Home

Context Navigation

source: palm/trunk/SOURCE/data_output_2d.f90 @ 4001

Last change on this file since 4001 was 3994, checked in by suehring, 6 years ago
new module for diagnostic output quantities added + output of turbulence intensity
Property svn:keywords set to `Id`
File size: 97.5 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 1997-2019 Leibniz Universitaet Hannover
18	!------------------------------------------------------------------------------!
19	!
20	! Current revisions:
21	! ------------------
22	!
23	!
24	! Former revisions:
25	! -----------------
26	! $Id: data_output_2d.f90 3994 2019-05-22 18:08:09Z Giersch $
27	! output of turbulence intensity added
28	!
29	! 3987 2019-05-22 09:52:13Z kanani
30	! Introduce alternative switch for debug output during timestepping
31	!
32	! 3943 2019-05-02 09:50:41Z maronga
33	! Added output of qsws for green roofs.
34	!
35	! 3885 2019-04-11 11:29:34Z kanani
36	! Changes related to global restructuring of location messages and introduction
37	! of additional debug messages
38	!
39	! 3766 2019-02-26 16:23:41Z raasch
40	! unused variables removed
41	!
42	! 3655 2019-01-07 16:51:22Z knoop
43	! Bugfix: use time_since_reference_point instead of simulated_time (relevant
44	! when using wall/soil spinup)
45	!
46	! 3637 2018-12-20 01:51:36Z knoop
47	! Implementation of the PALM module interface
48	!
49	! 3597 2018-12-04 08:40:18Z maronga
50	! Added theta_2m
51	!
52	! 3589 2018-11-30 15:09:51Z suehring
53	! Move the control parameter "salsa" from salsa_mod to control_parameters
54	! (M. Kurppa)
55	!
56	! 3582 2018-11-29 19:16:36Z suehring
57	! Remove fill_value from bio_data_output_2d call
58	! dom_dwd_user, Schrempf:
59	! Clean up of biometeorology calls,
60	! remove uv exposure model code, this is now part of biometeorology_mod.
61	!
62	! 3554 2018-11-22 11:24:52Z gronemeier
63	! - add variable description
64	! - rename variable 'if' into 'ivar'
65	! - removed namelist LOCAL
66	! - removed variable rtext
67	!
68	! 3525 2018-11-14 16:06:14Z kanani
69	! Changes related to clean-up of biometeorology (dom_dwd_user)
70	!
71	! 3467 2018-10-30 19:05:21Z suehring
72	! Implementation of a new aerosol module salsa.
73	!
74	! 3448 2018-10-29 18:14:31Z kanani
75	! Add biometeorology
76	!
77	! 3421 2018-10-24 18:39:32Z gronemeier
78	! Renamed output variables
79	!
80	! 3419 2018-10-24 17:27:31Z gronemeier
81	! minor formatting (kanani)
82	! chem_data_output_2d subroutine added (basit)
83	!
84	! 3294 2018-10-01 02:37:10Z raasch
85	! changes concerning modularization of ocean option
86	!
87	! 3274 2018-09-24 15:42:55Z knoop
88	! Modularization of all bulk cloud physics code components
89	!
90	! 3241 2018-09-12 15:02:00Z raasch
91	! unused variables removed
92	!
93	! 3176 2018-07-26 17:12:48Z suehring
94	! Remove output of latent heat flux at urban-surfaces and set fill values
95	! instead
96	!
97	! 3052 2018-05-31 06:11:20Z raasch
98	! Do not open FORTRAN binary files in case of parallel netCDF I/O
99	!
100	! 3049 2018-05-29 13:52:36Z Giersch
101	! Error messages revised
102	!
103	! 3045 2018-05-28 07:55:41Z Giersch
104	! Error messages revised
105	!
106	! 3014 2018-05-09 08:42:38Z maronga
107	! Added nzb_do and nzt_do for some modules for 2d output
108	!
109	! 3004 2018-04-27 12:33:25Z Giersch
110	! precipitation_rate removed, case prr*_xy removed, to_be_resorted have to point
111	! to ql_vp_av and not to ql_vp, allocation checks implemented (averaged data
112	! will be assigned to fill values if no allocation happened so far)
113	!
114	! 2963 2018-04-12 14:47:44Z suehring
115	! Introduce index for vegetation/wall, pavement/green-wall and water/window
116	! surfaces, for clearer access of surface fraction, albedo, emissivity, etc. .
117	!
118	! 2817 2018-02-19 16:32:21Z knoop
119	! Preliminary gust module interface implemented
120	!
121	! 2805 2018-02-14 17:00:09Z suehring
122	! Consider also default-type surfaces for surface temperature output.
123	!
124	! 2797 2018-02-08 13:24:35Z suehring
125	! Enable output of ground-heat flux also at urban surfaces.
126	!
127	! 2743 2018-01-12 16:03:39Z suehring
128	! In case of natural- and urban-type surfaces output surfaces fluxes in W/m2.
129	!
130	! 2742 2018-01-12 14:59:47Z suehring
131	! Enable output of surface temperature
132	!
133	! 2735 2018-01-11 12:01:27Z suehring
134	! output of r_a moved from land-surface to consider also urban-type surfaces
135	!
136	! 2718 2018-01-02 08:49:38Z maronga
137	! Corrected "Former revisions" section
138	!
139	! 2696 2017-12-14 17:12:51Z kanani
140	! Change in file header (GPL part)
141	! Implementation of uv exposure model (FK)
142	! Implementation of turbulence_closure_mod (TG)
143	! Set fill values at topography grid points or e.g. non-natural-type surface
144	! in case of LSM output (MS)
145	!
146	! 2512 2017-10-04 08:26:59Z raasch
147	! upper bounds of cross section output changed from nx+1,ny+1 to nx,ny
148	! no output of ghost layer data
149	!
150	! 2292 2017-06-20 09:51:42Z schwenkel
151	! Implementation of new microphysic scheme: cloud_scheme = 'morrison'
152	! includes two more prognostic equations for cloud drop concentration (nc)
153	! and cloud water content (qc).
154	!
155	! 2277 2017-06-12 10:47:51Z kanani
156	! Removed unused variables do2d_xy_n, do2d_xz_n, do2d_yz_n
157	!
158	! 2233 2017-05-30 18:08:54Z suehring
159	!
160	! 2232 2017-05-30 17:47:52Z suehring
161	! Adjustments to new surface concept
162	!
163	!
164	! 2190 2017-03-21 12:16:43Z raasch
165	! bugfix for r2031: string rho replaced by rho_ocean
166	!
167	! 2031 2016-10-21 15:11:58Z knoop
168	! renamed variable rho to rho_ocean and rho_av to rho_ocean_av
169	!
170	! 2000 2016-08-20 18:09:15Z knoop
171	! Forced header and separation lines into 80 columns
172	!
173	! 1980 2016-07-29 15:51:57Z suehring
174	! Bugfix, in order to steer user-defined output, setting flag found explicitly
175	! to .F.
176	!
177	! 1976 2016-07-27 13:28:04Z maronga
178	! Output of radiation quantities is now done directly in the respective module
179	!
180	! 1972 2016-07-26 07:52:02Z maronga
181	! Output of land surface quantities is now done directly in the respective
182	! module
183	!
184	! 1960 2016-07-12 16:34:24Z suehring
185	! Scalar surface flux added
186	! Rename INTEGER variable s into s_ind, as s is already assigned to scalar
187	!
188	! 1849 2016-04-08 11:33:18Z hoffmann
189	! precipitation_amount, precipitation_rate, prr moved to arrays_3d
190	!
191	! 1822 2016-04-07 07:49:42Z hoffmann
192	! Output of bulk cloud physics simplified.
193	!
194	! 1788 2016-03-10 11:01:04Z maronga
195	! Added output of z0q
196	!
197	! 1783 2016-03-06 18:36:17Z raasch
198	! name change of netcdf routines and module + related changes
199	!
200	! 1745 2016-02-05 13:06:51Z gronemeier
201	! Bugfix: test if time axis limit exceeds moved to point after call of check_open
202	!
203	! 1703 2015-11-02 12:38:44Z raasch
204	! bugfix for output of single (*) xy-sections in case of parallel netcdf I/O
205	!
206	! 1701 2015-11-02 07:43:04Z maronga
207	! Bugfix in output of RRTGM data
208	!
209	! 1691 2015-10-26 16:17:44Z maronga
210	! Added output of Obukhov length (ol) and radiative heating rates for RRTMG.
211	! Formatting corrections.
212	!
213	! 1682 2015-10-07 23:56:08Z knoop
214	! Code annotations made doxygen readable
215	!
216	! 1585 2015-04-30 07:05:52Z maronga
217	! Added support for RRTMG
218	!
219	! 1555 2015-03-04 17:44:27Z maronga
220	! Added output of r_a and r_s
221	!
222	! 1551 2015-03-03 14:18:16Z maronga
223	! Added suppport for land surface model and radiation model output. In the course
224	! of this action, the limits for vertical loops have been changed (from nzb and
225	! nzt+1 to nzb_do and nzt_do, respectively in order to allow soil model output).
226	! Moreover, a new vertical grid zs was introduced.
227	!
228	! 1359 2014-04-11 17:15:14Z hoffmann
229	! New particle structure integrated.
230	!
231	! 1353 2014-04-08 15:21:23Z heinze
232	! REAL constants provided with KIND-attribute
233	!
234	! 1327 2014-03-21 11:00:16Z raasch
235	! parts concerning iso2d output removed,
236	! -netcdf output queries
237	!
238	! 1320 2014-03-20 08:40:49Z raasch
239	! ONLY-attribute added to USE-statements,
240	! kind-parameters added to all INTEGER and REAL declaration statements,
241	! kinds are defined in new module kinds,
242	! revision history before 2012 removed,
243	! comment fields (!:) to be used for variable explanations added to
244	! all variable declaration statements
245	!
246	! 1318 2014-03-17 13:35:16Z raasch
247	! barrier argument removed from cpu_log.
248	! module interfaces removed
249	!
250	! 1311 2014-03-14 12:13:39Z heinze
251	! bugfix: close #if defined( __netcdf )
252	!
253	! 1308 2014-03-13 14:58:42Z fricke
254	! +local_2d_sections, local_2d_sections_l, ns
255	! Check, if the limit of the time dimension is exceeded for parallel output
256	! To increase the performance for parallel output, the following is done:
257	! - Update of time axis is only done by PE0
258	! - Cross sections are first stored on a local array and are written
259	! collectively to the output file by all PEs.
260	!
261	! 1115 2013-03-26 18:16:16Z hoffmann
262	! ql is calculated by calc_liquid_water_content
263	!
264	! 1076 2012-12-05 08:30:18Z hoffmann
265	! Bugfix in output of ql
266	!
267	! 1065 2012-11-22 17:42:36Z hoffmann
268	! Bugfix: Output of cross sections of ql
269	!
270	! 1053 2012-11-13 17:11:03Z hoffmann
271	! +qr, nr, qc and cross sections
272	!
273	! 1036 2012-10-22 13:43:42Z raasch
274	! code put under GPL (PALM 3.9)
275	!
276	! 1031 2012-10-19 14:35:30Z raasch
277	! netCDF4 without parallel file support implemented
278	!
279	! 1007 2012-09-19 14:30:36Z franke
280	! Bugfix: missing calculation of ql_vp added
281	!
282	! 978 2012-08-09 08:28:32Z fricke
283	! +z0h
284	!
285	! Revision 1.1 1997/08/11 06:24:09 raasch
286	! Initial revision
287	!
288	!
289	! Description:
290	! ------------
291	!> Data output of cross-sections in netCDF format or binary format
292	!> to be later converted to NetCDF by helper routine combine_plot_fields.
293	!> Attention: The position of the sectional planes is still not always computed
294	!> --------- correctly. (zu is used always)!
295	!------------------------------------------------------------------------------!
296	SUBROUTINE data_output_2d( mode, av )
297
298
299	USE arrays_3d, &
300	ONLY: dzw, d_exner, e, heatflux_output_conversion, p, pt, q, ql, ql_c, ql_v, s, tend, u, &
301	v, vpt, w, waterflux_output_conversion, zu, zw
302
303	USE averaging
304
305	USE basic_constants_and_equations_mod, &
306	ONLY: c_p, lv_d_cp, l_v
307
308	USE bulk_cloud_model_mod, &
309	ONLY: bulk_cloud_model
310
311	USE control_parameters, &
312	ONLY: data_output_2d_on_each_pe, &
313	data_output_xy, data_output_xz, data_output_yz, &
314	debug_output_timestep, &
315	do2d, &
316	do2d_xy_last_time, do2d_xy_time_count, &
317	do2d_xz_last_time, do2d_xz_time_count, &
318	do2d_yz_last_time, do2d_yz_time_count, &
319	ibc_uv_b, io_blocks, io_group, message_string, &
320	ntdim_2d_xy, ntdim_2d_xz, ntdim_2d_yz, &
321	psolver, section, &
322	time_since_reference_point
323
324	USE cpulog, &
325	ONLY: cpu_log, log_point
326
327	USE diagnostic_output_quantities_mod, &
328	ONLY: ti, ti_av
329
330	USE indices, &
331	ONLY: nbgp, nx, nxl, nxlg, nxr, nxrg, ny, nyn, nyng, nys, nysg, &
332	nzb, nzt, wall_flags_0
333
334	USE kinds
335
336	USE land_surface_model_mod, &
337	ONLY: zs
338
339	USE module_interface, &
340	ONLY: module_interface_data_output_2d
341
342	#if defined( __netcdf )
343	USE NETCDF
344	#endif
345
346	USE netcdf_interface, &
347	ONLY: fill_value, id_set_xy, id_set_xz, id_set_yz, id_var_do2d, &
348	id_var_time_xy, id_var_time_xz, id_var_time_yz, nc_stat, &
349	netcdf_data_format, netcdf_handle_error
350
351	USE particle_attributes, &
352	ONLY: grid_particles, number_of_particles, particle_advection_start, &
353	particles, prt_count
354
355	USE pegrid
356
357	USE surface_mod, &
358	ONLY: ind_pav_green, ind_veg_wall, ind_wat_win, surf_def_h, &
359	surf_lsm_h, surf_usm_h
360
361	USE turbulence_closure_mod, &
362	ONLY: tcm_data_output_2d
363
364
365	IMPLICIT NONE
366
367	CHARACTER (LEN=2) :: do2d_mode !< output mode of variable ('xy', 'xz', 'yz')
368	CHARACTER (LEN=2) :: mode !< mode with which the routine is called ('xy', 'xz', 'yz')
369	CHARACTER (LEN=4) :: grid !< string defining the vertical grid
370
371	INTEGER(iwp) :: av !< flag for (non-)average output
372	INTEGER(iwp) :: ngp !< number of grid points of an output slice
373	INTEGER(iwp) :: file_id !< id of output files
374	INTEGER(iwp) :: flag_nr !< number of masking flag
375	INTEGER(iwp) :: i !< loop index
376	INTEGER(iwp) :: iis !< vertical index of a xy slice in array 'local_2d_sections'
377	INTEGER(iwp) :: is !< slice index
378	INTEGER(iwp) :: ivar !< variable index
379	INTEGER(iwp) :: j !< loop index
380	INTEGER(iwp) :: k !< loop index
381	INTEGER(iwp) :: l !< loop index
382	INTEGER(iwp) :: layer_xy !< vertical index of a xy slice in array 'local_pf'
383	INTEGER(iwp) :: m !< loop index
384	INTEGER(iwp) :: n !< loop index
385	INTEGER(iwp) :: nis !< number of vertical slices to be written via parallel NetCDF output
386	INTEGER(iwp) :: ns !< number of output slices
387	INTEGER(iwp) :: nzb_do !< lower limit of the data field (usually nzb)
388	INTEGER(iwp) :: nzt_do !< upper limit of the data field (usually nzt+1)
389	INTEGER(iwp) :: s_ind !< index of slice types (xy=1, xz=2, yz=3)
390	INTEGER(iwp) :: sender !< PE id of sending PE
391	INTEGER(iwp) :: ind(4) !< index limits (lower/upper bounds) of array 'local_2d'
392
393	LOGICAL :: found !< true if output variable was found
394	LOGICAL :: resorted !< true if variable is resorted
395	LOGICAL :: two_d !< true if variable is only two dimensional
396
397	REAL(wp) :: mean_r !< mean particle radius
398	REAL(wp) :: s_r2 !< sum( particle-radius**2 )
399	REAL(wp) :: s_r3 !< sum( particle-radius**3 )
400
401	REAL(wp), DIMENSION(:), ALLOCATABLE :: level_z !< z levels for output array
402	REAL(wp), DIMENSION(:,:), ALLOCATABLE :: local_2d !< local 2-dimensional array containing output values
403	REAL(wp), DIMENSION(:,:), ALLOCATABLE :: local_2d_l !< local 2-dimensional array containing output values
404
405	REAL(wp), DIMENSION(:,:,:), ALLOCATABLE :: local_pf !< output array
406	REAL(wp), DIMENSION(:,:,:), ALLOCATABLE :: local_2d_sections !< local array containing values at all slices
407	REAL(wp), DIMENSION(:,:,:), ALLOCATABLE :: local_2d_sections_l !< local array containing values at all slices
408
409	#if defined( __parallel )
410	REAL(wp), DIMENSION(:,:), ALLOCATABLE :: total_2d !< same as local_2d
411	#endif
412	REAL(wp), DIMENSION(:,:,:), POINTER :: to_be_resorted !< points to array which shall be output
413
414
415	IF ( debug_output_timestep ) CALL debug_message( 'data_output_2d', 'start' )
416	!
417	!-- Immediate return, if no output is requested (no respective sections
418	!-- found in parameter data_output)
419	IF ( mode == 'xy' .AND. .NOT. data_output_xy(av) ) RETURN
420	IF ( mode == 'xz' .AND. .NOT. data_output_xz(av) ) RETURN
421	IF ( mode == 'yz' .AND. .NOT. data_output_yz(av) ) RETURN
422
423	CALL cpu_log (log_point(3),'data_output_2d','start')
424
425	two_d = .FALSE. ! local variable to distinguish between output of pure 2D
426	! arrays and cross-sections of 3D arrays.
427
428	!
429	!-- Depending on the orientation of the cross-section, the respective output
430	!-- files have to be opened.
431	SELECT CASE ( mode )
432
433	CASE ( 'xy' )
434	s_ind = 1
435	ALLOCATE( level_z(nzb:nzt+1), local_2d(nxl:nxr,nys:nyn) )
436
437	IF ( netcdf_data_format > 4 ) THEN
438	ns = 1
439	DO WHILE ( section(ns,s_ind) /= -9999 .AND. ns <= 100 )
440	ns = ns + 1
441	ENDDO
442	ns = ns - 1
443	ALLOCATE( local_2d_sections(nxl:nxr,nys:nyn,1:ns) )
444	local_2d_sections = 0.0_wp
445	ENDIF
446
447	!
448	!-- Parallel netCDF4/HDF5 output is done on all PEs, all other on PE0 only
449	IF ( myid == 0 .OR. netcdf_data_format > 4 ) THEN
450	CALL check_open( 101+av*10 )
451	ENDIF
452	IF ( data_output_2d_on_each_pe .AND. netcdf_data_format < 5 ) THEN
453	CALL check_open( 21 )
454	ELSE
455	IF ( myid == 0 ) THEN
456	#if defined( __parallel )
457	ALLOCATE( total_2d(0:nx,0:ny) )
458	#endif
459	ENDIF
460	ENDIF
461
462	CASE ( 'xz' )
463	s_ind = 2
464	ALLOCATE( local_2d(nxl:nxr,nzb:nzt+1) )
465
466	IF ( netcdf_data_format > 4 ) THEN
467	ns = 1
468	DO WHILE ( section(ns,s_ind) /= -9999 .AND. ns <= 100 )
469	ns = ns + 1
470	ENDDO
471	ns = ns - 1
472	ALLOCATE( local_2d_sections(nxl:nxr,1:ns,nzb:nzt+1) )
473	ALLOCATE( local_2d_sections_l(nxl:nxr,1:ns,nzb:nzt+1) )
474	local_2d_sections = 0.0_wp; local_2d_sections_l = 0.0_wp
475	ENDIF
476
477	!
478	!-- Parallel netCDF4/HDF5 output is done on all PEs, all other on PE0 only
479	IF ( myid == 0 .OR. netcdf_data_format > 4 ) THEN
480	CALL check_open( 102+av*10 )
481	ENDIF
482
483	IF ( data_output_2d_on_each_pe .AND. netcdf_data_format < 5 ) THEN
484	CALL check_open( 22 )
485	ELSE
486	IF ( myid == 0 ) THEN
487	#if defined( __parallel )
488	ALLOCATE( total_2d(0:nx,nzb:nzt+1) )
489	#endif
490	ENDIF
491	ENDIF
492
493	CASE ( 'yz' )
494	s_ind = 3
495	ALLOCATE( local_2d(nys:nyn,nzb:nzt+1) )
496
497	IF ( netcdf_data_format > 4 ) THEN
498	ns = 1
499	DO WHILE ( section(ns,s_ind) /= -9999 .AND. ns <= 100 )
500	ns = ns + 1
501	ENDDO
502	ns = ns - 1
503	ALLOCATE( local_2d_sections(1:ns,nys:nyn,nzb:nzt+1) )
504	ALLOCATE( local_2d_sections_l(1:ns,nys:nyn,nzb:nzt+1) )
505	local_2d_sections = 0.0_wp; local_2d_sections_l = 0.0_wp
506	ENDIF
507
508	!
509	!-- Parallel netCDF4/HDF5 output is done on all PEs, all other on PE0 only
510	IF ( myid == 0 .OR. netcdf_data_format > 4 ) THEN
511	CALL check_open( 103+av*10 )
512	ENDIF
513
514	IF ( data_output_2d_on_each_pe .AND. netcdf_data_format < 5 ) THEN
515	CALL check_open( 23 )
516	ELSE
517	IF ( myid == 0 ) THEN
518	#if defined( __parallel )
519	ALLOCATE( total_2d(0:ny,nzb:nzt+1) )
520	#endif
521	ENDIF
522	ENDIF
523
524	CASE DEFAULT
525	message_string = 'unknown cross-section: ' // TRIM( mode )
526	CALL message( 'data_output_2d', 'PA0180', 1, 2, 0, 6, 0 )
527
528	END SELECT
529
530	!
531	!-- For parallel netcdf output the time axis must be limited. Return, if this
532	!-- limit is exceeded. This could be the case, if the simulated time exceeds
533	!-- the given end time by the length of the given output interval.
534	IF ( netcdf_data_format > 4 ) THEN
535	IF ( mode == 'xy' .AND. do2d_xy_time_count(av) + 1 > &
536	ntdim_2d_xy(av) ) THEN
537	WRITE ( message_string, * ) 'Output of xy cross-sections is not ', &
538	'given at t=', time_since_reference_point, 's because the', &
539	' maximum number of output time levels is exceeded.'
540	CALL message( 'data_output_2d', 'PA0384', 0, 1, 0, 6, 0 )
541	CALL cpu_log( log_point(3), 'data_output_2d', 'stop' )
542	RETURN
543	ENDIF
544	IF ( mode == 'xz' .AND. do2d_xz_time_count(av) + 1 > &
545	ntdim_2d_xz(av) ) THEN
546	WRITE ( message_string, * ) 'Output of xz cross-sections is not ', &
547	'given at t=', time_since_reference_point, 's because the', &
548	' maximum number of output time levels is exceeded.'
549	CALL message( 'data_output_2d', 'PA0385', 0, 1, 0, 6, 0 )
550	CALL cpu_log( log_point(3), 'data_output_2d', 'stop' )
551	RETURN
552	ENDIF
553	IF ( mode == 'yz' .AND. do2d_yz_time_count(av) + 1 > &
554	ntdim_2d_yz(av) ) THEN
555	WRITE ( message_string, * ) 'Output of yz cross-sections is not ', &
556	'given at t=', time_since_reference_point, 's because the', &
557	' maximum number of output time levels is exceeded.'
558	CALL message( 'data_output_2d', 'PA0386', 0, 1, 0, 6, 0 )
559	CALL cpu_log( log_point(3), 'data_output_2d', 'stop' )
560	RETURN
561	ENDIF
562	ENDIF
563
564	!
565	!-- Allocate a temporary array for resorting (kji -> ijk).
566	ALLOCATE( local_pf(nxl:nxr,nys:nyn,nzb:nzt+1) )
567	local_pf = 0.0
568
569	!
570	!-- Loop of all variables to be written.
571	!-- Output dimensions chosen
572	ivar = 1
573	l = MAX( 2, LEN_TRIM( do2d(av,ivar) ) )
574	do2d_mode = do2d(av,ivar)(l-1:l)
575
576	DO WHILE ( do2d(av,ivar)(1:1) /= ' ' )
577
578	IF ( do2d_mode == mode ) THEN
579	!
580	!-- Set flag to steer output of radiation, land-surface, or user-defined
581	!-- quantities
582	found = .FALSE.
583
584	nzb_do = nzb
585	nzt_do = nzt+1
586	!
587	!-- Before each output, set array local_pf to fill value
588	local_pf = fill_value
589	!
590	!-- Set masking flag for topography for not resorted arrays
591	flag_nr = 0
592
593	!
594	!-- Store the array chosen on the temporary array.
595	resorted = .FALSE.
596	SELECT CASE ( TRIM( do2d(av,ivar) ) )
597	CASE ( 'e_xy', 'e_xz', 'e_yz' )
598	IF ( av == 0 ) THEN
599	to_be_resorted => e
600	ELSE
601	IF ( .NOT. ALLOCATED( e_av ) ) THEN
602	ALLOCATE( e_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
603	e_av = REAL( fill_value, KIND = wp )
604	ENDIF
605	to_be_resorted => e_av
606	ENDIF
607	IF ( mode == 'xy' ) level_z = zu
608
609	CASE ( 'thetal_xy', 'thetal_xz', 'thetal_yz' )
610	IF ( av == 0 ) THEN
611	to_be_resorted => pt
612	ELSE
613	IF ( .NOT. ALLOCATED( lpt_av ) ) THEN
614	ALLOCATE( lpt_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
615	lpt_av = REAL( fill_value, KIND = wp )
616	ENDIF
617	to_be_resorted => lpt_av
618	ENDIF
619	IF ( mode == 'xy' ) level_z = zu
620
621	CASE ( 'lwp*_xy' ) ! 2d-array
622	IF ( av == 0 ) THEN
623	DO i = nxl, nxr
624	DO j = nys, nyn
625	local_pf(i,j,nzb+1) = SUM( ql(nzb:nzt,j,i) * &
626	dzw(1:nzt+1) )
627	ENDDO
628	ENDDO
629	ELSE
630	IF ( .NOT. ALLOCATED( lwp_av ) ) THEN
631	ALLOCATE( lwp_av(nysg:nyng,nxlg:nxrg) )
632	lwp_av = REAL( fill_value, KIND = wp )
633	ENDIF
634	DO i = nxl, nxr
635	DO j = nys, nyn
636	local_pf(i,j,nzb+1) = lwp_av(j,i)
637	ENDDO
638	ENDDO
639	ENDIF
640	resorted = .TRUE.
641	two_d = .TRUE.
642	level_z(nzb+1) = zu(nzb+1)
643
644	CASE ( 'ghf*_xy' ) ! 2d-array
645	IF ( av == 0 ) THEN
646	DO m = 1, surf_lsm_h%ns
647	i = surf_lsm_h%i(m)
648	j = surf_lsm_h%j(m)
649	local_pf(i,j,nzb+1) = surf_lsm_h%ghf(m)
650	ENDDO
651	DO m = 1, surf_usm_h%ns
652	i = surf_usm_h%i(m)
653	j = surf_usm_h%j(m)
654	local_pf(i,j,nzb+1) = surf_usm_h%frac(ind_veg_wall,m) * &
655	surf_usm_h%wghf_eb(m) + &
656	surf_usm_h%frac(ind_pav_green,m) * &
657	surf_usm_h%wghf_eb_green(m) + &
658	surf_usm_h%frac(ind_wat_win,m) * &
659	surf_usm_h%wghf_eb_window(m)
660	ENDDO
661	ELSE
662	IF ( .NOT. ALLOCATED( ghf_av ) ) THEN
663	ALLOCATE( ghf_av(nysg:nyng,nxlg:nxrg) )
664	ghf_av = REAL( fill_value, KIND = wp )
665	ENDIF
666	DO i = nxl, nxr
667	DO j = nys, nyn
668	local_pf(i,j,nzb+1) = ghf_av(j,i)
669	ENDDO
670	ENDDO
671	ENDIF
672
673	resorted = .TRUE.
674	two_d = .TRUE.
675	level_z(nzb+1) = zu(nzb+1)
676
677	CASE ( 'ol*_xy' ) ! 2d-array
678	IF ( av == 0 ) THEN
679	DO m = 1, surf_def_h(0)%ns
680	i = surf_def_h(0)%i(m)
681	j = surf_def_h(0)%j(m)
682	local_pf(i,j,nzb+1) = surf_def_h(0)%ol(m)
683	ENDDO
684	DO m = 1, surf_lsm_h%ns
685	i = surf_lsm_h%i(m)
686	j = surf_lsm_h%j(m)
687	local_pf(i,j,nzb+1) = surf_lsm_h%ol(m)
688	ENDDO
689	DO m = 1, surf_usm_h%ns
690	i = surf_usm_h%i(m)
691	j = surf_usm_h%j(m)
692	local_pf(i,j,nzb+1) = surf_usm_h%ol(m)
693	ENDDO
694	ELSE
695	IF ( .NOT. ALLOCATED( ol_av ) ) THEN
696	ALLOCATE( ol_av(nysg:nyng,nxlg:nxrg) )
697	ol_av = REAL( fill_value, KIND = wp )
698	ENDIF
699	DO i = nxl, nxr
700	DO j = nys, nyn
701	local_pf(i,j,nzb+1) = ol_av(j,i)
702	ENDDO
703	ENDDO
704	ENDIF
705	resorted = .TRUE.
706	two_d = .TRUE.
707	level_z(nzb+1) = zu(nzb+1)
708
709	CASE ( 'p_xy', 'p_xz', 'p_yz' )
710	IF ( av == 0 ) THEN
711	IF ( psolver /= 'sor' ) CALL exchange_horiz( p, nbgp )
712	to_be_resorted => p
713	ELSE
714	IF ( .NOT. ALLOCATED( p_av ) ) THEN
715	ALLOCATE( p_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
716	p_av = REAL( fill_value, KIND = wp )
717	ENDIF
718	IF ( psolver /= 'sor' ) CALL exchange_horiz( p_av, nbgp )
719	to_be_resorted => p_av
720	ENDIF
721	IF ( mode == 'xy' ) level_z = zu
722
723	CASE ( 'pc_xy', 'pc_xz', 'pc_yz' ) ! particle concentration
724	IF ( av == 0 ) THEN
725	IF ( time_since_reference_point >= particle_advection_start ) THEN
726	tend = prt_count
727	! CALL exchange_horiz( tend, nbgp )
728	ELSE
729	tend = 0.0_wp
730	ENDIF
731	DO i = nxl, nxr
732	DO j = nys, nyn
733	DO k = nzb, nzt+1
734	local_pf(i,j,k) = tend(k,j,i)
735	ENDDO
736	ENDDO
737	ENDDO
738	resorted = .TRUE.
739	ELSE
740	IF ( .NOT. ALLOCATED( pc_av ) ) THEN
741	ALLOCATE( pc_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
742	pc_av = REAL( fill_value, KIND = wp )
743	ENDIF
744	! CALL exchange_horiz( pc_av, nbgp )
745	to_be_resorted => pc_av
746	ENDIF
747
748	CASE ( 'pr_xy', 'pr_xz', 'pr_yz' ) ! mean particle radius (effective radius)
749	IF ( av == 0 ) THEN
750	IF ( time_since_reference_point >= particle_advection_start ) THEN
751	DO i = nxl, nxr
752	DO j = nys, nyn
753	DO k = nzb, nzt+1
754	number_of_particles = prt_count(k,j,i)
755	IF (number_of_particles <= 0) CYCLE
756	particles => grid_particles(k,j,i)%particles(1:number_of_particles)
757	s_r2 = 0.0_wp
758	s_r3 = 0.0_wp
759	DO n = 1, number_of_particles
760	IF ( particles(n)%particle_mask ) THEN
761	s_r2 = s_r2 + particles(n)%radius*2 &
762	particles(n)%weight_factor
763	s_r3 = s_r3 + particles(n)%radius*3 &
764	particles(n)%weight_factor
765	ENDIF
766	ENDDO
767	IF ( s_r2 > 0.0_wp ) THEN
768	mean_r = s_r3 / s_r2
769	ELSE
770	mean_r = 0.0_wp
771	ENDIF
772	tend(k,j,i) = mean_r
773	ENDDO
774	ENDDO
775	ENDDO
776	! CALL exchange_horiz( tend, nbgp )
777	ELSE
778	tend = 0.0_wp
779	ENDIF
780	DO i = nxl, nxr
781	DO j = nys, nyn
782	DO k = nzb, nzt+1
783	local_pf(i,j,k) = tend(k,j,i)
784	ENDDO
785	ENDDO
786	ENDDO
787	resorted = .TRUE.
788	ELSE
789	IF ( .NOT. ALLOCATED( pr_av ) ) THEN
790	ALLOCATE( pr_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
791	pr_av = REAL( fill_value, KIND = wp )
792	ENDIF
793	! CALL exchange_horiz( pr_av, nbgp )
794	to_be_resorted => pr_av
795	ENDIF
796
797	CASE ( 'theta_xy', 'theta_xz', 'theta_yz' )
798	IF ( av == 0 ) THEN
799	IF ( .NOT. bulk_cloud_model ) THEN
800	to_be_resorted => pt
801	ELSE
802	DO i = nxl, nxr
803	DO j = nys, nyn
804	DO k = nzb, nzt+1
805	local_pf(i,j,k) = pt(k,j,i) + lv_d_cp * &
806	d_exner(k) * &
807	ql(k,j,i)
808	ENDDO
809	ENDDO
810	ENDDO
811	resorted = .TRUE.
812	ENDIF
813	ELSE
814	IF ( .NOT. ALLOCATED( pt_av ) ) THEN
815	ALLOCATE( pt_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
816	pt_av = REAL( fill_value, KIND = wp )
817	ENDIF
818	to_be_resorted => pt_av
819	ENDIF
820	IF ( mode == 'xy' ) level_z = zu
821
822	CASE ( 'q_xy', 'q_xz', 'q_yz' )
823	IF ( av == 0 ) THEN
824	to_be_resorted => q
825	ELSE
826	IF ( .NOT. ALLOCATED( q_av ) ) THEN
827	ALLOCATE( q_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
828	q_av = REAL( fill_value, KIND = wp )
829	ENDIF
830	to_be_resorted => q_av
831	ENDIF
832	IF ( mode == 'xy' ) level_z = zu
833
834	CASE ( 'ql_xy', 'ql_xz', 'ql_yz' )
835	IF ( av == 0 ) THEN
836	to_be_resorted => ql
837	ELSE
838	IF ( .NOT. ALLOCATED( ql_av ) ) THEN
839	ALLOCATE( ql_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
840	ql_av = REAL( fill_value, KIND = wp )
841	ENDIF
842	to_be_resorted => ql_av
843	ENDIF
844	IF ( mode == 'xy' ) level_z = zu
845
846	CASE ( 'ql_c_xy', 'ql_c_xz', 'ql_c_yz' )
847	IF ( av == 0 ) THEN
848	to_be_resorted => ql_c
849	ELSE
850	IF ( .NOT. ALLOCATED( ql_c_av ) ) THEN
851	ALLOCATE( ql_c_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
852	ql_c_av = REAL( fill_value, KIND = wp )
853	ENDIF
854	to_be_resorted => ql_c_av
855	ENDIF
856	IF ( mode == 'xy' ) level_z = zu
857
858	CASE ( 'ql_v_xy', 'ql_v_xz', 'ql_v_yz' )
859	IF ( av == 0 ) THEN
860	to_be_resorted => ql_v
861	ELSE
862	IF ( .NOT. ALLOCATED( ql_v_av ) ) THEN
863	ALLOCATE( ql_v_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
864	ql_v_av = REAL( fill_value, KIND = wp )
865	ENDIF
866	to_be_resorted => ql_v_av
867	ENDIF
868	IF ( mode == 'xy' ) level_z = zu
869
870	CASE ( 'ql_vp_xy', 'ql_vp_xz', 'ql_vp_yz' )
871	IF ( av == 0 ) THEN
872	IF ( time_since_reference_point >= particle_advection_start ) THEN
873	DO i = nxl, nxr
874	DO j = nys, nyn
875	DO k = nzb, nzt+1
876	number_of_particles = prt_count(k,j,i)
877	IF (number_of_particles <= 0) CYCLE
878	particles => grid_particles(k,j,i)%particles(1:number_of_particles)
879	DO n = 1, number_of_particles
880	IF ( particles(n)%particle_mask ) THEN
881	tend(k,j,i) = tend(k,j,i) + &
882	particles(n)%weight_factor / &
883	prt_count(k,j,i)
884	ENDIF
885	ENDDO
886	ENDDO
887	ENDDO
888	ENDDO
889	! CALL exchange_horiz( tend, nbgp )
890	ELSE
891	tend = 0.0_wp
892	ENDIF
893	DO i = nxl, nxr
894	DO j = nys, nyn
895	DO k = nzb, nzt+1
896	local_pf(i,j,k) = tend(k,j,i)
897	ENDDO
898	ENDDO
899	ENDDO
900	resorted = .TRUE.
901	ELSE
902	IF ( .NOT. ALLOCATED( ql_vp_av ) ) THEN
903	ALLOCATE( ql_vp_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
904	ql_vp_av = REAL( fill_value, KIND = wp )
905	ENDIF
906	! CALL exchange_horiz( ql_vp_av, nbgp )
907	to_be_resorted => ql_vp_av
908	ENDIF
909	IF ( mode == 'xy' ) level_z = zu
910
911	CASE ( 'qsws*_xy' ) ! 2d-array
912	IF ( av == 0 ) THEN
913	local_pf(:,:,nzb+1) = REAL( fill_value, KIND = wp )
914	!
915	!-- In case of default surfaces, clean-up flux by density.
916	!-- In case of land-surfaces, convert fluxes into
917	!-- dynamic units
918	DO m = 1, surf_def_h(0)%ns
919	i = surf_def_h(0)%i(m)
920	j = surf_def_h(0)%j(m)
921	k = surf_def_h(0)%k(m)
922	local_pf(i,j,nzb+1) = surf_def_h(0)%qsws(m) * &
923	waterflux_output_conversion(k)
924	ENDDO
925	DO m = 1, surf_lsm_h%ns
926	i = surf_lsm_h%i(m)
927	j = surf_lsm_h%j(m)
928	k = surf_lsm_h%k(m)
929	local_pf(i,j,nzb+1) = surf_lsm_h%qsws(m) * l_v
930	ENDDO
931	DO m = 1, surf_usm_h%ns
932	i = surf_usm_h%i(m)
933	j = surf_usm_h%j(m)
934	k = surf_usm_h%k(m)
935	local_pf(i,j,nzb+1) = surf_usm_h%qsws(m) * l_v
936	ENDDO
937	ELSE
938	IF ( .NOT. ALLOCATED( qsws_av ) ) THEN
939	ALLOCATE( qsws_av(nysg:nyng,nxlg:nxrg) )
940	qsws_av = REAL( fill_value, KIND = wp )
941	ENDIF
942	DO i = nxl, nxr
943	DO j = nys, nyn
944	local_pf(i,j,nzb+1) = qsws_av(j,i)
945	ENDDO
946	ENDDO
947	ENDIF
948	resorted = .TRUE.
949	two_d = .TRUE.
950	level_z(nzb+1) = zu(nzb+1)
951
952	CASE ( 'qv_xy', 'qv_xz', 'qv_yz' )
953	IF ( av == 0 ) THEN
954	DO i = nxl, nxr
955	DO j = nys, nyn
956	DO k = nzb, nzt+1
957	local_pf(i,j,k) = q(k,j,i) - ql(k,j,i)
958	ENDDO
959	ENDDO
960	ENDDO
961	resorted = .TRUE.
962	ELSE
963	IF ( .NOT. ALLOCATED( qv_av ) ) THEN
964	ALLOCATE( qv_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
965	qv_av = REAL( fill_value, KIND = wp )
966	ENDIF
967	to_be_resorted => qv_av
968	ENDIF
969	IF ( mode == 'xy' ) level_z = zu
970
971	CASE ( 'r_a*_xy' ) ! 2d-array
972	IF ( av == 0 ) THEN
973	DO m = 1, surf_lsm_h%ns
974	i = surf_lsm_h%i(m)
975	j = surf_lsm_h%j(m)
976	local_pf(i,j,nzb+1) = surf_lsm_h%r_a(m)
977	ENDDO
978
979	DO m = 1, surf_usm_h%ns
980	i = surf_usm_h%i(m)
981	j = surf_usm_h%j(m)
982	local_pf(i,j,nzb+1) = &
983	( surf_usm_h%frac(ind_veg_wall,m) * &
984	surf_usm_h%r_a(m) + &
985	surf_usm_h%frac(ind_pav_green,m) * &
986	surf_usm_h%r_a_green(m) + &
987	surf_usm_h%frac(ind_wat_win,m) * &
988	surf_usm_h%r_a_window(m) )
989	ENDDO
990	ELSE
991	IF ( .NOT. ALLOCATED( r_a_av ) ) THEN
992	ALLOCATE( r_a_av(nysg:nyng,nxlg:nxrg) )
993	r_a_av = REAL( fill_value, KIND = wp )
994	ENDIF
995	DO i = nxl, nxr
996	DO j = nys, nyn
997	local_pf(i,j,nzb+1) = r_a_av(j,i)
998	ENDDO
999	ENDDO
1000	ENDIF
1001	resorted = .TRUE.
1002	two_d = .TRUE.
1003	level_z(nzb+1) = zu(nzb+1)
1004
1005	CASE ( 's_xy', 's_xz', 's_yz' )
1006	IF ( av == 0 ) THEN
1007	to_be_resorted => s
1008	ELSE
1009	IF ( .NOT. ALLOCATED( s_av ) ) THEN
1010	ALLOCATE( s_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
1011	s_av = REAL( fill_value, KIND = wp )
1012	ENDIF
1013	to_be_resorted => s_av
1014	ENDIF
1015
1016	CASE ( 'shf*_xy' ) ! 2d-array
1017	IF ( av == 0 ) THEN
1018	!
1019	!-- In case of default surfaces, clean-up flux by density.
1020	!-- In case of land- and urban-surfaces, convert fluxes into
1021	!-- dynamic units.
1022	DO m = 1, surf_def_h(0)%ns
1023	i = surf_def_h(0)%i(m)
1024	j = surf_def_h(0)%j(m)
1025	k = surf_def_h(0)%k(m)
1026	local_pf(i,j,nzb+1) = surf_def_h(0)%shf(m) * &
1027	heatflux_output_conversion(k)
1028	ENDDO
1029	DO m = 1, surf_lsm_h%ns
1030	i = surf_lsm_h%i(m)
1031	j = surf_lsm_h%j(m)
1032	k = surf_lsm_h%k(m)
1033	local_pf(i,j,nzb+1) = surf_lsm_h%shf(m) * c_p
1034	ENDDO
1035	DO m = 1, surf_usm_h%ns
1036	i = surf_usm_h%i(m)
1037	j = surf_usm_h%j(m)
1038	k = surf_usm_h%k(m)
1039	local_pf(i,j,nzb+1) = surf_usm_h%shf(m) * c_p
1040	ENDDO
1041	ELSE
1042	IF ( .NOT. ALLOCATED( shf_av ) ) THEN
1043	ALLOCATE( shf_av(nysg:nyng,nxlg:nxrg) )
1044	shf_av = REAL( fill_value, KIND = wp )
1045	ENDIF
1046	DO i = nxl, nxr
1047	DO j = nys, nyn
1048	local_pf(i,j,nzb+1) = shf_av(j,i)
1049	ENDDO
1050	ENDDO
1051	ENDIF
1052	resorted = .TRUE.
1053	two_d = .TRUE.
1054	level_z(nzb+1) = zu(nzb+1)
1055
1056	CASE ( 'ssws*_xy' ) ! 2d-array
1057	IF ( av == 0 ) THEN
1058	DO m = 1, surf_def_h(0)%ns
1059	i = surf_def_h(0)%i(m)
1060	j = surf_def_h(0)%j(m)
1061	local_pf(i,j,nzb+1) = surf_def_h(0)%ssws(m)
1062	ENDDO
1063	DO m = 1, surf_lsm_h%ns
1064	i = surf_lsm_h%i(m)
1065	j = surf_lsm_h%j(m)
1066	local_pf(i,j,nzb+1) = surf_lsm_h%ssws(m)
1067	ENDDO
1068	DO m = 1, surf_usm_h%ns
1069	i = surf_usm_h%i(m)
1070	j = surf_usm_h%j(m)
1071	local_pf(i,j,nzb+1) = surf_usm_h%ssws(m)
1072	ENDDO
1073	ELSE
1074	IF ( .NOT. ALLOCATED( ssws_av ) ) THEN
1075	ALLOCATE( ssws_av(nysg:nyng,nxlg:nxrg) )
1076	ssws_av = REAL( fill_value, KIND = wp )
1077	ENDIF
1078	DO i = nxl, nxr
1079	DO j = nys, nyn
1080	local_pf(i,j,nzb+1) = ssws_av(j,i)
1081	ENDDO
1082	ENDDO
1083	ENDIF
1084	resorted = .TRUE.
1085	two_d = .TRUE.
1086	level_z(nzb+1) = zu(nzb+1)
1087
1088	CASE ( 't*_xy' ) ! 2d-array
1089	IF ( av == 0 ) THEN
1090	DO m = 1, surf_def_h(0)%ns
1091	i = surf_def_h(0)%i(m)
1092	j = surf_def_h(0)%j(m)
1093	local_pf(i,j,nzb+1) = surf_def_h(0)%ts(m)
1094	ENDDO
1095	DO m = 1, surf_lsm_h%ns
1096	i = surf_lsm_h%i(m)
1097	j = surf_lsm_h%j(m)
1098	local_pf(i,j,nzb+1) = surf_lsm_h%ts(m)
1099	ENDDO
1100	DO m = 1, surf_usm_h%ns
1101	i = surf_usm_h%i(m)
1102	j = surf_usm_h%j(m)
1103	local_pf(i,j,nzb+1) = surf_usm_h%ts(m)
1104	ENDDO
1105	ELSE
1106	IF ( .NOT. ALLOCATED( ts_av ) ) THEN
1107	ALLOCATE( ts_av(nysg:nyng,nxlg:nxrg) )
1108	ts_av = REAL( fill_value, KIND = wp )
1109	ENDIF
1110	DO i = nxl, nxr
1111	DO j = nys, nyn
1112	local_pf(i,j,nzb+1) = ts_av(j,i)
1113	ENDDO
1114	ENDDO
1115	ENDIF
1116	resorted = .TRUE.
1117	two_d = .TRUE.
1118	level_z(nzb+1) = zu(nzb+1)
1119
1120	CASE ( 'tsurf*_xy' ) ! 2d-array
1121	IF ( av == 0 ) THEN
1122	DO m = 1, surf_def_h(0)%ns
1123	i = surf_def_h(0)%i(m)
1124	j = surf_def_h(0)%j(m)
1125	local_pf(i,j,nzb+1) = surf_def_h(0)%pt_surface(m)
1126	ENDDO
1127
1128	DO m = 1, surf_lsm_h%ns
1129	i = surf_lsm_h%i(m)
1130	j = surf_lsm_h%j(m)
1131	local_pf(i,j,nzb+1) = surf_lsm_h%pt_surface(m)
1132	ENDDO
1133
1134	DO m = 1, surf_usm_h%ns
1135	i = surf_usm_h%i(m)
1136	j = surf_usm_h%j(m)
1137	local_pf(i,j,nzb+1) = surf_usm_h%pt_surface(m)
1138	ENDDO
1139
1140	ELSE
1141	IF ( .NOT. ALLOCATED( tsurf_av ) ) THEN
1142	ALLOCATE( tsurf_av(nysg:nyng,nxlg:nxrg) )
1143	tsurf_av = REAL( fill_value, KIND = wp )
1144	ENDIF
1145	DO i = nxl, nxr
1146	DO j = nys, nyn
1147	local_pf(i,j,nzb+1) = tsurf_av(j,i)
1148	ENDDO
1149	ENDDO
1150	ENDIF
1151	resorted = .TRUE.
1152	two_d = .TRUE.
1153	level_z(nzb+1) = zu(nzb+1)
1154
1155	CASE ( 'u_xy', 'u_xz', 'u_yz' )
1156	flag_nr = 1
1157	IF ( av == 0 ) THEN
1158	to_be_resorted => u
1159	ELSE
1160	IF ( .NOT. ALLOCATED( u_av ) ) THEN
1161	ALLOCATE( u_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
1162	u_av = REAL( fill_value, KIND = wp )
1163	ENDIF
1164	to_be_resorted => u_av
1165	ENDIF
1166	IF ( mode == 'xy' ) level_z = zu
1167	!
1168	!-- Substitute the values generated by "mirror" boundary condition
1169	!-- at the bottom boundary by the real surface values.
1170	IF ( do2d(av,ivar) == 'u_xz' .OR. do2d(av,ivar) == 'u_yz' ) THEN
1171	IF ( ibc_uv_b == 0 ) local_pf(:,:,nzb) = 0.0_wp
1172	ENDIF
1173
1174	CASE ( 'us*_xy' ) ! 2d-array
1175	IF ( av == 0 ) THEN
1176	DO m = 1, surf_def_h(0)%ns
1177	i = surf_def_h(0)%i(m)
1178	j = surf_def_h(0)%j(m)
1179	local_pf(i,j,nzb+1) = surf_def_h(0)%us(m)
1180	ENDDO
1181	DO m = 1, surf_lsm_h%ns
1182	i = surf_lsm_h%i(m)
1183	j = surf_lsm_h%j(m)
1184	local_pf(i,j,nzb+1) = surf_lsm_h%us(m)
1185	ENDDO
1186	DO m = 1, surf_usm_h%ns
1187	i = surf_usm_h%i(m)
1188	j = surf_usm_h%j(m)
1189	local_pf(i,j,nzb+1) = surf_usm_h%us(m)
1190	ENDDO
1191	ELSE
1192	IF ( .NOT. ALLOCATED( us_av ) ) THEN
1193	ALLOCATE( us_av(nysg:nyng,nxlg:nxrg) )
1194	us_av = REAL( fill_value, KIND = wp )
1195	ENDIF
1196	DO i = nxl, nxr
1197	DO j = nys, nyn
1198	local_pf(i,j,nzb+1) = us_av(j,i)
1199	ENDDO
1200	ENDDO
1201	ENDIF
1202	resorted = .TRUE.
1203	two_d = .TRUE.
1204	level_z(nzb+1) = zu(nzb+1)
1205
1206	CASE ( 'v_xy', 'v_xz', 'v_yz' )
1207	flag_nr = 2
1208	IF ( av == 0 ) THEN
1209	to_be_resorted => v
1210	ELSE
1211	IF ( .NOT. ALLOCATED( v_av ) ) THEN
1212	ALLOCATE( v_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
1213	v_av = REAL( fill_value, KIND = wp )
1214	ENDIF
1215	to_be_resorted => v_av
1216	ENDIF
1217	IF ( mode == 'xy' ) level_z = zu
1218	!
1219	!-- Substitute the values generated by "mirror" boundary condition
1220	!-- at the bottom boundary by the real surface values.
1221	IF ( do2d(av,ivar) == 'v_xz' .OR. do2d(av,ivar) == 'v_yz' ) THEN
1222	IF ( ibc_uv_b == 0 ) local_pf(:,:,nzb) = 0.0_wp
1223	ENDIF
1224
1225	CASE ( 'thetav_xy', 'thetav_xz', 'thetav_yz' )
1226	IF ( av == 0 ) THEN
1227	to_be_resorted => vpt
1228	ELSE
1229	IF ( .NOT. ALLOCATED( vpt_av ) ) THEN
1230	ALLOCATE( vpt_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
1231	vpt_av = REAL( fill_value, KIND = wp )
1232	ENDIF
1233	to_be_resorted => vpt_av
1234	ENDIF
1235	IF ( mode == 'xy' ) level_z = zu
1236
1237	CASE ( 'theta_2m*_xy' ) ! 2d-array
1238	IF ( av == 0 ) THEN
1239	DO m = 1, surf_def_h(0)%ns
1240	i = surf_def_h(0)%i(m)
1241	j = surf_def_h(0)%j(m)
1242	local_pf(i,j,nzb+1) = surf_def_h(0)%pt_2m(m)
1243	ENDDO
1244	DO m = 1, surf_lsm_h%ns
1245	i = surf_lsm_h%i(m)
1246	j = surf_lsm_h%j(m)
1247	local_pf(i,j,nzb+1) = surf_lsm_h%pt_2m(m)
1248	ENDDO
1249	DO m = 1, surf_usm_h%ns
1250	i = surf_usm_h%i(m)
1251	j = surf_usm_h%j(m)
1252	local_pf(i,j,nzb+1) = surf_usm_h%pt_2m(m)
1253	ENDDO
1254	ELSE
1255	IF ( .NOT. ALLOCATED( pt_2m_av ) ) THEN
1256	ALLOCATE( pt_2m_av(nysg:nyng,nxlg:nxrg) )
1257	pt_2m_av = REAL( fill_value, KIND = wp )
1258	ENDIF
1259	DO i = nxl, nxr
1260	DO j = nys, nyn
1261	local_pf(i,j,nzb+1) = pt_2m_av(j,i)
1262	ENDDO
1263	ENDDO
1264	ENDIF
1265	resorted = .TRUE.
1266	two_d = .TRUE.
1267	level_z(nzb+1) = zu(nzb+1)
1268
1269	CASE ( 'ti_xy', 'ti_xz', 'ti_yz' )
1270	IF ( av == 0 ) THEN
1271	to_be_resorted => ti
1272	ELSE
1273	IF ( .NOT. ALLOCATED( ti_av ) ) THEN
1274	ALLOCATE( ti_av(nzb:nzt+1,nys:nyn,nxl:nxr) )
1275	ti_av = REAL( fill_value, KIND = wp )
1276	ENDIF
1277	to_be_resorted => ti_av
1278	ENDIF
1279	IF ( mode == 'xy' ) level_z = zu
1280
1281	CASE ( 'w_xy', 'w_xz', 'w_yz' )
1282	flag_nr = 3
1283	IF ( av == 0 ) THEN
1284	to_be_resorted => w
1285	ELSE
1286	IF ( .NOT. ALLOCATED( w_av ) ) THEN
1287	ALLOCATE( w_av(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
1288	w_av = REAL( fill_value, KIND = wp )
1289	ENDIF
1290	to_be_resorted => w_av
1291	ENDIF
1292	IF ( mode == 'xy' ) level_z = zw
1293
1294	CASE ( 'z0*_xy' ) ! 2d-array
1295	IF ( av == 0 ) THEN
1296	DO m = 1, surf_def_h(0)%ns
1297	i = surf_def_h(0)%i(m)
1298	j = surf_def_h(0)%j(m)
1299	local_pf(i,j,nzb+1) = surf_def_h(0)%z0(m)
1300	ENDDO
1301	DO m = 1, surf_lsm_h%ns
1302	i = surf_lsm_h%i(m)
1303	j = surf_lsm_h%j(m)
1304	local_pf(i,j,nzb+1) = surf_lsm_h%z0(m)
1305	ENDDO
1306	DO m = 1, surf_usm_h%ns
1307	i = surf_usm_h%i(m)
1308	j = surf_usm_h%j(m)
1309	local_pf(i,j,nzb+1) = surf_usm_h%z0(m)
1310	ENDDO
1311	ELSE
1312	IF ( .NOT. ALLOCATED( z0_av ) ) THEN
1313	ALLOCATE( z0_av(nysg:nyng,nxlg:nxrg) )
1314	z0_av = REAL( fill_value, KIND = wp )
1315	ENDIF
1316	DO i = nxl, nxr
1317	DO j = nys, nyn
1318	local_pf(i,j,nzb+1) = z0_av(j,i)
1319	ENDDO
1320	ENDDO
1321	ENDIF
1322	resorted = .TRUE.
1323	two_d = .TRUE.
1324	level_z(nzb+1) = zu(nzb+1)
1325
1326	CASE ( 'z0h*_xy' ) ! 2d-array
1327	IF ( av == 0 ) THEN
1328	DO m = 1, surf_def_h(0)%ns
1329	i = surf_def_h(0)%i(m)
1330	j = surf_def_h(0)%j(m)
1331	local_pf(i,j,nzb+1) = surf_def_h(0)%z0h(m)
1332	ENDDO
1333	DO m = 1, surf_lsm_h%ns
1334	i = surf_lsm_h%i(m)
1335	j = surf_lsm_h%j(m)
1336	local_pf(i,j,nzb+1) = surf_lsm_h%z0h(m)
1337	ENDDO
1338	DO m = 1, surf_usm_h%ns
1339	i = surf_usm_h%i(m)
1340	j = surf_usm_h%j(m)
1341	local_pf(i,j,nzb+1) = surf_usm_h%z0h(m)
1342	ENDDO
1343	ELSE
1344	IF ( .NOT. ALLOCATED( z0h_av ) ) THEN
1345	ALLOCATE( z0h_av(nysg:nyng,nxlg:nxrg) )
1346	z0h_av = REAL( fill_value, KIND = wp )
1347	ENDIF
1348	DO i = nxl, nxr
1349	DO j = nys, nyn
1350	local_pf(i,j,nzb+1) = z0h_av(j,i)
1351	ENDDO
1352	ENDDO
1353	ENDIF
1354	resorted = .TRUE.
1355	two_d = .TRUE.
1356	level_z(nzb+1) = zu(nzb+1)
1357
1358	CASE ( 'z0q*_xy' ) ! 2d-array
1359	IF ( av == 0 ) THEN
1360	DO m = 1, surf_def_h(0)%ns
1361	i = surf_def_h(0)%i(m)
1362	j = surf_def_h(0)%j(m)
1363	local_pf(i,j,nzb+1) = surf_def_h(0)%z0q(m)
1364	ENDDO
1365	DO m = 1, surf_lsm_h%ns
1366	i = surf_lsm_h%i(m)
1367	j = surf_lsm_h%j(m)
1368	local_pf(i,j,nzb+1) = surf_lsm_h%z0q(m)
1369	ENDDO
1370	DO m = 1, surf_usm_h%ns
1371	i = surf_usm_h%i(m)
1372	j = surf_usm_h%j(m)
1373	local_pf(i,j,nzb+1) = surf_usm_h%z0q(m)
1374	ENDDO
1375	ELSE
1376	IF ( .NOT. ALLOCATED( z0q_av ) ) THEN
1377	ALLOCATE( z0q_av(nysg:nyng,nxlg:nxrg) )
1378	z0q_av = REAL( fill_value, KIND = wp )
1379	ENDIF
1380	DO i = nxl, nxr
1381	DO j = nys, nyn
1382	local_pf(i,j,nzb+1) = z0q_av(j,i)
1383	ENDDO
1384	ENDDO
1385	ENDIF
1386	resorted = .TRUE.
1387	two_d = .TRUE.
1388	level_z(nzb+1) = zu(nzb+1)
1389
1390	CASE DEFAULT
1391
1392	!
1393	!-- Quantities of other modules
1394	IF ( .NOT. found ) THEN
1395	CALL module_interface_data_output_2d( &
1396	av, do2d(av,ivar), found, grid, mode, &
1397	local_pf, two_d, nzb_do, nzt_do, &
1398	fill_value &
1399	)
1400	ENDIF
1401
1402	resorted = .TRUE.
1403
1404	IF ( grid == 'zu' ) THEN
1405	IF ( mode == 'xy' ) level_z = zu
1406	ELSEIF ( grid == 'zw' ) THEN
1407	IF ( mode == 'xy' ) level_z = zw
1408	ELSEIF ( grid == 'zu1' ) THEN
1409	IF ( mode == 'xy' ) level_z(nzb+1) = zu(nzb+1)
1410	ELSEIF ( grid == 'zs' ) THEN
1411	IF ( mode == 'xy' ) level_z = zs
1412	ENDIF
1413
1414	IF ( .NOT. found ) THEN
1415	message_string = 'no output provided for: ' // &
1416	TRIM( do2d(av,ivar) )
1417	CALL message( 'data_output_2d', 'PA0181', 0, 0, 0, 6, 0 )
1418	ENDIF
1419
1420	END SELECT
1421
1422	!
1423	!-- Resort the array to be output, if not done above. Flag topography
1424	!-- grid points with fill values, using the corresponding maksing flag.
1425	IF ( .NOT. resorted ) THEN
1426	DO i = nxl, nxr
1427	DO j = nys, nyn
1428	DO k = nzb_do, nzt_do
1429	local_pf(i,j,k) = MERGE( to_be_resorted(k,j,i), &
1430	REAL( fill_value, KIND = wp ), &
1431	BTEST( wall_flags_0(k,j,i), &
1432	flag_nr ) )
1433	ENDDO
1434	ENDDO
1435	ENDDO
1436	ENDIF
1437
1438	!
1439	!-- Output of the individual cross-sections, depending on the cross-
1440	!-- section mode chosen.
1441	is = 1
1442	loop1: DO WHILE ( section(is,s_ind) /= -9999 .OR. two_d )
1443
1444	SELECT CASE ( mode )
1445
1446	CASE ( 'xy' )
1447	!
1448	!-- Determine the cross section index
1449	IF ( two_d ) THEN
1450	layer_xy = nzb+1
1451	ELSE
1452	layer_xy = section(is,s_ind)
1453	ENDIF
1454
1455	!
1456	!-- Exit the loop for layers beyond the data output domain
1457	!-- (used for soil model)
1458	IF ( layer_xy > nzt_do ) THEN
1459	EXIT loop1
1460	ENDIF
1461
1462	!
1463	!-- Update the netCDF xy cross section time axis.
1464	!-- In case of parallel output, this is only done by PE0
1465	!-- to increase the performance.
1466	IF ( time_since_reference_point /= do2d_xy_last_time(av) ) THEN
1467	do2d_xy_time_count(av) = do2d_xy_time_count(av) + 1
1468	do2d_xy_last_time(av) = time_since_reference_point
1469	IF ( myid == 0 ) THEN
1470	IF ( .NOT. data_output_2d_on_each_pe &
1471	.OR. netcdf_data_format > 4 ) &
1472	THEN
1473	#if defined( __netcdf )
1474	nc_stat = NF90_PUT_VAR( id_set_xy(av), &
1475	id_var_time_xy(av), &
1476	(/ time_since_reference_point /), &
1477	start = (/ do2d_xy_time_count(av) /), &
1478	count = (/ 1 /) )
1479	CALL netcdf_handle_error( 'data_output_2d', 53 )
1480	#endif
1481	ENDIF
1482	ENDIF
1483	ENDIF
1484	!
1485	!-- If required, carry out averaging along z
1486	IF ( section(is,s_ind) == -1 .AND. .NOT. two_d ) THEN
1487
1488	local_2d = 0.0_wp
1489	!
1490	!-- Carry out the averaging (all data are on the PE)
1491	DO k = nzb_do, nzt_do
1492	DO j = nys, nyn
1493	DO i = nxl, nxr
1494	local_2d(i,j) = local_2d(i,j) + local_pf(i,j,k)
1495	ENDDO
1496	ENDDO
1497	ENDDO
1498
1499	local_2d = local_2d / ( nzt_do - nzb_do + 1.0_wp)
1500
1501	ELSE
1502	!
1503	!-- Just store the respective section on the local array
1504	local_2d = local_pf(:,:,layer_xy)
1505
1506	ENDIF
1507
1508	#if defined( __parallel )
1509	IF ( netcdf_data_format > 4 ) THEN
1510	!
1511	!-- Parallel output in netCDF4/HDF5 format.
1512	IF ( two_d ) THEN
1513	iis = 1
1514	ELSE
1515	iis = is
1516	ENDIF
1517
1518	#if defined( __netcdf )
1519	!
1520	!-- For parallel output, all cross sections are first stored
1521	!-- here on a local array and will be written to the output
1522	!-- file afterwards to increase the performance.
1523	DO i = nxl, nxr
1524	DO j = nys, nyn
1525	local_2d_sections(i,j,iis) = local_2d(i,j)
1526	ENDDO
1527	ENDDO
1528	#endif
1529	ELSE
1530
1531	IF ( data_output_2d_on_each_pe ) THEN
1532	!
1533	!-- Output of partial arrays on each PE
1534	#if defined( __netcdf )
1535	IF ( myid == 0 ) THEN
1536	WRITE ( 21 ) time_since_reference_point, &
1537	do2d_xy_time_count(av), av
1538	ENDIF
1539	#endif
1540	DO i = 0, io_blocks-1
1541	IF ( i == io_group ) THEN
1542	WRITE ( 21 ) nxl, nxr, nys, nyn, nys, nyn
1543	WRITE ( 21 ) local_2d
1544	ENDIF
1545	#if defined( __parallel )
1546	CALL MPI_BARRIER( comm2d, ierr )
1547	#endif
1548	ENDDO
1549
1550	ELSE
1551	!
1552	!-- PE0 receives partial arrays from all processors and
1553	!-- then outputs them. Here a barrier has to be set,
1554	!-- because otherwise "-MPI- FATAL: Remote protocol queue
1555	!-- full" may occur.
1556	CALL MPI_BARRIER( comm2d, ierr )
1557
1558	ngp = ( nxr-nxl+1 ) * ( nyn-nys+1 )
1559	IF ( myid == 0 ) THEN
1560	!
1561	!-- Local array can be relocated directly.
1562	total_2d(nxl:nxr,nys:nyn) = local_2d
1563	!
1564	!-- Receive data from all other PEs.
1565	DO n = 1, numprocs-1
1566	!
1567	!-- Receive index limits first, then array.
1568	!-- Index limits are received in arbitrary order from
1569	!-- the PEs.
1570	CALL MPI_RECV( ind(1), 4, MPI_INTEGER, &
1571	MPI_ANY_SOURCE, 0, comm2d, &
1572	status, ierr )
1573	sender = status(MPI_SOURCE)
1574	DEALLOCATE( local_2d )
1575	ALLOCATE( local_2d(ind(1):ind(2),ind(3):ind(4)) )
1576	CALL MPI_RECV( local_2d(ind(1),ind(3)), ngp, &
1577	MPI_REAL, sender, 1, comm2d, &
1578	status, ierr )
1579	total_2d(ind(1):ind(2),ind(3):ind(4)) = local_2d
1580	ENDDO
1581	!
1582	!-- Relocate the local array for the next loop increment
1583	DEALLOCATE( local_2d )
1584	ALLOCATE( local_2d(nxl:nxr,nys:nyn) )
1585
1586	#if defined( __netcdf )
1587	IF ( two_d ) THEN
1588	nc_stat = NF90_PUT_VAR( id_set_xy(av), &
1589	id_var_do2d(av,ivar), &
1590	total_2d(0:nx,0:ny), &
1591	start = (/ 1, 1, 1, do2d_xy_time_count(av) /), &
1592	count = (/ nx+1, ny+1, 1, 1 /) )
1593	ELSE
1594	nc_stat = NF90_PUT_VAR( id_set_xy(av), &
1595	id_var_do2d(av,ivar), &
1596	total_2d(0:nx,0:ny), &
1597	start = (/ 1, 1, is, do2d_xy_time_count(av) /), &
1598	count = (/ nx+1, ny+1, 1, 1 /) )
1599	ENDIF
1600	CALL netcdf_handle_error( 'data_output_2d', 54 )
1601	#endif
1602
1603	ELSE
1604	!
1605	!-- First send the local index limits to PE0
1606	ind(1) = nxl; ind(2) = nxr
1607	ind(3) = nys; ind(4) = nyn
1608	CALL MPI_SEND( ind(1), 4, MPI_INTEGER, 0, 0, &
1609	comm2d, ierr )
1610	!
1611	!-- Send data to PE0
1612	CALL MPI_SEND( local_2d(nxl,nys), ngp, &
1613	MPI_REAL, 0, 1, comm2d, ierr )
1614	ENDIF
1615	!
1616	!-- A barrier has to be set, because otherwise some PEs may
1617	!-- proceed too fast so that PE0 may receive wrong data on
1618	!-- tag 0
1619	CALL MPI_BARRIER( comm2d, ierr )
1620	ENDIF
1621
1622	ENDIF
1623	#else
1624	#if defined( __netcdf )
1625	IF ( two_d ) THEN
1626	nc_stat = NF90_PUT_VAR( id_set_xy(av), &
1627	id_var_do2d(av,ivar), &
1628	local_2d(nxl:nxr,nys:nyn), &
1629	start = (/ 1, 1, 1, do2d_xy_time_count(av) /), &
1630	count = (/ nx+1, ny+1, 1, 1 /) )
1631	ELSE
1632	nc_stat = NF90_PUT_VAR( id_set_xy(av), &
1633	id_var_do2d(av,ivar), &
1634	local_2d(nxl:nxr,nys:nyn), &
1635	start = (/ 1, 1, is, do2d_xy_time_count(av) /), &
1636	count = (/ nx+1, ny+1, 1, 1 /) )
1637	ENDIF
1638	CALL netcdf_handle_error( 'data_output_2d', 447 )
1639	#endif
1640	#endif
1641
1642	!
1643	!-- For 2D-arrays (e.g. u*) only one cross-section is available.
1644	!-- Hence exit loop of output levels.
1645	IF ( two_d ) THEN
1646	IF ( netcdf_data_format < 5 ) two_d = .FALSE.
1647	EXIT loop1
1648	ENDIF
1649
1650	CASE ( 'xz' )
1651	!
1652	!-- Update the netCDF xz cross section time axis.
1653	!-- In case of parallel output, this is only done by PE0
1654	!-- to increase the performance.
1655	IF ( time_since_reference_point /= do2d_xz_last_time(av) ) THEN
1656	do2d_xz_time_count(av) = do2d_xz_time_count(av) + 1
1657	do2d_xz_last_time(av) = time_since_reference_point
1658	IF ( myid == 0 ) THEN
1659	IF ( .NOT. data_output_2d_on_each_pe &
1660	.OR. netcdf_data_format > 4 ) &
1661	THEN
1662	#if defined( __netcdf )
1663	nc_stat = NF90_PUT_VAR( id_set_xz(av), &
1664	id_var_time_xz(av), &
1665	(/ time_since_reference_point /), &
1666	start = (/ do2d_xz_time_count(av) /), &
1667	count = (/ 1 /) )
1668	CALL netcdf_handle_error( 'data_output_2d', 56 )
1669	#endif
1670	ENDIF
1671	ENDIF
1672	ENDIF
1673
1674	!
1675	!-- If required, carry out averaging along y
1676	IF ( section(is,s_ind) == -1 ) THEN
1677
1678	ALLOCATE( local_2d_l(nxl:nxr,nzb_do:nzt_do) )
1679	local_2d_l = 0.0_wp
1680	ngp = ( nxr-nxl + 1 ) * ( nzt_do-nzb_do + 1 )
1681	!
1682	!-- First local averaging on the PE
1683	DO k = nzb_do, nzt_do
1684	DO j = nys, nyn
1685	DO i = nxl, nxr
1686	local_2d_l(i,k) = local_2d_l(i,k) + &
1687	local_pf(i,j,k)
1688	ENDDO
1689	ENDDO
1690	ENDDO
1691	#if defined( __parallel )
1692	!
1693	!-- Now do the averaging over all PEs along y
1694	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
1695	CALL MPI_ALLREDUCE( local_2d_l(nxl,nzb_do), &
1696	local_2d(nxl,nzb_do), ngp, MPI_REAL, &
1697	MPI_SUM, comm1dy, ierr )
1698	#else
1699	local_2d = local_2d_l
1700	#endif
1701	local_2d = local_2d / ( ny + 1.0_wp )
1702
1703	DEALLOCATE( local_2d_l )
1704
1705	ELSE
1706	!
1707	!-- Just store the respective section on the local array
1708	!-- (but only if it is available on this PE!)
1709	IF ( section(is,s_ind) >= nys .AND. section(is,s_ind) <= nyn ) &
1710	THEN
1711	local_2d = local_pf(:,section(is,s_ind),nzb_do:nzt_do)
1712	ENDIF
1713
1714	ENDIF
1715
1716	#if defined( __parallel )
1717	IF ( netcdf_data_format > 4 ) THEN
1718	!
1719	!-- Output in netCDF4/HDF5 format.
1720	!-- Output only on those PEs where the respective cross
1721	!-- sections reside. Cross sections averaged along y are
1722	!-- output on the respective first PE along y (myidy=0).
1723	IF ( ( section(is,s_ind) >= nys .AND. &
1724	section(is,s_ind) <= nyn ) .OR. &
1725	( section(is,s_ind) == -1 .AND. myidy == 0 ) ) THEN
1726	#if defined( __netcdf )
1727	!
1728	!-- For parallel output, all cross sections are first
1729	!-- stored here on a local array and will be written to the
1730	!-- output file afterwards to increase the performance.
1731	DO i = nxl, nxr
1732	DO k = nzb_do, nzt_do
1733	local_2d_sections_l(i,is,k) = local_2d(i,k)
1734	ENDDO
1735	ENDDO
1736	#endif
1737	ENDIF
1738
1739	ELSE
1740
1741	IF ( data_output_2d_on_each_pe ) THEN
1742	!
1743	!-- Output of partial arrays on each PE. If the cross
1744	!-- section does not reside on the PE, output special
1745	!-- index values.
1746	#if defined( __netcdf )
1747	IF ( myid == 0 ) THEN
1748	WRITE ( 22 ) time_since_reference_point, &
1749	do2d_xz_time_count(av), av
1750	ENDIF
1751	#endif
1752	DO i = 0, io_blocks-1
1753	IF ( i == io_group ) THEN
1754	IF ( ( section(is,s_ind) >= nys .AND. &
1755	section(is,s_ind) <= nyn ) .OR. &
1756	( section(is,s_ind) == -1 .AND. &
1757	nys-1 == -1 ) ) &
1758	THEN
1759	WRITE (22) nxl, nxr, nzb_do, nzt_do, nzb, nzt+1
1760	WRITE (22) local_2d
1761	ELSE
1762	WRITE (22) -1, -1, -1, -1, -1, -1
1763	ENDIF
1764	ENDIF
1765	#if defined( __parallel )
1766	CALL MPI_BARRIER( comm2d, ierr )
1767	#endif
1768	ENDDO
1769
1770	ELSE
1771	!
1772	!-- PE0 receives partial arrays from all processors of the
1773	!-- respective cross section and outputs them. Here a
1774	!-- barrier has to be set, because otherwise
1775	!-- "-MPI- FATAL: Remote protocol queue full" may occur.
1776	CALL MPI_BARRIER( comm2d, ierr )
1777
1778	ngp = ( nxr-nxl + 1 ) * ( nzt_do-nzb_do + 1 )
1779	IF ( myid == 0 ) THEN
1780	!
1781	!-- Local array can be relocated directly.
1782	IF ( ( section(is,s_ind) >= nys .AND. &
1783	section(is,s_ind) <= nyn ) .OR. &
1784	( section(is,s_ind) == -1 .AND. &
1785	nys-1 == -1 ) ) THEN
1786	total_2d(nxl:nxr,nzb_do:nzt_do) = local_2d
1787	ENDIF
1788	!
1789	!-- Receive data from all other PEs.
1790	DO n = 1, numprocs-1
1791	!
1792	!-- Receive index limits first, then array.
1793	!-- Index limits are received in arbitrary order from
1794	!-- the PEs.
1795	CALL MPI_RECV( ind(1), 4, MPI_INTEGER, &
1796	MPI_ANY_SOURCE, 0, comm2d, &
1797	status, ierr )
1798	!
1799	!-- Not all PEs have data for XZ-cross-section.
1800	IF ( ind(1) /= -9999 ) THEN
1801	sender = status(MPI_SOURCE)
1802	DEALLOCATE( local_2d )
1803	ALLOCATE( local_2d(ind(1):ind(2), &
1804	ind(3):ind(4)) )
1805	CALL MPI_RECV( local_2d(ind(1),ind(3)), ngp, &
1806	MPI_REAL, sender, 1, comm2d, &
1807	status, ierr )
1808	total_2d(ind(1):ind(2),ind(3):ind(4)) = &
1809	local_2d
1810	ENDIF
1811	ENDDO
1812	!
1813	!-- Relocate the local array for the next loop increment
1814	DEALLOCATE( local_2d )
1815	ALLOCATE( local_2d(nxl:nxr,nzb_do:nzt_do) )
1816
1817	#if defined( __netcdf )
1818	nc_stat = NF90_PUT_VAR( id_set_xz(av), &
1819	id_var_do2d(av,ivar), &
1820	total_2d(0:nx,nzb_do:nzt_do), &
1821	start = (/ 1, is, 1, do2d_xz_time_count(av) /), &
1822	count = (/ nx+1, 1, nzt_do-nzb_do+1, 1 /) )
1823	CALL netcdf_handle_error( 'data_output_2d', 58 )
1824	#endif
1825
1826	ELSE
1827	!
1828	!-- If the cross section resides on the PE, send the
1829	!-- local index limits, otherwise send -9999 to PE0.
1830	IF ( ( section(is,s_ind) >= nys .AND. &
1831	section(is,s_ind) <= nyn ) .OR. &
1832	( section(is,s_ind) == -1 .AND. nys-1 == -1 ) ) &
1833	THEN
1834	ind(1) = nxl; ind(2) = nxr
1835	ind(3) = nzb_do; ind(4) = nzt_do
1836	ELSE
1837	ind(1) = -9999; ind(2) = -9999
1838	ind(3) = -9999; ind(4) = -9999
1839	ENDIF
1840	CALL MPI_SEND( ind(1), 4, MPI_INTEGER, 0, 0, &
1841	comm2d, ierr )
1842	!
1843	!-- If applicable, send data to PE0.
1844	IF ( ind(1) /= -9999 ) THEN
1845	CALL MPI_SEND( local_2d(nxl,nzb_do), ngp, &
1846	MPI_REAL, 0, 1, comm2d, ierr )
1847	ENDIF
1848	ENDIF
1849	!
1850	!-- A barrier has to be set, because otherwise some PEs may
1851	!-- proceed too fast so that PE0 may receive wrong data on
1852	!-- tag 0
1853	CALL MPI_BARRIER( comm2d, ierr )
1854	ENDIF
1855
1856	ENDIF
1857	#else
1858	#if defined( __netcdf )
1859	nc_stat = NF90_PUT_VAR( id_set_xz(av), &
1860	id_var_do2d(av,ivar), &
1861	local_2d(nxl:nxr,nzb_do:nzt_do), &
1862	start = (/ 1, is, 1, do2d_xz_time_count(av) /), &
1863	count = (/ nx+1, 1, nzt_do-nzb_do+1, 1 /) )
1864	CALL netcdf_handle_error( 'data_output_2d', 451 )
1865	#endif
1866	#endif
1867
1868	CASE ( 'yz' )
1869	!
1870	!-- Update the netCDF yz cross section time axis.
1871	!-- In case of parallel output, this is only done by PE0
1872	!-- to increase the performance.
1873	IF ( time_since_reference_point /= do2d_yz_last_time(av) ) THEN
1874	do2d_yz_time_count(av) = do2d_yz_time_count(av) + 1
1875	do2d_yz_last_time(av) = time_since_reference_point
1876	IF ( myid == 0 ) THEN
1877	IF ( .NOT. data_output_2d_on_each_pe &
1878	.OR. netcdf_data_format > 4 ) &
1879	THEN
1880	#if defined( __netcdf )
1881	nc_stat = NF90_PUT_VAR( id_set_yz(av), &
1882	id_var_time_yz(av), &
1883	(/ time_since_reference_point /), &
1884	start = (/ do2d_yz_time_count(av) /), &
1885	count = (/ 1 /) )
1886	CALL netcdf_handle_error( 'data_output_2d', 59 )
1887	#endif
1888	ENDIF
1889	ENDIF
1890	ENDIF
1891
1892	!
1893	!-- If required, carry out averaging along x
1894	IF ( section(is,s_ind) == -1 ) THEN
1895
1896	ALLOCATE( local_2d_l(nys:nyn,nzb_do:nzt_do) )
1897	local_2d_l = 0.0_wp
1898	ngp = ( nyn-nys+1 ) * ( nzt_do-nzb_do+1 )
1899	!
1900	!-- First local averaging on the PE
1901	DO k = nzb_do, nzt_do
1902	DO j = nys, nyn
1903	DO i = nxl, nxr
1904	local_2d_l(j,k) = local_2d_l(j,k) + &
1905	local_pf(i,j,k)
1906	ENDDO
1907	ENDDO
1908	ENDDO
1909	#if defined( __parallel )
1910	!
1911	!-- Now do the averaging over all PEs along x
1912	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
1913	CALL MPI_ALLREDUCE( local_2d_l(nys,nzb_do), &
1914	local_2d(nys,nzb_do), ngp, MPI_REAL, &
1915	MPI_SUM, comm1dx, ierr )
1916	#else
1917	local_2d = local_2d_l
1918	#endif
1919	local_2d = local_2d / ( nx + 1.0_wp )
1920
1921	DEALLOCATE( local_2d_l )
1922
1923	ELSE
1924	!
1925	!-- Just store the respective section on the local array
1926	!-- (but only if it is available on this PE!)
1927	IF ( section(is,s_ind) >= nxl .AND. section(is,s_ind) <= nxr ) &
1928	THEN
1929	local_2d = local_pf(section(is,s_ind),:,nzb_do:nzt_do)
1930	ENDIF
1931
1932	ENDIF
1933
1934	#if defined( __parallel )
1935	IF ( netcdf_data_format > 4 ) THEN
1936	!
1937	!-- Output in netCDF4/HDF5 format.
1938	!-- Output only on those PEs where the respective cross
1939	!-- sections reside. Cross sections averaged along x are
1940	!-- output on the respective first PE along x (myidx=0).
1941	IF ( ( section(is,s_ind) >= nxl .AND. &
1942	section(is,s_ind) <= nxr ) .OR. &
1943	( section(is,s_ind) == -1 .AND. myidx == 0 ) ) THEN
1944	#if defined( __netcdf )
1945	!
1946	!-- For parallel output, all cross sections are first
1947	!-- stored here on a local array and will be written to the
1948	!-- output file afterwards to increase the performance.
1949	DO j = nys, nyn
1950	DO k = nzb_do, nzt_do
1951	local_2d_sections_l(is,j,k) = local_2d(j,k)
1952	ENDDO
1953	ENDDO
1954	#endif
1955	ENDIF
1956
1957	ELSE
1958
1959	IF ( data_output_2d_on_each_pe ) THEN
1960	!
1961	!-- Output of partial arrays on each PE. If the cross
1962	!-- section does not reside on the PE, output special
1963	!-- index values.
1964	#if defined( __netcdf )
1965	IF ( myid == 0 ) THEN
1966	WRITE ( 23 ) time_since_reference_point, &
1967	do2d_yz_time_count(av), av
1968	ENDIF
1969	#endif
1970	DO i = 0, io_blocks-1
1971	IF ( i == io_group ) THEN
1972	IF ( ( section(is,s_ind) >= nxl .AND. &
1973	section(is,s_ind) <= nxr ) .OR. &
1974	( section(is,s_ind) == -1 .AND. &
1975	nxl-1 == -1 ) ) &
1976	THEN
1977	WRITE (23) nys, nyn, nzb_do, nzt_do, nzb, nzt+1
1978	WRITE (23) local_2d
1979	ELSE
1980	WRITE (23) -1, -1, -1, -1, -1, -1
1981	ENDIF
1982	ENDIF
1983	#if defined( __parallel )
1984	CALL MPI_BARRIER( comm2d, ierr )
1985	#endif
1986	ENDDO
1987
1988	ELSE
1989	!
1990	!-- PE0 receives partial arrays from all processors of the
1991	!-- respective cross section and outputs them. Here a
1992	!-- barrier has to be set, because otherwise
1993	!-- "-MPI- FATAL: Remote protocol queue full" may occur.
1994	CALL MPI_BARRIER( comm2d, ierr )
1995
1996	ngp = ( nyn-nys+1 ) * ( nzt_do-nzb_do+1 )
1997	IF ( myid == 0 ) THEN
1998	!
1999	!-- Local array can be relocated directly.
2000	IF ( ( section(is,s_ind) >= nxl .AND. &
2001	section(is,s_ind) <= nxr ) .OR. &
2002	( section(is,s_ind) == -1 .AND. nxl-1 == -1 ) ) &
2003	THEN
2004	total_2d(nys:nyn,nzb_do:nzt_do) = local_2d
2005	ENDIF
2006	!
2007	!-- Receive data from all other PEs.
2008	DO n = 1, numprocs-1
2009	!
2010	!-- Receive index limits first, then array.
2011	!-- Index limits are received in arbitrary order from
2012	!-- the PEs.
2013	CALL MPI_RECV( ind(1), 4, MPI_INTEGER, &
2014	MPI_ANY_SOURCE, 0, comm2d, &
2015	status, ierr )
2016	!
2017	!-- Not all PEs have data for YZ-cross-section.
2018	IF ( ind(1) /= -9999 ) THEN
2019	sender = status(MPI_SOURCE)
2020	DEALLOCATE( local_2d )
2021	ALLOCATE( local_2d(ind(1):ind(2), &
2022	ind(3):ind(4)) )
2023	CALL MPI_RECV( local_2d(ind(1),ind(3)), ngp, &
2024	MPI_REAL, sender, 1, comm2d, &
2025	status, ierr )
2026	total_2d(ind(1):ind(2),ind(3):ind(4)) = &
2027	local_2d
2028	ENDIF
2029	ENDDO
2030	!
2031	!-- Relocate the local array for the next loop increment
2032	DEALLOCATE( local_2d )
2033	ALLOCATE( local_2d(nys:nyn,nzb_do:nzt_do) )
2034
2035	#if defined( __netcdf )
2036	nc_stat = NF90_PUT_VAR( id_set_yz(av), &
2037	id_var_do2d(av,ivar), &
2038	total_2d(0:ny,nzb_do:nzt_do), &
2039	start = (/ is, 1, 1, do2d_yz_time_count(av) /), &
2040	count = (/ 1, ny+1, nzt_do-nzb_do+1, 1 /) )
2041	CALL netcdf_handle_error( 'data_output_2d', 61 )
2042	#endif
2043
2044	ELSE
2045	!
2046	!-- If the cross section resides on the PE, send the
2047	!-- local index limits, otherwise send -9999 to PE0.
2048	IF ( ( section(is,s_ind) >= nxl .AND. &
2049	section(is,s_ind) <= nxr ) .OR. &
2050	( section(is,s_ind) == -1 .AND. nxl-1 == -1 ) ) &
2051	THEN
2052	ind(1) = nys; ind(2) = nyn
2053	ind(3) = nzb_do; ind(4) = nzt_do
2054	ELSE
2055	ind(1) = -9999; ind(2) = -9999
2056	ind(3) = -9999; ind(4) = -9999
2057	ENDIF
2058	CALL MPI_SEND( ind(1), 4, MPI_INTEGER, 0, 0, &
2059	comm2d, ierr )
2060	!
2061	!-- If applicable, send data to PE0.
2062	IF ( ind(1) /= -9999 ) THEN
2063	CALL MPI_SEND( local_2d(nys,nzb_do), ngp, &
2064	MPI_REAL, 0, 1, comm2d, ierr )
2065	ENDIF
2066	ENDIF
2067	!
2068	!-- A barrier has to be set, because otherwise some PEs may
2069	!-- proceed too fast so that PE0 may receive wrong data on
2070	!-- tag 0
2071	CALL MPI_BARRIER( comm2d, ierr )
2072	ENDIF
2073
2074	ENDIF
2075	#else
2076	#if defined( __netcdf )
2077	nc_stat = NF90_PUT_VAR( id_set_yz(av), &
2078	id_var_do2d(av,ivar), &
2079	local_2d(nys:nyn,nzb_do:nzt_do), &
2080	start = (/ is, 1, 1, do2d_xz_time_count(av) /), &
2081	count = (/ 1, ny+1, nzt_do-nzb_do+1, 1 /) )
2082	CALL netcdf_handle_error( 'data_output_2d', 452 )
2083	#endif
2084	#endif
2085
2086	END SELECT
2087
2088	is = is + 1
2089	ENDDO loop1
2090
2091	!
2092	!-- For parallel output, all data were collected before on a local array
2093	!-- and are written now to the netcdf file. This must be done to increase
2094	!-- the performance of the parallel output.
2095	#if defined( __netcdf )
2096	IF ( netcdf_data_format > 4 ) THEN
2097
2098	SELECT CASE ( mode )
2099
2100	CASE ( 'xy' )
2101	IF ( two_d ) THEN
2102	nis = 1
2103	two_d = .FALSE.
2104	ELSE
2105	nis = ns
2106	ENDIF
2107	!
2108	!-- Do not output redundant ghost point data except for the
2109	!-- boundaries of the total domain.
2110	! IF ( nxr == nx .AND. nyn /= ny ) THEN
2111	! nc_stat = NF90_PUT_VAR( id_set_xy(av), &
2112	! id_var_do2d(av,ivar), &
2113	! local_2d_sections(nxl:nxr+1, &
2114	! nys:nyn,1:nis), &
2115	! start = (/ nxl+1, nys+1, 1, &
2116	! do2d_xy_time_count(av) /), &
2117	! count = (/ nxr-nxl+2, &
2118	! nyn-nys+1, nis, 1 &
2119	! /) )
2120	! ELSEIF ( nxr /= nx .AND. nyn == ny ) THEN
2121	! nc_stat = NF90_PUT_VAR( id_set_xy(av), &
2122	! id_var_do2d(av,ivar), &
2123	! local_2d_sections(nxl:nxr, &
2124	! nys:nyn+1,1:nis), &
2125	! start = (/ nxl+1, nys+1, 1, &
2126	! do2d_xy_time_count(av) /), &
2127	! count = (/ nxr-nxl+1, &
2128	! nyn-nys+2, nis, 1 &
2129	! /) )
2130	! ELSEIF ( nxr == nx .AND. nyn == ny ) THEN
2131	! nc_stat = NF90_PUT_VAR( id_set_xy(av), &
2132	! id_var_do2d(av,ivar), &
2133	! local_2d_sections(nxl:nxr+1, &
2134	! nys:nyn+1,1:nis), &
2135	! start = (/ nxl+1, nys+1, 1, &
2136	! do2d_xy_time_count(av) /), &
2137	! count = (/ nxr-nxl+2, &
2138	! nyn-nys+2, nis, 1 &
2139	! /) )
2140	! ELSE
2141	nc_stat = NF90_PUT_VAR( id_set_xy(av), &
2142	id_var_do2d(av,ivar), &
2143	local_2d_sections(nxl:nxr, &
2144	nys:nyn,1:nis), &
2145	start = (/ nxl+1, nys+1, 1, &
2146	do2d_xy_time_count(av) /), &
2147	count = (/ nxr-nxl+1, &
2148	nyn-nys+1, nis, 1 &
2149	/) )
2150	! ENDIF
2151
2152	CALL netcdf_handle_error( 'data_output_2d', 55 )
2153
2154	CASE ( 'xz' )
2155	!
2156	!-- First, all PEs get the information of all cross-sections.
2157	!-- Then the data are written to the output file by all PEs
2158	!-- while NF90_COLLECTIVE is set in subroutine
2159	!-- define_netcdf_header. Although redundant information are
2160	!-- written to the output file in that case, the performance
2161	!-- is significantly better compared to the case where only
2162	!-- the first row of PEs in x-direction (myidx = 0) is given
2163	!-- the output while NF90_INDEPENDENT is set.
2164	IF ( npey /= 1 ) THEN
2165
2166	#if defined( __parallel )
2167	!
2168	!-- Distribute data over all PEs along y
2169	ngp = ( nxr-nxl+1 ) * ( nzt_do-nzb_do+1 ) * ns
2170	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
2171	CALL MPI_ALLREDUCE( local_2d_sections_l(nxl,1,nzb_do), &
2172	local_2d_sections(nxl,1,nzb_do), &
2173	ngp, MPI_REAL, MPI_SUM, comm1dy, &
2174	ierr )
2175	#else
2176	local_2d_sections = local_2d_sections_l
2177	#endif
2178	ENDIF
2179	!
2180	!-- Do not output redundant ghost point data except for the
2181	!-- boundaries of the total domain.
2182	! IF ( nxr == nx ) THEN
2183	! nc_stat = NF90_PUT_VAR( id_set_xz(av), &
2184	! id_var_do2d(av,ivar), &
2185	! local_2d_sections(nxl:nxr+1,1:ns, &
2186	! nzb_do:nzt_do), &
2187	! start = (/ nxl+1, 1, 1, &
2188	! do2d_xz_time_count(av) /), &
2189	! count = (/ nxr-nxl+2, ns, nzt_do-nzb_do+1, &
2190	! 1 /) )
2191	! ELSE
2192	nc_stat = NF90_PUT_VAR( id_set_xz(av), &
2193	id_var_do2d(av,ivar), &
2194	local_2d_sections(nxl:nxr,1:ns, &
2195	nzb_do:nzt_do), &
2196	start = (/ nxl+1, 1, 1, &
2197	do2d_xz_time_count(av) /), &
2198	count = (/ nxr-nxl+1, ns, nzt_do-nzb_do+1, &
2199	1 /) )
2200	! ENDIF
2201
2202	CALL netcdf_handle_error( 'data_output_2d', 57 )
2203
2204	CASE ( 'yz' )
2205	!
2206	!-- First, all PEs get the information of all cross-sections.
2207	!-- Then the data are written to the output file by all PEs
2208	!-- while NF90_COLLECTIVE is set in subroutine
2209	!-- define_netcdf_header. Although redundant information are
2210	!-- written to the output file in that case, the performance
2211	!-- is significantly better compared to the case where only
2212	!-- the first row of PEs in y-direction (myidy = 0) is given
2213	!-- the output while NF90_INDEPENDENT is set.
2214	IF ( npex /= 1 ) THEN
2215
2216	#if defined( __parallel )
2217	!
2218	!-- Distribute data over all PEs along x
2219	ngp = ( nyn-nys+1 ) * ( nzt-nzb + 2 ) * ns
2220	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
2221	CALL MPI_ALLREDUCE( local_2d_sections_l(1,nys,nzb_do), &
2222	local_2d_sections(1,nys,nzb_do), &
2223	ngp, MPI_REAL, MPI_SUM, comm1dx, &
2224	ierr )
2225	#else
2226	local_2d_sections = local_2d_sections_l
2227	#endif
2228	ENDIF
2229	!
2230	!-- Do not output redundant ghost point data except for the
2231	!-- boundaries of the total domain.
2232	! IF ( nyn == ny ) THEN
2233	! nc_stat = NF90_PUT_VAR( id_set_yz(av), &
2234	! id_var_do2d(av,ivar), &
2235	! local_2d_sections(1:ns, &
2236	! nys:nyn+1,nzb_do:nzt_do), &
2237	! start = (/ 1, nys+1, 1, &
2238	! do2d_yz_time_count(av) /), &
2239	! count = (/ ns, nyn-nys+2, &
2240	! nzt_do-nzb_do+1, 1 /) )
2241	! ELSE
2242	nc_stat = NF90_PUT_VAR( id_set_yz(av), &
2243	id_var_do2d(av,ivar), &
2244	local_2d_sections(1:ns,nys:nyn, &
2245	nzb_do:nzt_do), &
2246	start = (/ 1, nys+1, 1, &
2247	do2d_yz_time_count(av) /), &
2248	count = (/ ns, nyn-nys+1, &
2249	nzt_do-nzb_do+1, 1 /) )
2250	! ENDIF
2251
2252	CALL netcdf_handle_error( 'data_output_2d', 60 )
2253
2254	CASE DEFAULT
2255	message_string = 'unknown cross-section: ' // TRIM( mode )
2256	CALL message( 'data_output_2d', 'PA0180', 1, 2, 0, 6, 0 )
2257
2258	END SELECT
2259
2260	ENDIF
2261	#endif
2262	ENDIF
2263
2264	ivar = ivar + 1
2265	l = MAX( 2, LEN_TRIM( do2d(av,ivar) ) )
2266	do2d_mode = do2d(av,ivar)(l-1:l)
2267
2268	ENDDO
2269
2270	!
2271	!-- Deallocate temporary arrays.
2272	IF ( ALLOCATED( level_z ) ) DEALLOCATE( level_z )
2273	IF ( netcdf_data_format > 4 ) THEN
2274	DEALLOCATE( local_pf, local_2d, local_2d_sections )
2275	IF( mode == 'xz' .OR. mode == 'yz' ) DEALLOCATE( local_2d_sections_l )
2276	ENDIF
2277	#if defined( __parallel )
2278	IF ( .NOT. data_output_2d_on_each_pe .AND. myid == 0 ) THEN
2279	DEALLOCATE( total_2d )
2280	ENDIF
2281	#endif
2282
2283	!
2284	!-- Close plot output file.
2285	file_id = 20 + s_ind
2286
2287	IF ( data_output_2d_on_each_pe ) THEN
2288	DO i = 0, io_blocks-1
2289	IF ( i == io_group ) THEN
2290	CALL close_file( file_id )
2291	ENDIF
2292	#if defined( __parallel )
2293	CALL MPI_BARRIER( comm2d, ierr )
2294	#endif
2295	ENDDO
2296	ELSE
2297	IF ( myid == 0 ) CALL close_file( file_id )
2298	ENDIF
2299
2300	CALL cpu_log( log_point(3), 'data_output_2d', 'stop' )
2301
2302	IF ( debug_output_timestep ) CALL debug_message( 'data_output_2d', 'end' )
2303
2304
2305	END SUBROUTINE data_output_2d

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

Download in other formats:

| Impressum | ©Leibniz Universität Hannover |