Home

Context Navigation

source: palm/trunk/SOURCE/pres.f90 @ 4454

Last change on this file since 4454 was 4360, checked in by suehring, 5 years ago
Bugfix in output of time-averaged plant-canopy quanities; Output of plant-canopy data only where tall canopy is defined; land-surface model: fix wrong location strings; tests: update urban test case; all source code files: copyright update
Property svn:keywords set to `Id`
File size: 30.1 KB

Rev	Line
[1682]	1	!> @file pres.f90
[2000]	2	!------------------------------------------------------------------------------!
[2696]	3	! This file is part of the PALM model system.
[1036]	4	!
[2000]	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.
[1036]	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	!
[4360]	17	! Copyright 1997-2020 Leibniz Universitaet Hannover
[2000]	18	!------------------------------------------------------------------------------!
[1036]	19	!
[484]	20	! Current revisions:
[1212]	21	! ------------------
[1932]	22	!
[3183]	23	!
[1919]	24	! Former revisions:
	25	! -----------------
[3016]	26	! $Id: pres.f90 4360 2020-01-07 11:25:50Z forkel $
[4346]	27	! Introduction of wall_flags_total_0, which currently sets bits based on static
	28	! topography information used in wall_flags_static_0
	29	!
	30	! 4329 2019-12-10 15:46:36Z motisi
[4329]	31	! Renamed wall_flags_0 to wall_flags_static_0
	32	!
	33	! 4182 2019-08-22 15:20:23Z scharf
[4182]	34	! Corrected "Former revisions" section
	35	!
	36	! 4015 2019-06-05 13:25:35Z raasch
[4015]	37	! variable child_domain_nvn eliminated
	38	!
	39	! 3849 2019-04-01 16:35:16Z knoop
[3634]	40	! OpenACC port for SPEC
[1919]	41	!
[4182]	42	! Revision 1.1 1997/07/24 11:24:44 raasch
	43	! Initial revision
	44	!
	45	!
[1]	46	! Description:
	47	! ------------
[1682]	48	!> Compute the divergence of the provisional velocity field. Solve the Poisson
	49	!> equation for the perturbation pressure. Compute the final velocities using
	50	!> this perturbation pressure. Compute the remaining divergence.
[1]	51	!------------------------------------------------------------------------------!
[1682]	52	SUBROUTINE pres
	53
[1]	54
[1320]	55	USE arrays_3d, &
[2037]	56	ONLY: d, ddzu, ddzu_pres, ddzw, dzw, p, p_loc, rho_air, rho_air_zw, &
	57	tend, u, v, w
[1320]	58
	59	USE control_parameters, &
[3182]	60	ONLY: bc_lr_cyc, bc_ns_cyc, bc_radiation_l, bc_radiation_n, &
	61	bc_radiation_r, bc_radiation_s, child_domain, &
	62	conserve_volume_flow, coupling_mode, &
[3012]	63	dt_3d, gathered_size, ibc_p_b, ibc_p_t, &
	64	intermediate_timestep_count, intermediate_timestep_count_max, &
[3347]	65	mg_switch_to_pe0_level, nesting_offline, &
	66	psolver, subdomain_size, &
[3182]	67	topography, volume_flow, volume_flow_area, volume_flow_initial
[1320]	68
	69	USE cpulog, &
	70	ONLY: cpu_log, log_point, log_point_s
	71
	72	USE grid_variables, &
	73	ONLY: ddx, ddy
	74
	75	USE indices, &
	76	ONLY: nbgp, ngp_2dh_outer, nx, nxl, nxlg, nxl_mg, nxr, nxrg, nxr_mg, &
[2232]	77	ny, nys, nysg, nys_mg, nyn, nyng, nyn_mg, nzb, nzt, nzt_mg, &
[4346]	78	wall_flags_total_0
[1320]	79
	80	USE kinds
	81
[1]	82	USE pegrid
[1933]	83
	84	USE pmc_interface, &
	85	ONLY: nesting_mode
[1]	86
[1320]	87	USE poisfft_mod, &
	88	ONLY: poisfft
	89
[1575]	90	USE poismg_mod
	91
[2696]	92	USE poismg_noopt_mod
	93
[1320]	94	USE statistics, &
	95	ONLY: statistic_regions, sums_divnew_l, sums_divold_l, weight_pres, &
	96	weight_substep
	97
[2232]	98	USE surface_mod, &
	99	ONLY : bc_h
	100
[1]	101	IMPLICIT NONE
	102
[1682]	103	INTEGER(iwp) :: i !<
	104	INTEGER(iwp) :: j !<
	105	INTEGER(iwp) :: k !<
[2232]	106	INTEGER(iwp) :: m !<
[1]	107
[1682]	108	REAL(wp) :: ddt_3d !<
[1918]	109	REAL(wp) :: d_weight_pres !<
[1682]	110	REAL(wp) :: localsum !<
	111	REAL(wp) :: threadsum !<
[1918]	112	REAL(wp) :: weight_pres_l !<
[1929]	113	REAL(wp) :: weight_substep_l !<
[1]	114
[1762]	115	REAL(wp), DIMENSION(1:3) :: volume_flow_l !<
	116	REAL(wp), DIMENSION(1:3) :: volume_flow_offset !<
[1682]	117	REAL(wp), DIMENSION(1:nzt) :: w_l !<
	118	REAL(wp), DIMENSION(1:nzt) :: w_l_l !<
[1]	119
	120
	121	CALL cpu_log( log_point(8), 'pres', 'start' )
	122
[1918]	123	!
	124	!-- Calculate quantities to be used locally
[1342]	125	ddt_3d = 1.0_wp / dt_3d
[1918]	126	IF ( intermediate_timestep_count == 0 ) THEN
	127	!
	128	!-- If pres is called before initial time step
[1929]	129	weight_pres_l = 1.0_wp
	130	d_weight_pres = 1.0_wp
	131	weight_substep_l = 1.0_wp
[1918]	132	ELSE
[1929]	133	weight_pres_l = weight_pres(intermediate_timestep_count)
	134	d_weight_pres = 1.0_wp / weight_pres(intermediate_timestep_count)
	135	weight_substep_l = weight_substep(intermediate_timestep_count)
[1918]	136	ENDIF
[85]	137
[1]	138	!
[707]	139	!-- Multigrid method expects array d to have one ghost layer.
	140	!--
[1575]	141	IF ( psolver(1:9) == 'multigrid' ) THEN
[667]	142
[1]	143	DEALLOCATE( d )
[667]	144	ALLOCATE( d(nzb:nzt+1,nys-1:nyn+1,nxl-1:nxr+1) )
[707]	145
	146	!
	147	!-- Since p is later used to hold the weighted average of the substeps, it
	148	!-- cannot be used in the iterative solver. Therefore, its initial value is
	149	!-- stored on p_loc, which is then iteratively advanced in every substep.
[1762]	150	IF ( intermediate_timestep_count <= 1 ) THEN
[707]	151	DO i = nxl-1, nxr+1
	152	DO j = nys-1, nyn+1
	153	DO k = nzb, nzt+1
	154	p_loc(k,j,i) = p(k,j,i)
	155	ENDDO
	156	ENDDO
	157	ENDDO
[667]	158	ENDIF
	159
[1918]	160	ELSEIF ( psolver == 'sor' .AND. intermediate_timestep_count <= 1 ) THEN
[707]	161
	162	!
	163	!-- Since p is later used to hold the weighted average of the substeps, it
	164	!-- cannot be used in the iterative solver. Therefore, its initial value is
	165	!-- stored on p_loc, which is then iteratively advanced in every substep.
	166	p_loc = p
	167
[1]	168	ENDIF
	169
	170	!
[75]	171	!-- Conserve the volume flow at the outflow in case of non-cyclic lateral
	172	!-- boundary conditions
[106]	173	!-- WARNING: so far, this conservation does not work at the left/south
	174	!-- boundary if the topography at the inflow differs from that at the
	175	!-- outflow! For this case, volume_flow_area needs adjustment!
	176	!
	177	!-- Left/right
[3182]	178	IF ( conserve_volume_flow .AND. ( bc_radiation_l .OR. &
	179	bc_radiation_r ) ) THEN
[680]	180
[1342]	181	volume_flow(1) = 0.0_wp
	182	volume_flow_l(1) = 0.0_wp
[106]	183
[3182]	184	IF ( bc_radiation_l ) THEN
[106]	185	i = 0
[3182]	186	ELSEIF ( bc_radiation_r ) THEN
[106]	187	i = nx+1
	188	ENDIF
	189
	190	DO j = nys, nyn
	191	!
	192	!-- Sum up the volume flow through the south/north boundary
[2232]	193	DO k = nzb+1, nzt
	194	volume_flow_l(1) = volume_flow_l(1) + u(k,j,i) * dzw(k) &
	195	* MERGE( 1.0_wp, 0.0_wp, &
[4346]	196	BTEST( wall_flags_total_0(k,j,i), 1 ) &
[2232]	197	)
[106]	198	ENDDO
	199	ENDDO
	200
	201	#if defined( __parallel )
[680]	202	IF ( collective_wait ) CALL MPI_BARRIER( comm1dy, ierr )
[106]	203	CALL MPI_ALLREDUCE( volume_flow_l(1), volume_flow(1), 1, MPI_REAL, &
	204	MPI_SUM, comm1dy, ierr )
	205	#else
	206	volume_flow = volume_flow_l
	207	#endif
[709]	208	volume_flow_offset(1) = ( volume_flow_initial(1) - volume_flow(1) ) &
[106]	209	/ volume_flow_area(1)
	210
[667]	211	DO j = nysg, nyng
[2232]	212	DO k = nzb+1, nzt
	213	u(k,j,i) = u(k,j,i) + volume_flow_offset(1) &
	214	* MERGE( 1.0_wp, 0.0_wp, &
[4346]	215	BTEST( wall_flags_total_0(k,j,i), 1 ) &
[2232]	216	)
[106]	217	ENDDO
	218	ENDDO
	219
	220	ENDIF
	221
	222	!
	223	!-- South/north
[3182]	224	IF ( conserve_volume_flow .AND. ( bc_radiation_n .OR. bc_radiation_s ) ) THEN
[106]	225
[1342]	226	volume_flow(2) = 0.0_wp
	227	volume_flow_l(2) = 0.0_wp
[75]	228
[3182]	229	IF ( bc_radiation_s ) THEN
[106]	230	j = 0
[3182]	231	ELSEIF ( bc_radiation_n ) THEN
[75]	232	j = ny+1
[106]	233	ENDIF
	234
	235	DO i = nxl, nxr
[75]	236	!
[106]	237	!-- Sum up the volume flow through the south/north boundary
[2232]	238	DO k = nzb+1, nzt
	239	volume_flow_l(2) = volume_flow_l(2) + v(k,j,i) * dzw(k) &
	240	* MERGE( 1.0_wp, 0.0_wp, &
[4346]	241	BTEST( wall_flags_total_0(k,j,i), 2 ) &
[2232]	242	)
[75]	243	ENDDO
[106]	244	ENDDO
	245
[75]	246	#if defined( __parallel )
[680]	247	IF ( collective_wait ) CALL MPI_BARRIER( comm1dx, ierr )
[75]	248	CALL MPI_ALLREDUCE( volume_flow_l(2), volume_flow(2), 1, MPI_REAL, &
	249	MPI_SUM, comm1dx, ierr )
	250	#else
	251	volume_flow = volume_flow_l
	252	#endif
	253	volume_flow_offset(2) = ( volume_flow_initial(2) - volume_flow(2) ) &
[106]	254	/ volume_flow_area(2)
[75]	255
[667]	256	DO i = nxlg, nxrg
[2232]	257	DO k = nzb+1, nzt
	258	v(k,j,i) = v(k,j,i) + volume_flow_offset(2) &
	259	* MERGE( 1.0_wp, 0.0_wp, &
[4346]	260	BTEST( wall_flags_total_0(k,j,i), 2 ) &
[2232]	261	)
[75]	262	ENDDO
[106]	263	ENDDO
[75]	264
	265	ENDIF
	266
[76]	267	!
[4015]	268	!-- Remove mean vertical velocity in case that Neumann conditions are used both at bottom and top
	269	!-- boundary. With Neumann conditions at both vertical boundaries, the solver cannot remove
	270	!-- mean vertical velocities. They should be removed, because incompressibility requires that
	271	!-- the vertical gradient of vertical velocity is zero. Since w=0 at the solid surface, it must be
	272	!-- zero everywhere.
	273	!-- This must not be done in case of a 3d-nesting child domain, because a mean vertical velocity
	274	!-- can physically exist in such a domain.
	275	!-- Also in case of offline nesting, mean vertical velocities may exist (and must not be removed),
	276	!-- caused by horizontal divergence/convergence of the large scale flow that is prescribed at the
	277	!-- side boundaries.
	278	!-- The removal cannot be done before the first initial time step because ngp_2dh_outer
[1918]	279	!-- is not yet known then.
[4015]	280	IF ( ibc_p_b == 1 .AND. ibc_p_t == 1 .AND. .NOT. nesting_offline &
	281	.AND. .NOT. ( child_domain .AND. nesting_mode /= 'vertical' ) &
	282	.AND. intermediate_timestep_count /= 0 ) &
[1918]	283	THEN
	284	w_l = 0.0_wp; w_l_l = 0.0_wp
	285	DO i = nxl, nxr
	286	DO j = nys, nyn
[2232]	287	DO k = nzb+1, nzt
[4015]	288	w_l_l(k) = w_l_l(k) + w(k,j,i) &
[4346]	289	* MERGE( 1.0_wp, 0.0_wp, BTEST( wall_flags_total_0(k,j,i), 3 ) )
[76]	290	ENDDO
	291	ENDDO
[1918]	292	ENDDO
[76]	293	#if defined( __parallel )
[1918]	294	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
[4015]	295	CALL MPI_ALLREDUCE( w_l_l(1), w_l(1), nzt, MPI_REAL, MPI_SUM, comm2d, ierr )
[76]	296	#else
[1918]	297	w_l = w_l_l
[76]	298	#endif
[1918]	299	DO k = 1, nzt
	300	w_l(k) = w_l(k) / ngp_2dh_outer(k,0)
	301	ENDDO
	302	DO i = nxlg, nxrg
	303	DO j = nysg, nyng
[2232]	304	DO k = nzb+1, nzt
[4015]	305	w(k,j,i) = w(k,j,i) - w_l(k) &
[4346]	306	* MERGE( 1.0_wp, 0.0_wp, BTEST( wall_flags_total_0(k,j,i), 3 ) )
[76]	307	ENDDO
	308	ENDDO
[1918]	309	ENDDO
[76]	310	ENDIF
[75]	311
	312	!
[1]	313	!-- Compute the divergence of the provisional velocity field.
	314	CALL cpu_log( log_point_s(1), 'divergence', 'start' )
	315
[1575]	316	IF ( psolver(1:9) == 'multigrid' ) THEN
[2232]	317	!$OMP PARALLEL DO SCHEDULE( STATIC ) PRIVATE (i,j,k)
[1]	318	DO i = nxl-1, nxr+1
	319	DO j = nys-1, nyn+1
	320	DO k = nzb, nzt+1
[1342]	321	d(k,j,i) = 0.0_wp
[1]	322	ENDDO
	323	ENDDO
	324	ENDDO
	325	ELSE
[2232]	326	!$OMP PARALLEL DO SCHEDULE( STATIC ) PRIVATE (i,j,k)
[3634]	327	!$ACC PARALLEL LOOP COLLAPSE(3) PRIVATE(i, j, k) &
	328	!$ACC PRESENT(d)
[1003]	329	DO i = nxl, nxr
	330	DO j = nys, nyn
	331	DO k = nzb+1, nzt
[1342]	332	d(k,j,i) = 0.0_wp
[1]	333	ENDDO
	334	ENDDO
	335	ENDDO
	336	ENDIF
	337
[1342]	338	localsum = 0.0_wp
	339	threadsum = 0.0_wp
[1]	340
	341	#if defined( __ibm )
	342	!$OMP PARALLEL PRIVATE (i,j,k) FIRSTPRIVATE(threadsum) REDUCTION(+:localsum)
	343	!$OMP DO SCHEDULE( STATIC )
	344	DO i = nxl, nxr
	345	DO j = nys, nyn
[2232]	346	DO k = nzb+1, nzt
[2037]	347	d(k,j,i) = ( ( u(k,j,i+1) - u(k,j,i) ) * rho_air(k) * ddx + &
	348	( v(k,j+1,i) - v(k,j,i) ) * rho_air(k) * ddy + &
	349	( w(k,j,i) * rho_air_zw(k) - &
	350	w(k-1,j,i) * rho_air_zw(k-1) ) * ddzw(k) &
[2232]	351	) * ddt_3d * d_weight_pres &
	352	* MERGE( 1.0_wp, 0.0_wp, &
[4346]	353	BTEST( wall_flags_total_0(k,j,i), 0 ) &
[2232]	354	)
[1]	355	ENDDO
	356	!
	357	!-- Compute possible PE-sum of divergences for flow_statistics
[2232]	358	DO k = nzb+1, nzt
	359	threadsum = threadsum + ABS( d(k,j,i) ) &
	360	* MERGE( 1.0_wp, 0.0_wp, &
[4346]	361	BTEST( wall_flags_total_0(k,j,i), 0 ) &
[2232]	362	)
[1]	363	ENDDO
	364
	365	ENDDO
	366	ENDDO
	367
[1918]	368	IF ( intermediate_timestep_count == intermediate_timestep_count_max .OR. &
	369	intermediate_timestep_count == 0 ) THEN
	370	localsum = localsum + threadsum * dt_3d * weight_pres_l
[1908]	371	ENDIF
[1]	372	!$OMP END PARALLEL
	373	#else
	374
[1111]	375	!$OMP PARALLEL PRIVATE (i,j,k)
	376	!$OMP DO SCHEDULE( STATIC )
[3634]	377	!$ACC PARALLEL LOOP COLLAPSE(3) PRIVATE(i, j, k) &
[4346]	378	!$ACC PRESENT(u, v, w, rho_air, rho_air_zw, ddzw, wall_flags_total_0) &
[3634]	379	!$ACC PRESENT(d)
[1111]	380	DO i = nxl, nxr
	381	DO j = nys, nyn
	382	DO k = 1, nzt
[2037]	383	d(k,j,i) = ( ( u(k,j,i+1) - u(k,j,i) ) * rho_air(k) * ddx + &
	384	( v(k,j+1,i) - v(k,j,i) ) * rho_air(k) * ddy + &
	385	( w(k,j,i) * rho_air_zw(k) - &
	386	w(k-1,j,i) * rho_air_zw(k-1) ) * ddzw(k) &
[2232]	387	) * ddt_3d * d_weight_pres &
	388	* MERGE( 1.0_wp, 0.0_wp, &
[4346]	389	BTEST( wall_flags_total_0(k,j,i), 0 ) &
[2232]	390	)
[1]	391	ENDDO
	392	ENDDO
[1111]	393	ENDDO
	394	!$OMP END PARALLEL
[1]	395
	396	!
[1908]	397	!-- Compute possible PE-sum of divergences for flow_statistics. Carry out
	398	!-- computation only at last Runge-Kutta substep.
[1918]	399	IF ( intermediate_timestep_count == intermediate_timestep_count_max .OR. &
	400	intermediate_timestep_count == 0 ) THEN
[1908]	401	!$OMP PARALLEL PRIVATE (i,j,k) FIRSTPRIVATE(threadsum) REDUCTION(+:localsum)
	402	!$OMP DO SCHEDULE( STATIC )
[3634]	403	!$ACC PARALLEL LOOP COLLAPSE(3) PRIVATE(i,j,k) &
	404	!$ACC REDUCTION(+:threadsum) COPY(threadsum) &
	405	!$ACC PRESENT(d)
[1908]	406	DO i = nxl, nxr
	407	DO j = nys, nyn
	408	DO k = nzb+1, nzt
	409	threadsum = threadsum + ABS( d(k,j,i) )
	410	ENDDO
[1]	411	ENDDO
	412	ENDDO
[1918]	413	localsum = localsum + threadsum * dt_3d * weight_pres_l
[1908]	414	!$OMP END PARALLEL
	415	ENDIF
[1]	416	#endif
	417
	418	!
	419	!-- For completeness, set the divergence sum of all statistic regions to those
	420	!-- of the total domain
[1918]	421	IF ( intermediate_timestep_count == intermediate_timestep_count_max .OR. &
	422	intermediate_timestep_count == 0 ) THEN
[1908]	423	sums_divold_l(0:statistic_regions) = localsum
[1918]	424	ENDIF
[1]	425
	426	CALL cpu_log( log_point_s(1), 'divergence', 'stop' )
	427
	428	!
	429	!-- Compute the pressure perturbation solving the Poisson equation
[1575]	430	IF ( psolver == 'poisfft' ) THEN
[1]	431
	432	!
	433	!-- Solve Poisson equation via FFT and solution of tridiagonal matrices
[1575]	434	CALL poisfft( d )
[1212]	435
[1]	436	!
	437	!-- Store computed perturbation pressure and set boundary condition in
	438	!-- z-direction
[2232]	439	!$OMP PARALLEL DO PRIVATE (i,j,k)
[3634]	440	!$ACC PARALLEL LOOP COLLAPSE(3) PRIVATE(i, j, k) &
	441	!$ACC PRESENT(d, tend)
[1]	442	DO i = nxl, nxr
	443	DO j = nys, nyn
	444	DO k = nzb+1, nzt
	445	tend(k,j,i) = d(k,j,i)
	446	ENDDO
	447	ENDDO
	448	ENDDO
	449
	450	!
	451	!-- Bottom boundary:
	452	!-- This condition is only required for internal output. The pressure
	453	!-- gradient (dp(nzb+1)-dp(nzb))/dz is not used anywhere else.
	454	IF ( ibc_p_b == 1 ) THEN
	455	!
[2232]	456	!-- Neumann (dp/dz = 0). Using surfae data type, first for non-natural
	457	!-- surfaces, then for natural and urban surfaces
	458	!-- Upward facing
	459	!$OMP PARALLEL DO PRIVATE( i, j, k )
[3634]	460	!$ACC PARALLEL LOOP PRIVATE(i, j, k) &
	461	!$ACC PRESENT(bc_h, tend)
[2232]	462	DO m = 1, bc_h(0)%ns
[2298]	463	i = bc_h(0)%i(m)
	464	j = bc_h(0)%j(m)
	465	k = bc_h(0)%k(m)
[2232]	466	tend(k-1,j,i) = tend(k,j,i)
[1]	467	ENDDO
[2232]	468	!
	469	!-- Downward facing
	470	!$OMP PARALLEL DO PRIVATE( i, j, k )
[3634]	471	!$ACC PARALLEL LOOP PRIVATE(i, j, k) &
	472	!$ACC PRESENT(bc_h, tend)
[2232]	473	DO m = 1, bc_h(1)%ns
[2298]	474	i = bc_h(1)%i(m)
	475	j = bc_h(1)%j(m)
	476	k = bc_h(1)%k(m)
[2232]	477	tend(k+1,j,i) = tend(k,j,i)
	478	ENDDO
[1]	479
	480	ELSE
	481	!
[2232]	482	!-- Dirichlet. Using surface data type, first for non-natural
	483	!-- surfaces, then for natural and urban surfaces
	484	!-- Upward facing
	485	!$OMP PARALLEL DO PRIVATE( i, j, k )
	486	DO m = 1, bc_h(0)%ns
[2298]	487	i = bc_h(0)%i(m)
	488	j = bc_h(0)%j(m)
	489	k = bc_h(0)%k(m)
[2232]	490	tend(k-1,j,i) = 0.0_wp
[1]	491	ENDDO
[2232]	492	!
	493	!-- Downward facing
	494	!$OMP PARALLEL DO PRIVATE( i, j, k )
	495	DO m = 1, bc_h(1)%ns
[2298]	496	i = bc_h(1)%i(m)
	497	j = bc_h(1)%j(m)
	498	k = bc_h(1)%k(m)
[2232]	499	tend(k+1,j,i) = 0.0_wp
	500	ENDDO
[1]	501
	502	ENDIF
	503
	504	!
	505	!-- Top boundary
	506	IF ( ibc_p_t == 1 ) THEN
	507	!
	508	!-- Neumann
[2232]	509	!$OMP PARALLEL DO PRIVATE (i,j,k)
[667]	510	DO i = nxlg, nxrg
	511	DO j = nysg, nyng
[1]	512	tend(nzt+1,j,i) = tend(nzt,j,i)
	513	ENDDO
	514	ENDDO
	515
	516	ELSE
	517	!
	518	!-- Dirichlet
[2232]	519	!$OMP PARALLEL DO PRIVATE (i,j,k)
[3634]	520	!$ACC PARALLEL LOOP COLLAPSE(2) PRIVATE(i, j) &
	521	!$ACC PRESENT(tend)
[667]	522	DO i = nxlg, nxrg
	523	DO j = nysg, nyng
[1342]	524	tend(nzt+1,j,i) = 0.0_wp
[1]	525	ENDDO
	526	ENDDO
	527
	528	ENDIF
	529
	530	!
	531	!-- Exchange boundaries for p
[667]	532	CALL exchange_horiz( tend, nbgp )
[1]	533
	534	ELSEIF ( psolver == 'sor' ) THEN
	535
	536	!
	537	!-- Solve Poisson equation for perturbation pressure using SOR-Red/Black
	538	!-- scheme
[707]	539	CALL sor( d, ddzu_pres, ddzw, p_loc )
	540	tend = p_loc
[1]	541
[1575]	542	ELSEIF ( psolver(1:9) == 'multigrid' ) THEN
[1]	543
	544	!
	545	!-- Solve Poisson equation for perturbation pressure using Multigrid scheme,
[2298]	546	!-- array tend is used to store the residuals.
[778]	547
	548	!-- If the number of grid points of the gathered grid, which is collected
	549	!-- on PE0, is larger than the number of grid points of an PE, than array
	550	!-- tend will be enlarged.
	551	IF ( gathered_size > subdomain_size ) THEN
	552	DEALLOCATE( tend )
	553	ALLOCATE( tend(nzb:nzt_mg(mg_switch_to_pe0_level)+1,nys_mg( &
	554	mg_switch_to_pe0_level)-1:nyn_mg(mg_switch_to_pe0_level)+1,&
	555	nxl_mg(mg_switch_to_pe0_level)-1:nxr_mg( &
	556	mg_switch_to_pe0_level)+1) )
	557	ENDIF
	558
[1575]	559	IF ( psolver == 'multigrid' ) THEN
	560	CALL poismg( tend )
	561	ELSE
[1931]	562	CALL poismg_noopt( tend )
[1575]	563	ENDIF
[707]	564
[778]	565	IF ( gathered_size > subdomain_size ) THEN
	566	DEALLOCATE( tend )
	567	ALLOCATE( tend(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
	568	ENDIF
	569
[1]	570	!
	571	!-- Restore perturbation pressure on tend because this array is used
	572	!-- further below to correct the velocity fields
[707]	573	DO i = nxl-1, nxr+1
	574	DO j = nys-1, nyn+1
	575	DO k = nzb, nzt+1
	576	tend(k,j,i) = p_loc(k,j,i)
	577	ENDDO
	578	ENDDO
	579	ENDDO
[667]	580
[1]	581	ENDIF
	582
	583	!
[707]	584	!-- Store perturbation pressure on array p, used for pressure data output.
	585	!-- Ghost layers are added in the output routines (except sor-method: see below)
[1918]	586	IF ( intermediate_timestep_count <= 1 ) THEN
[707]	587	!$OMP PARALLEL PRIVATE (i,j,k)
	588	!$OMP DO
[3634]	589	!$ACC PARALLEL LOOP COLLAPSE(3) PRIVATE(i, j, k) &
	590	!$ACC PRESENT(p, tend)
[707]	591	DO i = nxl-1, nxr+1
	592	DO j = nys-1, nyn+1
	593	DO k = nzb, nzt+1
	594	p(k,j,i) = tend(k,j,i) * &
[1929]	595	weight_substep_l
[673]	596	ENDDO
[1]	597	ENDDO
[707]	598	ENDDO
	599	!$OMP END PARALLEL
	600
[1762]	601	ELSEIF ( intermediate_timestep_count > 1 ) THEN
[707]	602	!$OMP PARALLEL PRIVATE (i,j,k)
	603	!$OMP DO
[3634]	604	!$ACC PARALLEL LOOP COLLAPSE(3) PRIVATE(i, j, k) &
	605	!$ACC PRESENT(p, tend)
[707]	606	DO i = nxl-1, nxr+1
	607	DO j = nys-1, nyn+1
	608	DO k = nzb, nzt+1
	609	p(k,j,i) = p(k,j,i) + tend(k,j,i) * &
[1929]	610	weight_substep_l
[673]	611	ENDDO
	612	ENDDO
[707]	613	ENDDO
	614	!$OMP END PARALLEL
	615
	616	ENDIF
[673]	617
[707]	618	!
	619	!-- SOR-method needs ghost layers for the next timestep
	620	IF ( psolver == 'sor' ) CALL exchange_horiz( p, nbgp )
[682]	621
[1]	622	!
	623	!-- Correction of the provisional velocities with the current perturbation
	624	!-- pressure just computed
[709]	625	IF ( conserve_volume_flow .AND. ( bc_lr_cyc .OR. bc_ns_cyc ) ) THEN
[1342]	626	volume_flow_l(1) = 0.0_wp
	627	volume_flow_l(2) = 0.0_wp
[1]	628	ENDIF
[3347]	629	!
	630	!-- Add pressure gradients to the velocity components. Note, the loops are
	631	!-- running over the entire model domain, even though, in case of non-cyclic
	632	!-- boundaries u- and v-component are not prognostic at i=0 and j=0,
	633	!-- respectiveley. However, in case of Dirichlet boundary conditions for the
	634	!-- velocities, zero-gradient conditions for the pressure are set, so that
	635	!-- no modification is imposed at the boundaries.
[1]	636	!$OMP PARALLEL PRIVATE (i,j,k)
	637	!$OMP DO
[3634]	638	!$ACC PARALLEL LOOP COLLAPSE(2) PRIVATE(i, j, k) &
[4346]	639	!$ACC PRESENT(u, v, w, tend, ddzu, wall_flags_total_0)
[673]	640	DO i = nxl, nxr
[1]	641	DO j = nys, nyn
[2118]	642
[2232]	643	DO k = nzb+1, nzt
	644	w(k,j,i) = w(k,j,i) - dt_3d * &
	645	( tend(k+1,j,i) - tend(k,j,i) ) * ddzu(k+1) &
	646	* weight_pres_l &
	647	* MERGE( 1.0_wp, 0.0_wp, &
[4346]	648	BTEST( wall_flags_total_0(k,j,i), 3 ) &
[2232]	649	)
[1]	650	ENDDO
[2118]	651
[2232]	652	DO k = nzb+1, nzt
	653	u(k,j,i) = u(k,j,i) - dt_3d * &
	654	( tend(k,j,i) - tend(k,j,i-1) ) * ddx &
	655	* weight_pres_l &
	656	* MERGE( 1.0_wp, 0.0_wp, &
[4346]	657	BTEST( wall_flags_total_0(k,j,i), 1 ) &
[2232]	658	)
[1]	659	ENDDO
[2118]	660
[2232]	661	DO k = nzb+1, nzt
	662	v(k,j,i) = v(k,j,i) - dt_3d * &
	663	( tend(k,j,i) - tend(k,j-1,i) ) * ddy &
	664	* weight_pres_l &
	665	* MERGE( 1.0_wp, 0.0_wp, &
[4346]	666	BTEST( wall_flags_total_0(k,j,i), 2 ) &
[2232]	667	)
[673]	668	ENDDO
[1]	669
	670	ENDDO
	671	ENDDO
	672	!$OMP END PARALLEL
[1113]	673
	674	!
[3012]	675	!-- The vertical velocity is not set to zero at nzt + 1 for nested domains
	676	!-- Instead it is set to the values of nzt (see routine vnest_boundary_conds
	677	!-- or pmci_interp_tril_t) BEFORE calling the pressure solver. To avoid jumps
	678	!-- while plotting profiles w at the top has to be set to the values in the
	679	!-- height nzt after above modifications. Hint: w level nzt+1 does not impact
	680	!-- results.
[3182]	681	IF ( child_domain .OR. coupling_mode == 'vnested_fine' ) THEN
[3012]	682	w(nzt+1,:,:) = w(nzt,:,:)
	683	ENDIF
	684
	685	!
[1113]	686	!-- Sum up the volume flow through the right and north boundary
[2232]	687	IF ( conserve_volume_flow .AND. bc_lr_cyc .AND. bc_ns_cyc .AND. &
[1113]	688	nxr == nx ) THEN
	689
	690	!$OMP PARALLEL PRIVATE (j,k)
	691	!$OMP DO
	692	DO j = nys, nyn
	693	!$OMP CRITICAL
[2232]	694	DO k = nzb+1, nzt
	695	volume_flow_l(1) = volume_flow_l(1) + u(k,j,nxr) * dzw(k) &
	696	* MERGE( 1.0_wp, 0.0_wp, &
[4346]	697	BTEST( wall_flags_total_0(k,j,nxr), 1 )&
[2232]	698	)
[1113]	699	ENDDO
	700	!$OMP END CRITICAL
	701	ENDDO
	702	!$OMP END PARALLEL
	703
	704	ENDIF
	705
[2232]	706	IF ( conserve_volume_flow .AND. bc_ns_cyc .AND. bc_lr_cyc .AND. &
[1113]	707	nyn == ny ) THEN
	708
	709	!$OMP PARALLEL PRIVATE (i,k)
	710	!$OMP DO
	711	DO i = nxl, nxr
	712	!$OMP CRITICAL
[2232]	713	DO k = nzb+1, nzt
	714	volume_flow_l(2) = volume_flow_l(2) + v(k,nyn,i) * dzw(k) &
	715	* MERGE( 1.0_wp, 0.0_wp, &
[4346]	716	BTEST( wall_flags_total_0(k,nyn,i), 2 )&
[2232]	717	)
[1113]	718	ENDDO
	719	!$OMP END CRITICAL
	720	ENDDO
	721	!$OMP END PARALLEL
	722
	723	ENDIF
[673]	724
[1]	725	!
	726	!-- Conserve the volume flow
[707]	727	IF ( conserve_volume_flow .AND. ( bc_lr_cyc .AND. bc_ns_cyc ) ) THEN
[1]	728
	729	#if defined( __parallel )
[622]	730	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
[1]	731	CALL MPI_ALLREDUCE( volume_flow_l(1), volume_flow(1), 2, MPI_REAL, &
	732	MPI_SUM, comm2d, ierr )
	733	#else
	734	volume_flow = volume_flow_l
	735	#endif
	736
[1799]	737	volume_flow_offset(1:2) = ( volume_flow_initial(1:2) - volume_flow(1:2) ) / &
	738	volume_flow_area(1:2)
[1]	739
	740	!$OMP PARALLEL PRIVATE (i,j,k)
	741	!$OMP DO
	742	DO i = nxl, nxr
	743	DO j = nys, nyn
[2232]	744	DO k = nzb+1, nzt
	745	u(k,j,i) = u(k,j,i) + volume_flow_offset(1) &
	746	* MERGE( 1.0_wp, 0.0_wp, &
[4346]	747	BTEST( wall_flags_total_0(k,j,i), 1 ) &
[2232]	748	)
[719]	749	ENDDO
[2232]	750	DO k = nzb+1, nzt
	751	v(k,j,i) = v(k,j,i) + volume_flow_offset(2) &
	752	* MERGE( 1.0_wp, 0.0_wp, &
[4346]	753	BTEST( wall_flags_total_0(k,j,i), 2 ) &
[2232]	754	)
[667]	755	ENDDO
[1]	756	ENDDO
	757	ENDDO
[667]	758
[1]	759	!$OMP END PARALLEL
	760
	761	ENDIF
	762
	763	!
	764	!-- Exchange of boundaries for the velocities
[667]	765	CALL exchange_horiz( u, nbgp )
	766	CALL exchange_horiz( v, nbgp )
	767	CALL exchange_horiz( w, nbgp )
[1]	768
	769	!
	770	!-- Compute the divergence of the corrected velocity field,
[1908]	771	!-- A possible PE-sum is computed in flow_statistics. Carry out computation
	772	!-- only at last Runge-Kutta step.
[1918]	773	IF ( intermediate_timestep_count == intermediate_timestep_count_max .OR. &
	774	intermediate_timestep_count == 0 ) THEN
[1908]	775	CALL cpu_log( log_point_s(1), 'divergence', 'start' )
	776	sums_divnew_l = 0.0_wp
[1]	777
	778	!
[1908]	779	!-- d must be reset to zero because it can contain nonzero values below the
	780	!-- topography
	781	IF ( topography /= 'flat' ) d = 0.0_wp
[1]	782
[1908]	783	localsum = 0.0_wp
	784	threadsum = 0.0_wp
[1]	785
[1908]	786	!$OMP PARALLEL PRIVATE (i,j,k) FIRSTPRIVATE(threadsum) REDUCTION(+:localsum)
[2073]	787	#if defined( __ibm )
[1908]	788	!$OMP DO SCHEDULE( STATIC )
	789	DO i = nxl, nxr
	790	DO j = nys, nyn
[2232]	791	DO k = nzb+1, nzt
	792	d(k,j,i) = ( ( u(k,j,i+1) - u(k,j,i) ) * rho_air(k) * ddx + &
	793	( v(k,j+1,i) - v(k,j,i) ) * rho_air(k) * ddy + &
	794	( w(k,j,i) * rho_air_zw(k) - &
	795	w(k-1,j,i) * rho_air_zw(k-1) ) * ddzw(k) &
	796	) * MERGE( 1.0_wp, 0.0_wp, &
[4346]	797	BTEST( wall_flags_total_0(k,j,i), 0 ) &
[2232]	798	)
[1908]	799	ENDDO
	800	DO k = nzb+1, nzt
	801	threadsum = threadsum + ABS( d(k,j,i) )
	802	ENDDO
[1]	803	ENDDO
	804	ENDDO
	805	#else
[2073]	806	!$OMP DO SCHEDULE( STATIC )
[3634]	807	!$ACC PARALLEL LOOP COLLAPSE(3) PRIVATE(i, j, k) &
[4346]	808	!$ACC PRESENT(u, v, w, rho_air, rho_air_zw, ddzw, wall_flags_total_0) &
[3634]	809	!$ACC PRESENT(d)
[1908]	810	DO i = nxl, nxr
	811	DO j = nys, nyn
[2232]	812	DO k = nzb+1, nzt
[2037]	813	d(k,j,i) = ( ( u(k,j,i+1) - u(k,j,i) ) * rho_air(k) * ddx + &
	814	( v(k,j+1,i) - v(k,j,i) ) * rho_air(k) * ddy + &
	815	( w(k,j,i) * rho_air_zw(k) - &
	816	w(k-1,j,i) * rho_air_zw(k-1) ) * ddzw(k) &
[2232]	817	) * MERGE( 1.0_wp, 0.0_wp, &
[4346]	818	BTEST( wall_flags_total_0(k,j,i), 0 ) &
[2232]	819	)
[1908]	820	ENDDO
[1113]	821	ENDDO
	822	ENDDO
	823	!
[1908]	824	!-- Compute possible PE-sum of divergences for flow_statistics
[2073]	825	!$OMP DO SCHEDULE( STATIC )
[3634]	826	!$ACC PARALLEL LOOP COLLAPSE(3) PRIVATE(i, j, k) &
	827	!$ACC REDUCTION(+:threadsum) COPY(threadsum) &
	828	!$ACC PRESENT(d)
[1908]	829	DO i = nxl, nxr
	830	DO j = nys, nyn
	831	DO k = nzb+1, nzt
	832	threadsum = threadsum + ABS( d(k,j,i) )
	833	ENDDO
[1]	834	ENDDO
	835	ENDDO
	836	#endif
[667]	837
[1908]	838	localsum = localsum + threadsum
	839	!$OMP END PARALLEL
[1]	840
	841	!
[1908]	842	!-- For completeness, set the divergence sum of all statistic regions to those
	843	!-- of the total domain
	844	sums_divnew_l(0:statistic_regions) = localsum
[1]	845
[1908]	846	CALL cpu_log( log_point_s(1), 'divergence', 'stop' )
[1]	847
[1908]	848	ENDIF
	849
[1]	850	CALL cpu_log( log_point(8), 'pres', 'stop' )
	851
	852	END SUBROUTINE pres

Note: See TracBrowser for help on using the repository browser.

Download in other formats:

| Impressum | ©Leibniz Universität Hannover |