Home

Context Navigation

source: palm/trunk/SOURCE/data_output_2d.f90 @ 2408

Last change on this file since 2408 was 2292, checked in by schwenkel, 7 years ago
implementation of new bulk microphysics scheme
Property svn:keywords set to `Id`
File size: 84.8 KB

Line
1	!> @file data_output_2d.f90
2	!------------------------------------------------------------------------------!
3	! This file is part of PALM.
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 1997-2017 Leibniz Universitaet Hannover
18	!------------------------------------------------------------------------------!
19	!
20	! Current revisions:
21	! -----------------
22	!
23	!
24	! Former revisions:
25	! -----------------
26	! $Id: data_output_2d.f90 2292 2017-06-20 09:51:42Z gronemeier $
27	! Implementation of new microphysic scheme: cloud_scheme = 'morrison'
28	! includes two more prognostic equations for cloud drop concentration (nc)
29	! and cloud water content (qc).
30	!
31	! 2277 2017-06-12 10:47:51Z kanani
32	! Removed unused variables do2d_xy_n, do2d_xz_n, do2d_yz_n
33	!
34	! 2233 2017-05-30 18:08:54Z suehring
35	!
36	! 2232 2017-05-30 17:47:52Z suehring
37	! Adjustments to new surface concept
38	!
39	!
40	! 2190 2017-03-21 12:16:43Z raasch
41	! bugfix for r2031: string rho replaced by rho_ocean
42	!
43	! 2031 2016-10-21 15:11:58Z knoop
44	! renamed variable rho to rho_ocean and rho_av to rho_ocean_av
45	!
46	! 2000 2016-08-20 18:09:15Z knoop
47	! Forced header and separation lines into 80 columns
48	!
49	! 1980 2016-07-29 15:51:57Z suehring
50	! Bugfix, in order to steer user-defined output, setting flag found explicitly
51	! to .F.
52	!
53	! 1976 2016-07-27 13:28:04Z maronga
54	! Output of radiation quantities is now done directly in the respective module
55	!
56	! 1972 2016-07-26 07:52:02Z maronga
57	! Output of land surface quantities is now done directly in the respective
58	! module
59	!
60	! 1960 2016-07-12 16:34:24Z suehring
61	! Scalar surface flux added
62	! Rename INTEGER variable s into s_ind, as s is already assigned to scalar
63	!
64	! 1849 2016-04-08 11:33:18Z hoffmann
65	! precipitation_amount, precipitation_rate, prr moved to arrays_3d
66	!
67	! 1822 2016-04-07 07:49:42Z hoffmann
68	! Output of bulk cloud physics simplified.
69	!
70	! 1788 2016-03-10 11:01:04Z maronga
71	! Added output of z0q
72	!
73	! 1783 2016-03-06 18:36:17Z raasch
74	! name change of netcdf routines and module + related changes
75	!
76	! 1745 2016-02-05 13:06:51Z gronemeier
77	! Bugfix: test if time axis limit exceeds moved to point after call of check_open
78	!
79	! 1703 2015-11-02 12:38:44Z raasch
80	! bugfix for output of single (*) xy-sections in case of parallel netcdf I/O
81	!
82	! 1701 2015-11-02 07:43:04Z maronga
83	! Bugfix in output of RRTGM data
84	!
85	! 1691 2015-10-26 16:17:44Z maronga
86	! Added output of Obukhov length (ol) and radiative heating rates for RRTMG.
87	! Formatting corrections.
88	!
89	! 1682 2015-10-07 23:56:08Z knoop
90	! Code annotations made doxygen readable
91	!
92	! 1585 2015-04-30 07:05:52Z maronga
93	! Added support for RRTMG
94	!
95	! 1555 2015-03-04 17:44:27Z maronga
96	! Added output of r_a and r_s
97	!
98	! 1551 2015-03-03 14:18:16Z maronga
99	! Added suppport for land surface model and radiation model output. In the course
100	! of this action, the limits for vertical loops have been changed (from nzb and
101	! nzt+1 to nzb_do and nzt_do, respectively in order to allow soil model output).
102	! Moreover, a new vertical grid zs was introduced.
103	!
104	! 1359 2014-04-11 17:15:14Z hoffmann
105	! New particle structure integrated.
106	!
107	! 1353 2014-04-08 15:21:23Z heinze
108	! REAL constants provided with KIND-attribute
109	!
110	! 1327 2014-03-21 11:00:16Z raasch
111	! parts concerning iso2d output removed,
112	! -netcdf output queries
113	!
114	! 1320 2014-03-20 08:40:49Z raasch
115	! ONLY-attribute added to USE-statements,
116	! kind-parameters added to all INTEGER and REAL declaration statements,
117	! kinds are defined in new module kinds,
118	! revision history before 2012 removed,
119	! comment fields (!:) to be used for variable explanations added to
120	! all variable declaration statements
121	!
122	! 1318 2014-03-17 13:35:16Z raasch
123	! barrier argument removed from cpu_log.
124	! module interfaces removed
125	!
126	! 1311 2014-03-14 12:13:39Z heinze
127	! bugfix: close #if defined( __netcdf )
128	!
129	! 1308 2014-03-13 14:58:42Z fricke
130	! +local_2d_sections, local_2d_sections_l, ns
131	! Check, if the limit of the time dimension is exceeded for parallel output
132	! To increase the performance for parallel output, the following is done:
133	! - Update of time axis is only done by PE0
134	! - Cross sections are first stored on a local array and are written
135	! collectively to the output file by all PEs.
136	!
137	! 1115 2013-03-26 18:16:16Z hoffmann
138	! ql is calculated by calc_liquid_water_content
139	!
140	! 1076 2012-12-05 08:30:18Z hoffmann
141	! Bugfix in output of ql
142	!
143	! 1065 2012-11-22 17:42:36Z hoffmann
144	! Bugfix: Output of cross sections of ql
145	!
146	! 1053 2012-11-13 17:11:03Z hoffmann
147	! +qr, nr, qc and cross sections
148	!
149	! 1036 2012-10-22 13:43:42Z raasch
150	! code put under GPL (PALM 3.9)
151	!
152	! 1031 2012-10-19 14:35:30Z raasch
153	! netCDF4 without parallel file support implemented
154	!
155	! 1007 2012-09-19 14:30:36Z franke
156	! Bugfix: missing calculation of ql_vp added
157	!
158	! 978 2012-08-09 08:28:32Z fricke
159	! +z0h
160	!
161	! Revision 1.1 1997/08/11 06:24:09 raasch
162	! Initial revision
163	!
164	!
165	! Description:
166	! ------------
167	!> Data output of horizontal cross-sections in netCDF format or binary format
168	!> compatible to old graphic software iso2d.
169	!> Attention: The position of the sectional planes is still not always computed
170	!> --------- correctly. (zu is used always)!
171	!------------------------------------------------------------------------------!
172	SUBROUTINE data_output_2d( mode, av )
173
174
175	USE arrays_3d, &
176	ONLY: dzw, e, nc, nr, p, pt, precipitation_amount, precipitation_rate,&
177	prr, q, qc, ql, ql_c, ql_v, ql_vp, qr, rho_ocean, s, sa, &
178	tend, u, v, vpt, w, zu, zw
179
180	USE averaging
181
182	USE cloud_parameters, &
183	ONLY: hyrho, l_d_cp, pt_d_t
184
185	USE control_parameters, &
186	ONLY: cloud_physics, data_output_2d_on_each_pe, data_output_xy, &
187	data_output_xz, data_output_yz, do2d, &
188	do2d_xy_last_time, do2d_xy_time_count, &
189	do2d_xz_last_time, do2d_xz_time_count, &
190	do2d_yz_last_time, do2d_yz_time_count, &
191	ibc_uv_b, io_blocks, io_group, land_surface, message_string, &
192	ntdim_2d_xy, ntdim_2d_xz, ntdim_2d_yz, &
193	psolver, section, simulated_time, simulated_time_chr, &
194	time_since_reference_point
195
196	USE cpulog, &
197	ONLY: cpu_log, log_point
198
199	USE grid_variables, &
200	ONLY: dx, dy
201
202	USE indices, &
203	ONLY: nbgp, nx, nxl, nxlg, nxr, nxrg, ny, nyn, nyng, nys, nysg, &
204	nz, nzb, nzt
205
206	USE kinds
207
208	USE land_surface_model_mod, &
209	ONLY: lsm_data_output_2d, zs
210
211	#if defined( __netcdf )
212	USE NETCDF
213	#endif
214
215	USE netcdf_interface, &
216	ONLY: id_set_xy, id_set_xz, id_set_yz, id_var_do2d, id_var_time_xy, &
217	id_var_time_xz, id_var_time_yz, nc_stat, netcdf_data_format, &
218	netcdf_handle_error
219
220	USE particle_attributes, &
221	ONLY: grid_particles, number_of_particles, particle_advection_start, &
222	particles, prt_count
223
224	USE pegrid
225
226	USE radiation_model_mod, &
227	ONLY: radiation, radiation_data_output_2d
228
229	USE surface_mod, &
230	ONLY: surf_def_h, surf_lsm_h, surf_usm_h
231
232	IMPLICIT NONE
233
234	CHARACTER (LEN=2) :: do2d_mode !<
235	CHARACTER (LEN=2) :: mode !<
236	CHARACTER (LEN=4) :: grid !<
237	CHARACTER (LEN=25) :: section_chr !<
238	CHARACTER (LEN=50) :: rtext !<
239
240	INTEGER(iwp) :: av !<
241	INTEGER(iwp) :: ngp !<
242	INTEGER(iwp) :: file_id !<
243	INTEGER(iwp) :: i !<
244	INTEGER(iwp) :: if !<
245	INTEGER(iwp) :: is !<
246	INTEGER(iwp) :: iis !<
247	INTEGER(iwp) :: j !<
248	INTEGER(iwp) :: k !<
249	INTEGER(iwp) :: l !<
250	INTEGER(iwp) :: layer_xy !<
251	INTEGER(iwp) :: m !<
252	INTEGER(iwp) :: n !<
253	INTEGER(iwp) :: nis !<
254	INTEGER(iwp) :: ns !<
255	INTEGER(iwp) :: nzb_do !< lower limit of the data field (usually nzb)
256	INTEGER(iwp) :: nzt_do !< upper limit of the data field (usually nzt+1)
257	INTEGER(iwp) :: psi !<
258	INTEGER(iwp) :: s_ind !<
259	INTEGER(iwp) :: sender !<
260	INTEGER(iwp) :: ind(4) !<
261
262	LOGICAL :: found !<
263	LOGICAL :: resorted !<
264	LOGICAL :: two_d !<
265
266	REAL(wp) :: mean_r !<
267	REAL(wp) :: s_r2 !<
268	REAL(wp) :: s_r3 !<
269
270	REAL(wp), DIMENSION(:), ALLOCATABLE :: level_z !<
271	REAL(wp), DIMENSION(:,:), ALLOCATABLE :: local_2d !<
272	REAL(wp), DIMENSION(:,:), ALLOCATABLE :: local_2d_l !<
273	REAL(wp), DIMENSION(:,:), ALLOCATABLE :: tmp_2d !< temporary field used to exchange surface-related quantities
274
275	REAL(wp), DIMENSION(:,:,:), ALLOCATABLE :: local_pf !<
276	REAL(wp), DIMENSION(:,:,:), ALLOCATABLE :: local_2d_sections !<
277	REAL(wp), DIMENSION(:,:,:), ALLOCATABLE :: local_2d_sections_l !<
278
279	#if defined( __parallel )
280	REAL(wp), DIMENSION(:,:), ALLOCATABLE :: total_2d !<
281	#endif
282	REAL(wp), DIMENSION(:,:,:), POINTER :: to_be_resorted !<
283
284	NAMELIST /LOCAL/ rtext
285
286	!
287	!-- Immediate return, if no output is requested (no respective sections
288	!-- found in parameter data_output)
289	IF ( mode == 'xy' .AND. .NOT. data_output_xy(av) ) RETURN
290	IF ( mode == 'xz' .AND. .NOT. data_output_xz(av) ) RETURN
291	IF ( mode == 'yz' .AND. .NOT. data_output_yz(av) ) RETURN
292
293	CALL cpu_log (log_point(3),'data_output_2d','start')
294
295	two_d = .FALSE. ! local variable to distinguish between output of pure 2D
296	! arrays and cross-sections of 3D arrays.
297
298	!
299	!-- Depending on the orientation of the cross-section, the respective output
300	!-- files have to be opened.
301	SELECT CASE ( mode )
302
303	CASE ( 'xy' )
304	s_ind = 1
305	ALLOCATE( level_z(nzb:nzt+1), local_2d(nxlg:nxrg,nysg:nyng) )
306
307	IF ( netcdf_data_format > 4 ) THEN
308	ns = 1
309	DO WHILE ( section(ns,s_ind) /= -9999 .AND. ns <= 100 )
310	ns = ns + 1
311	ENDDO
312	ns = ns - 1
313	ALLOCATE( local_2d_sections(nxlg:nxrg,nysg:nyng,1:ns) )
314	local_2d_sections = 0.0_wp
315	ENDIF
316
317	!
318	!-- Parallel netCDF4/HDF5 output is done on all PEs, all other on PE0 only
319	IF ( myid == 0 .OR. netcdf_data_format > 4 ) THEN
320	CALL check_open( 101+av*10 )
321	ENDIF
322	IF ( data_output_2d_on_each_pe ) THEN
323	CALL check_open( 21 )
324	ELSE
325	IF ( myid == 0 ) THEN
326	#if defined( __parallel )
327	ALLOCATE( total_2d(-nbgp:nx+nbgp,-nbgp:ny+nbgp) )
328	#endif
329	ENDIF
330	ENDIF
331
332	CASE ( 'xz' )
333	s_ind = 2
334	ALLOCATE( local_2d(nxlg:nxrg,nzb:nzt+1) )
335
336	IF ( netcdf_data_format > 4 ) THEN
337	ns = 1
338	DO WHILE ( section(ns,s_ind) /= -9999 .AND. ns <= 100 )
339	ns = ns + 1
340	ENDDO
341	ns = ns - 1
342	ALLOCATE( local_2d_sections(nxlg:nxrg,1:ns,nzb:nzt+1) )
343	ALLOCATE( local_2d_sections_l(nxlg:nxrg,1:ns,nzb:nzt+1) )
344	local_2d_sections = 0.0_wp; local_2d_sections_l = 0.0_wp
345	ENDIF
346
347	!
348	!-- Parallel netCDF4/HDF5 output is done on all PEs, all other on PE0 only
349	IF ( myid == 0 .OR. netcdf_data_format > 4 ) THEN
350	CALL check_open( 102+av*10 )
351	ENDIF
352
353	IF ( data_output_2d_on_each_pe ) THEN
354	CALL check_open( 22 )
355	ELSE
356	IF ( myid == 0 ) THEN
357	#if defined( __parallel )
358	ALLOCATE( total_2d(-nbgp:nx+nbgp,nzb:nzt+1) )
359	#endif
360	ENDIF
361	ENDIF
362
363	CASE ( 'yz' )
364	s_ind = 3
365	ALLOCATE( local_2d(nysg:nyng,nzb:nzt+1) )
366
367	IF ( netcdf_data_format > 4 ) THEN
368	ns = 1
369	DO WHILE ( section(ns,s_ind) /= -9999 .AND. ns <= 100 )
370	ns = ns + 1
371	ENDDO
372	ns = ns - 1
373	ALLOCATE( local_2d_sections(1:ns,nysg:nyng,nzb:nzt+1) )
374	ALLOCATE( local_2d_sections_l(1:ns,nysg:nyng,nzb:nzt+1) )
375	local_2d_sections = 0.0_wp; local_2d_sections_l = 0.0_wp
376	ENDIF
377
378	!
379	!-- Parallel netCDF4/HDF5 output is done on all PEs, all other on PE0 only
380	IF ( myid == 0 .OR. netcdf_data_format > 4 ) THEN
381	CALL check_open( 103+av*10 )
382	ENDIF
383
384	IF ( data_output_2d_on_each_pe ) THEN
385	CALL check_open( 23 )
386	ELSE
387	IF ( myid == 0 ) THEN
388	#if defined( __parallel )
389	ALLOCATE( total_2d(-nbgp:ny+nbgp,nzb:nzt+1) )
390	#endif
391	ENDIF
392	ENDIF
393
394	CASE DEFAULT
395	message_string = 'unknown cross-section: ' // TRIM( mode )
396	CALL message( 'data_output_2d', 'PA0180', 1, 2, 0, 6, 0 )
397
398	END SELECT
399
400	!
401	!-- For parallel netcdf output the time axis must be limited. Return, if this
402	!-- limit is exceeded. This could be the case, if the simulated time exceeds
403	!-- the given end time by the length of the given output interval.
404	IF ( netcdf_data_format > 4 ) THEN
405	IF ( mode == 'xy' .AND. do2d_xy_time_count(av) + 1 > &
406	ntdim_2d_xy(av) ) THEN
407	WRITE ( message_string, * ) 'Output of xy cross-sections is not ', &
408	'given at t=', simulated_time, '&because the', &
409	' maximum number of output time levels is exceeded.'
410	CALL message( 'data_output_2d', 'PA0384', 0, 1, 0, 6, 0 )
411	CALL cpu_log( log_point(3), 'data_output_2d', 'stop' )
412	RETURN
413	ENDIF
414	IF ( mode == 'xz' .AND. do2d_xz_time_count(av) + 1 > &
415	ntdim_2d_xz(av) ) THEN
416	WRITE ( message_string, * ) 'Output of xz cross-sections is not ', &
417	'given at t=', simulated_time, '&because the', &
418	' maximum number of output time levels is exceeded.'
419	CALL message( 'data_output_2d', 'PA0385', 0, 1, 0, 6, 0 )
420	CALL cpu_log( log_point(3), 'data_output_2d', 'stop' )
421	RETURN
422	ENDIF
423	IF ( mode == 'yz' .AND. do2d_yz_time_count(av) + 1 > &
424	ntdim_2d_yz(av) ) THEN
425	WRITE ( message_string, * ) 'Output of yz cross-sections is not ', &
426	'given at t=', simulated_time, '&because the', &
427	' maximum number of output time levels is exceeded.'
428	CALL message( 'data_output_2d', 'PA0386', 0, 1, 0, 6, 0 )
429	CALL cpu_log( log_point(3), 'data_output_2d', 'stop' )
430	RETURN
431	ENDIF
432	ENDIF
433
434	!
435	!-- Allocate a temporary array for resorting (kji -> ijk).
436	ALLOCATE( local_pf(nxlg:nxrg,nysg:nyng,nzb:nzt+1) )
437	local_pf = 0.0
438	!
439	!-- Allocate temporary array used for exchanging ghoist points of surface-data
440	ALLOCATE( tmp_2d(nysg:nyng,nxlg:nxrg) )
441	tmp_2d = 0.0
442
443	!
444	!-- Loop of all variables to be written.
445	!-- Output dimensions chosen
446	if = 1
447	l = MAX( 2, LEN_TRIM( do2d(av,if) ) )
448	do2d_mode = do2d(av,if)(l-1:l)
449
450	DO WHILE ( do2d(av,if)(1:1) /= ' ' )
451
452	IF ( do2d_mode == mode ) THEN
453	!
454	!-- Set flag to steer output of radiation, land-surface, or user-defined
455	!-- quantities
456	found = .FALSE.
457
458	nzb_do = nzb
459	nzt_do = nzt+1
460	!
461	!-- Store the array chosen on the temporary array.
462	resorted = .FALSE.
463	SELECT CASE ( TRIM( do2d(av,if) ) )
464
465	CASE ( 'e_xy', 'e_xz', 'e_yz' )
466	IF ( av == 0 ) THEN
467	to_be_resorted => e
468	ELSE
469	to_be_resorted => e_av
470	ENDIF
471	IF ( mode == 'xy' ) level_z = zu
472
473	CASE ( 'lpt_xy', 'lpt_xz', 'lpt_yz' )
474	IF ( av == 0 ) THEN
475	to_be_resorted => pt
476	ELSE
477	to_be_resorted => lpt_av
478	ENDIF
479	IF ( mode == 'xy' ) level_z = zu
480
481	CASE ( 'lwp*_xy' ) ! 2d-array
482	IF ( av == 0 ) THEN
483	DO i = nxlg, nxrg
484	DO j = nysg, nyng
485	local_pf(i,j,nzb+1) = SUM( ql(nzb:nzt,j,i) * &
486	dzw(1:nzt+1) )
487	ENDDO
488	ENDDO
489	ELSE
490	DO i = nxlg, nxrg
491	DO j = nysg, nyng
492	local_pf(i,j,nzb+1) = lwp_av(j,i)
493	ENDDO
494	ENDDO
495	ENDIF
496	resorted = .TRUE.
497	two_d = .TRUE.
498	level_z(nzb+1) = zu(nzb+1)
499
500	CASE ( 'nc_xy', 'nc_xz', 'nc_yz' )
501	IF ( av == 0 ) THEN
502	to_be_resorted => nc
503	ELSE
504	to_be_resorted => nc_av
505	ENDIF
506	IF ( mode == 'xy' ) level_z = zu
507
508	CASE ( 'nr_xy', 'nr_xz', 'nr_yz' )
509	IF ( av == 0 ) THEN
510	to_be_resorted => nr
511	ELSE
512	to_be_resorted => nr_av
513	ENDIF
514	IF ( mode == 'xy' ) level_z = zu
515
516	CASE ( 'ol*_xy' ) ! 2d-array
517	IF ( av == 0 ) THEN
518	DO m = 1, surf_def_h(0)%ns
519	i = surf_def_h(0)%i(m)
520	j = surf_def_h(0)%j(m)
521	tmp_2d(j,i) = surf_def_h(0)%ol(m)
522	ENDDO
523	DO m = 1, surf_lsm_h%ns
524	i = surf_lsm_h%i(m)
525	j = surf_lsm_h%j(m)
526	tmp_2d(j,i) = surf_lsm_h%ol(m)
527	ENDDO
528	DO m = 1, surf_usm_h%ns
529	i = surf_usm_h%i(m)
530	j = surf_usm_h%j(m)
531	tmp_2d(j,i) = surf_usm_h%ol(m)
532	ENDDO
533
534	CALL exchange_horiz_2d( tmp_2d, nbgp )
535
536	DO i = nxlg, nxrg
537	DO j = nysg, nyng
538	local_pf(i,j,nzb+1) = tmp_2d(j,i)
539	ENDDO
540	ENDDO
541
542	ELSE
543	DO i = nxlg, nxrg
544	DO j = nysg, nyng
545	local_pf(i,j,nzb+1) = ol_av(j,i)
546	ENDDO
547	ENDDO
548	ENDIF
549	resorted = .TRUE.
550	two_d = .TRUE.
551	level_z(nzb+1) = zu(nzb+1)
552
553	CASE ( 'p_xy', 'p_xz', 'p_yz' )
554	IF ( av == 0 ) THEN
555	IF ( psolver /= 'sor' ) CALL exchange_horiz( p, nbgp )
556	to_be_resorted => p
557	ELSE
558	IF ( psolver /= 'sor' ) CALL exchange_horiz( p_av, nbgp )
559	to_be_resorted => p_av
560	ENDIF
561	IF ( mode == 'xy' ) level_z = zu
562
563	CASE ( 'pc_xy', 'pc_xz', 'pc_yz' ) ! particle concentration
564	IF ( av == 0 ) THEN
565	IF ( simulated_time >= particle_advection_start ) THEN
566	tend = prt_count
567	CALL exchange_horiz( tend, nbgp )
568	ELSE
569	tend = 0.0_wp
570	ENDIF
571	DO i = nxlg, nxrg
572	DO j = nysg, nyng
573	DO k = nzb, nzt+1
574	local_pf(i,j,k) = tend(k,j,i)
575	ENDDO
576	ENDDO
577	ENDDO
578	resorted = .TRUE.
579	ELSE
580	CALL exchange_horiz( pc_av, nbgp )
581	to_be_resorted => pc_av
582	ENDIF
583
584	CASE ( 'pr_xy', 'pr_xz', 'pr_yz' ) ! mean particle radius (effective radius)
585	IF ( av == 0 ) THEN
586	IF ( simulated_time >= particle_advection_start ) THEN
587	DO i = nxl, nxr
588	DO j = nys, nyn
589	DO k = nzb, nzt+1
590	number_of_particles = prt_count(k,j,i)
591	IF (number_of_particles <= 0) CYCLE
592	particles => grid_particles(k,j,i)%particles(1:number_of_particles)
593	s_r2 = 0.0_wp
594	s_r3 = 0.0_wp
595	DO n = 1, number_of_particles
596	IF ( particles(n)%particle_mask ) THEN
597	s_r2 = s_r2 + particles(n)%radius*2 &
598	particles(n)%weight_factor
599	s_r3 = s_r3 + particles(n)%radius*3 &
600	particles(n)%weight_factor
601	ENDIF
602	ENDDO
603	IF ( s_r2 > 0.0_wp ) THEN
604	mean_r = s_r3 / s_r2
605	ELSE
606	mean_r = 0.0_wp
607	ENDIF
608	tend(k,j,i) = mean_r
609	ENDDO
610	ENDDO
611	ENDDO
612	CALL exchange_horiz( tend, nbgp )
613	ELSE
614	tend = 0.0_wp
615	ENDIF
616	DO i = nxlg, nxrg
617	DO j = nysg, nyng
618	DO k = nzb, nzt+1
619	local_pf(i,j,k) = tend(k,j,i)
620	ENDDO
621	ENDDO
622	ENDDO
623	resorted = .TRUE.
624	ELSE
625	CALL exchange_horiz( pr_av, nbgp )
626	to_be_resorted => pr_av
627	ENDIF
628
629	CASE ( 'pra*_xy' ) ! 2d-array / integral quantity => no av
630	CALL exchange_horiz_2d( precipitation_amount )
631	DO i = nxlg, nxrg
632	DO j = nysg, nyng
633	local_pf(i,j,nzb+1) = precipitation_amount(j,i)
634	ENDDO
635	ENDDO
636	precipitation_amount = 0.0_wp ! reset for next integ. interval
637	resorted = .TRUE.
638	two_d = .TRUE.
639	level_z(nzb+1) = zu(nzb+1)
640
641	CASE ( 'prr*_xy' ) ! 2d-array
642	IF ( av == 0 ) THEN
643	CALL exchange_horiz_2d( prr(nzb+1,:,:) )
644	DO i = nxlg, nxrg
645	DO j = nysg, nyng
646	local_pf(i,j,nzb+1) = prr(nzb+1,j,i) * hyrho(nzb+1)
647	ENDDO
648	ENDDO
649	ELSE
650	CALL exchange_horiz_2d( prr_av(nzb+1,:,:) )
651	DO i = nxlg, nxrg
652	DO j = nysg, nyng
653	local_pf(i,j,nzb+1) = prr_av(nzb+1,j,i) * hyrho(nzb+1)
654	ENDDO
655	ENDDO
656	ENDIF
657	resorted = .TRUE.
658	two_d = .TRUE.
659	level_z(nzb+1) = zu(nzb+1)
660
661	CASE ( 'prr_xy', 'prr_xz', 'prr_yz' )
662	IF ( av == 0 ) THEN
663	CALL exchange_horiz( prr, nbgp )
664	DO i = nxlg, nxrg
665	DO j = nysg, nyng
666	DO k = nzb, nzt+1
667	local_pf(i,j,k) = prr(k,j,i) * hyrho(nzb+1)
668	ENDDO
669	ENDDO
670	ENDDO
671	ELSE
672	CALL exchange_horiz( prr_av, nbgp )
673	DO i = nxlg, nxrg
674	DO j = nysg, nyng
675	DO k = nzb, nzt+1
676	local_pf(i,j,k) = prr_av(k,j,i) * hyrho(nzb+1)
677	ENDDO
678	ENDDO
679	ENDDO
680	ENDIF
681	resorted = .TRUE.
682	IF ( mode == 'xy' ) level_z = zu
683
684	CASE ( 'pt_xy', 'pt_xz', 'pt_yz' )
685	IF ( av == 0 ) THEN
686	IF ( .NOT. cloud_physics ) THEN
687	to_be_resorted => pt
688	ELSE
689	DO i = nxlg, nxrg
690	DO j = nysg, nyng
691	DO k = nzb, nzt+1
692	local_pf(i,j,k) = pt(k,j,i) + l_d_cp * &
693	pt_d_t(k) * &
694	ql(k,j,i)
695	ENDDO
696	ENDDO
697	ENDDO
698	resorted = .TRUE.
699	ENDIF
700	ELSE
701	to_be_resorted => pt_av
702	ENDIF
703	IF ( mode == 'xy' ) level_z = zu
704
705	CASE ( 'q_xy', 'q_xz', 'q_yz' )
706	IF ( av == 0 ) THEN
707	to_be_resorted => q
708	ELSE
709	to_be_resorted => q_av
710	ENDIF
711	IF ( mode == 'xy' ) level_z = zu
712
713	CASE ( 'qc_xy', 'qc_xz', 'qc_yz' )
714	IF ( av == 0 ) THEN
715	to_be_resorted => qc
716	ELSE
717	to_be_resorted => qc_av
718	ENDIF
719	IF ( mode == 'xy' ) level_z = zu
720
721	CASE ( 'ql_xy', 'ql_xz', 'ql_yz' )
722	IF ( av == 0 ) THEN
723	to_be_resorted => ql
724	ELSE
725	to_be_resorted => ql_av
726	ENDIF
727	IF ( mode == 'xy' ) level_z = zu
728
729	CASE ( 'ql_c_xy', 'ql_c_xz', 'ql_c_yz' )
730	IF ( av == 0 ) THEN
731	to_be_resorted => ql_c
732	ELSE
733	to_be_resorted => ql_c_av
734	ENDIF
735	IF ( mode == 'xy' ) level_z = zu
736
737	CASE ( 'ql_v_xy', 'ql_v_xz', 'ql_v_yz' )
738	IF ( av == 0 ) THEN
739	to_be_resorted => ql_v
740	ELSE
741	to_be_resorted => ql_v_av
742	ENDIF
743	IF ( mode == 'xy' ) level_z = zu
744
745	CASE ( 'ql_vp_xy', 'ql_vp_xz', 'ql_vp_yz' )
746	IF ( av == 0 ) THEN
747	IF ( simulated_time >= particle_advection_start ) THEN
748	DO i = nxl, nxr
749	DO j = nys, nyn
750	DO k = nzb, nzt+1
751	number_of_particles = prt_count(k,j,i)
752	IF (number_of_particles <= 0) CYCLE
753	particles => grid_particles(k,j,i)%particles(1:number_of_particles)
754	DO n = 1, number_of_particles
755	IF ( particles(n)%particle_mask ) THEN
756	tend(k,j,i) = tend(k,j,i) + &
757	particles(n)%weight_factor / &
758	prt_count(k,j,i)
759	ENDIF
760	ENDDO
761	ENDDO
762	ENDDO
763	ENDDO
764	CALL exchange_horiz( tend, nbgp )
765	ELSE
766	tend = 0.0_wp
767	ENDIF
768	DO i = nxlg, nxrg
769	DO j = nysg, nyng
770	DO k = nzb, nzt+1
771	local_pf(i,j,k) = tend(k,j,i)
772	ENDDO
773	ENDDO
774	ENDDO
775	resorted = .TRUE.
776	ELSE
777	CALL exchange_horiz( ql_vp_av, nbgp )
778	to_be_resorted => ql_vp
779	ENDIF
780	IF ( mode == 'xy' ) level_z = zu
781
782	CASE ( 'qr_xy', 'qr_xz', 'qr_yz' )
783	IF ( av == 0 ) THEN
784	to_be_resorted => qr
785	ELSE
786	to_be_resorted => qr_av
787	ENDIF
788	IF ( mode == 'xy' ) level_z = zu
789
790	CASE ( 'qsws*_xy' ) ! 2d-array
791	IF ( av == 0 ) THEN
792	DO m = 1, surf_def_h(0)%ns
793	i = surf_def_h(0)%i(m)
794	j = surf_def_h(0)%j(m)
795	tmp_2d(j,i) = surf_def_h(0)%qsws(m)
796	ENDDO
797	DO m = 1, surf_lsm_h%ns
798	i = surf_lsm_h%i(m)
799	j = surf_lsm_h%j(m)
800	tmp_2d(j,i) = surf_lsm_h%qsws(m)
801	ENDDO
802	DO m = 1, surf_usm_h%ns
803	i = surf_usm_h%i(m)
804	j = surf_usm_h%j(m)
805	tmp_2d(j,i) = surf_usm_h%qsws(m)
806	ENDDO
807
808	CALL exchange_horiz_2d( tmp_2d, nbgp )
809
810	DO i = nxlg, nxrg
811	DO j = nysg, nyng
812	local_pf(i,j,nzb+1) = tmp_2d(j,i)
813	ENDDO
814	ENDDO
815	ELSE
816	DO i = nxlg, nxrg
817	DO j = nysg, nyng
818	local_pf(i,j,nzb+1) = qsws_av(j,i)
819	ENDDO
820	ENDDO
821	ENDIF
822	resorted = .TRUE.
823	two_d = .TRUE.
824	level_z(nzb+1) = zu(nzb+1)
825
826	CASE ( 'qv_xy', 'qv_xz', 'qv_yz' )
827	IF ( av == 0 ) THEN
828	DO i = nxlg, nxrg
829	DO j = nysg, nyng
830	DO k = nzb, nzt+1
831	local_pf(i,j,k) = q(k,j,i) - ql(k,j,i)
832	ENDDO
833	ENDDO
834	ENDDO
835	resorted = .TRUE.
836	ELSE
837	to_be_resorted => qv_av
838	ENDIF
839	IF ( mode == 'xy' ) level_z = zu
840
841
842
843	CASE ( 'rho_ocean_xy', 'rho_ocean_xz', 'rho_ocean_yz' )
844	IF ( av == 0 ) THEN
845	to_be_resorted => rho_ocean
846	ELSE
847	to_be_resorted => rho_ocean_av
848	ENDIF
849
850	CASE ( 's_xy', 's_xz', 's_yz' )
851	IF ( av == 0 ) THEN
852	to_be_resorted => s
853	ELSE
854	to_be_resorted => s_av
855	ENDIF
856
857	CASE ( 'sa_xy', 'sa_xz', 'sa_yz' )
858	IF ( av == 0 ) THEN
859	to_be_resorted => sa
860	ELSE
861	to_be_resorted => sa_av
862	ENDIF
863
864	CASE ( 'shf*_xy' ) ! 2d-array
865	IF ( av == 0 ) THEN
866	DO m = 1, surf_def_h(0)%ns
867	i = surf_def_h(0)%i(m)
868	j = surf_def_h(0)%j(m)
869	tmp_2d(j,i) = surf_def_h(0)%shf(m)
870	ENDDO
871	DO m = 1, surf_lsm_h%ns
872	i = surf_lsm_h%i(m)
873	j = surf_lsm_h%j(m)
874	tmp_2d(j,i) = surf_lsm_h%shf(m)
875	ENDDO
876	DO m = 1, surf_usm_h%ns
877	i = surf_usm_h%i(m)
878	j = surf_usm_h%j(m)
879	tmp_2d(j,i) = surf_usm_h%shf(m)
880	ENDDO
881
882	CALL exchange_horiz_2d( tmp_2d, nbgp )
883
884	DO i = nxlg, nxrg
885	DO j = nysg, nyng
886	local_pf(i,j,nzb+1) = tmp_2d(j,i)
887	ENDDO
888	ENDDO
889	ELSE
890	DO i = nxlg, nxrg
891	DO j = nysg, nyng
892	local_pf(i,j,nzb+1) = shf_av(j,i)
893	ENDDO
894	ENDDO
895	ENDIF
896	resorted = .TRUE.
897	two_d = .TRUE.
898	level_z(nzb+1) = zu(nzb+1)
899
900	CASE ( 'ssws*_xy' ) ! 2d-array
901	IF ( av == 0 ) THEN
902	DO m = 1, surf_def_h(0)%ns
903	i = surf_def_h(0)%i(m)
904	j = surf_def_h(0)%j(m)
905	tmp_2d(j,i) = surf_def_h(0)%ssws(m)
906	ENDDO
907	DO m = 1, surf_lsm_h%ns
908	i = surf_lsm_h%i(m)
909	j = surf_lsm_h%j(m)
910	tmp_2d(j,i) = surf_lsm_h%ssws(m)
911	ENDDO
912	DO m = 1, surf_usm_h%ns
913	i = surf_usm_h%i(m)
914	j = surf_usm_h%j(m)
915	tmp_2d(j,i) = surf_usm_h%ssws(m)
916	ENDDO
917
918	CALL exchange_horiz_2d( tmp_2d, nbgp )
919
920	DO i = nxlg, nxrg
921	DO j = nysg, nyng
922	local_pf(i,j,nzb+1) = tmp_2d(j,i)
923	ENDDO
924	ENDDO
925	ELSE
926	DO i = nxlg, nxrg
927	DO j = nysg, nyng
928	local_pf(i,j,nzb+1) = ssws_av(j,i)
929	ENDDO
930	ENDDO
931	ENDIF
932	resorted = .TRUE.
933	two_d = .TRUE.
934	level_z(nzb+1) = zu(nzb+1)
935
936	CASE ( 't*_xy' ) ! 2d-array
937	IF ( av == 0 ) THEN
938	DO m = 1, surf_def_h(0)%ns
939	i = surf_def_h(0)%i(m)
940	j = surf_def_h(0)%j(m)
941	tmp_2d(j,i) = surf_def_h(0)%ts(m)
942	ENDDO
943	DO m = 1, surf_lsm_h%ns
944	i = surf_lsm_h%i(m)
945	j = surf_lsm_h%j(m)
946	tmp_2d(j,i) = surf_lsm_h%ts(m)
947	ENDDO
948	DO m = 1, surf_usm_h%ns
949	i = surf_usm_h%i(m)
950	j = surf_usm_h%j(m)
951	tmp_2d(j,i) = surf_usm_h%ts(m)
952	ENDDO
953
954	CALL exchange_horiz_2d( tmp_2d, nbgp )
955
956	DO i = nxlg, nxrg
957	DO j = nysg, nyng
958	local_pf(i,j,nzb+1) = tmp_2d(j,i)
959	ENDDO
960	ENDDO
961
962	ELSE
963	DO i = nxlg, nxrg
964	DO j = nysg, nyng
965	local_pf(i,j,nzb+1) = ts_av(j,i)
966	ENDDO
967	ENDDO
968	ENDIF
969	resorted = .TRUE.
970	two_d = .TRUE.
971	level_z(nzb+1) = zu(nzb+1)
972
973	CASE ( 'u_xy', 'u_xz', 'u_yz' )
974	IF ( av == 0 ) THEN
975	to_be_resorted => u
976	ELSE
977	to_be_resorted => u_av
978	ENDIF
979	IF ( mode == 'xy' ) level_z = zu
980	!
981	!-- Substitute the values generated by "mirror" boundary condition
982	!-- at the bottom boundary by the real surface values.
983	IF ( do2d(av,if) == 'u_xz' .OR. do2d(av,if) == 'u_yz' ) THEN
984	IF ( ibc_uv_b == 0 ) local_pf(:,:,nzb) = 0.0_wp
985	ENDIF
986
987	CASE ( 'u*_xy' ) ! 2d-array
988	IF ( av == 0 ) THEN
989	DO m = 1, surf_def_h(0)%ns
990	i = surf_def_h(0)%i(m)
991	j = surf_def_h(0)%j(m)
992	tmp_2d(j,i) = surf_def_h(0)%us(m)
993	ENDDO
994	DO m = 1, surf_lsm_h%ns
995	i = surf_lsm_h%i(m)
996	j = surf_lsm_h%j(m)
997	tmp_2d(j,i) = surf_lsm_h%us(m)
998	ENDDO
999	DO m = 1, surf_usm_h%ns
1000	i = surf_usm_h%i(m)
1001	j = surf_usm_h%j(m)
1002	tmp_2d(j,i) = surf_usm_h%us(m)
1003	ENDDO
1004
1005	CALL exchange_horiz_2d( tmp_2d, nbgp )
1006
1007	DO i = nxlg, nxrg
1008	DO j = nysg, nyng
1009	local_pf(i,j,nzb+1) = tmp_2d(j,i)
1010	ENDDO
1011	ENDDO
1012	ELSE
1013	DO i = nxlg, nxrg
1014	DO j = nysg, nyng
1015	local_pf(i,j,nzb+1) = us_av(j,i)
1016	ENDDO
1017	ENDDO
1018	ENDIF
1019	resorted = .TRUE.
1020	two_d = .TRUE.
1021	level_z(nzb+1) = zu(nzb+1)
1022
1023	CASE ( 'v_xy', 'v_xz', 'v_yz' )
1024	IF ( av == 0 ) THEN
1025	to_be_resorted => v
1026	ELSE
1027	to_be_resorted => v_av
1028	ENDIF
1029	IF ( mode == 'xy' ) level_z = zu
1030	!
1031	!-- Substitute the values generated by "mirror" boundary condition
1032	!-- at the bottom boundary by the real surface values.
1033	IF ( do2d(av,if) == 'v_xz' .OR. do2d(av,if) == 'v_yz' ) THEN
1034	IF ( ibc_uv_b == 0 ) local_pf(:,:,nzb) = 0.0_wp
1035	ENDIF
1036
1037	CASE ( 'vpt_xy', 'vpt_xz', 'vpt_yz' )
1038	IF ( av == 0 ) THEN
1039	to_be_resorted => vpt
1040	ELSE
1041	to_be_resorted => vpt_av
1042	ENDIF
1043	IF ( mode == 'xy' ) level_z = zu
1044
1045	CASE ( 'w_xy', 'w_xz', 'w_yz' )
1046	IF ( av == 0 ) THEN
1047	to_be_resorted => w
1048	ELSE
1049	to_be_resorted => w_av
1050	ENDIF
1051	IF ( mode == 'xy' ) level_z = zw
1052
1053	CASE ( 'z0*_xy' ) ! 2d-array
1054	IF ( av == 0 ) THEN
1055	DO m = 1, surf_def_h(0)%ns
1056	i = surf_def_h(0)%i(m)
1057	j = surf_def_h(0)%j(m)
1058	tmp_2d(j,i) = surf_def_h(0)%z0(m)
1059	ENDDO
1060	DO m = 1, surf_lsm_h%ns
1061	i = surf_lsm_h%i(m)
1062	j = surf_lsm_h%j(m)
1063	tmp_2d(j,i) = surf_lsm_h%z0(m)
1064	ENDDO
1065	DO m = 1, surf_usm_h%ns
1066	i = surf_usm_h%i(m)
1067	j = surf_usm_h%j(m)
1068	tmp_2d(j,i) = surf_usm_h%z0(m)
1069	ENDDO
1070
1071	CALL exchange_horiz_2d( tmp_2d, nbgp )
1072
1073	DO i = nxlg, nxrg
1074	DO j = nysg, nyng
1075	local_pf(i,j,nzb+1) = tmp_2d(j,i)
1076	ENDDO
1077	ENDDO
1078
1079	ELSE
1080	DO i = nxlg, nxrg
1081	DO j = nysg, nyng
1082	local_pf(i,j,nzb+1) = z0_av(j,i)
1083	ENDDO
1084	ENDDO
1085	ENDIF
1086	resorted = .TRUE.
1087	two_d = .TRUE.
1088	level_z(nzb+1) = zu(nzb+1)
1089
1090	CASE ( 'z0h*_xy' ) ! 2d-array
1091	IF ( av == 0 ) THEN
1092	DO m = 1, surf_def_h(0)%ns
1093	i = surf_def_h(0)%i(m)
1094	j = surf_def_h(0)%j(m)
1095	tmp_2d(j,i) = surf_def_h(0)%z0h(m)
1096	ENDDO
1097	DO m = 1, surf_lsm_h%ns
1098	i = surf_lsm_h%i(m)
1099	j = surf_lsm_h%j(m)
1100	tmp_2d(j,i) = surf_lsm_h%z0h(m)
1101	ENDDO
1102	DO m = 1, surf_usm_h%ns
1103	i = surf_usm_h%i(m)
1104	j = surf_usm_h%j(m)
1105	tmp_2d(j,i) = surf_usm_h%z0h(m)
1106	ENDDO
1107
1108	CALL exchange_horiz_2d( tmp_2d, nbgp )
1109
1110	DO i = nxlg, nxrg
1111	DO j = nysg, nyng
1112	local_pf(i,j,nzb+1) = tmp_2d(j,i)
1113	ENDDO
1114	ENDDO
1115	ELSE
1116	DO i = nxlg, nxrg
1117	DO j = nysg, nyng
1118	local_pf(i,j,nzb+1) = z0h_av(j,i)
1119	ENDDO
1120	ENDDO
1121	ENDIF
1122	resorted = .TRUE.
1123	two_d = .TRUE.
1124	level_z(nzb+1) = zu(nzb+1)
1125
1126	CASE ( 'z0q*_xy' ) ! 2d-array
1127	IF ( av == 0 ) THEN
1128	DO m = 1, surf_def_h(0)%ns
1129	i = surf_def_h(0)%i(m)
1130	j = surf_def_h(0)%j(m)
1131	tmp_2d(j,i) = surf_def_h(0)%z0q(m)
1132	ENDDO
1133	DO m = 1, surf_lsm_h%ns
1134	i = surf_lsm_h%i(m)
1135	j = surf_lsm_h%j(m)
1136	tmp_2d(j,i) = surf_lsm_h%z0q(m)
1137	ENDDO
1138	DO m = 1, surf_usm_h%ns
1139	i = surf_usm_h%i(m)
1140	j = surf_usm_h%j(m)
1141	tmp_2d(j,i) = surf_usm_h%z0q(m)
1142	ENDDO
1143
1144	CALL exchange_horiz_2d( tmp_2d, nbgp )
1145
1146	DO i = nxlg, nxrg
1147	DO j = nysg, nyng
1148	local_pf(i,j,nzb+1) = tmp_2d(j,i)
1149	ENDDO
1150	ENDDO
1151	ELSE
1152	DO i = nxlg, nxrg
1153	DO j = nysg, nyng
1154	local_pf(i,j,nzb+1) = z0q_av(j,i)
1155	ENDDO
1156	ENDDO
1157	ENDIF
1158	resorted = .TRUE.
1159	two_d = .TRUE.
1160	level_z(nzb+1) = zu(nzb+1)
1161
1162	CASE DEFAULT
1163
1164	!
1165	!-- Land surface model quantity
1166	IF ( land_surface ) THEN
1167	CALL lsm_data_output_2d( av, do2d(av,if), found, grid, mode,&
1168	local_pf, two_d, nzb_do, nzt_do )
1169	ENDIF
1170
1171	!
1172	!-- Radiation quantity
1173	IF ( .NOT. found .AND. radiation ) THEN
1174	CALL radiation_data_output_2d( av, do2d(av,if), found, grid,&
1175	mode, local_pf, two_d )
1176	ENDIF
1177
1178	!
1179	!-- User defined quantity
1180	IF ( .NOT. found ) THEN
1181	CALL user_data_output_2d( av, do2d(av,if), found, grid, &
1182	local_pf, two_d, nzb_do, nzt_do )
1183	ENDIF
1184
1185	resorted = .TRUE.
1186
1187	IF ( grid == 'zu' ) THEN
1188	IF ( mode == 'xy' ) level_z = zu
1189	ELSEIF ( grid == 'zw' ) THEN
1190	IF ( mode == 'xy' ) level_z = zw
1191	ELSEIF ( grid == 'zu1' ) THEN
1192	IF ( mode == 'xy' ) level_z(nzb+1) = zu(nzb+1)
1193	ELSEIF ( grid == 'zs' ) THEN
1194	IF ( mode == 'xy' ) level_z = zs
1195	ENDIF
1196
1197	IF ( .NOT. found ) THEN
1198	message_string = 'no output provided for: ' // &
1199	TRIM( do2d(av,if) )
1200	CALL message( 'data_output_2d', 'PA0181', 0, 0, 0, 6, 0 )
1201	ENDIF
1202
1203	END SELECT
1204
1205	!
1206	!-- Resort the array to be output, if not done above
1207	IF ( .NOT. resorted ) THEN
1208	DO i = nxlg, nxrg
1209	DO j = nysg, nyng
1210	DO k = nzb_do, nzt_do
1211	local_pf(i,j,k) = to_be_resorted(k,j,i)
1212	ENDDO
1213	ENDDO
1214	ENDDO
1215	ENDIF
1216
1217	!
1218	!-- Output of the individual cross-sections, depending on the cross-
1219	!-- section mode chosen.
1220	is = 1
1221	loop1: DO WHILE ( section(is,s_ind) /= -9999 .OR. two_d )
1222
1223	SELECT CASE ( mode )
1224
1225	CASE ( 'xy' )
1226	!
1227	!-- Determine the cross section index
1228	IF ( two_d ) THEN
1229	layer_xy = nzb+1
1230	ELSE
1231	layer_xy = section(is,s_ind)
1232	ENDIF
1233
1234	!
1235	!-- Exit the loop for layers beyond the data output domain
1236	!-- (used for soil model)
1237	IF ( layer_xy > nzt_do ) THEN
1238	EXIT loop1
1239	ENDIF
1240
1241	!
1242	!-- Update the netCDF xy cross section time axis.
1243	!-- In case of parallel output, this is only done by PE0
1244	!-- to increase the performance.
1245	IF ( simulated_time /= do2d_xy_last_time(av) ) THEN
1246	do2d_xy_time_count(av) = do2d_xy_time_count(av) + 1
1247	do2d_xy_last_time(av) = simulated_time
1248	IF ( myid == 0 ) THEN
1249	IF ( .NOT. data_output_2d_on_each_pe &
1250	.OR. netcdf_data_format > 4 ) &
1251	THEN
1252	#if defined( __netcdf )
1253	nc_stat = NF90_PUT_VAR( id_set_xy(av), &
1254	id_var_time_xy(av), &
1255	(/ time_since_reference_point /), &
1256	start = (/ do2d_xy_time_count(av) /), &
1257	count = (/ 1 /) )
1258	CALL netcdf_handle_error( 'data_output_2d', 53 )
1259	#endif
1260	ENDIF
1261	ENDIF
1262	ENDIF
1263	!
1264	!-- If required, carry out averaging along z
1265	IF ( section(is,s_ind) == -1 .AND. .NOT. two_d ) THEN
1266
1267	local_2d = 0.0_wp
1268	!
1269	!-- Carry out the averaging (all data are on the PE)
1270	DO k = nzb_do, nzt_do
1271	DO j = nysg, nyng
1272	DO i = nxlg, nxrg
1273	local_2d(i,j) = local_2d(i,j) + local_pf(i,j,k)
1274	ENDDO
1275	ENDDO
1276	ENDDO
1277
1278	local_2d = local_2d / ( nzt_do - nzb_do + 1.0_wp)
1279
1280	ELSE
1281	!
1282	!-- Just store the respective section on the local array
1283	local_2d = local_pf(:,:,layer_xy)
1284
1285	ENDIF
1286
1287	#if defined( __parallel )
1288	IF ( netcdf_data_format > 4 ) THEN
1289	!
1290	!-- Parallel output in netCDF4/HDF5 format.
1291	IF ( two_d ) THEN
1292	iis = 1
1293	ELSE
1294	iis = is
1295	ENDIF
1296
1297	#if defined( __netcdf )
1298	!
1299	!-- For parallel output, all cross sections are first stored
1300	!-- here on a local array and will be written to the output
1301	!-- file afterwards to increase the performance.
1302	DO i = nxlg, nxrg
1303	DO j = nysg, nyng
1304	local_2d_sections(i,j,iis) = local_2d(i,j)
1305	ENDDO
1306	ENDDO
1307	#endif
1308	ELSE
1309
1310	IF ( data_output_2d_on_each_pe ) THEN
1311	!
1312	!-- Output of partial arrays on each PE
1313	#if defined( __netcdf )
1314	IF ( myid == 0 ) THEN
1315	WRITE ( 21 ) time_since_reference_point, &
1316	do2d_xy_time_count(av), av
1317	ENDIF
1318	#endif
1319	DO i = 0, io_blocks-1
1320	IF ( i == io_group ) THEN
1321	WRITE ( 21 ) nxlg, nxrg, nysg, nyng, nysg, nyng
1322	WRITE ( 21 ) local_2d
1323	ENDIF
1324	#if defined( __parallel )
1325	CALL MPI_BARRIER( comm2d, ierr )
1326	#endif
1327	ENDDO
1328
1329	ELSE
1330	!
1331	!-- PE0 receives partial arrays from all processors and
1332	!-- then outputs them. Here a barrier has to be set,
1333	!-- because otherwise "-MPI- FATAL: Remote protocol queue
1334	!-- full" may occur.
1335	CALL MPI_BARRIER( comm2d, ierr )
1336
1337	ngp = ( nxrg-nxlg+1 ) * ( nyng-nysg+1 )
1338	IF ( myid == 0 ) THEN
1339	!
1340	!-- Local array can be relocated directly.
1341	total_2d(nxlg:nxrg,nysg:nyng) = local_2d
1342	!
1343	!-- Receive data from all other PEs.
1344	DO n = 1, numprocs-1
1345	!
1346	!-- Receive index limits first, then array.
1347	!-- Index limits are received in arbitrary order from
1348	!-- the PEs.
1349	CALL MPI_RECV( ind(1), 4, MPI_INTEGER, &
1350	MPI_ANY_SOURCE, 0, comm2d, &
1351	status, ierr )
1352	sender = status(MPI_SOURCE)
1353	DEALLOCATE( local_2d )
1354	ALLOCATE( local_2d(ind(1):ind(2),ind(3):ind(4)) )
1355	CALL MPI_RECV( local_2d(ind(1),ind(3)), ngp, &
1356	MPI_REAL, sender, 1, comm2d, &
1357	status, ierr )
1358	total_2d(ind(1):ind(2),ind(3):ind(4)) = local_2d
1359	ENDDO
1360	!
1361	!-- Relocate the local array for the next loop increment
1362	DEALLOCATE( local_2d )
1363	ALLOCATE( local_2d(nxlg:nxrg,nysg:nyng) )
1364
1365	#if defined( __netcdf )
1366	IF ( two_d ) THEN
1367	nc_stat = NF90_PUT_VAR( id_set_xy(av), &
1368	id_var_do2d(av,if), &
1369	total_2d(0:nx+1,0:ny+1), &
1370	start = (/ 1, 1, 1, do2d_xy_time_count(av) /), &
1371	count = (/ nx+2, ny+2, 1, 1 /) )
1372	ELSE
1373	nc_stat = NF90_PUT_VAR( id_set_xy(av), &
1374	id_var_do2d(av,if), &
1375	total_2d(0:nx+1,0:ny+1), &
1376	start = (/ 1, 1, is, do2d_xy_time_count(av) /), &
1377	count = (/ nx+2, ny+2, 1, 1 /) )
1378	ENDIF
1379	CALL netcdf_handle_error( 'data_output_2d', 54 )
1380	#endif
1381
1382	ELSE
1383	!
1384	!-- First send the local index limits to PE0
1385	ind(1) = nxlg; ind(2) = nxrg
1386	ind(3) = nysg; ind(4) = nyng
1387	CALL MPI_SEND( ind(1), 4, MPI_INTEGER, 0, 0, &
1388	comm2d, ierr )
1389	!
1390	!-- Send data to PE0
1391	CALL MPI_SEND( local_2d(nxlg,nysg), ngp, &
1392	MPI_REAL, 0, 1, comm2d, ierr )
1393	ENDIF
1394	!
1395	!-- A barrier has to be set, because otherwise some PEs may
1396	!-- proceed too fast so that PE0 may receive wrong data on
1397	!-- tag 0
1398	CALL MPI_BARRIER( comm2d, ierr )
1399	ENDIF
1400
1401	ENDIF
1402	#else
1403	#if defined( __netcdf )
1404	IF ( two_d ) THEN
1405	nc_stat = NF90_PUT_VAR( id_set_xy(av), &
1406	id_var_do2d(av,if), &
1407	local_2d(nxl:nxr+1,nys:nyn+1), &
1408	start = (/ 1, 1, 1, do2d_xy_time_count(av) /), &
1409	count = (/ nx+2, ny+2, 1, 1 /) )
1410	ELSE
1411	nc_stat = NF90_PUT_VAR( id_set_xy(av), &
1412	id_var_do2d(av,if), &
1413	local_2d(nxl:nxr+1,nys:nyn+1), &
1414	start = (/ 1, 1, is, do2d_xy_time_count(av) /), &
1415	count = (/ nx+2, ny+2, 1, 1 /) )
1416	ENDIF
1417	CALL netcdf_handle_error( 'data_output_2d', 447 )
1418	#endif
1419	#endif
1420
1421	!
1422	!-- For 2D-arrays (e.g. u*) only one cross-section is available.
1423	!-- Hence exit loop of output levels.
1424	IF ( two_d ) THEN
1425	IF ( netcdf_data_format < 5 ) two_d = .FALSE.
1426	EXIT loop1
1427	ENDIF
1428
1429	CASE ( 'xz' )
1430	!
1431	!-- Update the netCDF xz cross section time axis.
1432	!-- In case of parallel output, this is only done by PE0
1433	!-- to increase the performance.
1434	IF ( simulated_time /= do2d_xz_last_time(av) ) THEN
1435	do2d_xz_time_count(av) = do2d_xz_time_count(av) + 1
1436	do2d_xz_last_time(av) = simulated_time
1437	IF ( myid == 0 ) THEN
1438	IF ( .NOT. data_output_2d_on_each_pe &
1439	.OR. netcdf_data_format > 4 ) &
1440	THEN
1441	#if defined( __netcdf )
1442	nc_stat = NF90_PUT_VAR( id_set_xz(av), &
1443	id_var_time_xz(av), &
1444	(/ time_since_reference_point /), &
1445	start = (/ do2d_xz_time_count(av) /), &
1446	count = (/ 1 /) )
1447	CALL netcdf_handle_error( 'data_output_2d', 56 )
1448	#endif
1449	ENDIF
1450	ENDIF
1451	ENDIF
1452
1453	!
1454	!-- If required, carry out averaging along y
1455	IF ( section(is,s_ind) == -1 ) THEN
1456
1457	ALLOCATE( local_2d_l(nxlg:nxrg,nzb_do:nzt_do) )
1458	local_2d_l = 0.0_wp
1459	ngp = ( nxrg-nxlg + 1 ) * ( nzt_do-nzb_do + 1 )
1460	!
1461	!-- First local averaging on the PE
1462	DO k = nzb_do, nzt_do
1463	DO j = nys, nyn
1464	DO i = nxlg, nxrg
1465	local_2d_l(i,k) = local_2d_l(i,k) + &
1466	local_pf(i,j,k)
1467	ENDDO
1468	ENDDO
1469	ENDDO
1470	#if defined( __parallel )
1471	!
1472	!-- Now do the averaging over all PEs along y
1473	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
1474	CALL MPI_ALLREDUCE( local_2d_l(nxlg,nzb_do), &
1475	local_2d(nxlg,nzb_do), ngp, MPI_REAL, &
1476	MPI_SUM, comm1dy, ierr )
1477	#else
1478	local_2d = local_2d_l
1479	#endif
1480	local_2d = local_2d / ( ny + 1.0_wp )
1481
1482	DEALLOCATE( local_2d_l )
1483
1484	ELSE
1485	!
1486	!-- Just store the respective section on the local array
1487	!-- (but only if it is available on this PE!)
1488	IF ( section(is,s_ind) >= nys .AND. section(is,s_ind) <= nyn ) &
1489	THEN
1490	local_2d = local_pf(:,section(is,s_ind),nzb_do:nzt_do)
1491	ENDIF
1492
1493	ENDIF
1494
1495	#if defined( __parallel )
1496	IF ( netcdf_data_format > 4 ) THEN
1497	!
1498	!-- Output in netCDF4/HDF5 format.
1499	!-- Output only on those PEs where the respective cross
1500	!-- sections reside. Cross sections averaged along y are
1501	!-- output on the respective first PE along y (myidy=0).
1502	IF ( ( section(is,s_ind) >= nys .AND. &
1503	section(is,s_ind) <= nyn ) .OR. &
1504	( section(is,s_ind) == -1 .AND. myidy == 0 ) ) THEN
1505	#if defined( __netcdf )
1506	!
1507	!-- For parallel output, all cross sections are first
1508	!-- stored here on a local array and will be written to the
1509	!-- output file afterwards to increase the performance.
1510	DO i = nxlg, nxrg
1511	DO k = nzb_do, nzt_do
1512	local_2d_sections_l(i,is,k) = local_2d(i,k)
1513	ENDDO
1514	ENDDO
1515	#endif
1516	ENDIF
1517
1518	ELSE
1519
1520	IF ( data_output_2d_on_each_pe ) THEN
1521	!
1522	!-- Output of partial arrays on each PE. If the cross
1523	!-- section does not reside on the PE, output special
1524	!-- index values.
1525	#if defined( __netcdf )
1526	IF ( myid == 0 ) THEN
1527	WRITE ( 22 ) time_since_reference_point, &
1528	do2d_xz_time_count(av), av
1529	ENDIF
1530	#endif
1531	DO i = 0, io_blocks-1
1532	IF ( i == io_group ) THEN
1533	IF ( ( section(is,s_ind) >= nys .AND. &
1534	section(is,s_ind) <= nyn ) .OR. &
1535	( section(is,s_ind) == -1 .AND. &
1536	nys-1 == -1 ) ) &
1537	THEN
1538	WRITE (22) nxlg, nxrg, nzb_do, nzt_do, nzb, nzt+1
1539	WRITE (22) local_2d
1540	ELSE
1541	WRITE (22) -1, -1, -1, -1, -1, -1
1542	ENDIF
1543	ENDIF
1544	#if defined( __parallel )
1545	CALL MPI_BARRIER( comm2d, ierr )
1546	#endif
1547	ENDDO
1548
1549	ELSE
1550	!
1551	!-- PE0 receives partial arrays from all processors of the
1552	!-- respective cross section and outputs them. Here a
1553	!-- barrier has to be set, because otherwise
1554	!-- "-MPI- FATAL: Remote protocol queue full" may occur.
1555	CALL MPI_BARRIER( comm2d, ierr )
1556
1557	ngp = ( nxrg-nxlg + 1 ) * ( nzt_do-nzb_do + 1 )
1558	IF ( myid == 0 ) THEN
1559	!
1560	!-- Local array can be relocated directly.
1561	IF ( ( section(is,s_ind) >= nys .AND. &
1562	section(is,s_ind) <= nyn ) .OR. &
1563	( section(is,s_ind) == -1 .AND. &
1564	nys-1 == -1 ) ) THEN
1565	total_2d(nxlg:nxrg,nzb_do:nzt_do) = local_2d
1566	ENDIF
1567	!
1568	!-- Receive data from all other PEs.
1569	DO n = 1, numprocs-1
1570	!
1571	!-- Receive index limits first, then array.
1572	!-- Index limits are received in arbitrary order from
1573	!-- the PEs.
1574	CALL MPI_RECV( ind(1), 4, MPI_INTEGER, &
1575	MPI_ANY_SOURCE, 0, comm2d, &
1576	status, ierr )
1577	!
1578	!-- Not all PEs have data for XZ-cross-section.
1579	IF ( ind(1) /= -9999 ) THEN
1580	sender = status(MPI_SOURCE)
1581	DEALLOCATE( local_2d )
1582	ALLOCATE( local_2d(ind(1):ind(2), &
1583	ind(3):ind(4)) )
1584	CALL MPI_RECV( local_2d(ind(1),ind(3)), ngp, &
1585	MPI_REAL, sender, 1, comm2d, &
1586	status, ierr )
1587	total_2d(ind(1):ind(2),ind(3):ind(4)) = &
1588	local_2d
1589	ENDIF
1590	ENDDO
1591	!
1592	!-- Relocate the local array for the next loop increment
1593	DEALLOCATE( local_2d )
1594	ALLOCATE( local_2d(nxlg:nxrg,nzb_do:nzt_do) )
1595
1596	#if defined( __netcdf )
1597	nc_stat = NF90_PUT_VAR( id_set_xz(av), &
1598	id_var_do2d(av,if), &
1599	total_2d(0:nx+1,nzb_do:nzt_do),&
1600	start = (/ 1, is, 1, do2d_xz_time_count(av) /), &
1601	count = (/ nx+2, 1, nzt_do-nzb_do+1, 1 /) )
1602	CALL netcdf_handle_error( 'data_output_2d', 58 )
1603	#endif
1604
1605	ELSE
1606	!
1607	!-- If the cross section resides on the PE, send the
1608	!-- local index limits, otherwise send -9999 to PE0.
1609	IF ( ( section(is,s_ind) >= nys .AND. &
1610	section(is,s_ind) <= nyn ) .OR. &
1611	( section(is,s_ind) == -1 .AND. nys-1 == -1 ) ) &
1612	THEN
1613	ind(1) = nxlg; ind(2) = nxrg
1614	ind(3) = nzb_do; ind(4) = nzt_do
1615	ELSE
1616	ind(1) = -9999; ind(2) = -9999
1617	ind(3) = -9999; ind(4) = -9999
1618	ENDIF
1619	CALL MPI_SEND( ind(1), 4, MPI_INTEGER, 0, 0, &
1620	comm2d, ierr )
1621	!
1622	!-- If applicable, send data to PE0.
1623	IF ( ind(1) /= -9999 ) THEN
1624	CALL MPI_SEND( local_2d(nxlg,nzb_do), ngp, &
1625	MPI_REAL, 0, 1, comm2d, ierr )
1626	ENDIF
1627	ENDIF
1628	!
1629	!-- A barrier has to be set, because otherwise some PEs may
1630	!-- proceed too fast so that PE0 may receive wrong data on
1631	!-- tag 0
1632	CALL MPI_BARRIER( comm2d, ierr )
1633	ENDIF
1634
1635	ENDIF
1636	#else
1637	#if defined( __netcdf )
1638	nc_stat = NF90_PUT_VAR( id_set_xz(av), &
1639	id_var_do2d(av,if), &
1640	local_2d(nxl:nxr+1,nzb_do:nzt_do), &
1641	start = (/ 1, is, 1, do2d_xz_time_count(av) /), &
1642	count = (/ nx+2, 1, nzt_do-nzb_do+1, 1 /) )
1643	CALL netcdf_handle_error( 'data_output_2d', 451 )
1644	#endif
1645	#endif
1646
1647	CASE ( 'yz' )
1648	!
1649	!-- Update the netCDF yz cross section time axis.
1650	!-- In case of parallel output, this is only done by PE0
1651	!-- to increase the performance.
1652	IF ( simulated_time /= do2d_yz_last_time(av) ) THEN
1653	do2d_yz_time_count(av) = do2d_yz_time_count(av) + 1
1654	do2d_yz_last_time(av) = simulated_time
1655	IF ( myid == 0 ) THEN
1656	IF ( .NOT. data_output_2d_on_each_pe &
1657	.OR. netcdf_data_format > 4 ) &
1658	THEN
1659	#if defined( __netcdf )
1660	nc_stat = NF90_PUT_VAR( id_set_yz(av), &
1661	id_var_time_yz(av), &
1662	(/ time_since_reference_point /), &
1663	start = (/ do2d_yz_time_count(av) /), &
1664	count = (/ 1 /) )
1665	CALL netcdf_handle_error( 'data_output_2d', 59 )
1666	#endif
1667	ENDIF
1668	ENDIF
1669	ENDIF
1670
1671	!
1672	!-- If required, carry out averaging along x
1673	IF ( section(is,s_ind) == -1 ) THEN
1674
1675	ALLOCATE( local_2d_l(nysg:nyng,nzb_do:nzt_do) )
1676	local_2d_l = 0.0_wp
1677	ngp = ( nyng-nysg+1 ) * ( nzt_do-nzb_do+1 )
1678	!
1679	!-- First local averaging on the PE
1680	DO k = nzb_do, nzt_do
1681	DO j = nysg, nyng
1682	DO i = nxl, nxr
1683	local_2d_l(j,k) = local_2d_l(j,k) + &
1684	local_pf(i,j,k)
1685	ENDDO
1686	ENDDO
1687	ENDDO
1688	#if defined( __parallel )
1689	!
1690	!-- Now do the averaging over all PEs along x
1691	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
1692	CALL MPI_ALLREDUCE( local_2d_l(nysg,nzb_do), &
1693	local_2d(nysg,nzb_do), ngp, MPI_REAL, &
1694	MPI_SUM, comm1dx, ierr )
1695	#else
1696	local_2d = local_2d_l
1697	#endif
1698	local_2d = local_2d / ( nx + 1.0_wp )
1699
1700	DEALLOCATE( local_2d_l )
1701
1702	ELSE
1703	!
1704	!-- Just store the respective section on the local array
1705	!-- (but only if it is available on this PE!)
1706	IF ( section(is,s_ind) >= nxl .AND. section(is,s_ind) <= nxr ) &
1707	THEN
1708	local_2d = local_pf(section(is,s_ind),:,nzb_do:nzt_do)
1709	ENDIF
1710
1711	ENDIF
1712
1713	#if defined( __parallel )
1714	IF ( netcdf_data_format > 4 ) THEN
1715	!
1716	!-- Output in netCDF4/HDF5 format.
1717	!-- Output only on those PEs where the respective cross
1718	!-- sections reside. Cross sections averaged along x are
1719	!-- output on the respective first PE along x (myidx=0).
1720	IF ( ( section(is,s_ind) >= nxl .AND. &
1721	section(is,s_ind) <= nxr ) .OR. &
1722	( section(is,s_ind) == -1 .AND. myidx == 0 ) ) THEN
1723	#if defined( __netcdf )
1724	!
1725	!-- For parallel output, all cross sections are first
1726	!-- stored here on a local array and will be written to the
1727	!-- output file afterwards to increase the performance.
1728	DO j = nysg, nyng
1729	DO k = nzb_do, nzt_do
1730	local_2d_sections_l(is,j,k) = local_2d(j,k)
1731	ENDDO
1732	ENDDO
1733	#endif
1734	ENDIF
1735
1736	ELSE
1737
1738	IF ( data_output_2d_on_each_pe ) THEN
1739	!
1740	!-- Output of partial arrays on each PE. If the cross
1741	!-- section does not reside on the PE, output special
1742	!-- index values.
1743	#if defined( __netcdf )
1744	IF ( myid == 0 ) THEN
1745	WRITE ( 23 ) time_since_reference_point, &
1746	do2d_yz_time_count(av), av
1747	ENDIF
1748	#endif
1749	DO i = 0, io_blocks-1
1750	IF ( i == io_group ) THEN
1751	IF ( ( section(is,s_ind) >= nxl .AND. &
1752	section(is,s_ind) <= nxr ) .OR. &
1753	( section(is,s_ind) == -1 .AND. &
1754	nxl-1 == -1 ) ) &
1755	THEN
1756	WRITE (23) nysg, nyng, nzb_do, nzt_do, nzb, nzt+1
1757	WRITE (23) local_2d
1758	ELSE
1759	WRITE (23) -1, -1, -1, -1, -1, -1
1760	ENDIF
1761	ENDIF
1762	#if defined( __parallel )
1763	CALL MPI_BARRIER( comm2d, ierr )
1764	#endif
1765	ENDDO
1766
1767	ELSE
1768	!
1769	!-- PE0 receives partial arrays from all processors of the
1770	!-- respective cross section and outputs them. Here a
1771	!-- barrier has to be set, because otherwise
1772	!-- "-MPI- FATAL: Remote protocol queue full" may occur.
1773	CALL MPI_BARRIER( comm2d, ierr )
1774
1775	ngp = ( nyng-nysg+1 ) * ( nzt_do-nzb_do+1 )
1776	IF ( myid == 0 ) THEN
1777	!
1778	!-- Local array can be relocated directly.
1779	IF ( ( section(is,s_ind) >= nxl .AND. &
1780	section(is,s_ind) <= nxr ) .OR. &
1781	( section(is,s_ind) == -1 .AND. nxl-1 == -1 ) ) &
1782	THEN
1783	total_2d(nysg:nyng,nzb_do:nzt_do) = local_2d
1784	ENDIF
1785	!
1786	!-- Receive data from all other PEs.
1787	DO n = 1, numprocs-1
1788	!
1789	!-- Receive index limits first, then array.
1790	!-- Index limits are received in arbitrary order from
1791	!-- the PEs.
1792	CALL MPI_RECV( ind(1), 4, MPI_INTEGER, &
1793	MPI_ANY_SOURCE, 0, comm2d, &
1794	status, ierr )
1795	!
1796	!-- Not all PEs have data for YZ-cross-section.
1797	IF ( ind(1) /= -9999 ) THEN
1798	sender = status(MPI_SOURCE)
1799	DEALLOCATE( local_2d )
1800	ALLOCATE( local_2d(ind(1):ind(2), &
1801	ind(3):ind(4)) )
1802	CALL MPI_RECV( local_2d(ind(1),ind(3)), ngp, &
1803	MPI_REAL, sender, 1, comm2d, &
1804	status, ierr )
1805	total_2d(ind(1):ind(2),ind(3):ind(4)) = &
1806	local_2d
1807	ENDIF
1808	ENDDO
1809	!
1810	!-- Relocate the local array for the next loop increment
1811	DEALLOCATE( local_2d )
1812	ALLOCATE( local_2d(nysg:nyng,nzb_do:nzt_do) )
1813
1814	#if defined( __netcdf )
1815	nc_stat = NF90_PUT_VAR( id_set_yz(av), &
1816	id_var_do2d(av,if), &
1817	total_2d(0:ny+1,nzb_do:nzt_do),&
1818	start = (/ is, 1, 1, do2d_yz_time_count(av) /), &
1819	count = (/ 1, ny+2, nzt_do-nzb_do+1, 1 /) )
1820	CALL netcdf_handle_error( 'data_output_2d', 61 )
1821	#endif
1822
1823	ELSE
1824	!
1825	!-- If the cross section resides on the PE, send the
1826	!-- local index limits, otherwise send -9999 to PE0.
1827	IF ( ( section(is,s_ind) >= nxl .AND. &
1828	section(is,s_ind) <= nxr ) .OR. &
1829	( section(is,s_ind) == -1 .AND. nxl-1 == -1 ) ) &
1830	THEN
1831	ind(1) = nysg; ind(2) = nyng
1832	ind(3) = nzb_do; ind(4) = nzt_do
1833	ELSE
1834	ind(1) = -9999; ind(2) = -9999
1835	ind(3) = -9999; ind(4) = -9999
1836	ENDIF
1837	CALL MPI_SEND( ind(1), 4, MPI_INTEGER, 0, 0, &
1838	comm2d, ierr )
1839	!
1840	!-- If applicable, send data to PE0.
1841	IF ( ind(1) /= -9999 ) THEN
1842	CALL MPI_SEND( local_2d(nysg,nzb_do), ngp, &
1843	MPI_REAL, 0, 1, comm2d, ierr )
1844	ENDIF
1845	ENDIF
1846	!
1847	!-- A barrier has to be set, because otherwise some PEs may
1848	!-- proceed too fast so that PE0 may receive wrong data on
1849	!-- tag 0
1850	CALL MPI_BARRIER( comm2d, ierr )
1851	ENDIF
1852
1853	ENDIF
1854	#else
1855	#if defined( __netcdf )
1856	nc_stat = NF90_PUT_VAR( id_set_yz(av), &
1857	id_var_do2d(av,if), &
1858	local_2d(nys:nyn+1,nzb_do:nzt_do), &
1859	start = (/ is, 1, 1, do2d_xz_time_count(av) /), &
1860	count = (/ 1, ny+2, nzt_do-nzb_do+1, 1 /) )
1861	CALL netcdf_handle_error( 'data_output_2d', 452 )
1862	#endif
1863	#endif
1864
1865	END SELECT
1866
1867	is = is + 1
1868	ENDDO loop1
1869
1870	!
1871	!-- For parallel output, all data were collected before on a local array
1872	!-- and are written now to the netcdf file. This must be done to increase
1873	!-- the performance of the parallel output.
1874	#if defined( __netcdf )
1875	IF ( netcdf_data_format > 4 ) THEN
1876
1877	SELECT CASE ( mode )
1878
1879	CASE ( 'xy' )
1880	IF ( two_d ) THEN
1881	nis = 1
1882	two_d = .FALSE.
1883	ELSE
1884	nis = ns
1885	ENDIF
1886	!
1887	!-- Do not output redundant ghost point data except for the
1888	!-- boundaries of the total domain.
1889	IF ( nxr == nx .AND. nyn /= ny ) THEN
1890	nc_stat = NF90_PUT_VAR( id_set_xy(av), &
1891	id_var_do2d(av,if), &
1892	local_2d_sections(nxl:nxr+1, &
1893	nys:nyn,1:nis), &
1894	start = (/ nxl+1, nys+1, 1, &
1895	do2d_xy_time_count(av) /), &
1896	count = (/ nxr-nxl+2, &
1897	nyn-nys+1, nis, 1 &
1898	/) )
1899	ELSEIF ( nxr /= nx .AND. nyn == ny ) THEN
1900	nc_stat = NF90_PUT_VAR( id_set_xy(av), &
1901	id_var_do2d(av,if), &
1902	local_2d_sections(nxl:nxr, &
1903	nys:nyn+1,1:nis), &
1904	start = (/ nxl+1, nys+1, 1, &
1905	do2d_xy_time_count(av) /), &
1906	count = (/ nxr-nxl+1, &
1907	nyn-nys+2, nis, 1 &
1908	/) )
1909	ELSEIF ( nxr == nx .AND. nyn == ny ) THEN
1910	nc_stat = NF90_PUT_VAR( id_set_xy(av), &
1911	id_var_do2d(av,if), &
1912	local_2d_sections(nxl:nxr+1, &
1913	nys:nyn+1,1:nis), &
1914	start = (/ nxl+1, nys+1, 1, &
1915	do2d_xy_time_count(av) /), &
1916	count = (/ nxr-nxl+2, &
1917	nyn-nys+2, nis, 1 &
1918	/) )
1919	ELSE
1920	nc_stat = NF90_PUT_VAR( id_set_xy(av), &
1921	id_var_do2d(av,if), &
1922	local_2d_sections(nxl:nxr, &
1923	nys:nyn,1:nis), &
1924	start = (/ nxl+1, nys+1, 1, &
1925	do2d_xy_time_count(av) /), &
1926	count = (/ nxr-nxl+1, &
1927	nyn-nys+1, nis, 1 &
1928	/) )
1929	ENDIF
1930
1931	CALL netcdf_handle_error( 'data_output_2d', 55 )
1932
1933	CASE ( 'xz' )
1934	!
1935	!-- First, all PEs get the information of all cross-sections.
1936	!-- Then the data are written to the output file by all PEs
1937	!-- while NF90_COLLECTIVE is set in subroutine
1938	!-- define_netcdf_header. Although redundant information are
1939	!-- written to the output file in that case, the performance
1940	!-- is significantly better compared to the case where only
1941	!-- the first row of PEs in x-direction (myidx = 0) is given
1942	!-- the output while NF90_INDEPENDENT is set.
1943	IF ( npey /= 1 ) THEN
1944
1945	#if defined( __parallel )
1946	!
1947	!-- Distribute data over all PEs along y
1948	ngp = ( nxrg-nxlg+1 ) * ( nzt_do-nzb_do+1 ) * ns
1949	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
1950	CALL MPI_ALLREDUCE( local_2d_sections_l(nxlg,1,nzb_do), &
1951	local_2d_sections(nxlg,1,nzb_do), &
1952	ngp, MPI_REAL, MPI_SUM, comm1dy, &
1953	ierr )
1954	#else
1955	local_2d_sections = local_2d_sections_l
1956	#endif
1957	ENDIF
1958	!
1959	!-- Do not output redundant ghost point data except for the
1960	!-- boundaries of the total domain.
1961	IF ( nxr == nx ) THEN
1962	nc_stat = NF90_PUT_VAR( id_set_xz(av), &
1963	id_var_do2d(av,if), &
1964	local_2d_sections(nxl:nxr+1,1:ns, &
1965	nzb_do:nzt_do), &
1966	start = (/ nxl+1, 1, 1, &
1967	do2d_xz_time_count(av) /), &
1968	count = (/ nxr-nxl+2, ns, nzt_do-nzb_do+1, &
1969	1 /) )
1970	ELSE
1971	nc_stat = NF90_PUT_VAR( id_set_xz(av), &
1972	id_var_do2d(av,if), &
1973	local_2d_sections(nxl:nxr,1:ns, &
1974	nzb_do:nzt_do), &
1975	start = (/ nxl+1, 1, 1, &
1976	do2d_xz_time_count(av) /), &
1977	count = (/ nxr-nxl+1, ns, nzt_do-nzb_do+1, &
1978	1 /) )
1979	ENDIF
1980
1981	CALL netcdf_handle_error( 'data_output_2d', 57 )
1982
1983	CASE ( 'yz' )
1984	!
1985	!-- First, all PEs get the information of all cross-sections.
1986	!-- Then the data are written to the output file by all PEs
1987	!-- while NF90_COLLECTIVE is set in subroutine
1988	!-- define_netcdf_header. Although redundant information are
1989	!-- written to the output file in that case, the performance
1990	!-- is significantly better compared to the case where only
1991	!-- the first row of PEs in y-direction (myidy = 0) is given
1992	!-- the output while NF90_INDEPENDENT is set.
1993	IF ( npex /= 1 ) THEN
1994
1995	#if defined( __parallel )
1996	!
1997	!-- Distribute data over all PEs along x
1998	ngp = ( nyng-nysg+1 ) * ( nzt-nzb + 2 ) * ns
1999	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
2000	CALL MPI_ALLREDUCE( local_2d_sections_l(1,nysg,nzb_do), &
2001	local_2d_sections(1,nysg,nzb_do), &
2002	ngp, MPI_REAL, MPI_SUM, comm1dx, &
2003	ierr )
2004	#else
2005	local_2d_sections = local_2d_sections_l
2006	#endif
2007	ENDIF
2008	!
2009	!-- Do not output redundant ghost point data except for the
2010	!-- boundaries of the total domain.
2011	IF ( nyn == ny ) THEN
2012	nc_stat = NF90_PUT_VAR( id_set_yz(av), &
2013	id_var_do2d(av,if), &
2014	local_2d_sections(1:ns, &
2015	nys:nyn+1,nzb_do:nzt_do), &
2016	start = (/ 1, nys+1, 1, &
2017	do2d_yz_time_count(av) /), &
2018	count = (/ ns, nyn-nys+2, &
2019	nzt_do-nzb_do+1, 1 /) )
2020	ELSE
2021	nc_stat = NF90_PUT_VAR( id_set_yz(av), &
2022	id_var_do2d(av,if), &
2023	local_2d_sections(1:ns,nys:nyn, &
2024	nzb_do:nzt_do), &
2025	start = (/ 1, nys+1, 1, &
2026	do2d_yz_time_count(av) /), &
2027	count = (/ ns, nyn-nys+1, &
2028	nzt_do-nzb_do+1, 1 /) )
2029	ENDIF
2030
2031	CALL netcdf_handle_error( 'data_output_2d', 60 )
2032
2033	CASE DEFAULT
2034	message_string = 'unknown cross-section: ' // TRIM( mode )
2035	CALL message( 'data_output_2d', 'PA0180', 1, 2, 0, 6, 0 )
2036
2037	END SELECT
2038
2039	ENDIF
2040	#endif
2041	ENDIF
2042
2043	if = if + 1
2044	l = MAX( 2, LEN_TRIM( do2d(av,if) ) )
2045	do2d_mode = do2d(av,if)(l-1:l)
2046
2047	ENDDO
2048
2049	!
2050	!-- Deallocate temporary arrays.
2051	IF ( ALLOCATED( level_z ) ) DEALLOCATE( level_z )
2052	IF ( netcdf_data_format > 4 ) THEN
2053	DEALLOCATE( local_pf, local_2d, local_2d_sections )
2054	IF( mode == 'xz' .OR. mode == 'yz' ) DEALLOCATE( local_2d_sections_l )
2055	ENDIF
2056	#if defined( __parallel )
2057	IF ( .NOT. data_output_2d_on_each_pe .AND. myid == 0 ) THEN
2058	DEALLOCATE( total_2d )
2059	ENDIF
2060	#endif
2061
2062	!
2063	!-- Close plot output file.
2064	file_id = 20 + s_ind
2065
2066	IF ( data_output_2d_on_each_pe ) THEN
2067	DO i = 0, io_blocks-1
2068	IF ( i == io_group ) THEN
2069	CALL close_file( file_id )
2070	ENDIF
2071	#if defined( __parallel )
2072	CALL MPI_BARRIER( comm2d, ierr )
2073	#endif
2074	ENDDO
2075	ELSE
2076	IF ( myid == 0 ) CALL close_file( file_id )
2077	ENDIF
2078
2079	CALL cpu_log( log_point(3), 'data_output_2d', 'stop' )
2080
2081	END SUBROUTINE data_output_2d

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

Download in other formats:

| Impressum | ©Leibniz Universität Hannover |