Home

Context Navigation

source: palm/trunk/SOURCE/data_output_2d.f90 @ 3012

Last change on this file since 3012 was 3004, checked in by Giersch, 6 years ago
precipitation_rate removed, further allocation checks for data output of averaged quantities implemented, double CALL of flow_statistics at the beginning of time_integration removed, further minor bugfixes, comments added
Property svn:keywords set to `Id`
File size: 97.6 KB

Line
1	!> @file data_output_2d.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 1997-2018 Leibniz Universitaet Hannover
18	!------------------------------------------------------------------------------!
19	!
20	! Current revisions:
21	! -----------------
22	!
23	!
24	! Former revisions:
25	! -----------------
26	! $Id: data_output_2d.f90 3004 2018-04-27 12:33:25Z Giersch $
27	! precipitation_rate removed, case prr*_xy removed, to_be_resorted have to point
28	! to ql_vp_av and not to ql_vp, allocation checks implemented (averaged data
29	! will be assigned to fill values if no allocation happened so far)
30	!
31	! 2963 2018-04-12 14:47:44Z suehring
32	! Introduce index for vegetation/wall, pavement/green-wall and water/window
33	! surfaces, for clearer access of surface fraction, albedo, emissivity, etc. .
34	!
35	! 2817 2018-02-19 16:32:21Z knoop
36	! Preliminary gust module interface implemented
37	!
38	! 2805 2018-02-14 17:00:09Z suehring
39	! Consider also default-type surfaces for surface temperature output.
40	!
41	! 2797 2018-02-08 13:24:35Z suehring
42	! Enable output of ground-heat flux also at urban surfaces.
43	!
44	! 2743 2018-01-12 16:03:39Z suehring
45	! In case of natural- and urban-type surfaces output surfaces fluxes in W/m2.
46	!
47	! 2742 2018-01-12 14:59:47Z suehring
48	! Enable output of surface temperature
49	!
50	! 2735 2018-01-11 12:01:27Z suehring
51	! output of r_a moved from land-surface to consider also urban-type surfaces
52	!
53	! 2718 2018-01-02 08:49:38Z maronga
54	! Corrected "Former revisions" section
55	!
56	! 2696 2017-12-14 17:12:51Z kanani
57	! Change in file header (GPL part)
58	! Implementation of uv exposure model (FK)
59	! Implementation of turbulence_closure_mod (TG)
60	! Set fill values at topography grid points or e.g. non-natural-type surface
61	! in case of LSM output (MS)
62	!
63	! 2512 2017-10-04 08:26:59Z raasch
64	! upper bounds of cross section output changed from nx+1,ny+1 to nx,ny
65	! no output of ghost layer data
66	!
67	! 2292 2017-06-20 09:51:42Z schwenkel
68	! Implementation of new microphysic scheme: cloud_scheme = 'morrison'
69	! includes two more prognostic equations for cloud drop concentration (nc)
70	! and cloud water content (qc).
71	!
72	! 2277 2017-06-12 10:47:51Z kanani
73	! Removed unused variables do2d_xy_n, do2d_xz_n, do2d_yz_n
74	!
75	! 2233 2017-05-30 18:08:54Z suehring
76	!
77	! 2232 2017-05-30 17:47:52Z suehring
78	! Adjustments to new surface concept
79	!
80	!
81	! 2190 2017-03-21 12:16:43Z raasch
82	! bugfix for r2031: string rho replaced by rho_ocean
83	!
84	! 2031 2016-10-21 15:11:58Z knoop
85	! renamed variable rho to rho_ocean and rho_av to rho_ocean_av
86	!
87	! 2000 2016-08-20 18:09:15Z knoop
88	! Forced header and separation lines into 80 columns
89	!
90	! 1980 2016-07-29 15:51:57Z suehring
91	! Bugfix, in order to steer user-defined output, setting flag found explicitly
92	! to .F.
93	!
94	! 1976 2016-07-27 13:28:04Z maronga
95	! Output of radiation quantities is now done directly in the respective module
96	!
97	! 1972 2016-07-26 07:52:02Z maronga
98	! Output of land surface quantities is now done directly in the respective
99	! module
100	!
101	! 1960 2016-07-12 16:34:24Z suehring
102	! Scalar surface flux added
103	! Rename INTEGER variable s into s_ind, as s is already assigned to scalar
104	!
105	! 1849 2016-04-08 11:33:18Z hoffmann
106	! precipitation_amount, precipitation_rate, prr moved to arrays_3d
107	!
108	! 1822 2016-04-07 07:49:42Z hoffmann
109	! Output of bulk cloud physics simplified.
110	!
111	! 1788 2016-03-10 11:01:04Z maronga
112	! Added output of z0q
113	!
114	! 1783 2016-03-06 18:36:17Z raasch
115	! name change of netcdf routines and module + related changes
116	!
117	! 1745 2016-02-05 13:06:51Z gronemeier
118	! Bugfix: test if time axis limit exceeds moved to point after call of check_open
119	!
120	! 1703 2015-11-02 12:38:44Z raasch
121	! bugfix for output of single (*) xy-sections in case of parallel netcdf I/O
122	!
123	! 1701 2015-11-02 07:43:04Z maronga
124	! Bugfix in output of RRTGM data
125	!
126	! 1691 2015-10-26 16:17:44Z maronga
127	! Added output of Obukhov length (ol) and radiative heating rates for RRTMG.
128	! Formatting corrections.
129	!
130	! 1682 2015-10-07 23:56:08Z knoop
131	! Code annotations made doxygen readable
132	!
133	! 1585 2015-04-30 07:05:52Z maronga
134	! Added support for RRTMG
135	!
136	! 1555 2015-03-04 17:44:27Z maronga
137	! Added output of r_a and r_s
138	!
139	! 1551 2015-03-03 14:18:16Z maronga
140	! Added suppport for land surface model and radiation model output. In the course
141	! of this action, the limits for vertical loops have been changed (from nzb and
142	! nzt+1 to nzb_do and nzt_do, respectively in order to allow soil model output).
143	! Moreover, a new vertical grid zs was introduced.
144	!
145	! 1359 2014-04-11 17:15:14Z hoffmann
146	! New particle structure integrated.
147	!
148	! 1353 2014-04-08 15:21:23Z heinze
149	! REAL constants provided with KIND-attribute
150	!
151	! 1327 2014-03-21 11:00:16Z raasch
152	! parts concerning iso2d output removed,
153	! -netcdf output queries
154	!
155	! 1320 2014-03-20 08:40:49Z raasch
156	! ONLY-attribute added to USE-statements,
157	! kind-parameters added to all INTEGER and REAL declaration statements,
158	! kinds are defined in new module kinds,
159	! revision history before 2012 removed,
160	! comment fields (!:) to be used for variable explanations added to
161	! all variable declaration statements
162	!
163	! 1318 2014-03-17 13:35:16Z raasch
164	! barrier argument removed from cpu_log.
165	! module interfaces removed
166	!
167	! 1311 2014-03-14 12:13:39Z heinze
168	! bugfix: close #if defined( __netcdf )
169	!
170	! 1308 2014-03-13 14:58:42Z fricke
171	! +local_2d_sections, local_2d_sections_l, ns
172	! Check, if the limit of the time dimension is exceeded for parallel output
173	! To increase the performance for parallel output, the following is done:
174	! - Update of time axis is only done by PE0
175	! - Cross sections are first stored on a local array and are written
176	! collectively to the output file by all PEs.
177	!
178	! 1115 2013-03-26 18:16:16Z hoffmann
179	! ql is calculated by calc_liquid_water_content
180	!
181	! 1076 2012-12-05 08:30:18Z hoffmann
182	! Bugfix in output of ql
183	!
184	! 1065 2012-11-22 17:42:36Z hoffmann
185	! Bugfix: Output of cross sections of ql
186	!
187	! 1053 2012-11-13 17:11:03Z hoffmann
188	! +qr, nr, qc and cross sections
189	!
190	! 1036 2012-10-22 13:43:42Z raasch
191	! code put under GPL (PALM 3.9)
192	!
193	! 1031 2012-10-19 14:35:30Z raasch
194	! netCDF4 without parallel file support implemented
195	!
196	! 1007 2012-09-19 14:30:36Z franke
197	! Bugfix: missing calculation of ql_vp added
198	!
199	! 978 2012-08-09 08:28:32Z fricke
200	! +z0h
201	!
202	! Revision 1.1 1997/08/11 06:24:09 raasch
203	! Initial revision
204	!
205	!
206	! Description:
207	! ------------
208	!> Data output of cross-sections in netCDF format or binary format
209	!> to be later converted to NetCDF by helper routine combine_plot_fields.
210	!> Attention: The position of the sectional planes is still not always computed
211	!> --------- correctly. (zu is used always)!
212	!------------------------------------------------------------------------------!
213	SUBROUTINE data_output_2d( mode, av )
214
215
216	USE arrays_3d, &
217	ONLY: dzw, e, heatflux_output_conversion, nc, nr, p, pt, &
218	precipitation_amount, prr, q, qc, ql, ql_c, ql_v, ql_vp, qr, &
219	rho_ocean, s, sa, tend, u, v, vpt, w, zu, zw, &
220	waterflux_output_conversion
221
222	USE averaging
223
224	USE cloud_parameters, &
225	ONLY: cp, hyrho, l_d_cp, l_v, pt_d_t
226
227	USE control_parameters, &
228	ONLY: cloud_physics, data_output_2d_on_each_pe, data_output_xy, &
229	data_output_xz, data_output_yz, do2d, &
230	do2d_xy_last_time, do2d_xy_time_count, &
231	do2d_xz_last_time, do2d_xz_time_count, &
232	do2d_yz_last_time, do2d_yz_time_count, &
233	ibc_uv_b, io_blocks, io_group, land_surface, message_string, &
234	ntdim_2d_xy, ntdim_2d_xz, ntdim_2d_yz, &
235	psolver, section, simulated_time, simulated_time_chr, &
236	time_since_reference_point, uv_exposure
237
238	USE cpulog, &
239	ONLY: cpu_log, log_point
240
241	USE grid_variables, &
242	ONLY: dx, dy
243
244	USE gust_mod, &
245	ONLY: gust_data_output_2d, gust_module_enabled
246
247	USE indices, &
248	ONLY: nbgp, nx, nxl, nxlg, nxr, nxrg, ny, nyn, nyng, nys, nysg, nz, &
249	nzb, nzt, wall_flags_0
250
251	USE kinds
252
253	USE land_surface_model_mod, &
254	ONLY: lsm_data_output_2d, zs
255
256	#if defined( __netcdf )
257	USE NETCDF
258	#endif
259
260	USE netcdf_interface, &
261	ONLY: fill_value, id_set_xy, id_set_xz, id_set_yz, id_var_do2d, &
262	id_var_time_xy, id_var_time_xz, id_var_time_yz, nc_stat, &
263	netcdf_data_format, netcdf_handle_error
264
265	USE particle_attributes, &
266	ONLY: grid_particles, number_of_particles, particle_advection_start, &
267	particles, prt_count
268
269	USE pegrid
270
271	USE radiation_model_mod, &
272	ONLY: radiation, radiation_data_output_2d
273
274	USE surface_mod, &
275	ONLY: ind_pav_green, ind_veg_wall, ind_wat_win, surf_def_h, &
276	surf_lsm_h, surf_usm_h
277
278	USE turbulence_closure_mod, &
279	ONLY: tcm_data_output_2d
280
281	USE uv_exposure_model_mod, &
282	ONLY: uvem_data_output_2d
283
284
285	IMPLICIT NONE
286
287	CHARACTER (LEN=2) :: do2d_mode !<
288	CHARACTER (LEN=2) :: mode !<
289	CHARACTER (LEN=4) :: grid !<
290	CHARACTER (LEN=25) :: section_chr !<
291	CHARACTER (LEN=50) :: rtext !<
292
293	INTEGER(iwp) :: av !<
294	INTEGER(iwp) :: ngp !<
295	INTEGER(iwp) :: file_id !<
296	INTEGER(iwp) :: flag_nr !< number of masking flag
297	INTEGER(iwp) :: i !<
298	INTEGER(iwp) :: if !<
299	INTEGER(iwp) :: is !<
300	INTEGER(iwp) :: iis !<
301	INTEGER(iwp) :: j !<
302	INTEGER(iwp) :: k !<
303	INTEGER(iwp) :: l !<
304	INTEGER(iwp) :: layer_xy !<
305	INTEGER(iwp) :: m !<
306	INTEGER(iwp) :: n !<
307	INTEGER(iwp) :: nis !<
308	INTEGER(iwp) :: ns !<
309	INTEGER(iwp) :: nzb_do !< lower limit of the data field (usually nzb)
310	INTEGER(iwp) :: nzt_do !< upper limit of the data field (usually nzt+1)
311	INTEGER(iwp) :: psi !<
312	INTEGER(iwp) :: s_ind !<
313	INTEGER(iwp) :: sender !<
314	INTEGER(iwp) :: ind(4) !<
315
316	LOGICAL :: found !<
317	LOGICAL :: resorted !<
318	LOGICAL :: two_d !<
319
320	REAL(wp) :: mean_r !<
321	REAL(wp) :: s_r2 !<
322	REAL(wp) :: s_r3 !<
323
324	REAL(wp), DIMENSION(:), ALLOCATABLE :: level_z !<
325	REAL(wp), DIMENSION(:,:), ALLOCATABLE :: local_2d !<
326	REAL(wp), DIMENSION(:,:), ALLOCATABLE :: local_2d_l !<
327
328	REAL(wp), DIMENSION(:,:,:), ALLOCATABLE :: local_pf !<
329	REAL(wp), DIMENSION(:,:,:), ALLOCATABLE :: local_2d_sections !<
330	REAL(wp), DIMENSION(:,:,:), ALLOCATABLE :: local_2d_sections_l !<
331
332	#if defined( __parallel )
333	REAL(wp), DIMENSION(:,:), ALLOCATABLE :: total_2d !<
334	#endif
335	REAL(wp), DIMENSION(:,:,:), POINTER :: to_be_resorted !<
336
337	NAMELIST /LOCAL/ rtext
338
339	!
340	!-- Immediate return, if no output is requested (no respective sections
341	!-- found in parameter data_output)
342	IF ( mode == 'xy' .AND. .NOT. data_output_xy(av) ) RETURN
343	IF ( mode == 'xz' .AND. .NOT. data_output_xz(av) ) RETURN
344	IF ( mode == 'yz' .AND. .NOT. data_output_yz(av) ) RETURN
345
346	CALL cpu_log (log_point(3),'data_output_2d','start')
347
348	two_d = .FALSE. ! local variable to distinguish between output of pure 2D
349	! arrays and cross-sections of 3D arrays.
350
351	!
352	!-- Depending on the orientation of the cross-section, the respective output
353	!-- files have to be opened.
354	SELECT CASE ( mode )
355
356	CASE ( 'xy' )
357	s_ind = 1
358	ALLOCATE( level_z(nzb:nzt+1), local_2d(nxl:nxr,nys:nyn) )
359
360	IF ( netcdf_data_format > 4 ) THEN
361	ns = 1
362	DO WHILE ( section(ns,s_ind) /= -9999 .AND. ns <= 100 )
363	ns = ns + 1
364	ENDDO
365	ns = ns - 1
366	ALLOCATE( local_2d_sections(nxl:nxr,nys:nyn,1:ns) )
367	local_2d_sections = 0.0_wp
368	ENDIF
369
370	!
371	!-- Parallel netCDF4/HDF5 output is done on all PEs, all other on PE0 only
372	IF ( myid == 0 .OR. netcdf_data_format > 4 ) THEN
373	CALL check_open( 101+av*10 )
374	ENDIF
375	IF ( data_output_2d_on_each_pe ) THEN
376	CALL check_open( 21 )
377	ELSE
378	IF ( myid == 0 ) THEN
379	#if defined( __parallel )
380	ALLOCATE( total_2d(0:nx,0:ny) )
381	#endif
382	ENDIF
383	ENDIF
384
385	CASE ( 'xz' )
386	s_ind = 2
387	ALLOCATE( local_2d(nxl:nxr,nzb:nzt+1) )
388
389	IF ( netcdf_data_format > 4 ) THEN
390	ns = 1
391	DO WHILE ( section(ns,s_ind) /= -9999 .AND. ns <= 100 )
392	ns = ns + 1
393	ENDDO
394	ns = ns - 1
395	ALLOCATE( local_2d_sections(nxl:nxr,1:ns,nzb:nzt+1) )
396	ALLOCATE( local_2d_sections_l(nxl:nxr,1:ns,nzb:nzt+1) )
397	local_2d_sections = 0.0_wp; local_2d_sections_l = 0.0_wp
398	ENDIF
399
400	!
401	!-- Parallel netCDF4/HDF5 output is done on all PEs, all other on PE0 only
402	IF ( myid == 0 .OR. netcdf_data_format > 4 ) THEN
403	CALL check_open( 102+av*10 )
404	ENDIF
405
406	IF ( data_output_2d_on_each_pe ) THEN
407	CALL check_open( 22 )
408	ELSE
409	IF ( myid == 0 ) THEN
410	#if defined( __parallel )
411	ALLOCATE( total_2d(0:nx,nzb:nzt+1) )
412	#endif
413	ENDIF
414	ENDIF
415
416	CASE ( 'yz' )
417	s_ind = 3
418	ALLOCATE( local_2d(nys:nyn,nzb:nzt+1) )
419
420	IF ( netcdf_data_format > 4 ) THEN
421	ns = 1
422	DO WHILE ( section(ns,s_ind) /= -9999 .AND. ns <= 100 )
423	ns = ns + 1
424	ENDDO
425	ns = ns - 1
426	ALLOCATE( local_2d_sections(1:ns,nys:nyn,nzb:nzt+1) )
427	ALLOCATE( local_2d_sections_l(1:ns,nys:nyn,nzb:nzt+1) )
428	local_2d_sections = 0.0_wp; local_2d_sections_l = 0.0_wp
429	ENDIF
430
431	!
432	!-- Parallel netCDF4/HDF5 output is done on all PEs, all other on PE0 only
433	IF ( myid == 0 .OR. netcdf_data_format > 4 ) THEN
434	CALL check_open( 103+av*10 )
435	ENDIF
436
437	IF ( data_output_2d_on_each_pe ) THEN
438	CALL check_open( 23 )
439	ELSE
440	IF ( myid == 0 ) THEN
441	#if defined( __parallel )
442	ALLOCATE( total_2d(0:ny,nzb:nzt+1) )
443	#endif
444	ENDIF
445	ENDIF
446
447	CASE DEFAULT
448	message_string = 'unknown cross-section: ' // TRIM( mode )
449	CALL message( 'data_output_2d', 'PA0180', 1, 2, 0, 6, 0 )
450
451	END SELECT
452
453	!
454	!-- For parallel netcdf output the time axis must be limited. Return, if this
455	!-- limit is exceeded. This could be the case, if the simulated time exceeds
456	!-- the given end time by the length of the given output interval.
457	IF ( netcdf_data_format > 4 ) THEN
458	IF ( mode == 'xy' .AND. do2d_xy_time_count(av) + 1 > &
459	ntdim_2d_xy(av) ) THEN
460	WRITE ( message_string, * ) 'Output of xy cross-sections is not ', &
461	'given at t=', simulated_time, '&because the', &
462	' maximum number of output time levels is exceeded.'
463	CALL message( 'data_output_2d', 'PA0384', 0, 1, 0, 6, 0 )
464	CALL cpu_log( log_point(3), 'data_output_2d', 'stop' )
465	RETURN
466	ENDIF
467	IF ( mode == 'xz' .AND. do2d_xz_time_count(av) + 1 > &
468	ntdim_2d_xz(av) ) THEN
469	WRITE ( message_string, * ) 'Output of xz cross-sections is not ', &
470	'given at t=', simulated_time, '&because the', &
471	' maximum number of output time levels is exceeded.'
472	CALL message( 'data_output_2d', 'PA0385', 0, 1, 0, 6, 0 )
473	CALL cpu_log( log_point(3), 'data_output_2d', 'stop' )
474	RETURN
475	ENDIF
476	IF ( mode == 'yz' .AND. do2d_yz_time_count(av) + 1 > &
477	ntdim_2d_yz(av) ) THEN
478	WRITE ( message_string, * ) 'Output of yz cross-sections is not ', &
479	'given at t=', simulated_time, '&because the', &
480	' maximum number of output time levels is exceeded.'
481	CALL message( 'data_output_2d', 'PA0386', 0, 1, 0, 6, 0 )
482	CALL cpu_log( log_point(3), 'data_output_2d', 'stop' )
483	RETURN
484	ENDIF
485	ENDIF
486
487	!
488	!-- Allocate a temporary array for resorting (kji -> ijk).
489	ALLOCATE( local_pf(nxl:nxr,nys:nyn,nzb:nzt+1) )
490	local_pf = 0.0
491
492	!
493	!-- Loop of all variables to be written.
494	!-- Output dimensions chosen
495	if = 1
496	l = MAX( 2, LEN_TRIM( do2d(av,if) ) )
497	do2d_mode = do2d(av,if)(l-1:l)
498
499	DO WHILE ( do2d(av,if)(1:1) /= ' ' )
500
501	IF ( do2d_mode == mode ) THEN
502	!
503	!-- Set flag to steer output of radiation, land-surface, or user-defined
504	!-- quantities
505	found = .FALSE.
506
507	nzb_do = nzb
508	nzt_do = nzt+1
509	!
510	!-- Before each output, set array local_pf to fill value
511	local_pf = fill_value
512	!
513	!-- Set masking flag for topography for not resorted arrays
514	flag_nr = 0
515
516	!
517	!-- Store the array chosen on the temporary array.
518	resorted = .FALSE.
519	SELECT CASE ( TRIM( do2d(av,if) ) )
520	CASE ( 'e_xy', 'e_xz', 'e_yz' )
521	IF ( av == 0 ) THEN
522	to_be_resorted => e
523	ELSE
524	IF ( .NOT. ALLOCATED( e_av ) ) THEN
525	ALLOCATE( e_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
526	e_av = REAL( fill_value, KIND = wp )
527	ENDIF
528	to_be_resorted => e_av
529	ENDIF
530	IF ( mode == 'xy' ) level_z = zu
531
532	CASE ( 'lpt_xy', 'lpt_xz', 'lpt_yz' )
533	IF ( av == 0 ) THEN
534	to_be_resorted => pt
535	ELSE
536	IF ( .NOT. ALLOCATED( lpt_av ) ) THEN
537	ALLOCATE( lpt_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
538	lpt_av = REAL( fill_value, KIND = wp )
539	ENDIF
540	to_be_resorted => lpt_av
541	ENDIF
542	IF ( mode == 'xy' ) level_z = zu
543
544	CASE ( 'lwp*_xy' ) ! 2d-array
545	IF ( av == 0 ) THEN
546	DO i = nxl, nxr
547	DO j = nys, nyn
548	local_pf(i,j,nzb+1) = SUM( ql(nzb:nzt,j,i) * &
549	dzw(1:nzt+1) )
550	ENDDO
551	ENDDO
552	ELSE
553	IF ( .NOT. ALLOCATED( lwp_av ) ) THEN
554	ALLOCATE( lwp_av(nysg:nyng,nxlg:nxrg) )
555	lwp_av = REAL( fill_value, KIND = wp )
556	ENDIF
557	DO i = nxl, nxr
558	DO j = nys, nyn
559	local_pf(i,j,nzb+1) = lwp_av(j,i)
560	ENDDO
561	ENDDO
562	ENDIF
563	resorted = .TRUE.
564	two_d = .TRUE.
565	level_z(nzb+1) = zu(nzb+1)
566
567	CASE ( 'nc_xy', 'nc_xz', 'nc_yz' )
568	IF ( av == 0 ) THEN
569	to_be_resorted => nc
570	ELSE
571	IF ( .NOT. ALLOCATED( nc_av ) ) THEN
572	ALLOCATE( nc_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
573	nc_av = REAL( fill_value, KIND = wp )
574	ENDIF
575	to_be_resorted => nc_av
576	ENDIF
577	IF ( mode == 'xy' ) level_z = zu
578
579	CASE ( 'nr_xy', 'nr_xz', 'nr_yz' )
580	IF ( av == 0 ) THEN
581	to_be_resorted => nr
582	ELSE
583	IF ( .NOT. ALLOCATED( nr_av ) ) THEN
584	ALLOCATE( nr_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
585	nr_av = REAL( fill_value, KIND = wp )
586	ENDIF
587	to_be_resorted => nr_av
588	ENDIF
589	IF ( mode == 'xy' ) level_z = zu
590
591	CASE ( 'ghf*_xy' ) ! 2d-array
592	IF ( av == 0 ) THEN
593	DO m = 1, surf_lsm_h%ns
594	i = surf_lsm_h%i(m)
595	j = surf_lsm_h%j(m)
596	local_pf(i,j,nzb+1) = surf_lsm_h%ghf(m)
597	ENDDO
598	DO m = 1, surf_usm_h%ns
599	i = surf_usm_h%i(m)
600	j = surf_usm_h%j(m)
601	local_pf(i,j,nzb+1) = surf_usm_h%frac(ind_veg_wall,m) * &
602	surf_usm_h%wghf_eb(m) + &
603	surf_usm_h%frac(ind_pav_green,m) * &
604	surf_usm_h%wghf_eb_green(m) + &
605	surf_usm_h%frac(ind_wat_win,m) * &
606	surf_usm_h%wghf_eb_window(m)
607	ENDDO
608	ELSE
609	IF ( .NOT. ALLOCATED( ghf_av ) ) THEN
610	ALLOCATE( ghf_av(nysg:nyng,nxlg:nxrg) )
611	ghf_av = REAL( fill_value, KIND = wp )
612	ENDIF
613	DO i = nxl, nxr
614	DO j = nys, nyn
615	local_pf(i,j,nzb+1) = ghf_av(j,i)
616	ENDDO
617	ENDDO
618	ENDIF
619
620	resorted = .TRUE.
621	two_d = .TRUE.
622	level_z(nzb+1) = zu(nzb+1)
623
624	CASE ( 'ol*_xy' ) ! 2d-array
625	IF ( av == 0 ) THEN
626	DO m = 1, surf_def_h(0)%ns
627	i = surf_def_h(0)%i(m)
628	j = surf_def_h(0)%j(m)
629	local_pf(i,j,nzb+1) = surf_def_h(0)%ol(m)
630	ENDDO
631	DO m = 1, surf_lsm_h%ns
632	i = surf_lsm_h%i(m)
633	j = surf_lsm_h%j(m)
634	local_pf(i,j,nzb+1) = surf_lsm_h%ol(m)
635	ENDDO
636	DO m = 1, surf_usm_h%ns
637	i = surf_usm_h%i(m)
638	j = surf_usm_h%j(m)
639	local_pf(i,j,nzb+1) = surf_usm_h%ol(m)
640	ENDDO
641	ELSE
642	IF ( .NOT. ALLOCATED( ol_av ) ) THEN
643	ALLOCATE( ol_av(nysg:nyng,nxlg:nxrg) )
644	ol_av = REAL( fill_value, KIND = wp )
645	ENDIF
646	DO i = nxl, nxr
647	DO j = nys, nyn
648	local_pf(i,j,nzb+1) = ol_av(j,i)
649	ENDDO
650	ENDDO
651	ENDIF
652	resorted = .TRUE.
653	two_d = .TRUE.
654	level_z(nzb+1) = zu(nzb+1)
655
656	CASE ( 'p_xy', 'p_xz', 'p_yz' )
657	IF ( av == 0 ) THEN
658	IF ( psolver /= 'sor' ) CALL exchange_horiz( p, nbgp )
659	to_be_resorted => p
660	ELSE
661	IF ( .NOT. ALLOCATED( p_av ) ) THEN
662	ALLOCATE( p_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
663	p_av = REAL( fill_value, KIND = wp )
664	ENDIF
665	IF ( psolver /= 'sor' ) CALL exchange_horiz( p_av, nbgp )
666	to_be_resorted => p_av
667	ENDIF
668	IF ( mode == 'xy' ) level_z = zu
669
670	CASE ( 'pc_xy', 'pc_xz', 'pc_yz' ) ! particle concentration
671	IF ( av == 0 ) THEN
672	IF ( simulated_time >= particle_advection_start ) THEN
673	tend = prt_count
674	! CALL exchange_horiz( tend, nbgp )
675	ELSE
676	tend = 0.0_wp
677	ENDIF
678	DO i = nxl, nxr
679	DO j = nys, nyn
680	DO k = nzb, nzt+1
681	local_pf(i,j,k) = tend(k,j,i)
682	ENDDO
683	ENDDO
684	ENDDO
685	resorted = .TRUE.
686	ELSE
687	IF ( .NOT. ALLOCATED( pc_av ) ) THEN
688	ALLOCATE( pc_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
689	pc_av = REAL( fill_value, KIND = wp )
690	ENDIF
691	! CALL exchange_horiz( pc_av, nbgp )
692	to_be_resorted => pc_av
693	ENDIF
694
695	CASE ( 'pr_xy', 'pr_xz', 'pr_yz' ) ! mean particle radius (effective radius)
696	IF ( av == 0 ) THEN
697	IF ( simulated_time >= particle_advection_start ) THEN
698	DO i = nxl, nxr
699	DO j = nys, nyn
700	DO k = nzb, nzt+1
701	number_of_particles = prt_count(k,j,i)
702	IF (number_of_particles <= 0) CYCLE
703	particles => grid_particles(k,j,i)%particles(1:number_of_particles)
704	s_r2 = 0.0_wp
705	s_r3 = 0.0_wp
706	DO n = 1, number_of_particles
707	IF ( particles(n)%particle_mask ) THEN
708	s_r2 = s_r2 + particles(n)%radius*2 &
709	particles(n)%weight_factor
710	s_r3 = s_r3 + particles(n)%radius*3 &
711	particles(n)%weight_factor
712	ENDIF
713	ENDDO
714	IF ( s_r2 > 0.0_wp ) THEN
715	mean_r = s_r3 / s_r2
716	ELSE
717	mean_r = 0.0_wp
718	ENDIF
719	tend(k,j,i) = mean_r
720	ENDDO
721	ENDDO
722	ENDDO
723	! CALL exchange_horiz( tend, nbgp )
724	ELSE
725	tend = 0.0_wp
726	ENDIF
727	DO i = nxl, nxr
728	DO j = nys, nyn
729	DO k = nzb, nzt+1
730	local_pf(i,j,k) = tend(k,j,i)
731	ENDDO
732	ENDDO
733	ENDDO
734	resorted = .TRUE.
735	ELSE
736	IF ( .NOT. ALLOCATED( pr_av ) ) THEN
737	ALLOCATE( pr_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
738	pr_av = REAL( fill_value, KIND = wp )
739	ENDIF
740	! CALL exchange_horiz( pr_av, nbgp )
741	to_be_resorted => pr_av
742	ENDIF
743
744	CASE ( 'pra*_xy' ) ! 2d-array / integral quantity => no av
745	! CALL exchange_horiz_2d( precipitation_amount )
746	DO i = nxl, nxr
747	DO j = nys, nyn
748	local_pf(i,j,nzb+1) = precipitation_amount(j,i)
749	ENDDO
750	ENDDO
751	precipitation_amount = 0.0_wp ! reset for next integ. interval
752	resorted = .TRUE.
753	two_d = .TRUE.
754	level_z(nzb+1) = zu(nzb+1)
755
756	CASE ( 'prr_xy', 'prr_xz', 'prr_yz' )
757	IF ( av == 0 ) THEN
758	! CALL exchange_horiz( prr, nbgp )
759	DO i = nxl, nxr
760	DO j = nys, nyn
761	DO k = nzb, nzt+1
762	local_pf(i,j,k) = prr(k,j,i) * hyrho(nzb+1)
763	ENDDO
764	ENDDO
765	ENDDO
766	ELSE
767	IF ( .NOT. ALLOCATED( prr_av ) ) THEN
768	ALLOCATE( prr_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
769	prr_av = REAL( fill_value, KIND = wp )
770	ENDIF
771	! CALL exchange_horiz( prr_av, nbgp )
772	DO i = nxl, nxr
773	DO j = nys, nyn
774	DO k = nzb, nzt+1
775	local_pf(i,j,k) = prr_av(k,j,i) * hyrho(nzb+1)
776	ENDDO
777	ENDDO
778	ENDDO
779	ENDIF
780	resorted = .TRUE.
781	IF ( mode == 'xy' ) level_z = zu
782
783	CASE ( 'pt_xy', 'pt_xz', 'pt_yz' )
784	IF ( av == 0 ) THEN
785	IF ( .NOT. cloud_physics ) THEN
786	to_be_resorted => pt
787	ELSE
788	DO i = nxl, nxr
789	DO j = nys, nyn
790	DO k = nzb, nzt+1
791	local_pf(i,j,k) = pt(k,j,i) + l_d_cp * &
792	pt_d_t(k) * &
793	ql(k,j,i)
794	ENDDO
795	ENDDO
796	ENDDO
797	resorted = .TRUE.
798	ENDIF
799	ELSE
800	IF ( .NOT. ALLOCATED( pt_av ) ) THEN
801	ALLOCATE( pt_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
802	pt_av = REAL( fill_value, KIND = wp )
803	ENDIF
804	to_be_resorted => pt_av
805	ENDIF
806	IF ( mode == 'xy' ) level_z = zu
807
808	CASE ( 'q_xy', 'q_xz', 'q_yz' )
809	IF ( av == 0 ) THEN
810	to_be_resorted => q
811	ELSE
812	IF ( .NOT. ALLOCATED( q_av ) ) THEN
813	ALLOCATE( q_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
814	q_av = REAL( fill_value, KIND = wp )
815	ENDIF
816	to_be_resorted => q_av
817	ENDIF
818	IF ( mode == 'xy' ) level_z = zu
819
820	CASE ( 'qc_xy', 'qc_xz', 'qc_yz' )
821	IF ( av == 0 ) THEN
822	to_be_resorted => qc
823	ELSE
824	IF ( .NOT. ALLOCATED( qc_av ) ) THEN
825	ALLOCATE( qc_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
826	qc_av = REAL( fill_value, KIND = wp )
827	ENDIF
828	to_be_resorted => qc_av
829	ENDIF
830	IF ( mode == 'xy' ) level_z = zu
831
832	CASE ( 'ql_xy', 'ql_xz', 'ql_yz' )
833	IF ( av == 0 ) THEN
834	to_be_resorted => ql
835	ELSE
836	IF ( .NOT. ALLOCATED( ql_av ) ) THEN
837	ALLOCATE( ql_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
838	ql_av = REAL( fill_value, KIND = wp )
839	ENDIF
840	to_be_resorted => ql_av
841	ENDIF
842	IF ( mode == 'xy' ) level_z = zu
843
844	CASE ( 'ql_c_xy', 'ql_c_xz', 'ql_c_yz' )
845	IF ( av == 0 ) THEN
846	to_be_resorted => ql_c
847	ELSE
848	IF ( .NOT. ALLOCATED( ql_c_av ) ) THEN
849	ALLOCATE( ql_c_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
850	ql_c_av = REAL( fill_value, KIND = wp )
851	ENDIF
852	to_be_resorted => ql_c_av
853	ENDIF
854	IF ( mode == 'xy' ) level_z = zu
855
856	CASE ( 'ql_v_xy', 'ql_v_xz', 'ql_v_yz' )
857	IF ( av == 0 ) THEN
858	to_be_resorted => ql_v
859	ELSE
860	IF ( .NOT. ALLOCATED( ql_v_av ) ) THEN
861	ALLOCATE( ql_v_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
862	ql_v_av = REAL( fill_value, KIND = wp )
863	ENDIF
864	to_be_resorted => ql_v_av
865	ENDIF
866	IF ( mode == 'xy' ) level_z = zu
867
868	CASE ( 'ql_vp_xy', 'ql_vp_xz', 'ql_vp_yz' )
869	IF ( av == 0 ) THEN
870	IF ( simulated_time >= particle_advection_start ) THEN
871	DO i = nxl, nxr
872	DO j = nys, nyn
873	DO k = nzb, nzt+1
874	number_of_particles = prt_count(k,j,i)
875	IF (number_of_particles <= 0) CYCLE
876	particles => grid_particles(k,j,i)%particles(1:number_of_particles)
877	DO n = 1, number_of_particles
878	IF ( particles(n)%particle_mask ) THEN
879	tend(k,j,i) = tend(k,j,i) + &
880	particles(n)%weight_factor / &
881	prt_count(k,j,i)
882	ENDIF
883	ENDDO
884	ENDDO
885	ENDDO
886	ENDDO
887	! CALL exchange_horiz( tend, nbgp )
888	ELSE
889	tend = 0.0_wp
890	ENDIF
891	DO i = nxl, nxr
892	DO j = nys, nyn
893	DO k = nzb, nzt+1
894	local_pf(i,j,k) = tend(k,j,i)
895	ENDDO
896	ENDDO
897	ENDDO
898	resorted = .TRUE.
899	ELSE
900	IF ( .NOT. ALLOCATED( ql_vp_av ) ) THEN
901	ALLOCATE( ql_vp_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
902	ql_vp_av = REAL( fill_value, KIND = wp )
903	ENDIF
904	! CALL exchange_horiz( ql_vp_av, nbgp )
905	to_be_resorted => ql_vp_av
906	ENDIF
907	IF ( mode == 'xy' ) level_z = zu
908
909	CASE ( 'qr_xy', 'qr_xz', 'qr_yz' )
910	IF ( av == 0 ) THEN
911	to_be_resorted => qr
912	ELSE
913	IF ( .NOT. ALLOCATED( qr_av ) ) THEN
914	ALLOCATE( qr_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
915	qr_av = REAL( fill_value, KIND = wp )
916	ENDIF
917	to_be_resorted => qr_av
918	ENDIF
919	IF ( mode == 'xy' ) level_z = zu
920
921	CASE ( 'qsws*_xy' ) ! 2d-array
922	IF ( av == 0 ) THEN
923	!
924	!-- In case of default surfaces, clean-up flux by density.
925	!-- In case of land- and urban-surfaces, convert fluxes into
926	!-- dynamic units
927	DO m = 1, surf_def_h(0)%ns
928	i = surf_def_h(0)%i(m)
929	j = surf_def_h(0)%j(m)
930	k = surf_def_h(0)%k(m)
931	local_pf(i,j,nzb+1) = surf_def_h(0)%qsws(m) * &
932	waterflux_output_conversion(k)
933	ENDDO
934	DO m = 1, surf_lsm_h%ns
935	i = surf_lsm_h%i(m)
936	j = surf_lsm_h%j(m)
937	k = surf_lsm_h%k(m)
938	local_pf(i,j,nzb+1) = surf_lsm_h%qsws(m) * l_v
939	ENDDO
940	DO m = 1, surf_usm_h%ns
941	i = surf_usm_h%i(m)
942	j = surf_usm_h%j(m)
943	k = surf_usm_h%k(m)
944	local_pf(i,j,nzb+1) = surf_usm_h%qsws(m) * l_v
945	ENDDO
946	ELSE
947	IF ( .NOT. ALLOCATED( qsws_av ) ) THEN
948	ALLOCATE( qsws_av(nysg:nyng,nxlg:nxrg) )
949	qsws_av = REAL( fill_value, KIND = wp )
950	ENDIF
951	DO i = nxl, nxr
952	DO j = nys, nyn
953	local_pf(i,j,nzb+1) = qsws_av(j,i)
954	ENDDO
955	ENDDO
956	ENDIF
957	resorted = .TRUE.
958	two_d = .TRUE.
959	level_z(nzb+1) = zu(nzb+1)
960
961	CASE ( 'qv_xy', 'qv_xz', 'qv_yz' )
962	IF ( av == 0 ) THEN
963	DO i = nxl, nxr
964	DO j = nys, nyn
965	DO k = nzb, nzt+1
966	local_pf(i,j,k) = q(k,j,i) - ql(k,j,i)
967	ENDDO
968	ENDDO
969	ENDDO
970	resorted = .TRUE.
971	ELSE
972	IF ( .NOT. ALLOCATED( qv_av ) ) THEN
973	ALLOCATE( qv_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
974	qv_av = REAL( fill_value, KIND = wp )
975	ENDIF
976	to_be_resorted => qv_av
977	ENDIF
978	IF ( mode == 'xy' ) level_z = zu
979
980	CASE ( 'r_a*_xy' ) ! 2d-array
981	IF ( av == 0 ) THEN
982	DO m = 1, surf_lsm_h%ns
983	i = surf_lsm_h%i(m)
984	j = surf_lsm_h%j(m)
985	local_pf(i,j,nzb+1) = surf_lsm_h%r_a(m)
986	ENDDO
987
988	DO m = 1, surf_usm_h%ns
989	i = surf_usm_h%i(m)
990	j = surf_usm_h%j(m)
991	local_pf(i,j,nzb+1) = &
992	( surf_usm_h%frac(ind_veg_wall,m) * &
993	surf_usm_h%r_a(m) + &
994	surf_usm_h%frac(ind_pav_green,m) * &
995	surf_usm_h%r_a_green(m) + &
996	surf_usm_h%frac(ind_wat_win,m) * &
997	surf_usm_h%r_a_window(m) )
998	ENDDO
999	ELSE
1000	IF ( .NOT. ALLOCATED( r_a_av ) ) THEN
1001	ALLOCATE( r_a_av(nysg:nyng,nxlg:nxrg) )
1002	r_a_av = REAL( fill_value, KIND = wp )
1003	ENDIF
1004	DO i = nxl, nxr
1005	DO j = nys, nyn
1006	local_pf(i,j,nzb+1) = r_a_av(j,i)
1007	ENDDO
1008	ENDDO
1009	ENDIF
1010	resorted = .TRUE.
1011	two_d = .TRUE.
1012	level_z(nzb+1) = zu(nzb+1)
1013
1014	CASE ( 'rho_ocean_xy', 'rho_ocean_xz', 'rho_ocean_yz' )
1015	IF ( av == 0 ) THEN
1016	to_be_resorted => rho_ocean
1017	ELSE
1018	IF ( .NOT. ALLOCATED( rho_ocean_av ) ) THEN
1019	ALLOCATE( rho_ocean_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
1020	rho_ocean_av = REAL( fill_value, KIND = wp )
1021	ENDIF
1022	to_be_resorted => rho_ocean_av
1023	ENDIF
1024
1025	CASE ( 's_xy', 's_xz', 's_yz' )
1026	IF ( av == 0 ) THEN
1027	to_be_resorted => s
1028	ELSE
1029	IF ( .NOT. ALLOCATED( s_av ) ) THEN
1030	ALLOCATE( s_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
1031	s_av = REAL( fill_value, KIND = wp )
1032	ENDIF
1033	to_be_resorted => s_av
1034	ENDIF
1035
1036	CASE ( 'sa_xy', 'sa_xz', 'sa_yz' )
1037	IF ( av == 0 ) THEN
1038	to_be_resorted => sa
1039	ELSE
1040	IF ( .NOT. ALLOCATED( sa_av ) ) THEN
1041	ALLOCATE( sa_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
1042	sa_av = REAL( fill_value, KIND = wp )
1043	ENDIF
1044	to_be_resorted => sa_av
1045	ENDIF
1046
1047	CASE ( 'shf*_xy' ) ! 2d-array
1048	IF ( av == 0 ) THEN
1049	!
1050	!-- In case of default surfaces, clean-up flux by density.
1051	!-- In case of land- and urban-surfaces, convert fluxes into
1052	!-- dynamic units.
1053	DO m = 1, surf_def_h(0)%ns
1054	i = surf_def_h(0)%i(m)
1055	j = surf_def_h(0)%j(m)
1056	k = surf_def_h(0)%k(m)
1057	local_pf(i,j,nzb+1) = surf_def_h(0)%shf(m) * &
1058	heatflux_output_conversion(k)
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	k = surf_lsm_h%k(m)
1064	local_pf(i,j,nzb+1) = surf_lsm_h%shf(m) * cp
1065	ENDDO
1066	DO m = 1, surf_usm_h%ns
1067	i = surf_usm_h%i(m)
1068	j = surf_usm_h%j(m)
1069	k = surf_usm_h%k(m)
1070	local_pf(i,j,nzb+1) = surf_usm_h%shf(m) * cp
1071	ENDDO
1072	ELSE
1073	IF ( .NOT. ALLOCATED( shf_av ) ) THEN
1074	ALLOCATE( shf_av(nysg:nyng,nxlg:nxrg) )
1075	shf_av = REAL( fill_value, KIND = wp )
1076	ENDIF
1077	DO i = nxl, nxr
1078	DO j = nys, nyn
1079	local_pf(i,j,nzb+1) = shf_av(j,i)
1080	ENDDO
1081	ENDDO
1082	ENDIF
1083	resorted = .TRUE.
1084	two_d = .TRUE.
1085	level_z(nzb+1) = zu(nzb+1)
1086
1087	CASE ( 'ssws*_xy' ) ! 2d-array
1088	IF ( av == 0 ) THEN
1089	DO m = 1, surf_def_h(0)%ns
1090	i = surf_def_h(0)%i(m)
1091	j = surf_def_h(0)%j(m)
1092	local_pf(i,j,nzb+1) = surf_def_h(0)%ssws(m)
1093	ENDDO
1094	DO m = 1, surf_lsm_h%ns
1095	i = surf_lsm_h%i(m)
1096	j = surf_lsm_h%j(m)
1097	local_pf(i,j,nzb+1) = surf_lsm_h%ssws(m)
1098	ENDDO
1099	DO m = 1, surf_usm_h%ns
1100	i = surf_usm_h%i(m)
1101	j = surf_usm_h%j(m)
1102	local_pf(i,j,nzb+1) = surf_usm_h%ssws(m)
1103	ENDDO
1104	ELSE
1105	IF ( .NOT. ALLOCATED( ssws_av ) ) THEN
1106	ALLOCATE( ssws_av(nysg:nyng,nxlg:nxrg) )
1107	ssws_av = REAL( fill_value, KIND = wp )
1108	ENDIF
1109	DO i = nxl, nxr
1110	DO j = nys, nyn
1111	local_pf(i,j,nzb+1) = ssws_av(j,i)
1112	ENDDO
1113	ENDDO
1114	ENDIF
1115	resorted = .TRUE.
1116	two_d = .TRUE.
1117	level_z(nzb+1) = zu(nzb+1)
1118
1119	CASE ( 't*_xy' ) ! 2d-array
1120	IF ( av == 0 ) THEN
1121	DO m = 1, surf_def_h(0)%ns
1122	i = surf_def_h(0)%i(m)
1123	j = surf_def_h(0)%j(m)
1124	local_pf(i,j,nzb+1) = surf_def_h(0)%ts(m)
1125	ENDDO
1126	DO m = 1, surf_lsm_h%ns
1127	i = surf_lsm_h%i(m)
1128	j = surf_lsm_h%j(m)
1129	local_pf(i,j,nzb+1) = surf_lsm_h%ts(m)
1130	ENDDO
1131	DO m = 1, surf_usm_h%ns
1132	i = surf_usm_h%i(m)
1133	j = surf_usm_h%j(m)
1134	local_pf(i,j,nzb+1) = surf_usm_h%ts(m)
1135	ENDDO
1136	ELSE
1137	IF ( .NOT. ALLOCATED( ts_av ) ) THEN
1138	ALLOCATE( ts_av(nysg:nyng,nxlg:nxrg) )
1139	ts_av = REAL( fill_value, KIND = wp )
1140	ENDIF
1141	DO i = nxl, nxr
1142	DO j = nys, nyn
1143	local_pf(i,j,nzb+1) = ts_av(j,i)
1144	ENDDO
1145	ENDDO
1146	ENDIF
1147	resorted = .TRUE.
1148	two_d = .TRUE.
1149	level_z(nzb+1) = zu(nzb+1)
1150
1151	CASE ( 'tsurf*_xy' ) ! 2d-array
1152	IF ( av == 0 ) THEN
1153	DO m = 1, surf_def_h(0)%ns
1154	i = surf_def_h(0)%i(m)
1155	j = surf_def_h(0)%j(m)
1156	local_pf(i,j,nzb+1) = surf_def_h(0)%pt_surface(m)
1157	ENDDO
1158
1159	DO m = 1, surf_lsm_h%ns
1160	i = surf_lsm_h%i(m)
1161	j = surf_lsm_h%j(m)
1162	local_pf(i,j,nzb+1) = surf_lsm_h%pt_surface(m)
1163	ENDDO
1164
1165	DO m = 1, surf_usm_h%ns
1166	i = surf_usm_h%i(m)
1167	j = surf_usm_h%j(m)
1168	local_pf(i,j,nzb+1) = surf_usm_h%pt_surface(m)
1169	ENDDO
1170
1171	ELSE
1172	IF ( .NOT. ALLOCATED( tsurf_av ) ) THEN
1173	ALLOCATE( tsurf_av(nysg:nyng,nxlg:nxrg) )
1174	tsurf_av = REAL( fill_value, KIND = wp )
1175	ENDIF
1176	DO i = nxl, nxr
1177	DO j = nys, nyn
1178	local_pf(i,j,nzb+1) = tsurf_av(j,i)
1179	ENDDO
1180	ENDDO
1181	ENDIF
1182	resorted = .TRUE.
1183	two_d = .TRUE.
1184	level_z(nzb+1) = zu(nzb+1)
1185
1186	CASE ( 'u_xy', 'u_xz', 'u_yz' )
1187	flag_nr = 1
1188	IF ( av == 0 ) THEN
1189	to_be_resorted => u
1190	ELSE
1191	IF ( .NOT. ALLOCATED( u_av ) ) THEN
1192	ALLOCATE( u_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
1193	u_av = REAL( fill_value, KIND = wp )
1194	ENDIF
1195	to_be_resorted => u_av
1196	ENDIF
1197	IF ( mode == 'xy' ) level_z = zu
1198	!
1199	!-- Substitute the values generated by "mirror" boundary condition
1200	!-- at the bottom boundary by the real surface values.
1201	IF ( do2d(av,if) == 'u_xz' .OR. do2d(av,if) == 'u_yz' ) THEN
1202	IF ( ibc_uv_b == 0 ) local_pf(:,:,nzb) = 0.0_wp
1203	ENDIF
1204
1205	CASE ( 'u*_xy' ) ! 2d-array
1206	IF ( av == 0 ) THEN
1207	DO m = 1, surf_def_h(0)%ns
1208	i = surf_def_h(0)%i(m)
1209	j = surf_def_h(0)%j(m)
1210	local_pf(i,j,nzb+1) = surf_def_h(0)%us(m)
1211	ENDDO
1212	DO m = 1, surf_lsm_h%ns
1213	i = surf_lsm_h%i(m)
1214	j = surf_lsm_h%j(m)
1215	local_pf(i,j,nzb+1) = surf_lsm_h%us(m)
1216	ENDDO
1217	DO m = 1, surf_usm_h%ns
1218	i = surf_usm_h%i(m)
1219	j = surf_usm_h%j(m)
1220	local_pf(i,j,nzb+1) = surf_usm_h%us(m)
1221	ENDDO
1222	ELSE
1223	IF ( .NOT. ALLOCATED( us_av ) ) THEN
1224	ALLOCATE( us_av(nysg:nyng,nxlg:nxrg) )
1225	us_av = REAL( fill_value, KIND = wp )
1226	ENDIF
1227	DO i = nxl, nxr
1228	DO j = nys, nyn
1229	local_pf(i,j,nzb+1) = us_av(j,i)
1230	ENDDO
1231	ENDDO
1232	ENDIF
1233	resorted = .TRUE.
1234	two_d = .TRUE.
1235	level_z(nzb+1) = zu(nzb+1)
1236
1237	CASE ( 'v_xy', 'v_xz', 'v_yz' )
1238	flag_nr = 2
1239	IF ( av == 0 ) THEN
1240	to_be_resorted => v
1241	ELSE
1242	IF ( .NOT. ALLOCATED( v_av ) ) THEN
1243	ALLOCATE( v_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
1244	v_av = REAL( fill_value, KIND = wp )
1245	ENDIF
1246	to_be_resorted => v_av
1247	ENDIF
1248	IF ( mode == 'xy' ) level_z = zu
1249	!
1250	!-- Substitute the values generated by "mirror" boundary condition
1251	!-- at the bottom boundary by the real surface values.
1252	IF ( do2d(av,if) == 'v_xz' .OR. do2d(av,if) == 'v_yz' ) THEN
1253	IF ( ibc_uv_b == 0 ) local_pf(:,:,nzb) = 0.0_wp
1254	ENDIF
1255
1256	CASE ( 'vpt_xy', 'vpt_xz', 'vpt_yz' )
1257	IF ( av == 0 ) THEN
1258	to_be_resorted => vpt
1259	ELSE
1260	IF ( .NOT. ALLOCATED( vpt_av ) ) THEN
1261	ALLOCATE( vpt_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
1262	vpt_av = REAL( fill_value, KIND = wp )
1263	ENDIF
1264	to_be_resorted => vpt_av
1265	ENDIF
1266	IF ( mode == 'xy' ) level_z = zu
1267
1268	CASE ( 'w_xy', 'w_xz', 'w_yz' )
1269	flag_nr = 3
1270	IF ( av == 0 ) THEN
1271	to_be_resorted => w
1272	ELSE
1273	IF ( .NOT. ALLOCATED( w_av ) ) THEN
1274	ALLOCATE( w_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
1275	w_av = REAL( fill_value, KIND = wp )
1276	ENDIF
1277	to_be_resorted => w_av
1278	ENDIF
1279	IF ( mode == 'xy' ) level_z = zw
1280
1281	CASE ( 'z0*_xy' ) ! 2d-array
1282	IF ( av == 0 ) THEN
1283	DO m = 1, surf_def_h(0)%ns
1284	i = surf_def_h(0)%i(m)
1285	j = surf_def_h(0)%j(m)
1286	local_pf(i,j,nzb+1) = surf_def_h(0)%z0(m)
1287	ENDDO
1288	DO m = 1, surf_lsm_h%ns
1289	i = surf_lsm_h%i(m)
1290	j = surf_lsm_h%j(m)
1291	local_pf(i,j,nzb+1) = surf_lsm_h%z0(m)
1292	ENDDO
1293	DO m = 1, surf_usm_h%ns
1294	i = surf_usm_h%i(m)
1295	j = surf_usm_h%j(m)
1296	local_pf(i,j,nzb+1) = surf_usm_h%z0(m)
1297	ENDDO
1298	ELSE
1299	IF ( .NOT. ALLOCATED( z0_av ) ) THEN
1300	ALLOCATE( z0_av(nysg:nyng,nxlg:nxrg) )
1301	z0_av = REAL( fill_value, KIND = wp )
1302	ENDIF
1303	DO i = nxl, nxr
1304	DO j = nys, nyn
1305	local_pf(i,j,nzb+1) = z0_av(j,i)
1306	ENDDO
1307	ENDDO
1308	ENDIF
1309	resorted = .TRUE.
1310	two_d = .TRUE.
1311	level_z(nzb+1) = zu(nzb+1)
1312
1313	CASE ( 'z0h*_xy' ) ! 2d-array
1314	IF ( av == 0 ) THEN
1315	DO m = 1, surf_def_h(0)%ns
1316	i = surf_def_h(0)%i(m)
1317	j = surf_def_h(0)%j(m)
1318	local_pf(i,j,nzb+1) = surf_def_h(0)%z0h(m)
1319	ENDDO
1320	DO m = 1, surf_lsm_h%ns
1321	i = surf_lsm_h%i(m)
1322	j = surf_lsm_h%j(m)
1323	local_pf(i,j,nzb+1) = surf_lsm_h%z0h(m)
1324	ENDDO
1325	DO m = 1, surf_usm_h%ns
1326	i = surf_usm_h%i(m)
1327	j = surf_usm_h%j(m)
1328	local_pf(i,j,nzb+1) = surf_usm_h%z0h(m)
1329	ENDDO
1330	ELSE
1331	IF ( .NOT. ALLOCATED( z0h_av ) ) THEN
1332	ALLOCATE( z0h_av(nysg:nyng,nxlg:nxrg) )
1333	z0h_av = REAL( fill_value, KIND = wp )
1334	ENDIF
1335	DO i = nxl, nxr
1336	DO j = nys, nyn
1337	local_pf(i,j,nzb+1) = z0h_av(j,i)
1338	ENDDO
1339	ENDDO
1340	ENDIF
1341	resorted = .TRUE.
1342	two_d = .TRUE.
1343	level_z(nzb+1) = zu(nzb+1)
1344
1345	CASE ( 'z0q*_xy' ) ! 2d-array
1346	IF ( av == 0 ) THEN
1347	DO m = 1, surf_def_h(0)%ns
1348	i = surf_def_h(0)%i(m)
1349	j = surf_def_h(0)%j(m)
1350	local_pf(i,j,nzb+1) = surf_def_h(0)%z0q(m)
1351	ENDDO
1352	DO m = 1, surf_lsm_h%ns
1353	i = surf_lsm_h%i(m)
1354	j = surf_lsm_h%j(m)
1355	local_pf(i,j,nzb+1) = surf_lsm_h%z0q(m)
1356	ENDDO
1357	DO m = 1, surf_usm_h%ns
1358	i = surf_usm_h%i(m)
1359	j = surf_usm_h%j(m)
1360	local_pf(i,j,nzb+1) = surf_usm_h%z0q(m)
1361	ENDDO
1362	ELSE
1363	IF ( .NOT. ALLOCATED( z0q_av ) ) THEN
1364	ALLOCATE( z0q_av(nysg:nyng,nxlg:nxrg) )
1365	z0q_av = REAL( fill_value, KIND = wp )
1366	ENDIF
1367	DO i = nxl, nxr
1368	DO j = nys, nyn
1369	local_pf(i,j,nzb+1) = z0q_av(j,i)
1370	ENDDO
1371	ENDDO
1372	ENDIF
1373	resorted = .TRUE.
1374	two_d = .TRUE.
1375	level_z(nzb+1) = zu(nzb+1)
1376
1377	CASE DEFAULT
1378
1379	!
1380	!-- Land surface model quantity
1381	IF ( land_surface ) THEN
1382	CALL lsm_data_output_2d( av, do2d(av,if), found, grid, mode,&
1383	local_pf, two_d, nzb_do, nzt_do )
1384	ENDIF
1385
1386	!
1387	!-- Turbulence closure variables
1388	IF ( .NOT. found ) THEN
1389	CALL tcm_data_output_2d( av, do2d(av,if), found, grid, mode,&
1390	local_pf, two_d, nzb_do, nzt_do )
1391	ENDIF
1392
1393	!
1394	!-- Radiation quantity
1395	IF ( .NOT. found .AND. radiation ) THEN
1396	CALL radiation_data_output_2d( av, do2d(av,if), found, grid,&
1397	mode, local_pf, two_d )
1398	ENDIF
1399
1400	!
1401	!-- Gust module quantities
1402	IF ( .NOT. found .AND. gust_module_enabled ) THEN
1403	CALL gust_data_output_2d( av, do2d(av,if), found, grid, &
1404	local_pf, two_d, nzb_do, nzt_do )
1405	ENDIF
1406
1407	!
1408	!-- UV exposure model quantity
1409	IF ( uv_exposure ) THEN
1410	CALL uvem_data_output_2d( av, do2d(av,if), found, grid, mode,&
1411	local_pf, two_d, nzb_do, nzt_do )
1412	ENDIF
1413
1414	!
1415	!-- User defined quantity
1416	IF ( .NOT. found ) THEN
1417	CALL user_data_output_2d( av, do2d(av,if), found, grid, &
1418	local_pf, two_d, nzb_do, nzt_do )
1419	ENDIF
1420
1421	resorted = .TRUE.
1422
1423	IF ( grid == 'zu' ) THEN
1424	IF ( mode == 'xy' ) level_z = zu
1425	ELSEIF ( grid == 'zw' ) THEN
1426	IF ( mode == 'xy' ) level_z = zw
1427	ELSEIF ( grid == 'zu1' ) THEN
1428	IF ( mode == 'xy' ) level_z(nzb+1) = zu(nzb+1)
1429	ELSEIF ( grid == 'zs' ) THEN
1430	IF ( mode == 'xy' ) level_z = zs
1431	ENDIF
1432
1433	IF ( .NOT. found ) THEN
1434	message_string = 'no output provided for: ' // &
1435	TRIM( do2d(av,if) )
1436	CALL message( 'data_output_2d', 'PA0181', 0, 0, 0, 6, 0 )
1437	ENDIF
1438
1439	END SELECT
1440
1441	!
1442	!-- Resort the array to be output, if not done above. Flag topography
1443	!-- grid points with fill values, using the corresponding maksing flag.
1444	IF ( .NOT. resorted ) THEN
1445	DO i = nxl, nxr
1446	DO j = nys, nyn
1447	DO k = nzb_do, nzt_do
1448	local_pf(i,j,k) = MERGE( to_be_resorted(k,j,i), &
1449	REAL( fill_value, KIND = wp ), &
1450	BTEST( wall_flags_0(k,j,i), &
1451	flag_nr ) )
1452	ENDDO
1453	ENDDO
1454	ENDDO
1455	ENDIF
1456
1457	!
1458	!-- Output of the individual cross-sections, depending on the cross-
1459	!-- section mode chosen.
1460	is = 1
1461	loop1: DO WHILE ( section(is,s_ind) /= -9999 .OR. two_d )
1462
1463	SELECT CASE ( mode )
1464
1465	CASE ( 'xy' )
1466	!
1467	!-- Determine the cross section index
1468	IF ( two_d ) THEN
1469	layer_xy = nzb+1
1470	ELSE
1471	layer_xy = section(is,s_ind)
1472	ENDIF
1473
1474	!
1475	!-- Exit the loop for layers beyond the data output domain
1476	!-- (used for soil model)
1477	IF ( layer_xy > nzt_do ) THEN
1478	EXIT loop1
1479	ENDIF
1480
1481	!
1482	!-- Update the netCDF xy cross section time axis.
1483	!-- In case of parallel output, this is only done by PE0
1484	!-- to increase the performance.
1485	IF ( simulated_time /= do2d_xy_last_time(av) ) THEN
1486	do2d_xy_time_count(av) = do2d_xy_time_count(av) + 1
1487	do2d_xy_last_time(av) = simulated_time
1488	IF ( myid == 0 ) THEN
1489	IF ( .NOT. data_output_2d_on_each_pe &
1490	.OR. netcdf_data_format > 4 ) &
1491	THEN
1492	#if defined( __netcdf )
1493	nc_stat = NF90_PUT_VAR( id_set_xy(av), &
1494	id_var_time_xy(av), &
1495	(/ time_since_reference_point /), &
1496	start = (/ do2d_xy_time_count(av) /), &
1497	count = (/ 1 /) )
1498	CALL netcdf_handle_error( 'data_output_2d', 53 )
1499	#endif
1500	ENDIF
1501	ENDIF
1502	ENDIF
1503	!
1504	!-- If required, carry out averaging along z
1505	IF ( section(is,s_ind) == -1 .AND. .NOT. two_d ) THEN
1506
1507	local_2d = 0.0_wp
1508	!
1509	!-- Carry out the averaging (all data are on the PE)
1510	DO k = nzb_do, nzt_do
1511	DO j = nys, nyn
1512	DO i = nxl, nxr
1513	local_2d(i,j) = local_2d(i,j) + local_pf(i,j,k)
1514	ENDDO
1515	ENDDO
1516	ENDDO
1517
1518	local_2d = local_2d / ( nzt_do - nzb_do + 1.0_wp)
1519
1520	ELSE
1521	!
1522	!-- Just store the respective section on the local array
1523	local_2d = local_pf(:,:,layer_xy)
1524
1525	ENDIF
1526
1527	#if defined( __parallel )
1528	IF ( netcdf_data_format > 4 ) THEN
1529	!
1530	!-- Parallel output in netCDF4/HDF5 format.
1531	IF ( two_d ) THEN
1532	iis = 1
1533	ELSE
1534	iis = is
1535	ENDIF
1536
1537	#if defined( __netcdf )
1538	!
1539	!-- For parallel output, all cross sections are first stored
1540	!-- here on a local array and will be written to the output
1541	!-- file afterwards to increase the performance.
1542	DO i = nxl, nxr
1543	DO j = nys, nyn
1544	local_2d_sections(i,j,iis) = local_2d(i,j)
1545	ENDDO
1546	ENDDO
1547	#endif
1548	ELSE
1549
1550	IF ( data_output_2d_on_each_pe ) THEN
1551	!
1552	!-- Output of partial arrays on each PE
1553	#if defined( __netcdf )
1554	IF ( myid == 0 ) THEN
1555	WRITE ( 21 ) time_since_reference_point, &
1556	do2d_xy_time_count(av), av
1557	ENDIF
1558	#endif
1559	DO i = 0, io_blocks-1
1560	IF ( i == io_group ) THEN
1561	WRITE ( 21 ) nxl, nxr, nys, nyn, nys, nyn
1562	WRITE ( 21 ) local_2d
1563	ENDIF
1564	#if defined( __parallel )
1565	CALL MPI_BARRIER( comm2d, ierr )
1566	#endif
1567	ENDDO
1568
1569	ELSE
1570	!
1571	!-- PE0 receives partial arrays from all processors and
1572	!-- then outputs them. Here a barrier has to be set,
1573	!-- because otherwise "-MPI- FATAL: Remote protocol queue
1574	!-- full" may occur.
1575	CALL MPI_BARRIER( comm2d, ierr )
1576
1577	ngp = ( nxr-nxl+1 ) * ( nyn-nys+1 )
1578	IF ( myid == 0 ) THEN
1579	!
1580	!-- Local array can be relocated directly.
1581	total_2d(nxl:nxr,nys:nyn) = local_2d
1582	!
1583	!-- Receive data from all other PEs.
1584	DO n = 1, numprocs-1
1585	!
1586	!-- Receive index limits first, then array.
1587	!-- Index limits are received in arbitrary order from
1588	!-- the PEs.
1589	CALL MPI_RECV( ind(1), 4, MPI_INTEGER, &
1590	MPI_ANY_SOURCE, 0, comm2d, &
1591	status, ierr )
1592	sender = status(MPI_SOURCE)
1593	DEALLOCATE( local_2d )
1594	ALLOCATE( local_2d(ind(1):ind(2),ind(3):ind(4)) )
1595	CALL MPI_RECV( local_2d(ind(1),ind(3)), ngp, &
1596	MPI_REAL, sender, 1, comm2d, &
1597	status, ierr )
1598	total_2d(ind(1):ind(2),ind(3):ind(4)) = local_2d
1599	ENDDO
1600	!
1601	!-- Relocate the local array for the next loop increment
1602	DEALLOCATE( local_2d )
1603	ALLOCATE( local_2d(nxl:nxr,nys:nyn) )
1604
1605	#if defined( __netcdf )
1606	IF ( two_d ) THEN
1607	nc_stat = NF90_PUT_VAR( id_set_xy(av), &
1608	id_var_do2d(av,if), &
1609	total_2d(0:nx,0:ny), &
1610	start = (/ 1, 1, 1, do2d_xy_time_count(av) /), &
1611	count = (/ nx+1, ny+1, 1, 1 /) )
1612	ELSE
1613	nc_stat = NF90_PUT_VAR( id_set_xy(av), &
1614	id_var_do2d(av,if), &
1615	total_2d(0:nx,0:ny), &
1616	start = (/ 1, 1, is, do2d_xy_time_count(av) /), &
1617	count = (/ nx+1, ny+1, 1, 1 /) )
1618	ENDIF
1619	CALL netcdf_handle_error( 'data_output_2d', 54 )
1620	#endif
1621
1622	ELSE
1623	!
1624	!-- First send the local index limits to PE0
1625	ind(1) = nxl; ind(2) = nxr
1626	ind(3) = nys; ind(4) = nyn
1627	CALL MPI_SEND( ind(1), 4, MPI_INTEGER, 0, 0, &
1628	comm2d, ierr )
1629	!
1630	!-- Send data to PE0
1631	CALL MPI_SEND( local_2d(nxl,nys), ngp, &
1632	MPI_REAL, 0, 1, comm2d, ierr )
1633	ENDIF
1634	!
1635	!-- A barrier has to be set, because otherwise some PEs may
1636	!-- proceed too fast so that PE0 may receive wrong data on
1637	!-- tag 0
1638	CALL MPI_BARRIER( comm2d, ierr )
1639	ENDIF
1640
1641	ENDIF
1642	#else
1643	#if defined( __netcdf )
1644	IF ( two_d ) THEN
1645	nc_stat = NF90_PUT_VAR( id_set_xy(av), &
1646	id_var_do2d(av,if), &
1647	local_2d(nxl:nxr,nys:nyn), &
1648	start = (/ 1, 1, 1, do2d_xy_time_count(av) /), &
1649	count = (/ nx+1, ny+1, 1, 1 /) )
1650	ELSE
1651	nc_stat = NF90_PUT_VAR( id_set_xy(av), &
1652	id_var_do2d(av,if), &
1653	local_2d(nxl:nxr,nys:nyn), &
1654	start = (/ 1, 1, is, do2d_xy_time_count(av) /), &
1655	count = (/ nx+1, ny+1, 1, 1 /) )
1656	ENDIF
1657	CALL netcdf_handle_error( 'data_output_2d', 447 )
1658	#endif
1659	#endif
1660
1661	!
1662	!-- For 2D-arrays (e.g. u*) only one cross-section is available.
1663	!-- Hence exit loop of output levels.
1664	IF ( two_d ) THEN
1665	IF ( netcdf_data_format < 5 ) two_d = .FALSE.
1666	EXIT loop1
1667	ENDIF
1668
1669	CASE ( 'xz' )
1670	!
1671	!-- Update the netCDF xz cross section time axis.
1672	!-- In case of parallel output, this is only done by PE0
1673	!-- to increase the performance.
1674	IF ( simulated_time /= do2d_xz_last_time(av) ) THEN
1675	do2d_xz_time_count(av) = do2d_xz_time_count(av) + 1
1676	do2d_xz_last_time(av) = simulated_time
1677	IF ( myid == 0 ) THEN
1678	IF ( .NOT. data_output_2d_on_each_pe &
1679	.OR. netcdf_data_format > 4 ) &
1680	THEN
1681	#if defined( __netcdf )
1682	nc_stat = NF90_PUT_VAR( id_set_xz(av), &
1683	id_var_time_xz(av), &
1684	(/ time_since_reference_point /), &
1685	start = (/ do2d_xz_time_count(av) /), &
1686	count = (/ 1 /) )
1687	CALL netcdf_handle_error( 'data_output_2d', 56 )
1688	#endif
1689	ENDIF
1690	ENDIF
1691	ENDIF
1692
1693	!
1694	!-- If required, carry out averaging along y
1695	IF ( section(is,s_ind) == -1 ) THEN
1696
1697	ALLOCATE( local_2d_l(nxl:nxr,nzb_do:nzt_do) )
1698	local_2d_l = 0.0_wp
1699	ngp = ( nxr-nxl + 1 ) * ( nzt_do-nzb_do + 1 )
1700	!
1701	!-- First local averaging on the PE
1702	DO k = nzb_do, nzt_do
1703	DO j = nys, nyn
1704	DO i = nxl, nxr
1705	local_2d_l(i,k) = local_2d_l(i,k) + &
1706	local_pf(i,j,k)
1707	ENDDO
1708	ENDDO
1709	ENDDO
1710	#if defined( __parallel )
1711	!
1712	!-- Now do the averaging over all PEs along y
1713	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
1714	CALL MPI_ALLREDUCE( local_2d_l(nxl,nzb_do), &
1715	local_2d(nxl,nzb_do), ngp, MPI_REAL, &
1716	MPI_SUM, comm1dy, ierr )
1717	#else
1718	local_2d = local_2d_l
1719	#endif
1720	local_2d = local_2d / ( ny + 1.0_wp )
1721
1722	DEALLOCATE( local_2d_l )
1723
1724	ELSE
1725	!
1726	!-- Just store the respective section on the local array
1727	!-- (but only if it is available on this PE!)
1728	IF ( section(is,s_ind) >= nys .AND. section(is,s_ind) <= nyn ) &
1729	THEN
1730	local_2d = local_pf(:,section(is,s_ind),nzb_do:nzt_do)
1731	ENDIF
1732
1733	ENDIF
1734
1735	#if defined( __parallel )
1736	IF ( netcdf_data_format > 4 ) THEN
1737	!
1738	!-- Output in netCDF4/HDF5 format.
1739	!-- Output only on those PEs where the respective cross
1740	!-- sections reside. Cross sections averaged along y are
1741	!-- output on the respective first PE along y (myidy=0).
1742	IF ( ( section(is,s_ind) >= nys .AND. &
1743	section(is,s_ind) <= nyn ) .OR. &
1744	( section(is,s_ind) == -1 .AND. myidy == 0 ) ) THEN
1745	#if defined( __netcdf )
1746	!
1747	!-- For parallel output, all cross sections are first
1748	!-- stored here on a local array and will be written to the
1749	!-- output file afterwards to increase the performance.
1750	DO i = nxl, nxr
1751	DO k = nzb_do, nzt_do
1752	local_2d_sections_l(i,is,k) = local_2d(i,k)
1753	ENDDO
1754	ENDDO
1755	#endif
1756	ENDIF
1757
1758	ELSE
1759
1760	IF ( data_output_2d_on_each_pe ) THEN
1761	!
1762	!-- Output of partial arrays on each PE. If the cross
1763	!-- section does not reside on the PE, output special
1764	!-- index values.
1765	#if defined( __netcdf )
1766	IF ( myid == 0 ) THEN
1767	WRITE ( 22 ) time_since_reference_point, &
1768	do2d_xz_time_count(av), av
1769	ENDIF
1770	#endif
1771	DO i = 0, io_blocks-1
1772	IF ( i == io_group ) THEN
1773	IF ( ( section(is,s_ind) >= nys .AND. &
1774	section(is,s_ind) <= nyn ) .OR. &
1775	( section(is,s_ind) == -1 .AND. &
1776	nys-1 == -1 ) ) &
1777	THEN
1778	WRITE (22) nxl, nxr, nzb_do, nzt_do, nzb, nzt+1
1779	WRITE (22) local_2d
1780	ELSE
1781	WRITE (22) -1, -1, -1, -1, -1, -1
1782	ENDIF
1783	ENDIF
1784	#if defined( __parallel )
1785	CALL MPI_BARRIER( comm2d, ierr )
1786	#endif
1787	ENDDO
1788
1789	ELSE
1790	!
1791	!-- PE0 receives partial arrays from all processors of the
1792	!-- respective cross section and outputs them. Here a
1793	!-- barrier has to be set, because otherwise
1794	!-- "-MPI- FATAL: Remote protocol queue full" may occur.
1795	CALL MPI_BARRIER( comm2d, ierr )
1796
1797	ngp = ( nxr-nxl + 1 ) * ( nzt_do-nzb_do + 1 )
1798	IF ( myid == 0 ) THEN
1799	!
1800	!-- Local array can be relocated directly.
1801	IF ( ( section(is,s_ind) >= nys .AND. &
1802	section(is,s_ind) <= nyn ) .OR. &
1803	( section(is,s_ind) == -1 .AND. &
1804	nys-1 == -1 ) ) THEN
1805	total_2d(nxl:nxr,nzb_do:nzt_do) = local_2d
1806	ENDIF
1807	!
1808	!-- Receive data from all other PEs.
1809	DO n = 1, numprocs-1
1810	!
1811	!-- Receive index limits first, then array.
1812	!-- Index limits are received in arbitrary order from
1813	!-- the PEs.
1814	CALL MPI_RECV( ind(1), 4, MPI_INTEGER, &
1815	MPI_ANY_SOURCE, 0, comm2d, &
1816	status, ierr )
1817	!
1818	!-- Not all PEs have data for XZ-cross-section.
1819	IF ( ind(1) /= -9999 ) THEN
1820	sender = status(MPI_SOURCE)
1821	DEALLOCATE( local_2d )
1822	ALLOCATE( local_2d(ind(1):ind(2), &
1823	ind(3):ind(4)) )
1824	CALL MPI_RECV( local_2d(ind(1),ind(3)), ngp, &
1825	MPI_REAL, sender, 1, comm2d, &
1826	status, ierr )
1827	total_2d(ind(1):ind(2),ind(3):ind(4)) = &
1828	local_2d
1829	ENDIF
1830	ENDDO
1831	!
1832	!-- Relocate the local array for the next loop increment
1833	DEALLOCATE( local_2d )
1834	ALLOCATE( local_2d(nxl:nxr,nzb_do:nzt_do) )
1835
1836	#if defined( __netcdf )
1837	nc_stat = NF90_PUT_VAR( id_set_xz(av), &
1838	id_var_do2d(av,if), &
1839	total_2d(0:nx,nzb_do:nzt_do), &
1840	start = (/ 1, is, 1, do2d_xz_time_count(av) /), &
1841	count = (/ nx+1, 1, nzt_do-nzb_do+1, 1 /) )
1842	CALL netcdf_handle_error( 'data_output_2d', 58 )
1843	#endif
1844
1845	ELSE
1846	!
1847	!-- If the cross section resides on the PE, send the
1848	!-- local index limits, otherwise send -9999 to PE0.
1849	IF ( ( section(is,s_ind) >= nys .AND. &
1850	section(is,s_ind) <= nyn ) .OR. &
1851	( section(is,s_ind) == -1 .AND. nys-1 == -1 ) ) &
1852	THEN
1853	ind(1) = nxl; ind(2) = nxr
1854	ind(3) = nzb_do; ind(4) = nzt_do
1855	ELSE
1856	ind(1) = -9999; ind(2) = -9999
1857	ind(3) = -9999; ind(4) = -9999
1858	ENDIF
1859	CALL MPI_SEND( ind(1), 4, MPI_INTEGER, 0, 0, &
1860	comm2d, ierr )
1861	!
1862	!-- If applicable, send data to PE0.
1863	IF ( ind(1) /= -9999 ) THEN
1864	CALL MPI_SEND( local_2d(nxl,nzb_do), ngp, &
1865	MPI_REAL, 0, 1, comm2d, ierr )
1866	ENDIF
1867	ENDIF
1868	!
1869	!-- A barrier has to be set, because otherwise some PEs may
1870	!-- proceed too fast so that PE0 may receive wrong data on
1871	!-- tag 0
1872	CALL MPI_BARRIER( comm2d, ierr )
1873	ENDIF
1874
1875	ENDIF
1876	#else
1877	#if defined( __netcdf )
1878	nc_stat = NF90_PUT_VAR( id_set_xz(av), &
1879	id_var_do2d(av,if), &
1880	local_2d(nxl:nxr,nzb_do:nzt_do), &
1881	start = (/ 1, is, 1, do2d_xz_time_count(av) /), &
1882	count = (/ nx+1, 1, nzt_do-nzb_do+1, 1 /) )
1883	CALL netcdf_handle_error( 'data_output_2d', 451 )
1884	#endif
1885	#endif
1886
1887	CASE ( 'yz' )
1888	!
1889	!-- Update the netCDF yz cross section time axis.
1890	!-- In case of parallel output, this is only done by PE0
1891	!-- to increase the performance.
1892	IF ( simulated_time /= do2d_yz_last_time(av) ) THEN
1893	do2d_yz_time_count(av) = do2d_yz_time_count(av) + 1
1894	do2d_yz_last_time(av) = simulated_time
1895	IF ( myid == 0 ) THEN
1896	IF ( .NOT. data_output_2d_on_each_pe &
1897	.OR. netcdf_data_format > 4 ) &
1898	THEN
1899	#if defined( __netcdf )
1900	nc_stat = NF90_PUT_VAR( id_set_yz(av), &
1901	id_var_time_yz(av), &
1902	(/ time_since_reference_point /), &
1903	start = (/ do2d_yz_time_count(av) /), &
1904	count = (/ 1 /) )
1905	CALL netcdf_handle_error( 'data_output_2d', 59 )
1906	#endif
1907	ENDIF
1908	ENDIF
1909	ENDIF
1910
1911	!
1912	!-- If required, carry out averaging along x
1913	IF ( section(is,s_ind) == -1 ) THEN
1914
1915	ALLOCATE( local_2d_l(nys:nyn,nzb_do:nzt_do) )
1916	local_2d_l = 0.0_wp
1917	ngp = ( nyn-nys+1 ) * ( nzt_do-nzb_do+1 )
1918	!
1919	!-- First local averaging on the PE
1920	DO k = nzb_do, nzt_do
1921	DO j = nys, nyn
1922	DO i = nxl, nxr
1923	local_2d_l(j,k) = local_2d_l(j,k) + &
1924	local_pf(i,j,k)
1925	ENDDO
1926	ENDDO
1927	ENDDO
1928	#if defined( __parallel )
1929	!
1930	!-- Now do the averaging over all PEs along x
1931	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
1932	CALL MPI_ALLREDUCE( local_2d_l(nys,nzb_do), &
1933	local_2d(nys,nzb_do), ngp, MPI_REAL, &
1934	MPI_SUM, comm1dx, ierr )
1935	#else
1936	local_2d = local_2d_l
1937	#endif
1938	local_2d = local_2d / ( nx + 1.0_wp )
1939
1940	DEALLOCATE( local_2d_l )
1941
1942	ELSE
1943	!
1944	!-- Just store the respective section on the local array
1945	!-- (but only if it is available on this PE!)
1946	IF ( section(is,s_ind) >= nxl .AND. section(is,s_ind) <= nxr ) &
1947	THEN
1948	local_2d = local_pf(section(is,s_ind),:,nzb_do:nzt_do)
1949	ENDIF
1950
1951	ENDIF
1952
1953	#if defined( __parallel )
1954	IF ( netcdf_data_format > 4 ) THEN
1955	!
1956	!-- Output in netCDF4/HDF5 format.
1957	!-- Output only on those PEs where the respective cross
1958	!-- sections reside. Cross sections averaged along x are
1959	!-- output on the respective first PE along x (myidx=0).
1960	IF ( ( section(is,s_ind) >= nxl .AND. &
1961	section(is,s_ind) <= nxr ) .OR. &
1962	( section(is,s_ind) == -1 .AND. myidx == 0 ) ) THEN
1963	#if defined( __netcdf )
1964	!
1965	!-- For parallel output, all cross sections are first
1966	!-- stored here on a local array and will be written to the
1967	!-- output file afterwards to increase the performance.
1968	DO j = nys, nyn
1969	DO k = nzb_do, nzt_do
1970	local_2d_sections_l(is,j,k) = local_2d(j,k)
1971	ENDDO
1972	ENDDO
1973	#endif
1974	ENDIF
1975
1976	ELSE
1977
1978	IF ( data_output_2d_on_each_pe ) THEN
1979	!
1980	!-- Output of partial arrays on each PE. If the cross
1981	!-- section does not reside on the PE, output special
1982	!-- index values.
1983	#if defined( __netcdf )
1984	IF ( myid == 0 ) THEN
1985	WRITE ( 23 ) time_since_reference_point, &
1986	do2d_yz_time_count(av), av
1987	ENDIF
1988	#endif
1989	DO i = 0, io_blocks-1
1990	IF ( i == io_group ) THEN
1991	IF ( ( section(is,s_ind) >= nxl .AND. &
1992	section(is,s_ind) <= nxr ) .OR. &
1993	( section(is,s_ind) == -1 .AND. &
1994	nxl-1 == -1 ) ) &
1995	THEN
1996	WRITE (23) nys, nyn, nzb_do, nzt_do, nzb, nzt+1
1997	WRITE (23) local_2d
1998	ELSE
1999	WRITE (23) -1, -1, -1, -1, -1, -1
2000	ENDIF
2001	ENDIF
2002	#if defined( __parallel )
2003	CALL MPI_BARRIER( comm2d, ierr )
2004	#endif
2005	ENDDO
2006
2007	ELSE
2008	!
2009	!-- PE0 receives partial arrays from all processors of the
2010	!-- respective cross section and outputs them. Here a
2011	!-- barrier has to be set, because otherwise
2012	!-- "-MPI- FATAL: Remote protocol queue full" may occur.
2013	CALL MPI_BARRIER( comm2d, ierr )
2014
2015	ngp = ( nyn-nys+1 ) * ( nzt_do-nzb_do+1 )
2016	IF ( myid == 0 ) THEN
2017	!
2018	!-- Local array can be relocated directly.
2019	IF ( ( section(is,s_ind) >= nxl .AND. &
2020	section(is,s_ind) <= nxr ) .OR. &
2021	( section(is,s_ind) == -1 .AND. nxl-1 == -1 ) ) &
2022	THEN
2023	total_2d(nys:nyn,nzb_do:nzt_do) = local_2d
2024	ENDIF
2025	!
2026	!-- Receive data from all other PEs.
2027	DO n = 1, numprocs-1
2028	!
2029	!-- Receive index limits first, then array.
2030	!-- Index limits are received in arbitrary order from
2031	!-- the PEs.
2032	CALL MPI_RECV( ind(1), 4, MPI_INTEGER, &
2033	MPI_ANY_SOURCE, 0, comm2d, &
2034	status, ierr )
2035	!
2036	!-- Not all PEs have data for YZ-cross-section.
2037	IF ( ind(1) /= -9999 ) THEN
2038	sender = status(MPI_SOURCE)
2039	DEALLOCATE( local_2d )
2040	ALLOCATE( local_2d(ind(1):ind(2), &
2041	ind(3):ind(4)) )
2042	CALL MPI_RECV( local_2d(ind(1),ind(3)), ngp, &
2043	MPI_REAL, sender, 1, comm2d, &
2044	status, ierr )
2045	total_2d(ind(1):ind(2),ind(3):ind(4)) = &
2046	local_2d
2047	ENDIF
2048	ENDDO
2049	!
2050	!-- Relocate the local array for the next loop increment
2051	DEALLOCATE( local_2d )
2052	ALLOCATE( local_2d(nys:nyn,nzb_do:nzt_do) )
2053
2054	#if defined( __netcdf )
2055	nc_stat = NF90_PUT_VAR( id_set_yz(av), &
2056	id_var_do2d(av,if), &
2057	total_2d(0:ny,nzb_do:nzt_do), &
2058	start = (/ is, 1, 1, do2d_yz_time_count(av) /), &
2059	count = (/ 1, ny+1, nzt_do-nzb_do+1, 1 /) )
2060	CALL netcdf_handle_error( 'data_output_2d', 61 )
2061	#endif
2062
2063	ELSE
2064	!
2065	!-- If the cross section resides on the PE, send the
2066	!-- local index limits, otherwise send -9999 to PE0.
2067	IF ( ( section(is,s_ind) >= nxl .AND. &
2068	section(is,s_ind) <= nxr ) .OR. &
2069	( section(is,s_ind) == -1 .AND. nxl-1 == -1 ) ) &
2070	THEN
2071	ind(1) = nys; ind(2) = nyn
2072	ind(3) = nzb_do; ind(4) = nzt_do
2073	ELSE
2074	ind(1) = -9999; ind(2) = -9999
2075	ind(3) = -9999; ind(4) = -9999
2076	ENDIF
2077	CALL MPI_SEND( ind(1), 4, MPI_INTEGER, 0, 0, &
2078	comm2d, ierr )
2079	!
2080	!-- If applicable, send data to PE0.
2081	IF ( ind(1) /= -9999 ) THEN
2082	CALL MPI_SEND( local_2d(nys,nzb_do), ngp, &
2083	MPI_REAL, 0, 1, comm2d, ierr )
2084	ENDIF
2085	ENDIF
2086	!
2087	!-- A barrier has to be set, because otherwise some PEs may
2088	!-- proceed too fast so that PE0 may receive wrong data on
2089	!-- tag 0
2090	CALL MPI_BARRIER( comm2d, ierr )
2091	ENDIF
2092
2093	ENDIF
2094	#else
2095	#if defined( __netcdf )
2096	nc_stat = NF90_PUT_VAR( id_set_yz(av), &
2097	id_var_do2d(av,if), &
2098	local_2d(nys:nyn,nzb_do:nzt_do), &
2099	start = (/ is, 1, 1, do2d_xz_time_count(av) /), &
2100	count = (/ 1, ny+1, nzt_do-nzb_do+1, 1 /) )
2101	CALL netcdf_handle_error( 'data_output_2d', 452 )
2102	#endif
2103	#endif
2104
2105	END SELECT
2106
2107	is = is + 1
2108	ENDDO loop1
2109
2110	!
2111	!-- For parallel output, all data were collected before on a local array
2112	!-- and are written now to the netcdf file. This must be done to increase
2113	!-- the performance of the parallel output.
2114	#if defined( __netcdf )
2115	IF ( netcdf_data_format > 4 ) THEN
2116
2117	SELECT CASE ( mode )
2118
2119	CASE ( 'xy' )
2120	IF ( two_d ) THEN
2121	nis = 1
2122	two_d = .FALSE.
2123	ELSE
2124	nis = ns
2125	ENDIF
2126	!
2127	!-- Do not output redundant ghost point data except for the
2128	!-- boundaries of the total domain.
2129	! IF ( nxr == nx .AND. nyn /= ny ) THEN
2130	! nc_stat = NF90_PUT_VAR( id_set_xy(av), &
2131	! id_var_do2d(av,if), &
2132	! local_2d_sections(nxl:nxr+1, &
2133	! nys:nyn,1:nis), &
2134	! start = (/ nxl+1, nys+1, 1, &
2135	! do2d_xy_time_count(av) /), &
2136	! count = (/ nxr-nxl+2, &
2137	! nyn-nys+1, nis, 1 &
2138	! /) )
2139	! ELSEIF ( nxr /= nx .AND. nyn == ny ) THEN
2140	! nc_stat = NF90_PUT_VAR( id_set_xy(av), &
2141	! id_var_do2d(av,if), &
2142	! local_2d_sections(nxl:nxr, &
2143	! nys:nyn+1,1:nis), &
2144	! start = (/ nxl+1, nys+1, 1, &
2145	! do2d_xy_time_count(av) /), &
2146	! count = (/ nxr-nxl+1, &
2147	! nyn-nys+2, nis, 1 &
2148	! /) )
2149	! ELSEIF ( nxr == nx .AND. nyn == ny ) THEN
2150	! nc_stat = NF90_PUT_VAR( id_set_xy(av), &
2151	! id_var_do2d(av,if), &
2152	! local_2d_sections(nxl:nxr+1, &
2153	! nys:nyn+1,1:nis), &
2154	! start = (/ nxl+1, nys+1, 1, &
2155	! do2d_xy_time_count(av) /), &
2156	! count = (/ nxr-nxl+2, &
2157	! nyn-nys+2, nis, 1 &
2158	! /) )
2159	! ELSE
2160	nc_stat = NF90_PUT_VAR( id_set_xy(av), &
2161	id_var_do2d(av,if), &
2162	local_2d_sections(nxl:nxr, &
2163	nys:nyn,1:nis), &
2164	start = (/ nxl+1, nys+1, 1, &
2165	do2d_xy_time_count(av) /), &
2166	count = (/ nxr-nxl+1, &
2167	nyn-nys+1, nis, 1 &
2168	/) )
2169	! ENDIF
2170
2171	CALL netcdf_handle_error( 'data_output_2d', 55 )
2172
2173	CASE ( 'xz' )
2174	!
2175	!-- First, all PEs get the information of all cross-sections.
2176	!-- Then the data are written to the output file by all PEs
2177	!-- while NF90_COLLECTIVE is set in subroutine
2178	!-- define_netcdf_header. Although redundant information are
2179	!-- written to the output file in that case, the performance
2180	!-- is significantly better compared to the case where only
2181	!-- the first row of PEs in x-direction (myidx = 0) is given
2182	!-- the output while NF90_INDEPENDENT is set.
2183	IF ( npey /= 1 ) THEN
2184
2185	#if defined( __parallel )
2186	!
2187	!-- Distribute data over all PEs along y
2188	ngp = ( nxr-nxl+1 ) * ( nzt_do-nzb_do+1 ) * ns
2189	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
2190	CALL MPI_ALLREDUCE( local_2d_sections_l(nxl,1,nzb_do), &
2191	local_2d_sections(nxl,1,nzb_do), &
2192	ngp, MPI_REAL, MPI_SUM, comm1dy, &
2193	ierr )
2194	#else
2195	local_2d_sections = local_2d_sections_l
2196	#endif
2197	ENDIF
2198	!
2199	!-- Do not output redundant ghost point data except for the
2200	!-- boundaries of the total domain.
2201	! IF ( nxr == nx ) THEN
2202	! nc_stat = NF90_PUT_VAR( id_set_xz(av), &
2203	! id_var_do2d(av,if), &
2204	! local_2d_sections(nxl:nxr+1,1:ns, &
2205	! nzb_do:nzt_do), &
2206	! start = (/ nxl+1, 1, 1, &
2207	! do2d_xz_time_count(av) /), &
2208	! count = (/ nxr-nxl+2, ns, nzt_do-nzb_do+1, &
2209	! 1 /) )
2210	! ELSE
2211	nc_stat = NF90_PUT_VAR( id_set_xz(av), &
2212	id_var_do2d(av,if), &
2213	local_2d_sections(nxl:nxr,1:ns, &
2214	nzb_do:nzt_do), &
2215	start = (/ nxl+1, 1, 1, &
2216	do2d_xz_time_count(av) /), &
2217	count = (/ nxr-nxl+1, ns, nzt_do-nzb_do+1, &
2218	1 /) )
2219	! ENDIF
2220
2221	CALL netcdf_handle_error( 'data_output_2d', 57 )
2222
2223	CASE ( 'yz' )
2224	!
2225	!-- First, all PEs get the information of all cross-sections.
2226	!-- Then the data are written to the output file by all PEs
2227	!-- while NF90_COLLECTIVE is set in subroutine
2228	!-- define_netcdf_header. Although redundant information are
2229	!-- written to the output file in that case, the performance
2230	!-- is significantly better compared to the case where only
2231	!-- the first row of PEs in y-direction (myidy = 0) is given
2232	!-- the output while NF90_INDEPENDENT is set.
2233	IF ( npex /= 1 ) THEN
2234
2235	#if defined( __parallel )
2236	!
2237	!-- Distribute data over all PEs along x
2238	ngp = ( nyn-nys+1 ) * ( nzt-nzb + 2 ) * ns
2239	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
2240	CALL MPI_ALLREDUCE( local_2d_sections_l(1,nys,nzb_do), &
2241	local_2d_sections(1,nys,nzb_do), &
2242	ngp, MPI_REAL, MPI_SUM, comm1dx, &
2243	ierr )
2244	#else
2245	local_2d_sections = local_2d_sections_l
2246	#endif
2247	ENDIF
2248	!
2249	!-- Do not output redundant ghost point data except for the
2250	!-- boundaries of the total domain.
2251	! IF ( nyn == ny ) THEN
2252	! nc_stat = NF90_PUT_VAR( id_set_yz(av), &
2253	! id_var_do2d(av,if), &
2254	! local_2d_sections(1:ns, &
2255	! nys:nyn+1,nzb_do:nzt_do), &
2256	! start = (/ 1, nys+1, 1, &
2257	! do2d_yz_time_count(av) /), &
2258	! count = (/ ns, nyn-nys+2, &
2259	! nzt_do-nzb_do+1, 1 /) )
2260	! ELSE
2261	nc_stat = NF90_PUT_VAR( id_set_yz(av), &
2262	id_var_do2d(av,if), &
2263	local_2d_sections(1:ns,nys:nyn, &
2264	nzb_do:nzt_do), &
2265	start = (/ 1, nys+1, 1, &
2266	do2d_yz_time_count(av) /), &
2267	count = (/ ns, nyn-nys+1, &
2268	nzt_do-nzb_do+1, 1 /) )
2269	! ENDIF
2270
2271	CALL netcdf_handle_error( 'data_output_2d', 60 )
2272
2273	CASE DEFAULT
2274	message_string = 'unknown cross-section: ' // TRIM( mode )
2275	CALL message( 'data_output_2d', 'PA0180', 1, 2, 0, 6, 0 )
2276
2277	END SELECT
2278
2279	ENDIF
2280	#endif
2281	ENDIF
2282
2283	if = if + 1
2284	l = MAX( 2, LEN_TRIM( do2d(av,if) ) )
2285	do2d_mode = do2d(av,if)(l-1:l)
2286
2287	ENDDO
2288
2289	!
2290	!-- Deallocate temporary arrays.
2291	IF ( ALLOCATED( level_z ) ) DEALLOCATE( level_z )
2292	IF ( netcdf_data_format > 4 ) THEN
2293	DEALLOCATE( local_pf, local_2d, local_2d_sections )
2294	IF( mode == 'xz' .OR. mode == 'yz' ) DEALLOCATE( local_2d_sections_l )
2295	ENDIF
2296	#if defined( __parallel )
2297	IF ( .NOT. data_output_2d_on_each_pe .AND. myid == 0 ) THEN
2298	DEALLOCATE( total_2d )
2299	ENDIF
2300	#endif
2301
2302	!
2303	!-- Close plot output file.
2304	file_id = 20 + s_ind
2305
2306	IF ( data_output_2d_on_each_pe ) THEN
2307	DO i = 0, io_blocks-1
2308	IF ( i == io_group ) THEN
2309	CALL close_file( file_id )
2310	ENDIF
2311	#if defined( __parallel )
2312	CALL MPI_BARRIER( comm2d, ierr )
2313	#endif
2314	ENDDO
2315	ELSE
2316	IF ( myid == 0 ) CALL close_file( file_id )
2317	ENDIF
2318
2319	CALL cpu_log( log_point(3), 'data_output_2d', 'stop' )
2320
2321	END SUBROUTINE data_output_2d

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

Download in other formats:

| Impressum | ©Leibniz Universität Hannover |