, v, 'abs', 0.0_wp, & |
|---|
| 192 | v_max, v_max_ijk ) |
|---|
| 193 | CALL global_min_max( nzb, nzt+1, nysg, nyng, nxlg, nxrg, w, 'abs', 0.0_wp, & |
|---|
| 194 | w_max, w_max_ijk ) |
|---|
| 195 | |
|---|
| 196 | IF ( .NOT. dt_fixed ) THEN |
|---|
| 197 | ! |
|---|
| 198 | !-- Variable time step: |
|---|
| 199 | !-- Calculate the maximum time step according to the CFL-criterion |
|---|
| 200 | dt_u_l = 999999.9_wp |
|---|
| 201 | dt_v_l = 999999.9_wp |
|---|
| 202 | dt_w_l = 999999.9_wp |
|---|
| 203 | |
|---|
| 204 | IF ( .NOT. rans_mode ) THEN |
|---|
| 205 | ! |
|---|
| 206 | !-- Consider each velocity component individually |
|---|
| 207 | |
|---|
| 208 | !$ACC PARALLEL LOOP COLLAPSE(3) PRIVATE(i,j,k) & |
|---|
| 209 | !$ACC COPY(dt_u_l, dt_v_l, dt_w_l, u_stokes_zu, v_stokes_zu) & |
|---|
| 210 | !$ACC REDUCTION(MIN: dt_u_l, dt_v_l, dt_w_l) & |
|---|
| 211 | !$ACC PRESENT(u, v, w, dzu) |
|---|
| 212 | !$OMP PARALLEL DO PRIVATE(i,j,k) & |
|---|
| 213 | !$OMP REDUCTION(MIN: dt_u_l, dt_v_l, dt_w_l) |
|---|
| 214 | DO i = nxl, nxr |
|---|
| 215 | DO j = nys, nyn |
|---|
| 216 | DO k = nzb+1, nzt |
|---|
| 217 | dt_u_l = MIN( dt_u_l, ( dx / & |
|---|
| 218 | ( ABS( u(k,j,i) - u_gtrans + u_stokes_zu(k) ) & |
|---|
| 219 | + 1.0E-10_wp ) ) ) |
|---|
| 220 | dt_v_l = MIN( dt_v_l, ( dy / & |
|---|
| 221 | ( ABS( v(k,j,i) - v_gtrans + v_stokes_zu(k) ) & |
|---|
| 222 | + 1.0E-10_wp ) ) ) |
|---|
| 223 | dt_w_l = MIN( dt_w_l, ( dzu(k) / & |
|---|
| 224 | ( ABS( w(k,j,i) ) + 1.0E-10_wp ) ) ) |
|---|
| 225 | ENDDO |
|---|
| 226 | ENDDO |
|---|
| 227 | ENDDO |
|---|
| 228 | |
|---|
| 229 | ELSE |
|---|
| 230 | ! |
|---|
| 231 | !-- Consider the wind speed at the scalar-grid point |
|---|
| 232 | !-- !> @note considering the wind speed instead of each individual wind |
|---|
| 233 | !-- !> component is only a workaround so far. This might has to be |
|---|
| 234 | !-- !> changed in the future. |
|---|
| 235 | |
|---|
| 236 | !$ACC PARALLEL LOOP COLLAPSE(3) PRIVATE(i,j,k) & |
|---|
| 237 | !$ACC COPY(dt_u_l, u_stokes_zu, v_stokes_zu) & |
|---|
| 238 | !$ACC REDUCTION(MIN: dt_u_l) & |
|---|
| 239 | !$ACC PRESENT(u, v, w, dzu) |
|---|
| 240 | !$OMP PARALLEL DO PRIVATE(i,j,k) & |
|---|
| 241 | !$OMP REDUCTION(MIN: dt_u_l) |
|---|
| 242 | DO i = nxl, nxr |
|---|
| 243 | DO j = nys, nyn |
|---|
| 244 | DO k = nzb+1, nzt |
|---|
| 245 | dt_u_l = MIN( dt_u_l, ( MIN( dx, dy, dzu(k) ) / ( & |
|---|
| 246 | SQRT( ( 0.5 * ( u(k,j,i) + u(k,j,i+1) ) - u_gtrans + u_stokes_zu(k) )**2 & |
|---|
| 247 | + ( 0.5 * ( v(k,j,i) + v(k,j+1,i) ) - v_gtrans + v_stokes_zu(k) )**2 & |
|---|
| 248 | + ( 0.5 * ( w(k,j,i) + w(k-1,j,i) ) )**2 ) & |
|---|
| 249 | + 1.0E-10_wp ) ) ) |
|---|
| 250 | ENDDO |
|---|
| 251 | ENDDO |
|---|
| 252 | ENDDO |
|---|
| 253 | |
|---|
| 254 | dt_v_l = dt_u_l |
|---|
| 255 | dt_w_l = dt_u_l |
|---|
| 256 | |
|---|
| 257 | ENDIF |
|---|
| 258 | |
|---|
| 259 | #if defined( __parallel ) |
|---|
| 260 | reduce_l(1) = dt_u_l |
|---|
| 261 | reduce_l(2) = dt_v_l |
|---|
| 262 | reduce_l(3) = dt_w_l |
|---|
| 263 | IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr ) |
|---|
| 264 | CALL MPI_ALLREDUCE( reduce_l, reduce, 3, MPI_REAL, MPI_MIN, comm2d, ierr ) |
|---|
| 265 | dt_u = reduce(1) |
|---|
| 266 | dt_v = reduce(2) |
|---|
| 267 | dt_w = reduce(3) |
|---|
| 268 | #else |
|---|
| 269 | dt_u = dt_u_l |
|---|
| 270 | dt_v = dt_v_l |
|---|
| 271 | dt_w = dt_w_l |
|---|
| 272 | #endif |
|---|
| 273 | |
|---|
| 274 | ! |
|---|
| 275 | !-- Compute time step according to the diffusion criterion. |
|---|
| 276 | !-- First calculate minimum grid spacing which only depends on index k. |
|---|
| 277 | !-- When using the dynamic subgrid model, negative km are possible. |
|---|
| 278 | dt_diff_l = 999999.0_wp |
|---|
| 279 | |
|---|
| 280 | !$ACC PARALLEL LOOP PRESENT(dxyz2_min, dzw) |
|---|
| 281 | DO k = nzb+1, nzt |
|---|
| 282 | dxyz2_min(k) = MIN( dx2, dy2, dzw(k)*dzw(k) ) * 0.125_wp |
|---|
| 283 | ENDDO |
|---|
| 284 | |
|---|
| 285 | !$OMP PARALLEL private(i,j,k) reduction(MIN: dt_diff_l) |
|---|
| 286 | !$OMP DO |
|---|
| 287 | !$ACC PARALLEL LOOP COLLAPSE(3) PRIVATE(i,j,k) & |
|---|
| 288 | !$ACC COPY(dt_diff_l) REDUCTION(MIN: dt_diff_l) & |
|---|
| 289 | !$ACC PRESENT(dxyz2_min, kh, km) |
|---|
| 290 | DO i = nxl, nxr |
|---|
| 291 | DO j = nys, nyn |
|---|
| 292 | DO k = nzb+1, nzt |
|---|
| 293 | dt_diff_l = MIN( dt_diff_l, & |
|---|
| 294 | dxyz2_min(k) / & |
|---|
| 295 | ( MAX( kh(k,j,i), 2.0_wp * ABS( km(k,j,i) ) ) & |
|---|
| 296 | + 1E-20_wp ) ) |
|---|
| 297 | ENDDO |
|---|
| 298 | ENDDO |
|---|
| 299 | ENDDO |
|---|
| 300 | !$OMP END PARALLEL |
|---|
| 301 | #if defined( __parallel ) |
|---|
| 302 | IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr ) |
|---|
| 303 | CALL MPI_ALLREDUCE( dt_diff_l, dt_diff, 1, MPI_REAL, MPI_MIN, comm2d, & |
|---|
| 304 | ierr ) |
|---|
| 305 | #else |
|---|
| 306 | dt_diff = dt_diff_l |
|---|
| 307 | #endif |
|---|
| 308 | |
|---|
| 309 | ! |
|---|
| 310 | !-- The time step is the minimum of the 3-4 components and the diffusion time |
|---|
| 311 | !-- step minus a reduction (cfl_factor) to be on the safe side. |
|---|
| 312 | !-- The time step must not exceed the maximum allowed value. |
|---|
| 313 | dt_3d = cfl_factor * MIN( dt_diff, dt_u, dt_v, dt_w, dt_precipitation ) |
|---|
| 314 | dt_3d = MIN( dt_3d, dt_max ) |
|---|
| 315 | |
|---|
| 316 | ! |
|---|
| 317 | !-- Remember the restricting time step criterion for later output. |
|---|
| 318 | IF ( MIN( dt_u, dt_v, dt_w ) < dt_diff ) THEN |
|---|
| 319 | timestep_reason = 'A' |
|---|
| 320 | ELSE |
|---|
| 321 | timestep_reason = 'D' |
|---|
| 322 | ENDIF |
|---|
| 323 | |
|---|
| 324 | ! |
|---|
| 325 | !-- Set flag if the time step becomes too small. |
|---|
| 326 | IF ( dt_3d < ( 0.00001_wp * dt_max ) ) THEN |
|---|
| 327 | stop_dt = .TRUE. |
|---|
| 328 | |
|---|
| 329 | ! |
|---|
| 330 | !-- Determine the maxima of the diffusion coefficients, including their |
|---|
| 331 | !-- grid index positions. |
|---|
| 332 | CALL global_min_max( nzb, nzt+1, nysg, nyng, nxlg, nxrg, km, 'abs', & |
|---|
| 333 | 0.0_wp, km_max, km_max_ijk ) |
|---|
| 334 | CALL global_min_max( nzb, nzt+1, nysg, nyng, nxlg, nxrg, kh, 'abs', & |
|---|
| 335 | 0.0_wp, kh_max, kh_max_ijk ) |
|---|
| 336 | |
|---|
| 337 | WRITE( message_string, * ) 'Time step has reached minimum limit.', & |
|---|
| 338 | '&dt = ', dt_3d, ' s Simulation is terminated.', & |
|---|
| 339 | '&dt_u = ', dt_u, ' s', & |
|---|
| 340 | '&dt_v = ', dt_v, ' s', & |
|---|
| 341 | '&dt_w = ', dt_w, ' s', & |
|---|
| 342 | '&dt_diff = ', dt_diff, ' s', & |
|---|
| 343 | '&u_max = ', u_max, ' m/s k=', u_max_ijk(1), & |
|---|
| 344 | ' j=', u_max_ijk(2), ' i=', u_max_ijk(3), & |
|---|
| 345 | '&v_max = ', v_max, ' m/s k=', v_max_ijk(1), & |
|---|
| 346 | ' j=', v_max_ijk(2), ' i=', v_max_ijk(3), & |
|---|
| 347 | '&w_max = ', w_max, ' m/s k=', w_max_ijk(1), & |
|---|
| 348 | ' j=', w_max_ijk(2), ' i=', w_max_ijk(3), & |
|---|
| 349 | '&km_max = ', km_max, ' m2/s2 k=', km_max_ijk(1), & |
|---|
| 350 | ' j=', km_max_ijk(2), ' i=', km_max_ijk(3), & |
|---|
| 351 | '&kh_max = ', kh_max, ' m2/s2 k=', kh_max_ijk(1), & |
|---|
| 352 | ' j=', kh_max_ijk(2), ' i=', kh_max_ijk(3) |
|---|
| 353 | CALL message( 'timestep', 'PA0312', 0, 1, 0, 6, 0 ) |
|---|
| 354 | ! |
|---|
| 355 | !-- In case of coupled runs inform the remote model of the termination |
|---|
| 356 | !-- and its reason, provided the remote model has not already been |
|---|
| 357 | !-- informed of another termination reason (terminate_coupled > 0) before. |
|---|
| 358 | #if defined( __parallel ) |
|---|
| 359 | IF ( coupling_mode /= 'uncoupled' .AND. terminate_coupled == 0 ) THEN |
|---|
| 360 | terminate_coupled = 2 |
|---|
| 361 | IF ( myid == 0 ) THEN |
|---|
| 362 | CALL MPI_SENDRECV( & |
|---|
| 363 | terminate_coupled, 1, MPI_INTEGER, target_id, 0, & |
|---|
| 364 | terminate_coupled_remote, 1, MPI_INTEGER, target_id, 0, & |
|---|
| 365 | comm_inter, status, ierr ) |
|---|
| 366 | ENDIF |
|---|
| 367 | CALL MPI_BCAST( terminate_coupled_remote, 1, MPI_INTEGER, 0, & |
|---|
| 368 | comm2d, ierr) |
|---|
| 369 | ENDIF |
|---|
| 370 | #endif |
|---|
| 371 | ENDIF |
|---|
| 372 | |
|---|
| 373 | ! |
|---|
| 374 | !-- In case of nested runs all parent/child processes have to terminate if |
|---|
| 375 | !-- one process has set the stop flag, i.e. they need to set the stop flag |
|---|
| 376 | !-- too. |
|---|
| 377 | IF ( nested_run ) THEN |
|---|
| 378 | stop_dt_local = stop_dt |
|---|
| 379 | #if defined( __parallel ) |
|---|
| 380 | CALL MPI_ALLREDUCE( stop_dt_local, stop_dt, 1, MPI_LOGICAL, MPI_LOR, & |
|---|
| 381 | MPI_COMM_WORLD, ierr ) |
|---|
| 382 | #endif |
|---|
| 383 | ENDIF |
|---|
| 384 | |
|---|
| 385 | ! |
|---|
| 386 | !-- Ensure a smooth value (two significant digits) of the timestep. |
|---|
| 387 | div = 1000.0_wp |
|---|
| 388 | DO WHILE ( dt_3d < div ) |
|---|
| 389 | div = div / 10.0_wp |
|---|
| 390 | ENDDO |
|---|
| 391 | dt_3d = NINT( dt_3d * 100.0_wp / div ) * div / 100.0_wp |
|---|
| 392 | |
|---|
| 393 | ENDIF |
|---|
| 394 | |
|---|
| 395 | #if defined( __parallel ) |
|---|
| 396 | ! |
|---|
| 397 | !-- Vertical nesting: coarse and fine grid timestep has to be identical |
|---|
| 398 | IF ( vnested ) CALL vnest_timestep_sync |
|---|
| 399 | #endif |
|---|
| 400 | |
|---|
| 401 | CALL cpu_log( log_point(12), 'calculate_timestep', 'stop' ) |
|---|
| 402 | |
|---|
| 403 | END SUBROUTINE timestep |
|---|
