Home

Context Navigation

source: palm/trunk/SOURCE/flow_statistics.f90 @ 1557

Last change on this file since 1557 was 1557, checked in by suehring, 9 years ago
Enable monotone advection for scalars in combination with fifth-order scheme using monotonic limiter.
Property svn:keywords set to `Id`
File size: 138.8 KB

Line
1	#if ! defined( __openacc )
2	SUBROUTINE flow_statistics
3
4	!--------------------------------------------------------------------------------!
5	! This file is part of PALM.
6	!
7	! PALM is free software: you can redistribute it and/or modify it under the terms
8	! of the GNU General Public License as published by the Free Software Foundation,
9	! either version 3 of the License, or (at your option) any later version.
10	!
11	! PALM is distributed in the hope that it will be useful, but WITHOUT ANY
12	! WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR
13	! A PARTICULAR PURPOSE. See the GNU General Public License for more details.
14	!
15	! You should have received a copy of the GNU General Public License along with
16	! PALM. If not, see <http://www.gnu.org/licenses/>.
17	!
18	! Copyright 1997-2014 Leibniz Universitaet Hannover
19	!--------------------------------------------------------------------------------!
20	!
21	! Current revisions:
22	! -----------------
23	! Adjustments for monotonic limiter
24	!
25	! Former revisions:
26	! -----------------
27	! $Id: flow_statistics.f90 1557 2015-03-05 16:43:04Z suehring $
28	!
29	! 1555 2015-03-04 17:44:27Z maronga
30	! Added output of r_a and r_s.
31	!
32	! 1551 2015-03-03 14:18:16Z maronga
33	! Added suppport for land surface model and radiation model output.
34	!
35	! 1498 2014-12-03 14:09:51Z suehring
36	! Comments added
37	!
38	! 1482 2014-10-18 12:34:45Z raasch
39	! missing ngp_sums_ls added in accelerator version
40	!
41	! 1450 2014-08-21 07:31:51Z heinze
42	! bugfix: calculate fac only for simulated_time >= 0.0
43	!
44	! 1396 2014-05-06 13:37:41Z raasch
45	! bugfix: "copyin" replaced by "update device" in openacc-branch
46	!
47	! 1386 2014-05-05 13:55:30Z boeske
48	! bugfix: simulated time before the last timestep is needed for the correct
49	! calculation of the profiles of large scale forcing tendencies
50	!
51	! 1382 2014-04-30 12:15:41Z boeske
52	! Renamed variables which store large scale forcing tendencies
53	! pt_lsa -> td_lsa_lpt, pt_subs -> td_sub_lpt,
54	! q_lsa -> td_lsa_q, q_subs -> td_sub_q,
55	! added Neumann boundary conditions for profile data output of large scale
56	! advection and subsidence terms at nzt+1
57	!
58	! 1374 2014-04-25 12:55:07Z raasch
59	! bugfix: syntax errors removed from openacc-branch
60	! missing variables added to ONLY-lists
61	!
62	! 1365 2014-04-22 15:03:56Z boeske
63	! Output of large scale advection, large scale subsidence and nudging tendencies
64	! +sums_ls_l, ngp_sums_ls, use_subsidence_tendencies
65	!
66	! 1353 2014-04-08 15:21:23Z heinze
67	! REAL constants provided with KIND-attribute
68	!
69	! 1322 2014-03-20 16:38:49Z raasch
70	! REAL constants defined as wp-kind
71	!
72	! 1320 2014-03-20 08:40:49Z raasch
73	! ONLY-attribute added to USE-statements,
74	! kind-parameters added to all INTEGER and REAL declaration statements,
75	! kinds are defined in new module kinds,
76	! revision history before 2012 removed,
77	! comment fields (!:) to be used for variable explanations added to
78	! all variable declaration statements
79	!
80	! 1299 2014-03-06 13:15:21Z heinze
81	! Output of large scale vertical velocity w_subs
82	!
83	! 1257 2013-11-08 15:18:40Z raasch
84	! openacc "end parallel" replaced by "end parallel loop"
85	!
86	! 1241 2013-10-30 11:36:58Z heinze
87	! Output of ug and vg
88	!
89	! 1221 2013-09-10 08:59:13Z raasch
90	! ported for openACC in separate #else branch
91	!
92	! 1179 2013-06-14 05:57:58Z raasch
93	! comment for profile 77 added
94	!
95	! 1115 2013-03-26 18:16:16Z hoffmann
96	! ql is calculated by calc_liquid_water_content
97	!
98	! 1111 2013-03-08 23:54:10Z raasch
99	! openACC directive added
100	!
101	! 1053 2012-11-13 17:11:03Z hoffmann
102	! additions for two-moment cloud physics scheme:
103	! +nr, qr, qc, prr
104	!
105	! 1036 2012-10-22 13:43:42Z raasch
106	! code put under GPL (PALM 3.9)
107	!
108	! 1007 2012-09-19 14:30:36Z franke
109	! Calculation of buoyancy flux for humidity in case of WS-scheme is now using
110	! turbulent fluxes of WS-scheme
111	! Bugfix: Calculation of subgridscale buoyancy flux for humidity and cloud
112	! droplets at nzb and nzt added
113	!
114	! 801 2012-01-10 17:30:36Z suehring
115	! Calculation of turbulent fluxes in advec_ws is now thread-safe.
116	!
117	! Revision 1.1 1997/08/11 06:15:17 raasch
118	! Initial revision
119	!
120	!
121	! Description:
122	! ------------
123	! Compute average profiles and further average flow quantities for the different
124	! user-defined (sub-)regions. The region indexed 0 is the total model domain.
125	!
126	! NOTE: For simplicity, nzb_s_inner and nzb_diff_s_inner are being used as a
127	! ---- lower vertical index for k-loops for all variables, although strictly
128	! speaking the k-loops would have to be split up according to the staggered
129	! grid. However, this implies no error since staggered velocity components are
130	! zero at the walls and inside buildings.
131	!------------------------------------------------------------------------------!
132
133	USE arrays_3d, &
134	ONLY: ddzu, ddzw, e, hyp, km, kh, nr, p, prho, pt, q, qc, ql, qr, qs, &
135	qsws, qswst, rho, sa, saswsb, saswst, shf, td_lsa_lpt, td_lsa_q,&
136	td_sub_lpt, td_sub_q, time_vert, ts, tswst, u, ug, us, usws, &
137	uswst, vsws, v, vg, vpt, vswst, w, w_subs, zw
138
139	USE cloud_parameters, &
140	ONLY: l_d_cp, prr, pt_d_t
141
142	USE control_parameters, &
143	ONLY: average_count_pr, cloud_droplets, cloud_physics, do_sum, &
144	dt_3d, g, humidity, icloud_scheme, kappa, large_scale_forcing, &
145	large_scale_subsidence, max_pr_user, message_string, &
146	monotonic_adjustment, ocean, &
147	passive_scalar, precipitation, simulated_time, &
148	use_subsidence_tendencies, use_surface_fluxes, use_top_fluxes, &
149	ws_scheme_mom, ws_scheme_sca
150
151	USE cpulog, &
152	ONLY: cpu_log, log_point
153
154	USE grid_variables, &
155	ONLY: ddx, ddy
156
157	USE indices, &
158	ONLY: ngp_2dh, ngp_2dh_s_inner, ngp_3d, ngp_3d_inner, ngp_sums, &
159	ngp_sums_ls, nxl, nxr, nyn, nys, nzb, nzb_diff_s_inner, &
160	nzb_s_inner, nzt, nzt_diff
161
162	USE kinds
163
164	USE land_surface_model_mod, &
165	ONLY: dots_soil, ghf_eb, land_surface, m_soil, nzb_soil, nzt_soil, &
166	qsws_eb, qsws_liq_eb, qsws_soil_eb, qsws_veg_eb, r_a, r_s, &
167	shf_eb, t_soil
168
169	USE pegrid
170
171	USE radiation_model_mod, &
172	ONLY : dots_rad, rad_net, rad_sw_in, radiation
173
174	USE statistics
175
176	IMPLICIT NONE
177
178	INTEGER(iwp) :: i !:
179	INTEGER(iwp) :: j !:
180	INTEGER(iwp) :: k !:
181	INTEGER(iwp) :: nt !:
182	INTEGER(iwp) :: omp_get_thread_num !:
183	INTEGER(iwp) :: sr !:
184	INTEGER(iwp) :: tn !:
185
186	LOGICAL :: first !:
187
188	REAL(wp) :: dptdz_threshold !:
189	REAL(wp) :: fac !:
190	REAL(wp) :: height !:
191	REAL(wp) :: pts !:
192	REAL(wp) :: sums_l_eper !:
193	REAL(wp) :: sums_l_etot !:
194	REAL(wp) :: ust !:
195	REAL(wp) :: ust2 !:
196	REAL(wp) :: u2 !:
197	REAL(wp) :: vst !:
198	REAL(wp) :: vst2 !:
199	REAL(wp) :: v2 !:
200	REAL(wp) :: w2 !:
201	REAL(wp) :: z_i(2) !:
202
203	REAL(wp) :: dptdz(nzb+1:nzt+1) !:
204	REAL(wp) :: sums_ll(nzb:nzt+1,2) !:
205
206	CALL cpu_log( log_point(10), 'flow_statistics', 'start' )
207
208	!$acc update host( km, kh, e, pt, qs, qsws, rif, shf, ts, u, usws, v, vsws, w )
209
210	!
211	!-- To be on the safe side, check whether flow_statistics has already been
212	!-- called once after the current time step
213	IF ( flow_statistics_called ) THEN
214
215	message_string = 'flow_statistics is called two times within one ' // &
216	'timestep'
217	CALL message( 'flow_statistics', 'PA0190', 1, 2, 0, 6, 0 )
218
219	ENDIF
220
221	!
222	!-- Compute statistics for each (sub-)region
223	DO sr = 0, statistic_regions
224
225	!
226	!-- Initialize (local) summation array
227	sums_l = 0.0_wp
228
229	!
230	!-- Store sums that have been computed in other subroutines in summation
231	!-- array
232	sums_l(:,11,:) = sums_l_l(:,sr,:) ! mixing length from diffusivities
233	!-- WARNING: next line still has to be adjusted for OpenMP
234	sums_l(:,21,0) = sums_wsts_bc_l(:,sr) ! heat flux from advec_s_bc
235	sums_l(nzb+9,pr_palm,0) = sums_divold_l(sr) ! old divergence from pres
236	sums_l(nzb+10,pr_palm,0) = sums_divnew_l(sr) ! new divergence from pres
237
238	!
239	!-- When calcuating horizontally-averaged total (resolved- plus subgrid-
240	!-- scale) vertical fluxes and velocity variances by using commonly-
241	!-- applied Reynolds-based methods ( e.g. <w'pt'> = (w-<w>)*(pt-<pt>) )
242	!-- in combination with the 5th order advection scheme, pronounced
243	!-- artificial kinks could be observed in the vertical profiles near the
244	!-- surface. Please note: these kinks were not related to the model truth,
245	!-- i.e. these kinks are just related to an evaluation problem.
246	!-- In order avoid these kinks, vertical fluxes and horizontal as well
247	!-- vertical velocity variances are calculated directly within the advection
248	!-- routines, according to the numerical discretization, to evaluate the
249	!-- statistical quantities as they will appear within the prognostic
250	!-- equations.
251	!-- Copy the turbulent quantities, evaluated in the advection routines to
252	!-- the local array sums_l() for further computations.
253	IF ( ws_scheme_mom .AND. sr == 0 ) THEN
254
255	!
256	!-- According to the Neumann bc for the horizontal velocity components,
257	!-- the corresponding fluxes has to satisfiy the same bc.
258	IF ( ocean ) THEN
259	sums_us2_ws_l(nzt+1,:) = sums_us2_ws_l(nzt,:)
260	sums_vs2_ws_l(nzt+1,:) = sums_vs2_ws_l(nzt,:)
261	ENDIF
262
263	DO i = 0, threads_per_task-1
264	!
265	!-- Swap the turbulent quantities evaluated in advec_ws.
266	sums_l(:,13,i) = sums_wsus_ws_l(:,i) ! wu
267	sums_l(:,15,i) = sums_wsvs_ws_l(:,i) ! wv
268	sums_l(:,30,i) = sums_us2_ws_l(:,i) ! u*2
269	sums_l(:,31,i) = sums_vs2_ws_l(:,i) ! v*2
270	sums_l(:,32,i) = sums_ws2_ws_l(:,i) ! w*2
271	sums_l(:,34,i) = sums_l(:,34,i) + 0.5_wp * &
272	( sums_us2_ws_l(:,i) + sums_vs2_ws_l(:,i) + &
273	sums_ws2_ws_l(:,i) ) ! e*
274	DO k = nzb, nzt
275	sums_l(nzb+5,pr_palm,i) = sums_l(nzb+5,pr_palm,i) + 0.5_wp * ( &
276	sums_us2_ws_l(k,i) + &
277	sums_vs2_ws_l(k,i) + &
278	sums_ws2_ws_l(k,i) )
279	ENDDO
280	ENDDO
281
282	ENDIF
283
284	IF ( ws_scheme_sca .AND. .NOT. monotonic_adjustment &
285	.AND. sr == 0 ) THEN
286
287	DO i = 0, threads_per_task-1
288	sums_l(:,17,i) = sums_wspts_ws_l(:,i) ! wpt from advec_s_ws
289	IF ( ocean ) sums_l(:,66,i) = sums_wssas_ws_l(:,i) ! wsa
290	IF ( humidity .OR. passive_scalar ) sums_l(:,49,i) = &
291	sums_wsqs_ws_l(:,i) !wq
292	ENDDO
293
294	ENDIF
295	!
296	!-- Horizontally averaged profiles of horizontal velocities and temperature.
297	!-- They must have been computed before, because they are already required
298	!-- for other horizontal averages.
299	tn = 0
300
301	!$OMP PARALLEL PRIVATE( i, j, k, tn )
302	#if defined( __intel_openmp_bug )
303	tn = omp_get_thread_num()
304	#else
305	!$ tn = omp_get_thread_num()
306	#endif
307
308	!$OMP DO
309	DO i = nxl, nxr
310	DO j = nys, nyn
311	DO k = nzb_s_inner(j,i), nzt+1
312	sums_l(k,1,tn) = sums_l(k,1,tn) + u(k,j,i) * rmask(j,i,sr)
313	sums_l(k,2,tn) = sums_l(k,2,tn) + v(k,j,i) * rmask(j,i,sr)
314	sums_l(k,4,tn) = sums_l(k,4,tn) + pt(k,j,i) * rmask(j,i,sr)
315	ENDDO
316	ENDDO
317	ENDDO
318
319	!
320	!-- Horizontally averaged profile of salinity
321	IF ( ocean ) THEN
322	!$OMP DO
323	DO i = nxl, nxr
324	DO j = nys, nyn
325	DO k = nzb_s_inner(j,i), nzt+1
326	sums_l(k,23,tn) = sums_l(k,23,tn) + &
327	sa(k,j,i) * rmask(j,i,sr)
328	ENDDO
329	ENDDO
330	ENDDO
331	ENDIF
332
333	!
334	!-- Horizontally averaged profiles of virtual potential temperature,
335	!-- total water content, specific humidity and liquid water potential
336	!-- temperature
337	IF ( humidity ) THEN
338	!$OMP DO
339	DO i = nxl, nxr
340	DO j = nys, nyn
341	DO k = nzb_s_inner(j,i), nzt+1
342	sums_l(k,44,tn) = sums_l(k,44,tn) + &
343	vpt(k,j,i) * rmask(j,i,sr)
344	sums_l(k,41,tn) = sums_l(k,41,tn) + &
345	q(k,j,i) * rmask(j,i,sr)
346	ENDDO
347	ENDDO
348	ENDDO
349	IF ( cloud_physics ) THEN
350	!$OMP DO
351	DO i = nxl, nxr
352	DO j = nys, nyn
353	DO k = nzb_s_inner(j,i), nzt+1
354	sums_l(k,42,tn) = sums_l(k,42,tn) + &
355	( q(k,j,i) - ql(k,j,i) ) * rmask(j,i,sr)
356	sums_l(k,43,tn) = sums_l(k,43,tn) + ( &
357	pt(k,j,i) + l_d_cppt_d_t(k) ql(k,j,i) &
358	) * rmask(j,i,sr)
359	ENDDO
360	ENDDO
361	ENDDO
362	ENDIF
363	ENDIF
364
365	!
366	!-- Horizontally averaged profiles of passive scalar
367	IF ( passive_scalar ) THEN
368	!$OMP DO
369	DO i = nxl, nxr
370	DO j = nys, nyn
371	DO k = nzb_s_inner(j,i), nzt+1
372	sums_l(k,41,tn) = sums_l(k,41,tn) + q(k,j,i) * rmask(j,i,sr)
373	ENDDO
374	ENDDO
375	ENDDO
376	ENDIF
377	!$OMP END PARALLEL
378	!
379	!-- Summation of thread sums
380	IF ( threads_per_task > 1 ) THEN
381	DO i = 1, threads_per_task-1
382	sums_l(:,1,0) = sums_l(:,1,0) + sums_l(:,1,i)
383	sums_l(:,2,0) = sums_l(:,2,0) + sums_l(:,2,i)
384	sums_l(:,4,0) = sums_l(:,4,0) + sums_l(:,4,i)
385	IF ( ocean ) THEN
386	sums_l(:,23,0) = sums_l(:,23,0) + sums_l(:,23,i)
387	ENDIF
388	IF ( humidity ) THEN
389	sums_l(:,41,0) = sums_l(:,41,0) + sums_l(:,41,i)
390	sums_l(:,44,0) = sums_l(:,44,0) + sums_l(:,44,i)
391	IF ( cloud_physics ) THEN
392	sums_l(:,42,0) = sums_l(:,42,0) + sums_l(:,42,i)
393	sums_l(:,43,0) = sums_l(:,43,0) + sums_l(:,43,i)
394	ENDIF
395	ENDIF
396	IF ( passive_scalar ) THEN
397	sums_l(:,41,0) = sums_l(:,41,0) + sums_l(:,41,i)
398	ENDIF
399	ENDDO
400	ENDIF
401
402	#if defined( __parallel )
403	!
404	!-- Compute total sum from local sums
405	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
406	CALL MPI_ALLREDUCE( sums_l(nzb,1,0), sums(nzb,1), nzt+2-nzb, MPI_REAL, &
407	MPI_SUM, comm2d, ierr )
408	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
409	CALL MPI_ALLREDUCE( sums_l(nzb,2,0), sums(nzb,2), nzt+2-nzb, MPI_REAL, &
410	MPI_SUM, comm2d, ierr )
411	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
412	CALL MPI_ALLREDUCE( sums_l(nzb,4,0), sums(nzb,4), nzt+2-nzb, MPI_REAL, &
413	MPI_SUM, comm2d, ierr )
414	IF ( ocean ) THEN
415	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
416	CALL MPI_ALLREDUCE( sums_l(nzb,23,0), sums(nzb,23), nzt+2-nzb, &
417	MPI_REAL, MPI_SUM, comm2d, ierr )
418	ENDIF
419	IF ( humidity ) THEN
420	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
421	CALL MPI_ALLREDUCE( sums_l(nzb,44,0), sums(nzb,44), nzt+2-nzb, &
422	MPI_REAL, MPI_SUM, comm2d, ierr )
423	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
424	CALL MPI_ALLREDUCE( sums_l(nzb,41,0), sums(nzb,41), nzt+2-nzb, &
425	MPI_REAL, MPI_SUM, comm2d, ierr )
426	IF ( cloud_physics ) THEN
427	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
428	CALL MPI_ALLREDUCE( sums_l(nzb,42,0), sums(nzb,42), nzt+2-nzb, &
429	MPI_REAL, MPI_SUM, comm2d, ierr )
430	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
431	CALL MPI_ALLREDUCE( sums_l(nzb,43,0), sums(nzb,43), nzt+2-nzb, &
432	MPI_REAL, MPI_SUM, comm2d, ierr )
433	ENDIF
434	ENDIF
435
436	IF ( passive_scalar ) THEN
437	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
438	CALL MPI_ALLREDUCE( sums_l(nzb,41,0), sums(nzb,41), nzt+2-nzb, &
439	MPI_REAL, MPI_SUM, comm2d, ierr )
440	ENDIF
441	#else
442	sums(:,1) = sums_l(:,1,0)
443	sums(:,2) = sums_l(:,2,0)
444	sums(:,4) = sums_l(:,4,0)
445	IF ( ocean ) sums(:,23) = sums_l(:,23,0)
446	IF ( humidity ) THEN
447	sums(:,44) = sums_l(:,44,0)
448	sums(:,41) = sums_l(:,41,0)
449	IF ( cloud_physics ) THEN
450	sums(:,42) = sums_l(:,42,0)
451	sums(:,43) = sums_l(:,43,0)
452	ENDIF
453	ENDIF
454	IF ( passive_scalar ) sums(:,41) = sums_l(:,41,0)
455	#endif
456
457	!
458	!-- Final values are obtained by division by the total number of grid points
459	!-- used for summation. After that store profiles.
460	sums(:,1) = sums(:,1) / ngp_2dh(sr)
461	sums(:,2) = sums(:,2) / ngp_2dh(sr)
462	sums(:,4) = sums(:,4) / ngp_2dh_s_inner(:,sr)
463	hom(:,1,1,sr) = sums(:,1) ! u
464	hom(:,1,2,sr) = sums(:,2) ! v
465	hom(:,1,4,sr) = sums(:,4) ! pt
466
467
468	!
469	!-- Salinity
470	IF ( ocean ) THEN
471	sums(:,23) = sums(:,23) / ngp_2dh_s_inner(:,sr)
472	hom(:,1,23,sr) = sums(:,23) ! sa
473	ENDIF
474
475	!
476	!-- Humidity and cloud parameters
477	IF ( humidity ) THEN
478	sums(:,44) = sums(:,44) / ngp_2dh_s_inner(:,sr)
479	sums(:,41) = sums(:,41) / ngp_2dh_s_inner(:,sr)
480	hom(:,1,44,sr) = sums(:,44) ! vpt
481	hom(:,1,41,sr) = sums(:,41) ! qv (q)
482	IF ( cloud_physics ) THEN
483	sums(:,42) = sums(:,42) / ngp_2dh_s_inner(:,sr)
484	sums(:,43) = sums(:,43) / ngp_2dh_s_inner(:,sr)
485	hom(:,1,42,sr) = sums(:,42) ! qv
486	hom(:,1,43,sr) = sums(:,43) ! pt
487	ENDIF
488	ENDIF
489
490	!
491	!-- Passive scalar
492	IF ( passive_scalar ) hom(:,1,41,sr) = sums(:,41) / &
493	ngp_2dh_s_inner(:,sr) ! s (q)
494
495	!
496	!-- Horizontally averaged profiles of the remaining prognostic variables,
497	!-- variances, the total and the perturbation energy (single values in last
498	!-- column of sums_l) and some diagnostic quantities.
499	!-- NOTE: for simplicity, nzb_s_inner is used below, although strictly
500	!-- ---- speaking the following k-loop would have to be split up and
501	!-- rearranged according to the staggered grid.
502	!-- However, this implies no error since staggered velocity components
503	!-- are zero at the walls and inside buildings.
504	tn = 0
505	#if defined( __intel_openmp_bug )
506	!$OMP PARALLEL PRIVATE( i, j, k, pts, sums_ll, sums_l_eper, sums_l_etot, &
507	!$OMP tn, ust, ust2, u2, vst, vst2, v2, w2 )
508	tn = omp_get_thread_num()
509	#else
510	!$OMP PARALLEL PRIVATE( i, j, k, pts, sums_ll, sums_l_eper, sums_l_etot, tn, ust, ust2, u2, vst, vst2, v2, w2 )
511	!$ tn = omp_get_thread_num()
512	#endif
513	!$OMP DO
514	DO i = nxl, nxr
515	DO j = nys, nyn
516	sums_l_etot = 0.0_wp
517	DO k = nzb_s_inner(j,i), nzt+1
518	!
519	!-- Prognostic and diagnostic variables
520	sums_l(k,3,tn) = sums_l(k,3,tn) + w(k,j,i) * rmask(j,i,sr)
521	sums_l(k,8,tn) = sums_l(k,8,tn) + e(k,j,i) * rmask(j,i,sr)
522	sums_l(k,9,tn) = sums_l(k,9,tn) + km(k,j,i) * rmask(j,i,sr)
523	sums_l(k,10,tn) = sums_l(k,10,tn) + kh(k,j,i) * rmask(j,i,sr)
524	sums_l(k,40,tn) = sums_l(k,40,tn) + p(k,j,i)
525
526	sums_l(k,33,tn) = sums_l(k,33,tn) + &
527	( pt(k,j,i)-hom(k,1,4,sr) )*2 rmask(j,i,sr)
528
529	IF ( humidity ) THEN
530	sums_l(k,70,tn) = sums_l(k,70,tn) + &
531	( q(k,j,i)-hom(k,1,41,sr) )*2 rmask(j,i,sr)
532	ENDIF
533
534	!
535	!-- Higher moments
536	!-- (Computation of the skewness of w further below)
537	sums_l(k,38,tn) = sums_l(k,38,tn) + w(k,j,i)*3 rmask(j,i,sr)
538
539	sums_l_etot = sums_l_etot + &
540	0.5_wp * ( u(k,j,i)2 + v(k,j,i)2 + &
541	w(k,j,i)*2 ) rmask(j,i,sr)
542	ENDDO
543	!
544	!-- Total and perturbation energy for the total domain (being
545	!-- collected in the last column of sums_l). Summation of these
546	!-- quantities is seperated from the previous loop in order to
547	!-- allow vectorization of that loop.
548	sums_l(nzb+4,pr_palm,tn) = sums_l(nzb+4,pr_palm,tn) + sums_l_etot
549	!
550	!-- 2D-arrays (being collected in the last column of sums_l)
551	sums_l(nzb,pr_palm,tn) = sums_l(nzb,pr_palm,tn) + &
552	us(j,i) * rmask(j,i,sr)
553	sums_l(nzb+1,pr_palm,tn) = sums_l(nzb+1,pr_palm,tn) + &
554	usws(j,i) * rmask(j,i,sr)
555	sums_l(nzb+2,pr_palm,tn) = sums_l(nzb+2,pr_palm,tn) + &
556	vsws(j,i) * rmask(j,i,sr)
557	sums_l(nzb+3,pr_palm,tn) = sums_l(nzb+3,pr_palm,tn) + &
558	ts(j,i) * rmask(j,i,sr)
559	IF ( humidity ) THEN
560	sums_l(nzb+12,pr_palm,tn) = sums_l(nzb+12,pr_palm,tn) + &
561	qs(j,i) * rmask(j,i,sr)
562	ENDIF
563	ENDDO
564	ENDDO
565
566	!
567	!-- Computation of statistics when ws-scheme is not used. Else these
568	!-- quantities are evaluated in the advection routines.
569	IF ( .NOT. ws_scheme_mom .OR. sr /= 0 ) THEN
570	!$OMP DO
571	DO i = nxl, nxr
572	DO j = nys, nyn
573	sums_l_eper = 0.0_wp
574	DO k = nzb_s_inner(j,i), nzt+1
575	u2 = u(k,j,i)**2
576	v2 = v(k,j,i)**2
577	w2 = w(k,j,i)**2
578	ust2 = ( u(k,j,i) - hom(k,1,1,sr) )**2
579	vst2 = ( v(k,j,i) - hom(k,1,2,sr) )**2
580
581	sums_l(k,30,tn) = sums_l(k,30,tn) + ust2 * rmask(j,i,sr)
582	sums_l(k,31,tn) = sums_l(k,31,tn) + vst2 * rmask(j,i,sr)
583	sums_l(k,32,tn) = sums_l(k,32,tn) + w2 * rmask(j,i,sr)
584	!
585	!-- Perturbation energy
586
587	sums_l(k,34,tn) = sums_l(k,34,tn) + 0.5_wp * &
588	( ust2 + vst2 + w2 ) * rmask(j,i,sr)
589	sums_l_eper = sums_l_eper + &
590	0.5_wp * ( ust2+vst2+w2 ) * rmask(j,i,sr)
591
592	ENDDO
593	sums_l(nzb+5,pr_palm,tn) = sums_l(nzb+5,pr_palm,tn) &
594	+ sums_l_eper
595	ENDDO
596	ENDDO
597	ENDIF
598
599	!
600	!-- Horizontally averaged profiles of the vertical fluxes
601
602	!$OMP DO
603	DO i = nxl, nxr
604	DO j = nys, nyn
605	!
606	!-- Subgridscale fluxes (without Prandtl layer from k=nzb,
607	!-- oterwise from k=nzb+1)
608	!-- NOTE: for simplicity, nzb_diff_s_inner is used below, although
609	!-- ---- strictly speaking the following k-loop would have to be
610	!-- split up according to the staggered grid.
611	!-- However, this implies no error since staggered velocity
612	!-- components are zero at the walls and inside buildings.
613
614	DO k = nzb_diff_s_inner(j,i)-1, nzt_diff
615	!
616	!-- Momentum flux w"u"
617	sums_l(k,12,tn) = sums_l(k,12,tn) - 0.25_wp * ( &
618	km(k,j,i)+km(k+1,j,i)+km(k,j,i-1)+km(k+1,j,i-1) &
619	) * ( &
620	( u(k+1,j,i) - u(k,j,i) ) * ddzu(k+1) &
621	+ ( w(k,j,i) - w(k,j,i-1) ) * ddx &
622	) * rmask(j,i,sr)
623	!
624	!-- Momentum flux w"v"
625	sums_l(k,14,tn) = sums_l(k,14,tn) - 0.25_wp * ( &
626	km(k,j,i)+km(k+1,j,i)+km(k,j-1,i)+km(k+1,j-1,i) &
627	) * ( &
628	( v(k+1,j,i) - v(k,j,i) ) * ddzu(k+1) &
629	+ ( w(k,j,i) - w(k,j-1,i) ) * ddy &
630	) * rmask(j,i,sr)
631	!
632	!-- Heat flux w"pt"
633	sums_l(k,16,tn) = sums_l(k,16,tn) &
634	- 0.5_wp * ( kh(k,j,i) + kh(k+1,j,i) )&
635	* ( pt(k+1,j,i) - pt(k,j,i) ) &
636	* ddzu(k+1) * rmask(j,i,sr)
637
638
639	!
640	!-- Salinity flux w"sa"
641	IF ( ocean ) THEN
642	sums_l(k,65,tn) = sums_l(k,65,tn) &
643	- 0.5_wp * ( kh(k,j,i) + kh(k+1,j,i) )&
644	* ( sa(k+1,j,i) - sa(k,j,i) ) &
645	* ddzu(k+1) * rmask(j,i,sr)
646	ENDIF
647
648	!
649	!-- Buoyancy flux, water flux (humidity flux) w"q"
650	IF ( humidity ) THEN
651	sums_l(k,45,tn) = sums_l(k,45,tn) &
652	- 0.5_wp * ( kh(k,j,i) + kh(k+1,j,i) )&
653	* ( vpt(k+1,j,i) - vpt(k,j,i) ) &
654	* ddzu(k+1) * rmask(j,i,sr)
655	sums_l(k,48,tn) = sums_l(k,48,tn) &
656	- 0.5_wp * ( kh(k,j,i) + kh(k+1,j,i) )&
657	* ( q(k+1,j,i) - q(k,j,i) ) &
658	* ddzu(k+1) * rmask(j,i,sr)
659
660	IF ( cloud_physics ) THEN
661	sums_l(k,51,tn) = sums_l(k,51,tn) &
662	- 0.5_wp * ( kh(k,j,i) + kh(k+1,j,i) )&
663	* ( ( q(k+1,j,i) - ql(k+1,j,i) )&
664	- ( q(k,j,i) - ql(k,j,i) ) ) &
665	* ddzu(k+1) * rmask(j,i,sr)
666	ENDIF
667	ENDIF
668
669	!
670	!-- Passive scalar flux
671	IF ( passive_scalar ) THEN
672	sums_l(k,48,tn) = sums_l(k,48,tn) &
673	- 0.5_wp * ( kh(k,j,i) + kh(k+1,j,i) )&
674	* ( q(k+1,j,i) - q(k,j,i) ) &
675	* ddzu(k+1) * rmask(j,i,sr)
676	ENDIF
677
678	ENDDO
679
680	!
681	!-- Subgridscale fluxes in the Prandtl layer
682	IF ( use_surface_fluxes ) THEN
683	sums_l(nzb,12,tn) = sums_l(nzb,12,tn) + &
684	usws(j,i) * rmask(j,i,sr) ! w"u"
685	sums_l(nzb,14,tn) = sums_l(nzb,14,tn) + &
686	vsws(j,i) * rmask(j,i,sr) ! w"v"
687	sums_l(nzb,16,tn) = sums_l(nzb,16,tn) + &
688	shf(j,i) * rmask(j,i,sr) ! w"pt"
689	sums_l(nzb,58,tn) = sums_l(nzb,58,tn) + &
690	0.0_wp * rmask(j,i,sr) ! u"pt"
691	sums_l(nzb,61,tn) = sums_l(nzb,61,tn) + &
692	0.0_wp * rmask(j,i,sr) ! v"pt"
693	IF ( ocean ) THEN
694	sums_l(nzb,65,tn) = sums_l(nzb,65,tn) + &
695	saswsb(j,i) * rmask(j,i,sr) ! w"sa"
696	ENDIF
697	IF ( humidity ) THEN
698	sums_l(nzb,48,tn) = sums_l(nzb,48,tn) + &
699	qsws(j,i) * rmask(j,i,sr) ! w"q" (w"qv")
700	sums_l(nzb,45,tn) = sums_l(nzb,45,tn) + ( &
701	( 1.0_wp + 0.61_wp * q(nzb,j,i) ) * &
702	shf(j,i) + 0.61_wp * pt(nzb,j,i) * &
703	qsws(j,i) )
704	IF ( cloud_droplets ) THEN
705	sums_l(nzb,45,tn) = sums_l(nzb,45,tn) + ( &
706	( 1.0_wp + 0.61_wp * q(nzb,j,i) - &
707	ql(nzb,j,i) ) * shf(j,i) + &
708	0.61_wp * pt(nzb,j,i) * qsws(j,i) )
709	ENDIF
710	IF ( cloud_physics ) THEN
711	!
712	!-- Formula does not work if ql(nzb) /= 0.0
713	sums_l(nzb,51,tn) = sums_l(nzb,51,tn) + & ! w"q" (w"qv")
714	qsws(j,i) * rmask(j,i,sr)
715	ENDIF
716	ENDIF
717	IF ( passive_scalar ) THEN
718	sums_l(nzb,48,tn) = sums_l(nzb,48,tn) + &
719	qsws(j,i) * rmask(j,i,sr) ! w"q" (w"qv")
720	ENDIF
721	ENDIF
722
723	IF ( land_surface ) THEN
724	sums_l(nzb,93,tn) = sums_l(nzb,93,tn) + ghf_eb(j,i)
725	sums_l(nzb,94,tn) = sums_l(nzb,94,tn) + shf_eb(j,i)
726	sums_l(nzb,95,tn) = sums_l(nzb,95,tn) + qsws_eb(j,i)
727	sums_l(nzb,96,tn) = sums_l(nzb,96,tn) + qsws_liq_eb(j,i)
728	sums_l(nzb,97,tn) = sums_l(nzb,97,tn) + qsws_soil_eb(j,i)
729	sums_l(nzb,98,tn) = sums_l(nzb,98,tn) + qsws_veg_eb(j,i)
730	sums_l(nzb,99,tn) = sums_l(nzb,99,tn) + r_a(j,i)
731	sums_l(nzb,100,tn) = sums_l(nzb,100,tn)+ r_s(j,i)
732	ENDIF
733
734	IF ( radiation ) THEN
735	sums_l(nzb,101,tn) = sums_l(nzb,101,tn) + rad_net(j,i)
736	sums_l(nzb,102,tn) = sums_l(nzb,102,tn) + rad_sw_in(j,i)
737	ENDIF
738
739	!
740	!-- Subgridscale fluxes at the top surface
741	IF ( use_top_fluxes ) THEN
742	sums_l(nzt:nzt+1,12,tn) = sums_l(nzt:nzt+1,12,tn) + &
743	uswst(j,i) * rmask(j,i,sr) ! w"u"
744	sums_l(nzt:nzt+1,14,tn) = sums_l(nzt:nzt+1,14,tn) + &
745	vswst(j,i) * rmask(j,i,sr) ! w"v"
746	sums_l(nzt:nzt+1,16,tn) = sums_l(nzt:nzt+1,16,tn) + &
747	tswst(j,i) * rmask(j,i,sr) ! w"pt"
748	sums_l(nzt:nzt+1,58,tn) = sums_l(nzt:nzt+1,58,tn) + &
749	0.0_wp * rmask(j,i,sr) ! u"pt"
750	sums_l(nzt:nzt+1,61,tn) = sums_l(nzt:nzt+1,61,tn) + &
751	0.0_wp * rmask(j,i,sr) ! v"pt"
752
753	IF ( ocean ) THEN
754	sums_l(nzt,65,tn) = sums_l(nzt,65,tn) + &
755	saswst(j,i) * rmask(j,i,sr) ! w"sa"
756	ENDIF
757	IF ( humidity ) THEN
758	sums_l(nzt,48,tn) = sums_l(nzt,48,tn) + &
759	qswst(j,i) * rmask(j,i,sr) ! w"q" (w"qv")
760	sums_l(nzt,45,tn) = sums_l(nzt,45,tn) + ( &
761	( 1.0_wp + 0.61_wp * q(nzt,j,i) ) * &
762	tswst(j,i) + 0.61_wp * pt(nzt,j,i) * &
763	qswst(j,i) )
764	IF ( cloud_droplets ) THEN
765	sums_l(nzt,45,tn) = sums_l(nzt,45,tn) + ( &
766	( 1.0_wp + 0.61_wp * q(nzt,j,i) - &
767	ql(nzt,j,i) ) * tswst(j,i) + &
768	0.61_wp * pt(nzt,j,i) * qswst(j,i) )
769	ENDIF
770	IF ( cloud_physics ) THEN
771	!
772	!-- Formula does not work if ql(nzb) /= 0.0
773	sums_l(nzt,51,tn) = sums_l(nzt,51,tn) + & ! w"q" (w"qv")
774	qswst(j,i) * rmask(j,i,sr)
775	ENDIF
776	ENDIF
777	IF ( passive_scalar ) THEN
778	sums_l(nzt,48,tn) = sums_l(nzt,48,tn) + &
779	qswst(j,i) * rmask(j,i,sr) ! w"q" (w"qv")
780	ENDIF
781	ENDIF
782
783	!
784	!-- Resolved fluxes (can be computed for all horizontal points)
785	!-- NOTE: for simplicity, nzb_s_inner is used below, although strictly
786	!-- ---- speaking the following k-loop would have to be split up and
787	!-- rearranged according to the staggered grid.
788	DO k = nzb_s_inner(j,i), nzt
789	ust = 0.5_wp * ( u(k,j,i) - hom(k,1,1,sr) + &
790	u(k+1,j,i) - hom(k+1,1,1,sr) )
791	vst = 0.5_wp * ( v(k,j,i) - hom(k,1,2,sr) + &
792	v(k+1,j,i) - hom(k+1,1,2,sr) )
793	pts = 0.5_wp * ( pt(k,j,i) - hom(k,1,4,sr) + &
794	pt(k+1,j,i) - hom(k+1,1,4,sr) )
795
796	!-- Higher moments
797	sums_l(k,35,tn) = sums_l(k,35,tn) + pts * w(k,j,i)*2 &
798	rmask(j,i,sr)
799	sums_l(k,36,tn) = sums_l(k,36,tn) + pts*2 w(k,j,i) * &
800	rmask(j,i,sr)
801
802	!
803	!-- Salinity flux and density (density does not belong to here,
804	!-- but so far there is no other suitable place to calculate)
805	IF ( ocean ) THEN
806	IF( .NOT. ws_scheme_sca .OR. monotonic_adjustment .OR. &
807	sr /= 0 ) THEN
808	pts = 0.5_wp * ( sa(k,j,i) - hom(k,1,23,sr) + &
809	sa(k+1,j,i) - hom(k+1,1,23,sr) )
810	sums_l(k,66,tn) = sums_l(k,66,tn) + pts * w(k,j,i) * &
811	rmask(j,i,sr)
812	ENDIF
813	sums_l(k,64,tn) = sums_l(k,64,tn) + rho(k,j,i) * &
814	rmask(j,i,sr)
815	sums_l(k,71,tn) = sums_l(k,71,tn) + prho(k,j,i) * &
816	rmask(j,i,sr)
817	ENDIF
818
819	!
820	!-- Buoyancy flux, water flux, humidity flux, liquid water
821	!-- content, rain drop concentration and rain water content
822	IF ( humidity ) THEN
823	IF ( cloud_physics .OR. cloud_droplets ) THEN
824	pts = 0.5_wp * ( vpt(k,j,i) - hom(k,1,44,sr) + &
825	vpt(k+1,j,i) - hom(k+1,1,44,sr) )
826	sums_l(k,46,tn) = sums_l(k,46,tn) + pts * w(k,j,i) * &
827	rmask(j,i,sr)
828	IF ( .NOT. cloud_droplets ) THEN
829	pts = 0.5_wp * &
830	( ( q(k,j,i) - ql(k,j,i) ) - &
831	hom(k,1,42,sr) + &
832	( q(k+1,j,i) - ql(k+1,j,i) ) - &
833	hom(k+1,1,42,sr) )
834	sums_l(k,52,tn) = sums_l(k,52,tn) + pts * w(k,j,i) * &
835	rmask(j,i,sr)
836	IF ( icloud_scheme == 0 ) THEN
837	sums_l(k,54,tn) = sums_l(k,54,tn) + ql(k,j,i) * &
838	rmask(j,i,sr)
839	sums_l(k,75,tn) = sums_l(k,75,tn) + qc(k,j,i) * &
840	rmask(j,i,sr)
841	IF ( precipitation ) THEN
842	sums_l(k,73,tn) = sums_l(k,73,tn) + nr(k,j,i) * &
843	rmask(j,i,sr)
844	sums_l(k,74,tn) = sums_l(k,74,tn) + qr(k,j,i) * &
845	rmask(j,i,sr)
846	sums_l(k,76,tn) = sums_l(k,76,tn) + prr(k,j,i) *&
847	rmask(j,i,sr)
848	ENDIF
849	ELSE
850	sums_l(k,54,tn) = sums_l(k,54,tn) + ql(k,j,i) * &
851	rmask(j,i,sr)
852	ENDIF
853	ELSE
854	sums_l(k,54,tn) = sums_l(k,54,tn) + ql(k,j,i) * &
855	rmask(j,i,sr)
856	ENDIF
857	ELSE
858	IF( .NOT. ws_scheme_sca .OR. monotonic_adjustment .OR. &
859	sr /= 0 ) THEN
860	pts = 0.5_wp * ( vpt(k,j,i) - hom(k,1,44,sr) + &
861	vpt(k+1,j,i) - hom(k+1,1,44,sr) )
862	sums_l(k,46,tn) = sums_l(k,46,tn) + pts * w(k,j,i) * &
863	rmask(j,i,sr)
864	ELSE IF ( ws_scheme_sca .AND. .NOT. monotonic_adjustment &
865	.AND. sr == 0 ) THEN
866	sums_l(k,46,tn) = ( 1.0_wp + 0.61_wp * &
867	hom(k,1,41,sr) ) * &
868	sums_l(k,17,tn) + &
869	0.61_wp * hom(k,1,4,sr) * &
870	sums_l(k,49,tn)
871	END IF
872	END IF
873	ENDIF
874	!
875	!-- Passive scalar flux
876	IF ( passive_scalar .AND. ( .NOT. ws_scheme_sca &
877	.OR. monotonic_adjustment .OR. sr /= 0 ) ) THEN
878	pts = 0.5_wp * ( q(k,j,i) - hom(k,1,41,sr) + &
879	q(k+1,j,i) - hom(k+1,1,41,sr) )
880	sums_l(k,49,tn) = sums_l(k,49,tn) + pts * w(k,j,i) * &
881	rmask(j,i,sr)
882	ENDIF
883
884	!
885	!-- Energy flux we
886	!-- has to be adjusted
887	sums_l(k,37,tn) = sums_l(k,37,tn) + w(k,j,i) * 0.5_wp * &
888	( ust2 + vst2 + w(k,j,i)**2 ) &
889	* rmask(j,i,sr)
890	ENDDO
891	ENDDO
892	ENDDO
893	!
894	!-- For speed optimization fluxes which have been computed in part directly
895	!-- inside the WS advection routines are treated seperatly
896	!-- Momentum fluxes first:
897	IF ( .NOT. ws_scheme_mom .OR. sr /= 0 ) THEN
898	!$OMP DO
899	DO i = nxl, nxr
900	DO j = nys, nyn
901	DO k = nzb_diff_s_inner(j,i)-1, nzt_diff
902	ust = 0.5_wp * ( u(k,j,i) - hom(k,1,1,sr) + &
903	u(k+1,j,i) - hom(k+1,1,1,sr) )
904	vst = 0.5_wp * ( v(k,j,i) - hom(k,1,2,sr) + &
905	v(k+1,j,i) - hom(k+1,1,2,sr) )
906	!
907	!-- Momentum flux wu
908	sums_l(k,13,tn) = sums_l(k,13,tn) + 0.5_wp * &
909	( w(k,j,i-1) + w(k,j,i) ) &
910	* ust * rmask(j,i,sr)
911	!
912	!-- Momentum flux wv
913	sums_l(k,15,tn) = sums_l(k,15,tn) + 0.5_wp * &
914	( w(k,j-1,i) + w(k,j,i) ) &
915	* vst * rmask(j,i,sr)
916	ENDDO
917	ENDDO
918	ENDDO
919
920	ENDIF
921	IF ( .NOT. ws_scheme_sca .OR. monotonic_adjustment .OR. sr /= 0 ) THEN
922	!$OMP DO
923	DO i = nxl, nxr
924	DO j = nys, nyn
925	DO k = nzb_diff_s_inner(j,i)-1, nzt_diff
926	!
927	!-- Vertical heat flux
928	sums_l(k,17,tn) = sums_l(k,17,tn) + 0.5_wp * &
929	( pt(k,j,i) - hom(k,1,4,sr) + &
930	pt(k+1,j,i) - hom(k+1,1,4,sr) ) &
931	* w(k,j,i) * rmask(j,i,sr)
932	IF ( humidity ) THEN
933	pts = 0.5_wp * ( q(k,j,i) - hom(k,1,41,sr) + &
934	q(k+1,j,i) - hom(k+1,1,41,sr) )
935	sums_l(k,49,tn) = sums_l(k,49,tn) + pts * w(k,j,i) * &
936	rmask(j,i,sr)
937	ENDIF
938	ENDDO
939	ENDDO
940	ENDDO
941
942	ENDIF
943
944	!
945	!-- Density at top follows Neumann condition
946	IF ( ocean ) THEN
947	sums_l(nzt+1,64,tn) = sums_l(nzt,64,tn)
948	sums_l(nzt+1,71,tn) = sums_l(nzt,71,tn)
949	ENDIF
950
951	!
952	!-- Divergence of vertical flux of resolved scale energy and pressure
953	!-- fluctuations as well as flux of pressure fluctuation itself (68).
954	!-- First calculate the products, then the divergence.
955	!-- Calculation is time consuming. Do it only, if profiles shall be plotted.
956	IF ( hom(nzb+1,2,55,0) /= 0.0_wp .OR. hom(nzb+1,2,68,0) /= 0.0_wp ) THEN
957
958	sums_ll = 0.0_wp ! local array
959
960	!$OMP DO
961	DO i = nxl, nxr
962	DO j = nys, nyn
963	DO k = nzb_s_inner(j,i)+1, nzt
964
965	sums_ll(k,1) = sums_ll(k,1) + 0.5_wp * w(k,j,i) * ( &
966	( 0.25_wp * ( u(k,j,i)+u(k+1,j,i)+u(k,j,i+1)+u(k+1,j,i+1) &
967	- 0.5_wp * ( hom(k,1,1,sr) + hom(k+1,1,1,sr) ) &
968	) )**2 &
969	+ ( 0.25_wp * ( v(k,j,i)+v(k+1,j,i)+v(k,j+1,i)+v(k+1,j+1,i) &
970	- 0.5_wp * ( hom(k,1,2,sr) + hom(k+1,1,2,sr) ) &
971	) )**2 &
972	+ w(k,j,i)**2 )
973
974	sums_ll(k,2) = sums_ll(k,2) + 0.5_wp * w(k,j,i) &
975	* ( p(k,j,i) + p(k+1,j,i) )
976
977	ENDDO
978	ENDDO
979	ENDDO
980	sums_ll(0,1) = 0.0_wp ! because w is zero at the bottom
981	sums_ll(nzt+1,1) = 0.0_wp
982	sums_ll(0,2) = 0.0_wp
983	sums_ll(nzt+1,2) = 0.0_wp
984
985	DO k = nzb+1, nzt
986	sums_l(k,55,tn) = ( sums_ll(k,1) - sums_ll(k-1,1) ) * ddzw(k)
987	sums_l(k,56,tn) = ( sums_ll(k,2) - sums_ll(k-1,2) ) * ddzw(k)
988	sums_l(k,68,tn) = sums_ll(k,2)
989	ENDDO
990	sums_l(nzb,55,tn) = sums_l(nzb+1,55,tn)
991	sums_l(nzb,56,tn) = sums_l(nzb+1,56,tn)
992	sums_l(nzb,68,tn) = 0.0_wp ! because w* = 0 at nzb
993
994	ENDIF
995
996	!
997	!-- Divergence of vertical flux of SGS TKE and the flux itself (69)
998	IF ( hom(nzb+1,2,57,0) /= 0.0_wp .OR. hom(nzb+1,2,69,0) /= 0.0_wp ) THEN
999
1000	!$OMP DO
1001	DO i = nxl, nxr
1002	DO j = nys, nyn
1003	DO k = nzb_s_inner(j,i)+1, nzt
1004
1005	sums_l(k,57,tn) = sums_l(k,57,tn) - 0.5_wp * ( &
1006	(km(k,j,i)+km(k+1,j,i)) * (e(k+1,j,i)-e(k,j,i)) * ddzu(k+1) &
1007	- (km(k-1,j,i)+km(k,j,i)) * (e(k,j,i)-e(k-1,j,i)) * ddzu(k) &
1008	) * ddzw(k)
1009
1010	sums_l(k,69,tn) = sums_l(k,69,tn) - 0.5_wp * ( &
1011	(km(k,j,i)+km(k+1,j,i)) * (e(k+1,j,i)-e(k,j,i)) * ddzu(k+1) &
1012	)
1013
1014	ENDDO
1015	ENDDO
1016	ENDDO
1017	sums_l(nzb,57,tn) = sums_l(nzb+1,57,tn)
1018	sums_l(nzb,69,tn) = sums_l(nzb+1,69,tn)
1019
1020	ENDIF
1021
1022	!
1023	!-- Horizontal heat fluxes (subgrid, resolved, total).
1024	!-- Do it only, if profiles shall be plotted.
1025	IF ( hom(nzb+1,2,58,0) /= 0.0_wp ) THEN
1026
1027	!$OMP DO
1028	DO i = nxl, nxr
1029	DO j = nys, nyn
1030	DO k = nzb_s_inner(j,i)+1, nzt
1031	!
1032	!-- Subgrid horizontal heat fluxes u"pt", v"pt"
1033	sums_l(k,58,tn) = sums_l(k,58,tn) - 0.5_wp * &
1034	( kh(k,j,i) + kh(k,j,i-1) ) &
1035	* ( pt(k,j,i-1) - pt(k,j,i) ) &
1036	* ddx * rmask(j,i,sr)
1037	sums_l(k,61,tn) = sums_l(k,61,tn) - 0.5_wp * &
1038	( kh(k,j,i) + kh(k,j-1,i) ) &
1039	* ( pt(k,j-1,i) - pt(k,j,i) ) &
1040	* ddy * rmask(j,i,sr)
1041	!
1042	!-- Resolved horizontal heat fluxes upt, vpt
1043	sums_l(k,59,tn) = sums_l(k,59,tn) + &
1044	( u(k,j,i) - hom(k,1,1,sr) ) &
1045	* 0.5_wp * ( pt(k,j,i-1) - hom(k,1,4,sr) + &
1046	pt(k,j,i) - hom(k,1,4,sr) )
1047	pts = 0.5_wp * ( pt(k,j-1,i) - hom(k,1,4,sr) + &
1048	pt(k,j,i) - hom(k,1,4,sr) )
1049	sums_l(k,62,tn) = sums_l(k,62,tn) + &
1050	( v(k,j,i) - hom(k,1,2,sr) ) &
1051	* 0.5_wp * ( pt(k,j-1,i) - hom(k,1,4,sr) + &
1052	pt(k,j,i) - hom(k,1,4,sr) )
1053	ENDDO
1054	ENDDO
1055	ENDDO
1056	!
1057	!-- Fluxes at the surface must be zero (e.g. due to the Prandtl-layer)
1058	sums_l(nzb,58,tn) = 0.0_wp
1059	sums_l(nzb,59,tn) = 0.0_wp
1060	sums_l(nzb,60,tn) = 0.0_wp
1061	sums_l(nzb,61,tn) = 0.0_wp
1062	sums_l(nzb,62,tn) = 0.0_wp
1063	sums_l(nzb,63,tn) = 0.0_wp
1064
1065	ENDIF
1066
1067	!
1068	!-- Collect current large scale advection and subsidence tendencies for
1069	!-- data output
1070	IF ( large_scale_forcing .AND. ( simulated_time .GT. 0.0_wp ) ) THEN
1071	!
1072	!-- Interpolation in time of LSF_DATA
1073	nt = 1
1074	DO WHILE ( simulated_time - dt_3d > time_vert(nt) )
1075	nt = nt + 1
1076	ENDDO
1077	IF ( simulated_time - dt_3d /= time_vert(nt) ) THEN
1078	nt = nt - 1
1079	ENDIF
1080
1081	fac = ( simulated_time - dt_3d - time_vert(nt) ) &
1082	/ ( time_vert(nt+1)-time_vert(nt) )
1083
1084
1085	DO k = nzb, nzt
1086	sums_ls_l(k,0) = td_lsa_lpt(k,nt) &
1087	+ fac * ( td_lsa_lpt(k,nt+1) - td_lsa_lpt(k,nt) )
1088	sums_ls_l(k,1) = td_lsa_q(k,nt) &
1089	+ fac * ( td_lsa_q(k,nt+1) - td_lsa_q(k,nt) )
1090	ENDDO
1091
1092	sums_ls_l(nzt+1,0) = sums_ls_l(nzt,0)
1093	sums_ls_l(nzt+1,1) = sums_ls_l(nzt,1)
1094
1095	IF ( large_scale_subsidence .AND. use_subsidence_tendencies ) THEN
1096
1097	DO k = nzb, nzt
1098	sums_ls_l(k,2) = td_sub_lpt(k,nt) + fac * &
1099	( td_sub_lpt(k,nt+1) - td_sub_lpt(k,nt) )
1100	sums_ls_l(k,3) = td_sub_q(k,nt) + fac * &
1101	( td_sub_q(k,nt+1) - td_sub_q(k,nt) )
1102	ENDDO
1103
1104	sums_ls_l(nzt+1,2) = sums_ls_l(nzt,2)
1105	sums_ls_l(nzt+1,3) = sums_ls_l(nzt,3)
1106
1107	ENDIF
1108
1109	ENDIF
1110
1111
1112	IF ( land_surface ) THEN
1113	!$OMP DO
1114	DO i = nxl, nxr
1115	DO j = nys, nyn
1116	DO k = nzb_soil, nzt_soil
1117	sums_l(k,89,tn) = sums_l(k,89,tn) + t_soil(k,j,i) * rmask(j,i,sr)
1118	sums_l(k,91,tn) = sums_l(k,91,tn) + m_soil(k,j,i) * rmask(j,i,sr)
1119	ENDDO
1120	ENDDO
1121	ENDDO
1122	ENDIF
1123
1124
1125	!
1126	!-- Calculate the user-defined profiles
1127	CALL user_statistics( 'profiles', sr, tn )
1128	!$OMP END PARALLEL
1129
1130	!
1131	!-- Summation of thread sums
1132	IF ( threads_per_task > 1 ) THEN
1133	DO i = 1, threads_per_task-1
1134	sums_l(:,3,0) = sums_l(:,3,0) + sums_l(:,3,i)
1135	sums_l(:,4:40,0) = sums_l(:,4:40,0) + sums_l(:,4:40,i)
1136	sums_l(:,45:pr_palm,0) = sums_l(:,45:pr_palm,0) + &
1137	sums_l(:,45:pr_palm,i)
1138	IF ( max_pr_user > 0 ) THEN
1139	sums_l(:,pr_palm+1:pr_palm+max_pr_user,0) = &
1140	sums_l(:,pr_palm+1:pr_palm+max_pr_user,0) + &
1141	sums_l(:,pr_palm+1:pr_palm+max_pr_user,i)
1142	ENDIF
1143	ENDDO
1144	ENDIF
1145
1146	#if defined( __parallel )
1147
1148	!
1149	!-- Compute total sum from local sums
1150	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
1151	CALL MPI_ALLREDUCE( sums_l(nzb,1,0), sums(nzb,1), ngp_sums, MPI_REAL, &
1152	MPI_SUM, comm2d, ierr )
1153	IF ( large_scale_forcing ) THEN
1154	CALL MPI_ALLREDUCE( sums_ls_l(nzb,2), sums(nzb,83), ngp_sums_ls, &
1155	MPI_REAL, MPI_SUM, comm2d, ierr )
1156	ENDIF
1157	#else
1158	sums = sums_l(:,:,0)
1159	IF ( large_scale_forcing ) THEN
1160	sums(:,81:88) = sums_ls_l
1161	ENDIF
1162	#endif
1163
1164	!
1165	!-- Final values are obtained by division by the total number of grid points
1166	!-- used for summation. After that store profiles.
1167	!-- Profiles:
1168	DO k = nzb, nzt+1
1169	sums(k,3) = sums(k,3) / ngp_2dh(sr)
1170	sums(k,8:11) = sums(k,8:11) / ngp_2dh_s_inner(k,sr)
1171	sums(k,12:22) = sums(k,12:22) / ngp_2dh(sr)
1172	sums(k,23:29) = sums(k,23:29) / ngp_2dh_s_inner(k,sr)
1173	sums(k,30:32) = sums(k,30:32) / ngp_2dh(sr)
1174	sums(k,33:34) = sums(k,33:34) / ngp_2dh_s_inner(k,sr)
1175	sums(k,35:39) = sums(k,35:39) / ngp_2dh(sr)
1176	sums(k,40) = sums(k,40) / ngp_2dh_s_inner(k,sr)
1177	sums(k,45:53) = sums(k,45:53) / ngp_2dh(sr)
1178	sums(k,54) = sums(k,54) / ngp_2dh_s_inner(k,sr)
1179	sums(k,55:63) = sums(k,55:63) / ngp_2dh(sr)
1180	sums(k,64) = sums(k,64) / ngp_2dh_s_inner(k,sr)
1181	sums(k,65:69) = sums(k,65:69) / ngp_2dh(sr)
1182	sums(k,70:80) = sums(k,70:80) / ngp_2dh_s_inner(k,sr)
1183	sums(k,81:88) = sums(k,81:88) / ngp_2dh(sr)
1184	sums(k,89:100) = sums(k,89:100) / ngp_2dh(sr)
1185	sums(k,101:pr_palm-2) = sums(k,101:pr_palm-2)/ ngp_2dh_s_inner(k,sr)
1186	ENDDO
1187
1188	!-- Upstream-parts
1189	sums(nzb:nzb+11,pr_palm-1) = sums(nzb:nzb+11,pr_palm-1) / ngp_3d(sr)
1190	!-- u* and so on
1191	!-- As sums(nzb:nzb+3,pr_palm) are full 2D arrays (us, usws, vsws, ts) whose
1192	!-- size is always ( nx + 1 ) * ( ny + 1 ), defined at the first grid layer
1193	!-- above the topography, they are being divided by ngp_2dh(sr)
1194	sums(nzb:nzb+3,pr_palm) = sums(nzb:nzb+3,pr_palm) / &
1195	ngp_2dh(sr)
1196	sums(nzb+12,pr_palm) = sums(nzb+12,pr_palm) / & ! qs
1197	ngp_2dh(sr)
1198	!-- eges, e*
1199	sums(nzb+4:nzb+5,pr_palm) = sums(nzb+4:nzb+5,pr_palm) / &
1200	ngp_3d(sr)
1201	!-- Old and new divergence
1202	sums(nzb+9:nzb+10,pr_palm) = sums(nzb+9:nzb+10,pr_palm) / &
1203	ngp_3d_inner(sr)
1204
1205	!-- User-defined profiles
1206	IF ( max_pr_user > 0 ) THEN
1207	DO k = nzb, nzt+1
1208	sums(k,pr_palm+1:pr_palm+max_pr_user) = &
1209	sums(k,pr_palm+1:pr_palm+max_pr_user) / &
1210	ngp_2dh_s_inner(k,sr)
1211	ENDDO
1212	ENDIF
1213
1214	!
1215	!-- Collect horizontal average in hom.
1216	!-- Compute deduced averages (e.g. total heat flux)
1217	hom(:,1,3,sr) = sums(:,3) ! w
1218	hom(:,1,8,sr) = sums(:,8) ! e profiles 5-7 are initial profiles
1219	hom(:,1,9,sr) = sums(:,9) ! km
1220	hom(:,1,10,sr) = sums(:,10) ! kh
1221	hom(:,1,11,sr) = sums(:,11) ! l
1222	hom(:,1,12,sr) = sums(:,12) ! w"u"
1223	hom(:,1,13,sr) = sums(:,13) ! wu
1224	hom(:,1,14,sr) = sums(:,14) ! w"v"
1225	hom(:,1,15,sr) = sums(:,15) ! wv
1226	hom(:,1,16,sr) = sums(:,16) ! w"pt"
1227	hom(:,1,17,sr) = sums(:,17) ! wpt
1228	hom(:,1,18,sr) = sums(:,16) + sums(:,17) ! wpt
1229	hom(:,1,19,sr) = sums(:,12) + sums(:,13) ! wu
1230	hom(:,1,20,sr) = sums(:,14) + sums(:,15) ! wv
1231	hom(:,1,21,sr) = sums(:,21) ! wptBC
1232	hom(:,1,22,sr) = sums(:,16) + sums(:,21) ! wptBC
1233	! profile 24 is initial profile (sa)
1234	! profiles 25-29 left empty for initial
1235	! profiles
1236	hom(:,1,30,sr) = sums(:,30) ! u*2
1237	hom(:,1,31,sr) = sums(:,31) ! v*2
1238	hom(:,1,32,sr) = sums(:,32) ! w*2
1239	hom(:,1,33,sr) = sums(:,33) ! pt*2
1240	hom(:,1,34,sr) = sums(:,34) ! e*
1241	hom(:,1,35,sr) = sums(:,35) ! w2pt
1242	hom(:,1,36,sr) = sums(:,36) ! wpt2
1243	hom(:,1,37,sr) = sums(:,37) ! we
1244	hom(:,1,38,sr) = sums(:,38) ! w*3
1245	hom(:,1,39,sr) = sums(:,38) / ( abs( sums(:,32) ) + 1E-20_wp )**1.5_wp ! Sw
1246	hom(:,1,40,sr) = sums(:,40) ! p
1247	hom(:,1,45,sr) = sums(:,45) ! w"vpt"
1248	hom(:,1,46,sr) = sums(:,46) ! wvpt
1249	hom(:,1,47,sr) = sums(:,45) + sums(:,46) ! wvpt
1250	hom(:,1,48,sr) = sums(:,48) ! w"q" (w"qv")
1251	hom(:,1,49,sr) = sums(:,49) ! wq (wqv)
1252	hom(:,1,50,sr) = sums(:,48) + sums(:,49) ! wq (wqv)
1253	hom(:,1,51,sr) = sums(:,51) ! w"qv"
1254	hom(:,1,52,sr) = sums(:,52) ! wqv
1255	hom(:,1,53,sr) = sums(:,52) + sums(:,51) ! wq (wqv)
1256	hom(:,1,54,sr) = sums(:,54) ! ql
1257	hom(:,1,55,sr) = sums(:,55) ! wuu*/dz
1258	hom(:,1,56,sr) = sums(:,56) ! wp/dz
1259	hom(:,1,57,sr) = sums(:,57) ! ( w"e + w"p"/rho )/dz
1260	hom(:,1,58,sr) = sums(:,58) ! u"pt"
1261	hom(:,1,59,sr) = sums(:,59) ! upt
1262	hom(:,1,60,sr) = sums(:,58) + sums(:,59) ! upt_t
1263	hom(:,1,61,sr) = sums(:,61) ! v"pt"
1264	hom(:,1,62,sr) = sums(:,62) ! vpt
1265	hom(:,1,63,sr) = sums(:,61) + sums(:,62) ! vpt_t
1266	hom(:,1,64,sr) = sums(:,64) ! rho
1267	hom(:,1,65,sr) = sums(:,65) ! w"sa"
1268	hom(:,1,66,sr) = sums(:,66) ! wsa
1269	hom(:,1,67,sr) = sums(:,65) + sums(:,66) ! wsa
1270	hom(:,1,68,sr) = sums(:,68) ! wp
1271	hom(:,1,69,sr) = sums(:,69) ! w"e + w"p"/rho
1272	hom(:,1,70,sr) = sums(:,70) ! q*2
1273	hom(:,1,71,sr) = sums(:,71) ! prho
1274	hom(:,1,72,sr) = hyp * 1E-4_wp ! hyp in dbar
1275	hom(:,1,73,sr) = sums(:,73) ! nr
1276	hom(:,1,74,sr) = sums(:,74) ! qr
1277	hom(:,1,75,sr) = sums(:,75) ! qc
1278	hom(:,1,76,sr) = sums(:,76) ! prr (precipitation rate)
1279	! 77 is initial density profile
1280	hom(:,1,78,sr) = ug ! ug
1281	hom(:,1,79,sr) = vg ! vg
1282	hom(:,1,80,sr) = w_subs ! w_subs
1283
1284	IF ( large_scale_forcing ) THEN
1285	hom(:,1,81,sr) = sums_ls_l(:,0) ! td_lsa_lpt
1286	hom(:,1,82,sr) = sums_ls_l(:,1) ! td_lsa_q
1287	IF ( use_subsidence_tendencies ) THEN
1288	hom(:,1,83,sr) = sums_ls_l(:,2) ! td_sub_lpt
1289	hom(:,1,84,sr) = sums_ls_l(:,3) ! td_sub_q
1290	ELSE
1291	hom(:,1,83,sr) = sums(:,83) ! td_sub_lpt
1292	hom(:,1,84,sr) = sums(:,84) ! td_sub_q
1293	ENDIF
1294	hom(:,1,85,sr) = sums(:,85) ! td_nud_lpt
1295	hom(:,1,86,sr) = sums(:,86) ! td_nud_q
1296	hom(:,1,87,sr) = sums(:,87) ! td_nud_u
1297	hom(:,1,88,sr) = sums(:,88) ! td_nud_v
1298	ENDIF
1299
1300	IF ( land_surface ) THEN
1301	hom(:,1,89,sr) = sums(:,89) ! t_soil
1302	! 90 is initial t_soil profile
1303	hom(:,1,91,sr) = sums(:,91) ! m_soil
1304	! 92 is initial m_soil profile
1305	hom(:,1,93,sr) = sums(:,93) ! ghf_eb
1306	hom(:,1,94,sr) = sums(:,94) ! shf_eb
1307	hom(:,1,95,sr) = sums(:,95) ! qsws_eb
1308	hom(:,1,96,sr) = sums(:,96) ! qsws_liq_eb
1309	hom(:,1,97,sr) = sums(:,97) ! qsws_soil_eb
1310	hom(:,1,98,sr) = sums(:,98) ! qsws_veg_eb
1311	hom(:,1,99,sr) = sums(:,99) ! r_a
1312	hom(:,1,100,sr) = sums(:,100) ! r_s
1313
1314	ENDIF
1315
1316	IF ( radiation ) THEN
1317	hom(:,1,101 ,sr) = sums(:,101) ! rad_net
1318	hom(:,1,102,sr) = sums(:,102) ! rad_sw_in
1319	ENDIF
1320
1321	hom(:,1,pr_palm-1,sr) = sums(:,pr_palm-1)
1322	! upstream-parts u_x, u_y, u_z, v_x,
1323	! v_y, usw. (in last but one profile)
1324	hom(:,1,pr_palm,sr) = sums(:,pr_palm)
1325	! u, w'u', w'v', t (in last profile)
1326
1327	IF ( max_pr_user > 0 ) THEN ! user-defined profiles
1328	hom(:,1,pr_palm+1:pr_palm+max_pr_user,sr) = &
1329	sums(:,pr_palm+1:pr_palm+max_pr_user)
1330	ENDIF
1331
1332	!
1333	!-- Determine the boundary layer height using two different schemes.
1334	!-- First scheme: Starting from the Earth's (Ocean's) surface, look for the
1335	!-- first relative minimum (maximum) of the total heat flux.
1336	!-- The corresponding height is assumed as the boundary layer height, if it
1337	!-- is less than 1.5 times the height where the heat flux becomes negative
1338	!-- (positive) for the first time.
1339	z_i(1) = 0.0_wp
1340	first = .TRUE.
1341
1342	IF ( ocean ) THEN
1343	DO k = nzt, nzb+1, -1
1344	IF ( first .AND. hom(k,1,18,sr) < 0.0_wp &
1345	.AND. abs(hom(k,1,18,sr)) > 1.0E-8_wp) THEN
1346	first = .FALSE.
1347	height = zw(k)
1348	ENDIF
1349	IF ( hom(k,1,18,sr) < 0.0_wp .AND. &
1350	abs(hom(k,1,18,sr)) > 1.0E-8_wp .AND. &
1351	hom(k-1,1,18,sr) > hom(k,1,18,sr) ) THEN
1352	IF ( zw(k) < 1.5_wp * height ) THEN
1353	z_i(1) = zw(k)
1354	ELSE
1355	z_i(1) = height
1356	ENDIF
1357	EXIT
1358	ENDIF
1359	ENDDO
1360	ELSE
1361	DO k = nzb, nzt-1
1362	IF ( first .AND. hom(k,1,18,sr) < 0.0_wp &
1363	.AND. abs(hom(k,1,18,sr)) > 1.0E-8_wp ) THEN
1364	first = .FALSE.
1365	height = zw(k)
1366	ENDIF
1367	IF ( hom(k,1,18,sr) < 0.0_wp .AND. &
1368	abs(hom(k,1,18,sr)) > 1.0E-8_wp .AND. &
1369	hom(k+1,1,18,sr) > hom(k,1,18,sr) ) THEN
1370	IF ( zw(k) < 1.5_wp * height ) THEN
1371	z_i(1) = zw(k)
1372	ELSE
1373	z_i(1) = height
1374	ENDIF
1375	EXIT
1376	ENDIF
1377	ENDDO
1378	ENDIF
1379
1380	!
1381	!-- Second scheme: Gradient scheme from Sullivan et al. (1998), modified
1382	!-- by Uhlenbrock(2006). The boundary layer height is the height with the
1383	!-- maximal local temperature gradient: starting from the second (the last
1384	!-- but one) vertical gridpoint, the local gradient must be at least
1385	!-- 0.2K/100m and greater than the next four gradients.
1386	!-- WARNING: The threshold value of 0.2K/100m must be adjusted for the
1387	!-- ocean case!
1388	z_i(2) = 0.0_wp
1389	DO k = nzb+1, nzt+1
1390	dptdz(k) = ( hom(k,1,4,sr) - hom(k-1,1,4,sr) ) * ddzu(k)
1391	ENDDO
1392	dptdz_threshold = 0.2_wp / 100.0_wp
1393
1394	IF ( ocean ) THEN
1395	DO k = nzt+1, nzb+5, -1
1396	IF ( dptdz(k) > dptdz_threshold .AND. &
1397	dptdz(k) > dptdz(k-1) .AND. dptdz(k) > dptdz(k-2) .AND. &
1398	dptdz(k) > dptdz(k-3) .AND. dptdz(k) > dptdz(k-4) ) THEN
1399	z_i(2) = zw(k-1)
1400	EXIT
1401	ENDIF
1402	ENDDO
1403	ELSE
1404	DO k = nzb+1, nzt-3
1405	IF ( dptdz(k) > dptdz_threshold .AND. &
1406	dptdz(k) > dptdz(k+1) .AND. dptdz(k) > dptdz(k+2) .AND. &
1407	dptdz(k) > dptdz(k+3) .AND. dptdz(k) > dptdz(k+4) ) THEN
1408	z_i(2) = zw(k-1)
1409	EXIT
1410	ENDIF
1411	ENDDO
1412	ENDIF
1413
1414	hom(nzb+6,1,pr_palm,sr) = z_i(1)
1415	hom(nzb+7,1,pr_palm,sr) = z_i(2)
1416
1417	!
1418	!-- Computation of both the characteristic vertical velocity and
1419	!-- the characteristic convective boundary layer temperature.
1420	!-- The horizontal average at nzb+1 is input for the average temperature.
1421	IF ( hom(nzb,1,18,sr) > 0.0_wp .AND. abs(hom(nzb,1,18,sr)) > 1.0E-8_wp &
1422	.AND. z_i(1) /= 0.0_wp ) THEN
1423	hom(nzb+8,1,pr_palm,sr) = ( g / hom(nzb+1,1,4,sr) * &
1424	hom(nzb,1,18,sr) * &
1425	ABS( z_i(1) ) )**0.333333333_wp
1426	!-- so far this only works if Prandtl layer is used
1427	hom(nzb+11,1,pr_palm,sr) = hom(nzb,1,16,sr) / hom(nzb+8,1,pr_palm,sr)
1428	ELSE
1429	hom(nzb+8,1,pr_palm,sr) = 0.0_wp
1430	hom(nzb+11,1,pr_palm,sr) = 0.0_wp
1431	ENDIF
1432
1433	!
1434	!-- Collect the time series quantities
1435	ts_value(1,sr) = hom(nzb+4,1,pr_palm,sr) ! E
1436	ts_value(2,sr) = hom(nzb+5,1,pr_palm,sr) ! E*
1437	ts_value(3,sr) = dt_3d
1438	ts_value(4,sr) = hom(nzb,1,pr_palm,sr) ! u*
1439	ts_value(5,sr) = hom(nzb+3,1,pr_palm,sr) ! th*
1440	ts_value(6,sr) = u_max
1441	ts_value(7,sr) = v_max
1442	ts_value(8,sr) = w_max
1443	ts_value(9,sr) = hom(nzb+10,1,pr_palm,sr) ! new divergence
1444	ts_value(10,sr) = hom(nzb+9,1,pr_palm,sr) ! old Divergence
1445	ts_value(11,sr) = hom(nzb+6,1,pr_palm,sr) ! z_i(1)
1446	ts_value(12,sr) = hom(nzb+7,1,pr_palm,sr) ! z_i(2)
1447	ts_value(13,sr) = hom(nzb+8,1,pr_palm,sr) ! w*
1448	ts_value(14,sr) = hom(nzb,1,16,sr) ! w'pt' at k=0
1449	ts_value(15,sr) = hom(nzb+1,1,16,sr) ! w'pt' at k=1
1450	ts_value(16,sr) = hom(nzb+1,1,18,sr) ! wpt at k=1
1451	ts_value(17,sr) = hom(nzb,1,4,sr) ! pt(0)
1452	ts_value(18,sr) = hom(nzb+1,1,4,sr) ! pt(zp)
1453	ts_value(19,sr) = hom(nzb+1,1,pr_palm,sr) ! u'w' at k=0
1454	ts_value(20,sr) = hom(nzb+2,1,pr_palm,sr) ! v'w' at k=0
1455	ts_value(21,sr) = hom(nzb,1,48,sr) ! w"q" at k=0
1456
1457	IF ( ts_value(5,sr) /= 0.0_wp ) THEN
1458	ts_value(22,sr) = ts_value(4,sr)**2 / &
1459	( kappa * g * ts_value(5,sr) / ts_value(18,sr) ) ! L
1460	ELSE
1461	ts_value(22,sr) = 10000.0_wp
1462	ENDIF
1463
1464	ts_value(23,sr) = hom(nzb+12,1,pr_palm,sr) ! q*
1465
1466	!
1467	!-- Collect land surface model timeseries
1468	IF ( land_surface ) THEN
1469	ts_value(dots_soil ,sr) = hom(nzb,1,93,sr) ! ghf_eb
1470	ts_value(dots_soil+1,sr) = hom(nzb,1,94,sr) ! shf_eb
1471	ts_value(dots_soil+2,sr) = hom(nzb,1,95,sr) ! qsws_eb
1472	ts_value(dots_soil+3,sr) = hom(nzb,1,96,sr) ! qsws_liq_eb
1473	ts_value(dots_soil+4,sr) = hom(nzb,1,97,sr) ! qsws_soil_eb
1474	ts_value(dots_soil+5,sr) = hom(nzb,1,98,sr) ! qsws_veg_eb
1475	ts_value(dots_soil+6,sr) = hom(nzb,1,99,sr) ! r_a
1476	ts_value(dots_soil+7,sr) = hom(nzb,1,100,sr) ! r_s
1477	ENDIF
1478	!
1479	!-- Collect radiation model timeseries
1480	IF ( radiation ) THEN
1481	ts_value(dots_rad,sr) = hom(nzb,1,99,sr) ! rad_net
1482	ts_value(dots_rad+1,sr) = hom(nzb,1,100,sr) ! rad_sw_in
1483	ENDIF
1484
1485	!
1486	!-- Calculate additional statistics provided by the user interface
1487	CALL user_statistics( 'time_series', sr, 0 )
1488
1489	ENDDO ! loop of the subregions
1490
1491	!
1492	!-- If required, sum up horizontal averages for subsequent time averaging
1493	IF ( do_sum ) THEN
1494	IF ( average_count_pr == 0 ) hom_sum = 0.0_wp
1495	hom_sum = hom_sum + hom(:,1,:,:)
1496	average_count_pr = average_count_pr + 1
1497	do_sum = .FALSE.
1498	ENDIF
1499
1500	!
1501	!-- Set flag for other UPs (e.g. output routines, but also buoyancy).
1502	!-- This flag is reset after each time step in time_integration.
1503	flow_statistics_called = .TRUE.
1504
1505	CALL cpu_log( log_point(10), 'flow_statistics', 'stop' )
1506
1507
1508	END SUBROUTINE flow_statistics
1509
1510
1511	#else
1512
1513
1514	!------------------------------------------------------------------------------!
1515	! flow statistics - accelerator version
1516	!------------------------------------------------------------------------------!
1517	SUBROUTINE flow_statistics
1518
1519	USE arrays_3d, &
1520	ONLY: ddzu, ddzw, e, hyp, km, kh, nr, p, prho, pt, q, qc, ql, qr, qs, &
1521	qsws, qswst, rho, sa, saswsb, saswst, shf, td_lsa_lpt, td_lsa_q,&
1522	td_sub_lpt, td_sub_q, time_vert, ts, tswst, u, ug, us, usws, &
1523	uswst, vsws, v, vg, vpt, vswst, w, w_subs, zw
1524
1525
1526	USE cloud_parameters, &
1527	ONLY: l_d_cp, prr, pt_d_t
1528
1529	USE control_parameters, &
1530	ONLY : average_count_pr, cloud_droplets, cloud_physics, do_sum, &
1531	dt_3d, g, humidity, icloud_scheme, kappa, large_scale_forcing, &
1532	large_scale_subsidence, max_pr_user, message_string, &
1533	monotonic_adjustment, ocean, &
1534	passive_scalar, precipitation, simulated_time, &
1535	use_subsidence_tendencies, use_surface_fluxes, use_top_fluxes, &
1536	ws_scheme_mom, ws_scheme_sca
1537
1538	USE cpulog, &
1539	ONLY: cpu_log, log_point
1540
1541	USE grid_variables, &
1542	ONLY: ddx, ddy
1543
1544	USE indices, &
1545	ONLY: ngp_2dh, ngp_2dh_s_inner, ngp_3d, ngp_3d_inner, ngp_sums, &
1546	ngp_sums_ls, nxl, nxr, nyn, nys, nzb, nzb_diff_s_inner, &
1547	nzb_s_inner, nzt, nzt_diff, rflags_invers
1548
1549	USE kinds
1550
1551	USE pegrid
1552
1553	USE statistics
1554
1555	IMPLICIT NONE
1556
1557	INTEGER(iwp) :: i !:
1558	INTEGER(iwp) :: j !:
1559	INTEGER(iwp) :: k !:
1560	INTEGER(iwp) :: nt !:
1561	INTEGER(iwp) :: omp_get_thread_num !:
1562	INTEGER(iwp) :: sr !:
1563	INTEGER(iwp) :: tn !:
1564
1565	LOGICAL :: first !:
1566
1567	REAL(wp) :: dptdz_threshold !:
1568	REAL(wp) :: fac !:
1569	REAL(wp) :: height !:
1570	REAL(wp) :: pts !:
1571	REAL(wp) :: sums_l_eper !:
1572	REAL(wp) :: sums_l_etot !:
1573	REAL(wp) :: s1 !:
1574	REAL(wp) :: s2 !:
1575	REAL(wp) :: s3 !:
1576	REAL(wp) :: s4 !:
1577	REAL(wp) :: s5 !:
1578	REAL(wp) :: s6 !:
1579	REAL(wp) :: s7 !:
1580	REAL(wp) :: ust !:
1581	REAL(wp) :: ust2 !:
1582	REAL(wp) :: u2 !:
1583	REAL(wp) :: vst !:
1584	REAL(wp) :: vst2 !:
1585	REAL(wp) :: v2 !:
1586	REAL(wp) :: w2 !:
1587	REAL(wp) :: z_i(2) !:
1588
1589	REAL(wp) :: dptdz(nzb+1:nzt+1) !:
1590	REAL(wp) :: sums_ll(nzb:nzt+1,2) !:
1591
1592	CALL cpu_log( log_point(10), 'flow_statistics', 'start' )
1593
1594	!
1595	!-- To be on the safe side, check whether flow_statistics has already been
1596	!-- called once after the current time step
1597	IF ( flow_statistics_called ) THEN
1598
1599	message_string = 'flow_statistics is called two times within one ' // &
1600	'timestep'
1601	CALL message( 'flow_statistics', 'PA0190', 1, 2, 0, 6, 0 )
1602
1603	ENDIF
1604
1605	!$acc data create( sums, sums_l )
1606	!$acc update device( hom )
1607
1608	!
1609	!-- Compute statistics for each (sub-)region
1610	DO sr = 0, statistic_regions
1611
1612	!
1613	!-- Initialize (local) summation array
1614	sums_l = 0.0_wp
1615
1616	!
1617	!-- Store sums that have been computed in other subroutines in summation
1618	!-- array
1619	sums_l(:,11,:) = sums_l_l(:,sr,:) ! mixing length from diffusivities
1620	!-- WARNING: next line still has to be adjusted for OpenMP
1621	sums_l(:,21,0) = sums_wsts_bc_l(:,sr) ! heat flux from advec_s_bc
1622	sums_l(nzb+9,pr_palm,0) = sums_divold_l(sr) ! old divergence from pres
1623	sums_l(nzb+10,pr_palm,0) = sums_divnew_l(sr) ! new divergence from pres
1624
1625	!
1626	!-- When calcuating horizontally-averaged total (resolved- plus subgrid-
1627	!-- scale) vertical fluxes and velocity variances by using commonly-
1628	!-- applied Reynolds-based methods ( e.g. <w'pt'> = (w-<w>)*(pt-<pt>) )
1629	!-- in combination with the 5th order advection scheme, pronounced
1630	!-- artificial kinks could be observed in the vertical profiles near the
1631	!-- surface. Please note: these kinks were not related to the model truth,
1632	!-- i.e. these kinks are just related to an evaluation problem.
1633	!-- In order avoid these kinks, vertical fluxes and horizontal as well
1634	!-- vertical velocity variances are calculated directly within the advection
1635	!-- routines, according to the numerical discretization, to evaluate the
1636	!-- statistical quantities as they will appear within the prognostic
1637	!-- equations.
1638	!-- Copy the turbulent quantities, evaluated in the advection routines to
1639	!-- the local array sums_l() for further computations.
1640	IF ( ws_scheme_mom .AND. sr == 0 ) THEN
1641
1642	!
1643	!-- According to the Neumann bc for the horizontal velocity components,
1644	!-- the corresponding fluxes has to satisfiy the same bc.
1645	IF ( ocean ) THEN
1646	sums_us2_ws_l(nzt+1,:) = sums_us2_ws_l(nzt,:)
1647	sums_vs2_ws_l(nzt+1,:) = sums_vs2_ws_l(nzt,:)
1648	ENDIF
1649
1650	DO i = 0, threads_per_task-1
1651	!
1652	!-- Swap the turbulent quantities evaluated in advec_ws.
1653	sums_l(:,13,i) = sums_wsus_ws_l(:,i) ! wu
1654	sums_l(:,15,i) = sums_wsvs_ws_l(:,i) ! wv
1655	sums_l(:,30,i) = sums_us2_ws_l(:,i) ! u*2
1656	sums_l(:,31,i) = sums_vs2_ws_l(:,i) ! v*2
1657	sums_l(:,32,i) = sums_ws2_ws_l(:,i) ! w*2
1658	sums_l(:,34,i) = sums_l(:,34,i) + 0.5_wp * &
1659	( sums_us2_ws_l(:,i) + sums_vs2_ws_l(:,i) + &
1660	sums_ws2_ws_l(:,i) ) ! e*
1661	DO k = nzb, nzt
1662	sums_l(nzb+5,pr_palm,i) = sums_l(nzb+5,pr_palm,i) + 0.5_wp * ( &
1663	sums_us2_ws_l(k,i) + &
1664	sums_vs2_ws_l(k,i) + &
1665	sums_ws2_ws_l(k,i) )
1666	ENDDO
1667	ENDDO
1668
1669	ENDIF
1670
1671	IF ( ws_scheme_sca .AND. .NOT. monotonic_adjustment .AND. sr == 0 ) THEN
1672
1673	DO i = 0, threads_per_task-1
1674	sums_l(:,17,i) = sums_wspts_ws_l(:,i) ! wpt from advec_s_ws
1675	IF ( ocean ) sums_l(:,66,i) = sums_wssas_ws_l(:,i) ! wsa
1676	IF ( humidity .OR. passive_scalar ) sums_l(:,49,i) = &
1677	sums_wsqs_ws_l(:,i) !wq
1678	ENDDO
1679
1680	ENDIF
1681	!
1682	!-- Horizontally averaged profiles of horizontal velocities and temperature.
1683	!-- They must have been computed before, because they are already required
1684	!-- for other horizontal averages.
1685	tn = 0
1686
1687	!$OMP PARALLEL PRIVATE( i, j, k, tn )
1688	#if defined( __intel_openmp_bug )
1689	tn = omp_get_thread_num()
1690	#else
1691	!$ tn = omp_get_thread_num()
1692	#endif
1693
1694	!$acc update device( sums_l )
1695
1696	!$OMP DO
1697	!$acc parallel loop gang present( pt, rflags_invers, rmask, sums_l, u, v ) create( s1, s2, s3 )
1698	DO k = nzb, nzt+1
1699	!$acc loop vector collapse( 2 ) reduction( +: s1, s2, s3 )
1700	DO i = nxl, nxr
1701	DO j = nys, nyn
1702	!
1703	!-- k+1 is used in rflags since rflags is set 0 at surface points
1704	s1 = s1 + u(k,j,i) * rmask(j,i,sr) * rflags_invers(j,i,k+1)
1705	s2 = s2 + v(k,j,i) * rmask(j,i,sr) * rflags_invers(j,i,k+1)
1706	s3 = s3 + pt(k,j,i) * rmask(j,i,sr) * rflags_invers(j,i,k+1)
1707	ENDDO
1708	ENDDO
1709	sums_l(k,1,tn) = s1
1710	sums_l(k,2,tn) = s2
1711	sums_l(k,4,tn) = s3
1712	ENDDO
1713	!$acc end parallel loop
1714
1715	!
1716	!-- Horizontally averaged profile of salinity
1717	IF ( ocean ) THEN
1718	!$OMP DO
1719	!$acc parallel loop gang present( rflags_invers, rmask, sums_l, sa ) create( s1 )
1720	DO k = nzb, nzt+1
1721	!$acc loop vector collapse( 2 ) reduction( +: s1 )
1722	DO i = nxl, nxr
1723	DO j = nys, nyn
1724	s1 = s1 + sa(k,j,i) * rmask(j,i,sr) * rflags_invers(j,i,k+1)
1725	ENDDO
1726	ENDDO
1727	sums_l(k,23,tn) = s1
1728	ENDDO
1729	!$acc end parallel loop
1730	ENDIF
1731
1732	!
1733	!-- Horizontally averaged profiles of virtual potential temperature,
1734	!-- total water content, specific humidity and liquid water potential
1735	!-- temperature
1736	IF ( humidity ) THEN
1737
1738	!$OMP DO
1739	!$acc parallel loop gang present( q, rflags_invers, rmask, sums_l, vpt ) create( s1, s2 )
1740	DO k = nzb, nzt+1
1741	!$acc loop vector collapse( 2 ) reduction( +: s1, s2 )
1742	DO i = nxl, nxr
1743	DO j = nys, nyn
1744	s1 = s1 + q(k,j,i) * rmask(j,i,sr) * rflags_invers(j,i,k+1)
1745	s2 = s2 + vpt(k,j,i) * rmask(j,i,sr) * rflags_invers(j,i,k+1)
1746	ENDDO
1747	ENDDO
1748	sums_l(k,41,tn) = s1
1749	sums_l(k,44,tn) = s2
1750	ENDDO
1751	!$acc end parallel loop
1752
1753	IF ( cloud_physics ) THEN
1754	!$OMP DO
1755	!$acc parallel loop gang present( pt, q, ql, rflags_invers, rmask, sums_l ) create( s1, s2 )
1756	DO k = nzb, nzt+1
1757	!$acc loop vector collapse( 2 ) reduction( +: s1, s2 )
1758	DO i = nxl, nxr
1759	DO j = nys, nyn
1760	s1 = s1 + ( q(k,j,i) - ql(k,j,i) ) * &
1761	rmask(j,i,sr) * rflags_invers(j,i,k+1)
1762	s2 = s2 + ( pt(k,j,i) + l_d_cppt_d_t(k) ql(k,j,i) ) * &
1763	rmask(j,i,sr) * rflags_invers(j,i,k+1)
1764	ENDDO
1765	ENDDO
1766	sums_l(k,42,tn) = s1
1767	sums_l(k,43,tn) = s2
1768	ENDDO
1769	!$acc end parallel loop
1770	ENDIF
1771	ENDIF
1772
1773	!
1774	!-- Horizontally averaged profiles of passive scalar
1775	IF ( passive_scalar ) THEN
1776	!$OMP DO
1777	!$acc parallel loop gang present( q, rflags_invers, rmask, sums_l ) create( s1 )
1778	DO k = nzb, nzt+1
1779	!$acc loop vector collapse( 2 ) reduction( +: s1 )
1780	DO i = nxl, nxr
1781	DO j = nys, nyn
1782	s1 = s1 + q(k,j,i) * rmask(j,i,sr) * rflags_invers(j,i,k+1)
1783	ENDDO
1784	ENDDO
1785	sums_l(k,41,tn) = s1
1786	ENDDO
1787	!$acc end parallel loop
1788	ENDIF
1789	!$OMP END PARALLEL
1790
1791	!
1792	!-- Summation of thread sums
1793	IF ( threads_per_task > 1 ) THEN
1794	DO i = 1, threads_per_task-1
1795	!$acc parallel present( sums_l )
1796	sums_l(:,1,0) = sums_l(:,1,0) + sums_l(:,1,i)
1797	sums_l(:,2,0) = sums_l(:,2,0) + sums_l(:,2,i)
1798	sums_l(:,4,0) = sums_l(:,4,0) + sums_l(:,4,i)
1799	!$acc end parallel
1800	IF ( ocean ) THEN
1801	!$acc parallel present( sums_l )
1802	sums_l(:,23,0) = sums_l(:,23,0) + sums_l(:,23,i)
1803	!$acc end parallel
1804	ENDIF
1805	IF ( humidity ) THEN
1806	!$acc parallel present( sums_l )
1807	sums_l(:,41,0) = sums_l(:,41,0) + sums_l(:,41,i)
1808	sums_l(:,44,0) = sums_l(:,44,0) + sums_l(:,44,i)
1809	!$acc end parallel
1810	IF ( cloud_physics ) THEN
1811	!$acc parallel present( sums_l )
1812	sums_l(:,42,0) = sums_l(:,42,0) + sums_l(:,42,i)
1813	sums_l(:,43,0) = sums_l(:,43,0) + sums_l(:,43,i)
1814	!$acc end parallel
1815	ENDIF
1816	ENDIF
1817	IF ( passive_scalar ) THEN
1818	!$acc parallel present( sums_l )
1819	sums_l(:,41,0) = sums_l(:,41,0) + sums_l(:,41,i)
1820	!$acc end parallel
1821	ENDIF
1822	ENDDO
1823	ENDIF
1824
1825	#if defined( __parallel )
1826	!
1827	!-- Compute total sum from local sums
1828	!$acc update host( sums_l )
1829	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
1830	CALL MPI_ALLREDUCE( sums_l(nzb,1,0), sums(nzb,1), nzt+2-nzb, MPI_REAL, &
1831	MPI_SUM, comm2d, ierr )
1832	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
1833	CALL MPI_ALLREDUCE( sums_l(nzb,2,0), sums(nzb,2), nzt+2-nzb, MPI_REAL, &
1834	MPI_SUM, comm2d, ierr )
1835	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
1836	CALL MPI_ALLREDUCE( sums_l(nzb,4,0), sums(nzb,4), nzt+2-nzb, MPI_REAL, &
1837	MPI_SUM, comm2d, ierr )
1838	IF ( ocean ) THEN
1839	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
1840	CALL MPI_ALLREDUCE( sums_l(nzb,23,0), sums(nzb,23), nzt+2-nzb, &
1841	MPI_REAL, MPI_SUM, comm2d, ierr )
1842	ENDIF
1843	IF ( humidity ) THEN
1844	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
1845	CALL MPI_ALLREDUCE( sums_l(nzb,44,0), sums(nzb,44), nzt+2-nzb, &
1846	MPI_REAL, MPI_SUM, comm2d, ierr )
1847	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
1848	CALL MPI_ALLREDUCE( sums_l(nzb,41,0), sums(nzb,41), nzt+2-nzb, &
1849	MPI_REAL, MPI_SUM, comm2d, ierr )
1850	IF ( cloud_physics ) THEN
1851	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
1852	CALL MPI_ALLREDUCE( sums_l(nzb,42,0), sums(nzb,42), nzt+2-nzb, &
1853	MPI_REAL, MPI_SUM, comm2d, ierr )
1854	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
1855	CALL MPI_ALLREDUCE( sums_l(nzb,43,0), sums(nzb,43), nzt+2-nzb, &
1856	MPI_REAL, MPI_SUM, comm2d, ierr )
1857	ENDIF
1858	ENDIF
1859
1860	IF ( passive_scalar ) THEN
1861	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
1862	CALL MPI_ALLREDUCE( sums_l(nzb,41,0), sums(nzb,41), nzt+2-nzb, &
1863	MPI_REAL, MPI_SUM, comm2d, ierr )
1864	ENDIF
1865	!$acc update device( sums )
1866	#else
1867	!$acc parallel present( sums, sums_l )
1868	sums(:,1) = sums_l(:,1,0)
1869	sums(:,2) = sums_l(:,2,0)
1870	sums(:,4) = sums_l(:,4,0)
1871	!$acc end parallel
1872	IF ( ocean ) THEN
1873	!$acc parallel present( sums, sums_l )
1874	sums(:,23) = sums_l(:,23,0)
1875	!$acc end parallel
1876	ENDIF
1877	IF ( humidity ) THEN
1878	!$acc parallel present( sums, sums_l )
1879	sums(:,44) = sums_l(:,44,0)
1880	sums(:,41) = sums_l(:,41,0)
1881	!$acc end parallel
1882	IF ( cloud_physics ) THEN
1883	!$acc parallel present( sums, sums_l )
1884	sums(:,42) = sums_l(:,42,0)
1885	sums(:,43) = sums_l(:,43,0)
1886	!$acc end parallel
1887	ENDIF
1888	ENDIF
1889	IF ( passive_scalar ) THEN
1890	!$acc parallel present( sums, sums_l )
1891	sums(:,41) = sums_l(:,41,0)
1892	!$acc end parallel
1893	ENDIF
1894	#endif
1895
1896	!
1897	!-- Final values are obtained by division by the total number of grid points
1898	!-- used for summation. After that store profiles.
1899	!$acc parallel present( hom, ngp_2dh, ngp_2dh_s_inner, sums )
1900	sums(:,1) = sums(:,1) / ngp_2dh(sr)
1901	sums(:,2) = sums(:,2) / ngp_2dh(sr)
1902	sums(:,4) = sums(:,4) / ngp_2dh_s_inner(:,sr)
1903	hom(:,1,1,sr) = sums(:,1) ! u
1904	hom(:,1,2,sr) = sums(:,2) ! v
1905	hom(:,1,4,sr) = sums(:,4) ! pt
1906	!$acc end parallel
1907
1908	!
1909	!-- Salinity
1910	IF ( ocean ) THEN
1911	!$acc parallel present( hom, ngp_2dh_s_inner, sums )
1912	sums(:,23) = sums(:,23) / ngp_2dh_s_inner(:,sr)
1913	hom(:,1,23,sr) = sums(:,23) ! sa
1914	!$acc end parallel
1915	ENDIF
1916
1917	!
1918	!-- Humidity and cloud parameters
1919	IF ( humidity ) THEN
1920	!$acc parallel present( hom, ngp_2dh_s_inner, sums )
1921	sums(:,44) = sums(:,44) / ngp_2dh_s_inner(:,sr)
1922	sums(:,41) = sums(:,41) / ngp_2dh_s_inner(:,sr)
1923	hom(:,1,44,sr) = sums(:,44) ! vpt
1924	hom(:,1,41,sr) = sums(:,41) ! qv (q)
1925	!$acc end parallel
1926	IF ( cloud_physics ) THEN
1927	!$acc parallel present( hom, ngp_2dh_s_inner, sums )
1928	sums(:,42) = sums(:,42) / ngp_2dh_s_inner(:,sr)
1929	sums(:,43) = sums(:,43) / ngp_2dh_s_inner(:,sr)
1930	hom(:,1,42,sr) = sums(:,42) ! qv
1931	hom(:,1,43,sr) = sums(:,43) ! pt
1932	!$acc end parallel
1933	ENDIF
1934	ENDIF
1935
1936	!
1937	!-- Passive scalar
1938	IF ( passive_scalar ) THEN
1939	!$acc parallel present( hom, ngp_2dh_s_inner, sums )
1940	sums(:,41) = sums(:,41) / ngp_2dh_s_inner(:,sr)
1941	hom(:,1,41,sr) = sums(:,41) ! s (q)
1942	!$acc end parallel
1943	ENDIF
1944
1945	!
1946	!-- Horizontally averaged profiles of the remaining prognostic variables,
1947	!-- variances, the total and the perturbation energy (single values in last
1948	!-- column of sums_l) and some diagnostic quantities.
1949	!-- NOTE: for simplicity, nzb_s_inner is used below, although strictly
1950	!-- ---- speaking the following k-loop would have to be split up and
1951	!-- rearranged according to the staggered grid.
1952	!-- However, this implies no error since staggered velocity components
1953	!-- are zero at the walls and inside buildings.
1954	tn = 0
1955	#if defined( __intel_openmp_bug )
1956	!$OMP PARALLEL PRIVATE( i, j, k, pts, sums_ll, sums_l_eper, sums_l_etot, &
1957	!$OMP tn, ust, ust2, u2, vst, vst2, v2, w2 )
1958	tn = omp_get_thread_num()
1959	#else
1960	!$OMP PARALLEL PRIVATE( i, j, k, pts, sums_ll, sums_l_eper, sums_l_etot, tn, ust, ust2, u2, vst, vst2, v2, w2 )
1961	!$ tn = omp_get_thread_num()
1962	#endif
1963	!$OMP DO
1964	!$acc parallel loop gang present( e, hom, kh, km, p, pt, w, rflags_invers, rmask, sums_l ) create( s1, s2, s3, s4, s5, s6, s7 )
1965	DO k = nzb, nzt+1
1966	!$acc loop vector collapse( 2 ) reduction( +: s1, s2, s3, s4, s5, s6, s7 )
1967	DO i = nxl, nxr
1968	DO j = nys, nyn
1969	!
1970	!-- Prognostic and diagnostic variables
1971	s1 = s1 + w(k,j,i) * rmask(j,i,sr) * rflags_invers(j,i,k+1)
1972	s2 = s2 + e(k,j,i) * rmask(j,i,sr) * rflags_invers(j,i,k+1)
1973	s3 = s3 + km(k,j,i) * rmask(j,i,sr) * rflags_invers(j,i,k+1)
1974	s4 = s4 + kh(k,j,i) * rmask(j,i,sr) * rflags_invers(j,i,k+1)
1975	s5 = s5 + p(k,j,i) * rmask(j,i,sr) * rflags_invers(j,i,k+1)
1976	s6 = s6 + ( pt(k,j,i)-hom(k,1,4,sr) )*2 rmask(j,i,sr) * &
1977	rflags_invers(j,i,k+1)
1978	!
1979	!-- Higher moments
1980	!-- (Computation of the skewness of w further below)
1981	s7 = s7 + w(k,j,i)*3 rmask(j,i,sr) * rflags_invers(j,i,k+1)
1982	ENDDO
1983	ENDDO
1984	sums_l(k,3,tn) = s1
1985	sums_l(k,8,tn) = s2
1986	sums_l(k,9,tn) = s3
1987	sums_l(k,10,tn) = s4
1988	sums_l(k,40,tn) = s5
1989	sums_l(k,33,tn) = s6
1990	sums_l(k,38,tn) = s7
1991	ENDDO
1992	!$acc end parallel loop
1993
1994	IF ( humidity ) THEN
1995	!$OMP DO
1996	!$acc parallel loop gang present( hom, q, rflags_invers, rmask, sums_l ) create( s1 )
1997	DO k = nzb, nzt+1
1998	!$acc loop vector collapse( 2 ) reduction( +: s1 )
1999	DO i = nxl, nxr
2000	DO j = nys, nyn
2001	s1 = s1 + ( q(k,j,i)-hom(k,1,41,sr) )*2 rmask(j,i,sr) * &
2002	rflags_invers(j,i,k+1)
2003	ENDDO
2004	ENDDO
2005	sums_l(k,70,tn) = s1
2006	ENDDO
2007	!$acc end parallel loop
2008	ENDIF
2009
2010	!
2011	!-- Total and perturbation energy for the total domain (being
2012	!-- collected in the last column of sums_l).
2013	!$OMP DO
2014	!$acc parallel loop collapse(3) present( rflags_invers, rmask, u, v, w ) reduction(+:s1)
2015	DO i = nxl, nxr
2016	DO j = nys, nyn
2017	DO k = nzb, nzt+1
2018	s1 = s1 + 0.5_wp * &
2019	( u(k,j,i)2 + v(k,j,i)2 + w(k,j,i)*2 ) &
2020	rmask(j,i,sr) * rflags_invers(j,i,k+1)
2021	ENDDO
2022	ENDDO
2023	ENDDO
2024	!$acc end parallel loop
2025	!$acc parallel present( sums_l )
2026	sums_l(nzb+4,pr_palm,tn) = s1
2027	!$acc end parallel
2028
2029	!$OMP DO
2030	!$acc parallel present( rmask, sums_l, us, usws, vsws, ts ) create( s1, s2, s3, s4 )
2031	!$acc loop vector collapse( 2 ) reduction( +: s1, s2, s3, s4 )
2032	DO i = nxl, nxr
2033	DO j = nys, nyn
2034	!
2035	!-- 2D-arrays (being collected in the last column of sums_l)
2036	s1 = s1 + us(j,i) * rmask(j,i,sr)
2037	s2 = s2 + usws(j,i) * rmask(j,i,sr)
2038	s3 = s3 + vsws(j,i) * rmask(j,i,sr)
2039	s4 = s4 + ts(j,i) * rmask(j,i,sr)
2040	ENDDO
2041	ENDDO
2042	sums_l(nzb,pr_palm,tn) = s1
2043	sums_l(nzb+1,pr_palm,tn) = s2
2044	sums_l(nzb+2,pr_palm,tn) = s3
2045	sums_l(nzb+3,pr_palm,tn) = s4
2046	!$acc end parallel
2047
2048	IF ( humidity ) THEN
2049	!$acc parallel present( qs, rmask, sums_l ) create( s1 )
2050	!$acc loop vector collapse( 2 ) reduction( +: s1 )
2051	DO i = nxl, nxr
2052	DO j = nys, nyn
2053	s1 = s1 + qs(j,i) * rmask(j,i,sr)
2054	ENDDO
2055	ENDDO
2056	sums_l(nzb+12,pr_palm,tn) = s1
2057	!$acc end parallel
2058	ENDIF
2059
2060	!
2061	!-- Computation of statistics when ws-scheme is not used. Else these
2062	!-- quantities are evaluated in the advection routines.
2063	IF ( .NOT. ws_scheme_mom .OR. sr /= 0 ) THEN
2064
2065	!$OMP DO
2066	!$acc parallel loop gang present( u, v, w, rflags_invers, rmask, sums_l ) create( s1, s2, s3, s4, ust2, vst2, w2 )
2067	DO k = nzb, nzt+1
2068	!$acc loop vector collapse( 2 ) reduction( +: s1, s2, s3, s4 )
2069	DO i = nxl, nxr
2070	DO j = nys, nyn
2071	ust2 = ( u(k,j,i) - hom(k,1,1,sr) )**2
2072	vst2 = ( v(k,j,i) - hom(k,1,2,sr) )**2
2073	w2 = w(k,j,i)**2
2074
2075	s1 = s1 + ust2 * rmask(j,i,sr) * rflags_invers(j,i,k+1)
2076	s2 = s2 + vst2 * rmask(j,i,sr) * rflags_invers(j,i,k+1)
2077	s3 = s3 + w2 * rmask(j,i,sr) * rflags_invers(j,i,k+1)
2078	!
2079	!-- Perturbation energy
2080	s4 = s4 + 0.5_wp * ( ust2 + vst2 + w2 ) * rmask(j,i,sr) * &
2081	rflags_invers(j,i,k+1)
2082	ENDDO
2083	ENDDO
2084	sums_l(k,30,tn) = s1
2085	sums_l(k,31,tn) = s2
2086	sums_l(k,32,tn) = s3
2087	sums_l(k,34,tn) = s4
2088	ENDDO
2089	!$acc end parallel loop
2090	!
2091	!-- Total perturbation TKE
2092	!$OMP DO
2093	!$acc parallel present( sums_l ) create( s1 )
2094	!$acc loop reduction( +: s1 )
2095	DO k = nzb, nzt+1
2096	s1 = s1 + sums_l(k,34,tn)
2097	ENDDO
2098	sums_l(nzb+5,pr_palm,tn) = s1
2099	!$acc end parallel
2100
2101	ENDIF
2102
2103	!
2104	!-- Horizontally averaged profiles of the vertical fluxes
2105
2106	!
2107	!-- Subgridscale fluxes.
2108	!-- WARNING: If a Prandtl-layer is used (k=nzb for flat terrain), the fluxes
2109	!-- ------- should be calculated there in a different way. This is done
2110	!-- in the next loop further below, where results from this loop are
2111	!-- overwritten. However, THIS WORKS IN CASE OF FLAT TERRAIN ONLY!
2112	!-- The non-flat case still has to be handled.
2113	!-- NOTE: for simplicity, nzb_s_inner is used below, although
2114	!-- ---- strictly speaking the following k-loop would have to be
2115	!-- split up according to the staggered grid.
2116	!-- However, this implies no error since staggered velocity
2117	!-- components are zero at the walls and inside buildings.
2118	!$OMP DO
2119	!$acc parallel loop gang present( ddzu, kh, km, pt, u, v, w, rflags_invers, rmask, sums_l ) create( s1, s2, s3 )
2120	DO k = nzb, nzt_diff
2121	!$acc loop vector collapse( 2 ) reduction( +: s1, s2, s3 )
2122	DO i = nxl, nxr
2123	DO j = nys, nyn
2124
2125	!
2126	!-- Momentum flux w"u"
2127	s1 = s1 - 0.25_wp * ( &
2128	km(k,j,i)+km(k+1,j,i)+km(k,j,i-1)+km(k+1,j,i-1) &
2129	) * ( &
2130	( u(k+1,j,i) - u(k,j,i) ) * ddzu(k+1) &
2131	+ ( w(k,j,i) - w(k,j,i-1) ) * ddx &
2132	) &
2133	* rmask(j,i,sr) * rflags_invers(j,i,k+1)
2134	!
2135	!-- Momentum flux w"v"
2136	s2 = s2 - 0.25_wp * ( &
2137	km(k,j,i)+km(k+1,j,i)+km(k,j-1,i)+km(k+1,j-1,i) &
2138	) * ( &
2139	( v(k+1,j,i) - v(k,j,i) ) * ddzu(k+1) &
2140	+ ( w(k,j,i) - w(k,j-1,i) ) * ddy &
2141	) &
2142	* rmask(j,i,sr) * rflags_invers(j,i,k+1)
2143	!
2144	!-- Heat flux w"pt"
2145	s3 = s3 - 0.5_wp * ( kh(k,j,i) + kh(k+1,j,i) ) &
2146	* ( pt(k+1,j,i) - pt(k,j,i) ) &
2147	* ddzu(k+1) * rmask(j,i,sr) &
2148	* rflags_invers(j,i,k+1)
2149	ENDDO
2150	ENDDO
2151	sums_l(k,12,tn) = s1
2152	sums_l(k,14,tn) = s2
2153	sums_l(k,16,tn) = s3
2154	ENDDO
2155	!$acc end parallel loop
2156
2157	!
2158	!-- Salinity flux w"sa"
2159	IF ( ocean ) THEN
2160	!$acc parallel loop gang present( ddzu, kh, sa, rflags_invers, rmask, sums_l ) create( s1 )
2161	DO k = nzb, nzt_diff
2162	!$acc loop vector collapse( 2 ) reduction( +: s1 )
2163	DO i = nxl, nxr
2164	DO j = nys, nyn
2165	s1 = s1 - 0.5_wp * ( kh(k,j,i) + kh(k+1,j,i) ) &
2166	* ( sa(k+1,j,i) - sa(k,j,i) ) &
2167	* ddzu(k+1) * rmask(j,i,sr) &
2168	* rflags_invers(j,i,k+1)
2169	ENDDO
2170	ENDDO
2171	sums_l(k,65,tn) = s1
2172	ENDDO
2173	!$acc end parallel loop
2174	ENDIF
2175
2176	!
2177	!-- Buoyancy flux, water flux (humidity flux) w"q"
2178	IF ( humidity ) THEN
2179
2180	!$acc parallel loop gang present( ddzu, kh, q, vpt, rflags_invers, rmask, sums_l ) create( s1, s2 )
2181	DO k = nzb, nzt_diff
2182	!$acc loop vector collapse( 2 ) reduction( +: s1, s2 )
2183	DO i = nxl, nxr
2184	DO j = nys, nyn
2185	s1 = s1 - 0.5_wp * ( kh(k,j,i) + kh(k+1,j,i) ) &
2186	* ( vpt(k+1,j,i) - vpt(k,j,i) ) &
2187	* ddzu(k+1) * rmask(j,i,sr) &
2188	* rflags_invers(j,i,k+1)
2189	s2 = s2 - 0.5_wp * ( kh(k,j,i) + kh(k+1,j,i) ) &
2190	* ( q(k+1,j,i) - q(k,j,i) ) &
2191	* ddzu(k+1) * rmask(j,i,sr) &
2192	* rflags_invers(j,i,k+1)
2193	ENDDO
2194	ENDDO
2195	sums_l(k,45,tn) = s1
2196	sums_l(k,48,tn) = s2
2197	ENDDO
2198	!$acc end parallel loop
2199
2200	IF ( cloud_physics ) THEN
2201
2202	!$acc parallel loop gang present( ddzu, kh, q, ql, rflags_invers, rmask, sums_l ) create( s1 )
2203	DO k = nzb, nzt_diff
2204	!$acc loop vector collapse( 2 ) reduction( +: s1 )
2205	DO i = nxl, nxr
2206	DO j = nys, nyn
2207	s1 = s1 - 0.5_wp * ( kh(k,j,i) + kh(k+1,j,i) ) &
2208	* ( ( q(k+1,j,i) - ql(k+1,j,i) ) &
2209	- ( q(k,j,i) - ql(k,j,i) ) ) &
2210	* ddzu(k+1) * rmask(j,i,sr) &
2211	* rflags_invers(j,i,k+1)
2212	ENDDO
2213	ENDDO
2214	sums_l(k,51,tn) = s1
2215	ENDDO
2216	!$acc end parallel loop
2217
2218	ENDIF
2219
2220	ENDIF
2221	!
2222	!-- Passive scalar flux
2223	IF ( passive_scalar ) THEN
2224
2225	!$acc parallel loop gang present( ddzu, kh, q, rflags_invers, rmask, sums_l ) create( s1 )
2226	DO k = nzb, nzt_diff
2227	!$acc loop vector collapse( 2 ) reduction( +: s1 )
2228	DO i = nxl, nxr
2229	DO j = nys, nyn
2230	s1 = s1 - 0.5_wp * ( kh(k,j,i) + kh(k+1,j,i) ) &
2231	* ( q(k+1,j,i) - q(k,j,i) ) &
2232	* ddzu(k+1) * rmask(j,i,sr) &
2233	* rflags_invers(j,i,k+1)
2234	ENDDO
2235	ENDDO
2236	sums_l(k,48,tn) = s1
2237	ENDDO
2238	!$acc end parallel loop
2239
2240	ENDIF
2241
2242	IF ( use_surface_fluxes ) THEN
2243
2244	!$OMP DO
2245	!$acc parallel present( rmask, shf, sums_l, usws, vsws ) create( s1, s2, s3, s4, s5 )
2246	!$acc loop vector collapse( 2 ) reduction( +: s1, s2, s3, s4, s5 )
2247	DO i = nxl, nxr
2248	DO j = nys, nyn
2249	!
2250	!-- Subgridscale fluxes in the Prandtl layer
2251	s1 = s1 + usws(j,i) * rmask(j,i,sr) ! w"u"
2252	s2 = s2 + vsws(j,i) * rmask(j,i,sr) ! w"v"
2253	s3 = s3 + shf(j,i) * rmask(j,i,sr) ! w"pt"
2254	s4 = s4 + 0.0_wp * rmask(j,i,sr) ! u"pt"
2255	s5 = s5 + 0.0_wp * rmask(j,i,sr) ! v"pt"
2256	ENDDO
2257	ENDDO
2258	sums_l(nzb,12,tn) = s1
2259	sums_l(nzb,14,tn) = s2
2260	sums_l(nzb,16,tn) = s3
2261	sums_l(nzb,58,tn) = s4
2262	sums_l(nzb,61,tn) = s5
2263	!$acc end parallel
2264
2265	IF ( ocean ) THEN
2266
2267	!$OMP DO
2268	!$acc parallel present( rmask, saswsb, sums_l ) create( s1 )
2269	!$acc loop vector collapse( 2 ) reduction( +: s1 )
2270	DO i = nxl, nxr
2271	DO j = nys, nyn
2272	s1 = s1 + saswsb(j,i) * rmask(j,i,sr) ! w"sa"
2273	ENDDO
2274	ENDDO
2275	sums_l(nzb,65,tn) = s1
2276	!$acc end parallel
2277
2278	ENDIF
2279
2280	IF ( humidity ) THEN
2281
2282	!$OMP DO
2283	!$acc parallel present( pt, q, qsws, rmask, shf, sums_l ) create( s1, s2 )
2284	!$acc loop vector collapse( 2 ) reduction( +: s1, s2 )
2285	DO i = nxl, nxr
2286	DO j = nys, nyn
2287	s1 = s1 + qsws(j,i) * rmask(j,i,sr) ! w"q" (w"qv")
2288	s2 = s2 + ( ( 1.0_wp + 0.61_wp * q(nzb,j,i) ) * shf(j,i) &
2289	+ 0.61_wp * pt(nzb,j,i) * qsws(j,i) )
2290	ENDDO
2291	ENDDO
2292	sums_l(nzb,48,tn) = s1
2293	sums_l(nzb,45,tn) = s2
2294	!$acc end parallel
2295
2296	IF ( cloud_droplets ) THEN
2297
2298	!$OMP DO
2299	!$acc parallel present( pt, q, ql, qsws, rmask, shf, sums_l ) create( s1 )
2300	!$acc loop vector collapse( 2 ) reduction( +: s1 )
2301	DO i = nxl, nxr
2302	DO j = nys, nyn
2303	s1 = s1 + ( ( 1.0_wp + &
2304	0.61_wp * q(nzb,j,i) - ql(nzb,j,i) ) * &
2305	shf(j,i) + 0.61_wp * pt(nzb,j,i) * qsws(j,i) )
2306	ENDDO
2307	ENDDO
2308	sums_l(nzb,45,tn) = s1
2309	!$acc end parallel
2310
2311	ENDIF
2312
2313	IF ( cloud_physics ) THEN
2314
2315	!$OMP DO
2316	!$acc parallel present( qsws, rmask, sums_l ) create( s1 )
2317	!$acc loop vector collapse( 2 ) reduction( +: s1 )
2318	DO i = nxl, nxr
2319	DO j = nys, nyn
2320	!
2321	!-- Formula does not work if ql(nzb) /= 0.0
2322	s1 = s1 + qsws(j,i) * rmask(j,i,sr) ! w"q" (w"qv")
2323	ENDDO
2324	ENDDO
2325	sums_l(nzb,51,tn) = s1
2326	!$acc end parallel
2327
2328	ENDIF
2329
2330	ENDIF
2331
2332	IF ( passive_scalar ) THEN
2333
2334	!$OMP DO
2335	!$acc parallel present( qsws, rmask, sums_l ) create( s1 )
2336	!$acc loop vector collapse( 2 ) reduction( +: s1 )
2337	DO i = nxl, nxr
2338	DO j = nys, nyn
2339	s1 = s1 + qsws(j,i) * rmask(j,i,sr) ! w"q" (w"qv")
2340	ENDDO
2341	ENDDO
2342	sums_l(nzb,48,tn) = s1
2343	!$acc end parallel
2344
2345	ENDIF
2346
2347	ENDIF
2348
2349	!
2350	!-- Subgridscale fluxes at the top surface
2351	IF ( use_top_fluxes ) THEN
2352
2353	!$OMP DO
2354	!$acc parallel present( rmask, sums_l, tswst, uswst, vswst ) create( s1, s2, s3, s4, s5 )
2355	!$acc loop vector collapse( 2 ) reduction( +: s1, s2, s3, s4, s5 )
2356	DO i = nxl, nxr
2357	DO j = nys, nyn
2358	s1 = s1 + uswst(j,i) * rmask(j,i,sr) ! w"u"
2359	s2 = s2 + vswst(j,i) * rmask(j,i,sr) ! w"v"
2360	s3 = s3 + tswst(j,i) * rmask(j,i,sr) ! w"pt"
2361	s4 = s4 + 0.0_wp * rmask(j,i,sr) ! u"pt"
2362	s5 = s5 + 0.0_wp * rmask(j,i,sr) ! v"pt"
2363	ENDDO
2364	ENDDO
2365	sums_l(nzt:nzt+1,12,tn) = s1
2366	sums_l(nzt:nzt+1,14,tn) = s2
2367	sums_l(nzt:nzt+1,16,tn) = s3
2368	sums_l(nzt:nzt+1,58,tn) = s4
2369	sums_l(nzt:nzt+1,61,tn) = s5
2370	!$acc end parallel
2371
2372	IF ( ocean ) THEN
2373
2374	!$OMP DO
2375	!$acc parallel present( rmask, saswst, sums_l ) create( s1 )
2376	!$acc loop vector collapse( 2 ) reduction( +: s1 )
2377	DO i = nxl, nxr
2378	DO j = nys, nyn
2379	s1 = s1 + saswst(j,i) * rmask(j,i,sr) ! w"sa"
2380	ENDDO
2381	ENDDO
2382	sums_l(nzt,65,tn) = s1
2383	!$acc end parallel
2384
2385	ENDIF
2386
2387	IF ( humidity ) THEN
2388
2389	!$OMP DO
2390	!$acc parallel present( pt, q, qswst, rmask, tswst, sums_l ) create( s1, s2 )
2391	!$acc loop vector collapse( 2 ) reduction( +: s1, s2 )
2392	DO i = nxl, nxr
2393	DO j = nys, nyn
2394	s1 = s1 + qswst(j,i) * rmask(j,i,sr) ! w"q" (w"qv")
2395	s2 = s2 + ( ( 1.0_wp + 0.61_wp * q(nzt,j,i) ) * tswst(j,i) +&
2396	0.61_wp * pt(nzt,j,i) * qswst(j,i) )
2397	ENDDO
2398	ENDDO
2399	sums_l(nzt,48,tn) = s1
2400	sums_l(nzt,45,tn) = s2
2401	!$acc end parallel
2402
2403	IF ( cloud_droplets ) THEN
2404
2405	!$OMP DO
2406	!$acc parallel present( pt, q, ql, qswst, rmask, tswst, sums_l ) create( s1 )
2407	!$acc loop vector collapse( 2 ) reduction( +: s1 )
2408	DO i = nxl, nxr
2409	DO j = nys, nyn
2410	s1 = s1 + ( ( 1.0_wp + &
2411	0.61_wp * q(nzt,j,i) - ql(nzt,j,i) ) * &
2412	tswst(j,i) + &
2413	0.61_wp * pt(nzt,j,i) * qswst(j,i) )
2414	ENDDO
2415	ENDDO
2416	sums_l(nzt,45,tn) = s1
2417	!$acc end parallel
2418
2419	ENDIF
2420
2421	IF ( cloud_physics ) THEN
2422
2423	!$OMP DO
2424	!$acc parallel present( qswst, rmask, sums_l ) create( s1 )
2425	!$acc loop vector collapse( 2 ) reduction( +: s1 )
2426	DO i = nxl, nxr
2427	DO j = nys, nyn
2428	!
2429	!-- Formula does not work if ql(nzb) /= 0.0
2430	s1 = s1 + qswst(j,i) * rmask(j,i,sr) ! w"q" (w"qv")
2431	ENDDO
2432	ENDDO
2433	sums_l(nzt,51,tn) = s1
2434	!$acc end parallel
2435
2436	ENDIF
2437
2438	ENDIF
2439
2440	IF ( passive_scalar ) THEN
2441
2442	!$OMP DO
2443	!$acc parallel present( qswst, rmask, sums_l ) create( s1 )
2444	!$acc loop vector collapse( 2 ) reduction( +: s1 )
2445	DO i = nxl, nxr
2446	DO j = nys, nyn
2447	s1 = s1 + qswst(j,i) * rmask(j,i,sr) ! w"q" (w"qv")
2448	ENDDO
2449	ENDDO
2450	sums_l(nzt,48,tn) = s1
2451	!$acc end parallel
2452
2453	ENDIF
2454
2455	ENDIF
2456
2457	!
2458	!-- Resolved fluxes (can be computed for all horizontal points)
2459	!-- NOTE: for simplicity, nzb_s_inner is used below, although strictly
2460	!-- ---- speaking the following k-loop would have to be split up and
2461	!-- rearranged according to the staggered grid.
2462	!$acc parallel loop gang present( hom, pt, rflags_invers, rmask, sums_l, u, v, w ) create( s1, s2, s3 )
2463	DO k = nzb, nzt_diff
2464	!$acc loop vector collapse( 2 ) reduction( +: s1, s2, s3 )
2465	DO i = nxl, nxr
2466	DO j = nys, nyn
2467	ust = 0.5_wp * ( u(k,j,i) - hom(k,1,1,sr) + &
2468	u(k+1,j,i) - hom(k+1,1,1,sr) )
2469	vst = 0.5_wp * ( v(k,j,i) - hom(k,1,2,sr) + &
2470	v(k+1,j,i) - hom(k+1,1,2,sr) )
2471	pts = 0.5_wp * ( pt(k,j,i) - hom(k,1,4,sr) + &
2472	pt(k+1,j,i) - hom(k+1,1,4,sr) )
2473	!
2474	!-- Higher moments
2475	s1 = s1 + pts * w(k,j,i)*2 rmask(j,i,sr) * rflags_invers(j,i,k+1)
2476	s2 = s2 + pts*2 w(k,j,i) * rmask(j,i,sr) * rflags_invers(j,i,k+1)
2477	!
2478	!-- Energy flux we (has to be adjusted?)
2479	s3 = s3 + w(k,j,i) * 0.5_wp * ( ust2 + vst2 + w(k,j,i)**2 )&
2480	* rmask(j,i,sr) * rflags_invers(j,i,k+1)
2481	ENDDO
2482	ENDDO
2483	sums_l(k,35,tn) = s1
2484	sums_l(k,36,tn) = s2
2485	sums_l(k,37,tn) = s3
2486	ENDDO
2487	!$acc end parallel loop
2488
2489	!
2490	!-- Salinity flux and density (density does not belong to here,
2491	!-- but so far there is no other suitable place to calculate)
2492	IF ( ocean ) THEN
2493
2494	IF( .NOT. ws_scheme_sca .OR. monotonic_adjustment .OR. sr /= 0 ) THEN
2495
2496	!$acc parallel loop gang present( hom, rflags_invers, rmask, sa, sums_l, w ) create( s1 )
2497	DO k = nzb, nzt_diff
2498	!$acc loop vector collapse( 2 ) reduction( +: s1 )
2499	DO i = nxl, nxr
2500	DO j = nys, nyn
2501	s1 = s1 + 0.5_wp * ( sa(k,j,i) - hom(k,1,23,sr) + &
2502	sa(k+1,j,i) - hom(k+1,1,23,sr) ) &
2503	* w(k,j,i) * rmask(j,i,sr) &
2504	* rflags_invers(j,i,k+1)
2505	ENDDO
2506	ENDDO
2507	sums_l(k,66,tn) = s1
2508	ENDDO
2509	!$acc end parallel loop
2510
2511	ENDIF
2512
2513	!$acc parallel loop gang present( rflags_invers, rho, prho, rmask, sums_l ) create( s1, s2 )
2514	DO k = nzb, nzt_diff
2515	!$acc loop vector collapse( 2 ) reduction( +: s1, s2 )
2516	DO i = nxl, nxr
2517	DO j = nys, nyn
2518	s1 = s1 + rho(k,j,i) * rmask(j,i,sr) * rflags_invers(j,i,k+1)
2519	s2 = s2 + prho(k,j,i) * rmask(j,i,sr) * rflags_invers(j,i,k+1)
2520	ENDDO
2521	ENDDO
2522	sums_l(k,64,tn) = s1
2523	sums_l(k,71,tn) = s2
2524	ENDDO
2525	!$acc end parallel loop
2526
2527	ENDIF
2528
2529	!
2530	!-- Buoyancy flux, water flux, humidity flux, liquid water
2531	!-- content, rain drop concentration and rain water content
2532	IF ( humidity ) THEN
2533
2534	IF ( cloud_physics .OR. cloud_droplets ) THEN
2535
2536	!$acc parallel loop gang present( hom, rflags_invers, rmask, sums_l, vpt, w ) create( s1 )
2537	DO k = nzb, nzt_diff
2538	!$acc loop vector collapse( 2 ) reduction( +: s1 )
2539	DO i = nxl, nxr
2540	DO j = nys, nyn
2541	s1 = s1 + 0.5_wp * ( vpt(k,j,i) - hom(k,1,44,sr) + &
2542	vpt(k+1,j,i) - hom(k+1,1,44,sr) ) * &
2543	w(k,j,i) * rmask(j,i,sr) * rflags_invers(j,i,k+1)
2544	ENDDO
2545	ENDDO
2546	sums_l(k,46,tn) = s1
2547	ENDDO
2548	!$acc end parallel loop
2549
2550	IF ( .NOT. cloud_droplets ) THEN
2551
2552	!$acc parallel loop gang present( hom, q, ql, rflags_invers, rmask, sums_l, w ) create( s1 )
2553	DO k = nzb, nzt_diff
2554	!$acc loop vector collapse( 2 ) reduction( +: s1 )
2555	DO i = nxl, nxr
2556	DO j = nys, nyn
2557	s1 = s1 + 0.5_wp * ( ( q(k,j,i) - ql(k,j,i) ) - hom(k,1,42,sr) + &
2558	( q(k+1,j,i) - ql(k+1,j,i) ) - hom(k+1,1,42,sr) ) &
2559	* w(k,j,i) * rmask(j,i,sr) * rflags_invers(j,i,k+1)
2560	ENDDO
2561	ENDDO
2562	sums_l(k,52,tn) = s1
2563	ENDDO
2564	!$acc end parallel loop
2565
2566	IF ( icloud_scheme == 0 ) THEN
2567
2568	!$acc parallel loop gang present( qc, ql, rflags_invers, rmask, sums_l ) create( s1, s2 )
2569	DO k = nzb, nzt_diff
2570	!$acc loop vector collapse( 2 ) reduction( +: s1, s2 )
2571	DO i = nxl, nxr
2572	DO j = nys, nyn
2573	s1 = s1 + ql(k,j,i) * rmask(j,i,sr) * rflags_invers(j,i,k+1)
2574	s2 = s2 + qc(k,j,i) * rmask(j,i,sr) * rflags_invers(j,i,k+1)
2575	ENDDO
2576	ENDDO
2577	sums_l(k,54,tn) = s1
2578	sums_l(k,75,tn) = s2
2579	ENDDO
2580	!$acc end parallel loop
2581
2582	IF ( precipitation ) THEN
2583
2584	!$acc parallel loop gang present( nr, qr, prr, rflags_invers, rmask, sums_l ) create( s1, s2, s3 )
2585	DO k = nzb, nzt_diff
2586	!$acc loop vector collapse( 2 ) reduction( +: s1, s2, s3 )
2587	DO i = nxl, nxr
2588	DO j = nys, nyn
2589	s1 = s1 + nr(k,j,i) * rmask(j,i,sr) * rflags_invers(j,i,k+1)
2590	s2 = s2 + qr(k,j,i) * rmask(j,i,sr) * rflags_invers(j,i,k+1)
2591	s3 = s3 + prr(k,j,i) * rmask(j,i,sr) * rflags_invers(j,i,k+1)
2592	ENDDO
2593	ENDDO
2594	sums_l(k,73,tn) = s1
2595	sums_l(k,74,tn) = s2
2596	sums_l(k,76,tn) = s3
2597	ENDDO
2598	!$acc end parallel loop
2599
2600	ENDIF
2601
2602	ELSE
2603
2604	!$acc parallel loop gang present( ql, rflags_invers, rmask, sums_l ) create( s1 )
2605	DO k = nzb, nzt_diff
2606	!$acc loop vector collapse( 2 ) reduction( +: s1 )
2607	DO i = nxl, nxr
2608	DO j = nys, nyn
2609	s1 = s1 + ql(k,j,i) * rmask(j,i,sr) * rflags_invers(j,i,k+1)
2610	ENDDO
2611	ENDDO
2612	sums_l(k,54,tn) = s1
2613	ENDDO
2614	!$acc end parallel loop
2615
2616	ENDIF
2617
2618	ELSE
2619
2620	!$acc parallel loop gang present( ql, rflags_invers, rmask, sums_l ) create( s1 )
2621	DO k = nzb, nzt_diff
2622	!$acc loop vector collapse( 2 ) reduction( +: s1 )
2623	DO i = nxl, nxr
2624	DO j = nys, nyn
2625	s1 = s1 + ql(k,j,i) * rmask(j,i,sr) * rflags_invers(j,i,k+1)
2626	ENDDO
2627	ENDDO
2628	sums_l(k,54,tn) = s1
2629	ENDDO
2630	!$acc end parallel loop
2631
2632	ENDIF
2633
2634	ELSE
2635
2636	IF( .NOT. ws_scheme_sca .OR. monotonic_adjustment .OR. sr /= 0 ) THEN
2637
2638	!$acc parallel loop gang present( hom, rflags_invers, rmask, sums_l, vpt, w ) create( s1 )
2639	DO k = nzb, nzt_diff
2640	!$acc loop vector collapse( 2 ) reduction( +: s1 )
2641	DO i = nxl, nxr
2642	DO j = nys, nyn
2643	s1 = s1 + 0.5_wp * ( vpt(k,j,i) - hom(k,1,44,sr) + &
2644	vpt(k+1,j,i) - hom(k+1,1,44,sr) ) &
2645	* w(k,j,i) * rmask(j,i,sr) * rflags_invers(j,i,k+1)
2646	ENDDO
2647	ENDDO
2648	sums_l(k,46,tn) = s1
2649	ENDDO
2650	!$acc end parallel loop
2651
2652	ELSEIF ( ws_scheme_sca .AND. sr == 0 ) THEN
2653
2654	!$acc parallel loop present( hom, sums_l )
2655	DO k = nzb, nzt_diff
2656	sums_l(k,46,tn) = ( 1.0_wp + 0.61_wp * hom(k,1,41,sr) ) * sums_l(k,17,tn) + &
2657	0.61_wp * hom(k,1,4,sr) * sums_l(k,49,tn)
2658	ENDDO
2659	!$acc end parallel loop
2660
2661	ENDIF
2662
2663	ENDIF
2664
2665	ENDIF
2666	!
2667	!-- Passive scalar flux
2668	IF ( passive_scalar .AND. ( .NOT. ws_scheme_sca .OR. monotonic_adjustment &
2669	.OR. sr /= 0 ) ) THEN
2670
2671	!$acc parallel loop gang present( hom, q, rflags_invers, rmask, sums_l, w ) create( s1 )
2672	DO k = nzb, nzt_diff
2673	!$acc loop vector collapse( 2 ) reduction( +: s1 )
2674	DO i = nxl, nxr
2675	DO j = nys, nyn
2676	s1 = s1 + 0.5_wp * ( q(k,j,i) - hom(k,1,41,sr) + &
2677	q(k+1,j,i) - hom(k+1,1,41,sr) ) &
2678	* w(k,j,i) * rmask(j,i,sr) &
2679	* rflags_invers(j,i,k+1)
2680	ENDDO
2681	ENDDO
2682	sums_l(k,49,tn) = s1
2683	ENDDO
2684	!$acc end parallel loop
2685
2686	ENDIF
2687
2688	!
2689	!-- For speed optimization fluxes which have been computed in part directly
2690	!-- inside the WS advection routines are treated seperatly
2691	!-- Momentum fluxes first:
2692	IF ( .NOT. ws_scheme_mom .OR. sr /= 0 ) THEN
2693
2694	!$OMP DO
2695	!$acc parallel loop gang present( hom, rflags_invers, rmask, sums_l, u, v, w ) create( s1, s2 )
2696	DO k = nzb, nzt_diff
2697	!$acc loop vector collapse( 2 ) reduction( +: s1, s2 )
2698	DO i = nxl, nxr
2699	DO j = nys, nyn
2700	ust = 0.5_wp * ( u(k,j,i) - hom(k,1,1,sr) + &
2701	u(k+1,j,i) - hom(k+1,1,1,sr) )
2702	vst = 0.5_wp * ( v(k,j,i) - hom(k,1,2,sr) + &
2703	v(k+1,j,i) - hom(k+1,1,2,sr) )
2704	!
2705	!-- Momentum flux wu
2706	s1 = s1 + 0.5_wp * ( w(k,j,i-1) + w(k,j,i) ) &
2707	* ust * rmask(j,i,sr) &
2708	* rflags_invers(j,i,k+1)
2709	!
2710	!-- Momentum flux wv
2711	s2 = s2 + 0.5_wp * ( w(k,j-1,i) + w(k,j,i) ) &
2712	* vst * rmask(j,i,sr) &
2713	* rflags_invers(j,i,k+1)
2714	ENDDO
2715	ENDDO
2716	sums_l(k,13,tn) = s1
2717	sums_l(k,15,tn) = s1
2718	ENDDO
2719	!$acc end parallel loop
2720
2721	ENDIF
2722
2723	IF ( .NOT. ws_scheme_sca .OR. monotonic_adjustment .OR. sr /= 0 ) THEN
2724
2725	!$OMP DO
2726	!$acc parallel loop gang present( hom, pt, rflags_invers, rmask, sums_l, w ) create( s1 )
2727	DO k = nzb, nzt_diff
2728	!$acc loop vector collapse( 2 ) reduction( +: s1 )
2729	DO i = nxl, nxr
2730	DO j = nys, nyn
2731	!
2732	!-- Vertical heat flux
2733	s1 = s1 + 0.5_wp * ( pt(k,j,i) - hom(k,1,4,sr) + &
2734	pt(k+1,j,i) - hom(k+1,1,4,sr) ) &
2735	* w(k,j,i) * rmask(j,i,sr) &
2736	* rflags_invers(j,i,k+1)
2737	ENDDO
2738	ENDDO
2739	sums_l(k,17,tn) = s1
2740	ENDDO
2741	!$acc end parallel loop
2742
2743	IF ( humidity ) THEN
2744
2745	!$acc parallel loop gang present( hom, q, rflags_invers, rmask, sums_l, w ) create( s1 )
2746	DO k = nzb, nzt_diff
2747	!$acc loop vector collapse( 2 ) reduction( +: s1 )
2748	DO i = nxl, nxr
2749	DO j = nys, nyn
2750	s1 = s1 + 0.5_wp * ( q(k,j,i) - hom(k,1,41,sr) + &
2751	q(k+1,j,i) - hom(k+1,1,41,sr) ) &
2752	* w(k,j,i) * rmask(j,i,sr) &
2753	* rflags_invers(j,i,k+1)
2754	ENDDO
2755	ENDDO
2756	sums_l(k,49,tn) = s1
2757	ENDDO
2758	!$acc end parallel loop
2759
2760	ENDIF
2761
2762	ENDIF
2763
2764
2765	!
2766	!-- Density at top follows Neumann condition
2767	IF ( ocean ) THEN
2768	!$acc parallel present( sums_l )
2769	sums_l(nzt+1,64,tn) = sums_l(nzt,64,tn)
2770	sums_l(nzt+1,71,tn) = sums_l(nzt,71,tn)
2771	!$acc end parallel
2772	ENDIF
2773
2774	!
2775	!-- Divergence of vertical flux of resolved scale energy and pressure
2776	!-- fluctuations as well as flux of pressure fluctuation itself (68).
2777	!-- First calculate the products, then the divergence.
2778	!-- Calculation is time consuming. Do it only, if profiles shall be plotted.
2779	IF ( hom(nzb+1,2,55,0) /= 0.0_wp .OR. hom(nzb+1,2,68,0) /= 0.0_wp ) THEN
2780
2781	STOP '+++ openACC porting for vertical flux div of resolved scale TKE in flow_statistics is still missing'
2782	sums_ll = 0.0_wp ! local array
2783
2784	!$OMP DO
2785	DO i = nxl, nxr
2786	DO j = nys, nyn
2787	DO k = nzb_s_inner(j,i)+1, nzt
2788
2789	sums_ll(k,1) = sums_ll(k,1) + 0.5_wp * w(k,j,i) * ( &
2790	( 0.25_wp * ( u(k,j,i)+u(k+1,j,i)+u(k,j,i+1)+u(k+1,j,i+1) &
2791	- 0.5_wp * ( hom(k,1,1,sr) + hom(k+1,1,1,sr) ) &
2792	) )**2 &
2793	+ ( 0.25_wp * ( v(k,j,i)+v(k+1,j,i)+v(k,j+1,i)+v(k+1,j+1,i) &
2794	- 0.5_wp * ( hom(k,1,2,sr) + hom(k+1,1,2,sr) ) &
2795	) )**2 &
2796	+ w(k,j,i)**2 )
2797
2798	sums_ll(k,2) = sums_ll(k,2) + 0.5_wp * w(k,j,i) &
2799	* ( p(k,j,i) + p(k+1,j,i) )
2800
2801	ENDDO
2802	ENDDO
2803	ENDDO
2804	sums_ll(0,1) = 0.0_wp ! because w is zero at the bottom
2805	sums_ll(nzt+1,1) = 0.0_wp
2806	sums_ll(0,2) = 0.0_wp
2807	sums_ll(nzt+1,2) = 0.0_wp
2808
2809	DO k = nzb+1, nzt
2810	sums_l(k,55,tn) = ( sums_ll(k,1) - sums_ll(k-1,1) ) * ddzw(k)
2811	sums_l(k,56,tn) = ( sums_ll(k,2) - sums_ll(k-1,2) ) * ddzw(k)
2812	sums_l(k,68,tn) = sums_ll(k,2)
2813	ENDDO
2814	sums_l(nzb,55,tn) = sums_l(nzb+1,55,tn)
2815	sums_l(nzb,56,tn) = sums_l(nzb+1,56,tn)
2816	sums_l(nzb,68,tn) = 0.0_wp ! because w* = 0 at nzb
2817
2818	ENDIF
2819
2820	!
2821	!-- Divergence of vertical flux of SGS TKE and the flux itself (69)
2822	IF ( hom(nzb+1,2,57,0) /= 0.0_wp .OR. hom(nzb+1,2,69,0) /= 0.0_wp ) THEN
2823
2824	STOP '+++ openACC porting for vertical flux div of SGS TKE in flow_statistics is still missing'
2825	!$OMP DO
2826	DO i = nxl, nxr
2827	DO j = nys, nyn
2828	DO k = nzb_s_inner(j,i)+1, nzt
2829
2830	sums_l(k,57,tn) = sums_l(k,57,tn) - 0.5_wp * ( &
2831	(km(k,j,i)+km(k+1,j,i)) * (e(k+1,j,i)-e(k,j,i)) * ddzu(k+1) &
2832	- (km(k-1,j,i)+km(k,j,i)) * (e(k,j,i)-e(k-1,j,i)) * ddzu(k) &
2833	) * ddzw(k)
2834
2835	sums_l(k,69,tn) = sums_l(k,69,tn) - 0.5_wp * ( &
2836	(km(k,j,i)+km(k+1,j,i)) * (e(k+1,j,i)-e(k,j,i)) * ddzu(k+1) &
2837	)
2838
2839	ENDDO
2840	ENDDO
2841	ENDDO
2842	sums_l(nzb,57,tn) = sums_l(nzb+1,57,tn)
2843	sums_l(nzb,69,tn) = sums_l(nzb+1,69,tn)
2844
2845	ENDIF
2846
2847	!
2848	!-- Horizontal heat fluxes (subgrid, resolved, total).
2849	!-- Do it only, if profiles shall be plotted.
2850	IF ( hom(nzb+1,2,58,0) /= 0.0_wp ) THEN
2851
2852	STOP '+++ openACC porting for horizontal flux calculation in flow_statistics is still missing'
2853	!$OMP DO
2854	DO i = nxl, nxr
2855	DO j = nys, nyn
2856	DO k = nzb_s_inner(j,i)+1, nzt
2857	!
2858	!-- Subgrid horizontal heat fluxes u"pt", v"pt"
2859	sums_l(k,58,tn) = sums_l(k,58,tn) - 0.5_wp * &
2860	( kh(k,j,i) + kh(k,j,i-1) ) &
2861	* ( pt(k,j,i-1) - pt(k,j,i) ) &
2862	* ddx * rmask(j,i,sr)
2863	sums_l(k,61,tn) = sums_l(k,61,tn) - 0.5_wp * &
2864	( kh(k,j,i) + kh(k,j-1,i) ) &
2865	* ( pt(k,j-1,i) - pt(k,j,i) ) &
2866	* ddy * rmask(j,i,sr)
2867	!
2868	!-- Resolved horizontal heat fluxes upt, vpt
2869	sums_l(k,59,tn) = sums_l(k,59,tn) + &
2870	( u(k,j,i) - hom(k,1,1,sr) ) * 0.5_wp * &
2871	( pt(k,j,i-1) - hom(k,1,4,sr) + &
2872	pt(k,j,i) - hom(k,1,4,sr) )
2873	pts = 0.5_wp * ( pt(k,j-1,i) - hom(k,1,4,sr) + &
2874	pt(k,j,i) - hom(k,1,4,sr) )
2875	sums_l(k,62,tn) = sums_l(k,62,tn) + &
2876	( v(k,j,i) - hom(k,1,2,sr) ) * 0.5_wp * &
2877	( pt(k,j-1,i) - hom(k,1,4,sr) + &
2878	pt(k,j,i) - hom(k,1,4,sr) )
2879	ENDDO
2880	ENDDO
2881	ENDDO
2882	!
2883	!-- Fluxes at the surface must be zero (e.g. due to the Prandtl-layer)
2884	sums_l(nzb,58,tn) = 0.0_wp
2885	sums_l(nzb,59,tn) = 0.0_wp
2886	sums_l(nzb,60,tn) = 0.0_wp
2887	sums_l(nzb,61,tn) = 0.0_wp
2888	sums_l(nzb,62,tn) = 0.0_wp
2889	sums_l(nzb,63,tn) = 0.0_wp
2890
2891	ENDIF
2892
2893	!
2894	!-- Collect current large scale advection and subsidence tendencies for
2895	!-- data output
2896	IF ( large_scale_forcing .AND. ( simulated_time .GT. 0.0_wp ) ) THEN
2897	!
2898	!-- Interpolation in time of LSF_DATA
2899	nt = 1
2900	DO WHILE ( simulated_time - dt_3d > time_vert(nt) )
2901	nt = nt + 1
2902	ENDDO
2903	IF ( simulated_time - dt_3d /= time_vert(nt) ) THEN
2904	nt = nt - 1
2905	ENDIF
2906
2907	fac = ( simulated_time - dt_3d - time_vert(nt) ) &
2908	/ ( time_vert(nt+1)-time_vert(nt) )
2909
2910
2911	DO k = nzb, nzt
2912	sums_ls_l(k,0) = td_lsa_lpt(k,nt) &
2913	+ fac * ( td_lsa_lpt(k,nt+1) - td_lsa_lpt(k,nt) )
2914	sums_ls_l(k,1) = td_lsa_q(k,nt) &
2915	+ fac * ( td_lsa_q(k,nt+1) - td_lsa_q(k,nt) )
2916	ENDDO
2917
2918	sums_ls_l(nzt+1,0) = sums_ls_l(nzt,0)
2919	sums_ls_l(nzt+1,1) = sums_ls_l(nzt,1)
2920
2921	IF ( large_scale_subsidence .AND. use_subsidence_tendencies ) THEN
2922
2923	DO k = nzb, nzt
2924	sums_ls_l(k,2) = td_sub_lpt(k,nt) + fac * &
2925	( td_sub_lpt(k,nt+1) - td_sub_lpt(k,nt) )
2926	sums_ls_l(k,3) = td_sub_q(k,nt) + fac * &
2927	( td_sub_q(k,nt+1) - td_sub_q(k,nt) )
2928	ENDDO
2929
2930	sums_ls_l(nzt+1,2) = sums_ls_l(nzt,2)
2931	sums_ls_l(nzt+1,3) = sums_ls_l(nzt,3)
2932
2933	ENDIF
2934
2935	ENDIF
2936
2937
2938	IF ( land_surface ) THEN
2939	!$OMP DO
2940	DO i = nxl, nxr
2941	DO j = nys, nyn
2942	DO k = nzb_soil, nzt_soil
2943	sums_l(k,89,tn) = sums_l(k,89,tn) + t_soil(k,j,i) * rmask(j,i,sr)
2944	sums_l(k,91,tn) = sums_l(k,91,tn) + m_soil(k,j,i) * rmask(j,i,sr)
2945	ENDDO
2946	ENDDO
2947	ENDDO
2948	ENDIF
2949
2950
2951	!
2952	!-- Calculate the user-defined profiles
2953	CALL user_statistics( 'profiles', sr, tn )
2954	!$OMP END PARALLEL
2955
2956	!
2957	!-- Summation of thread sums
2958	IF ( threads_per_task > 1 ) THEN
2959	STOP '+++ openACC porting for threads_per_task > 1 in flow_statistics is still missing'
2960	DO i = 1, threads_per_task-1
2961	sums_l(:,3,0) = sums_l(:,3,0) + sums_l(:,3,i)
2962	sums_l(:,4:40,0) = sums_l(:,4:40,0) + sums_l(:,4:40,i)
2963	sums_l(:,45:pr_palm,0) = sums_l(:,45:pr_palm,0) + &
2964	sums_l(:,45:pr_palm,i)
2965	IF ( max_pr_user > 0 ) THEN
2966	sums_l(:,pr_palm+1:pr_palm+max_pr_user,0) = &
2967	sums_l(:,pr_palm+1:pr_palm+max_pr_user,0) + &
2968	sums_l(:,pr_palm+1:pr_palm+max_pr_user,i)
2969	ENDIF
2970	ENDDO
2971	ENDIF
2972
2973	!$acc update host( hom, sums, sums_l )
2974
2975	#if defined( __parallel )
2976
2977	!
2978	!-- Compute total sum from local sums
2979	IF ( collective_wait ) CALL MPI_BARRIER( comm2d, ierr )
2980	CALL MPI_ALLREDUCE( sums_l(nzb,1,0), sums(nzb,1), ngp_sums, MPI_REAL, &
2981	MPI_SUM, comm2d, ierr )
2982	IF ( large_scale_forcing ) THEN
2983	CALL MPI_ALLREDUCE( sums_ls_l(nzb,2), sums(nzb,83), ngp_sums_ls, &
2984	MPI_REAL, MPI_SUM, comm2d, ierr )
2985	ENDIF
2986	#else
2987	sums = sums_l(:,:,0)
2988	IF ( large_scale_forcing ) THEN
2989	sums(:,81:88) = sums_ls_l
2990	ENDIF
2991	#endif
2992
2993	!
2994	!-- Final values are obtained by division by the total number of grid points
2995	!-- used for summation. After that store profiles.
2996	!-- Profiles:
2997	DO k = nzb, nzt+1
2998	sums(k,3) = sums(k,3) / ngp_2dh(sr)
2999	sums(k,8:11) = sums(k,8:11) / ngp_2dh_s_inner(k,sr)
3000	sums(k,12:22) = sums(k,12:22) / ngp_2dh(sr)
3001	sums(k,23:29) = sums(k,23:29) / ngp_2dh_s_inner(k,sr)
3002	sums(k,30:32) = sums(k,30:32) / ngp_2dh(sr)
3003	sums(k,33:34) = sums(k,33:34) / ngp_2dh_s_inner(k,sr)
3004	sums(k,35:39) = sums(k,35:39) / ngp_2dh(sr)
3005	sums(k,40) = sums(k,40) / ngp_2dh_s_inner(k,sr)
3006	sums(k,45:53) = sums(k,45:53) / ngp_2dh(sr)
3007	sums(k,54) = sums(k,54) / ngp_2dh_s_inner(k,sr)
3008	sums(k,55:63) = sums(k,55:63) / ngp_2dh(sr)
3009	sums(k,64) = sums(k,64) / ngp_2dh_s_inner(k,sr)
3010	sums(k,70:80) = sums(k,70:80) / ngp_2dh_s_inner(k,sr)
3011	sums(k,81:88) = sums(k,81:88) / ngp_2dh(sr)
3012	sums(k,89:100) = sums(k,89:100) / ngp_2dh(sr)
3013	sums(k,101:pr_palm-2) = sums(k,101:pr_palm-2)/ ngp_2dh_s_inner(k,sr)
3014	ENDDO
3015
3016	!-- Upstream-parts
3017	sums(nzb:nzb+11,pr_palm-1) = sums(nzb:nzb+11,pr_palm-1) / ngp_3d(sr)
3018	!-- u* and so on
3019	!-- As sums(nzb:nzb+3,pr_palm) are full 2D arrays (us, usws, vsws, ts) whose
3020	!-- size is always ( nx + 1 ) * ( ny + 1 ), defined at the first grid layer
3021	!-- above the topography, they are being divided by ngp_2dh(sr)
3022	sums(nzb:nzb+3,pr_palm) = sums(nzb:nzb+3,pr_palm) / &
3023	ngp_2dh(sr)
3024	sums(nzb+12,pr_palm) = sums(nzb+12,pr_palm) / & ! qs
3025	ngp_2dh(sr)
3026	!-- eges, e*
3027	sums(nzb+4:nzb+5,pr_palm) = sums(nzb+4:nzb+5,pr_palm) / &
3028	ngp_3d(sr)
3029	!-- Old and new divergence
3030	sums(nzb+9:nzb+10,pr_palm) = sums(nzb+9:nzb+10,pr_palm) / &
3031	ngp_3d_inner(sr)
3032
3033	!-- User-defined profiles
3034	IF ( max_pr_user > 0 ) THEN
3035	DO k = nzb, nzt+1
3036	sums(k,pr_palm+1:pr_palm+max_pr_user) = &
3037	sums(k,pr_palm+1:pr_palm+max_pr_user) / &
3038	ngp_2dh_s_inner(k,sr)
3039	ENDDO
3040	ENDIF
3041
3042	!
3043	!-- Collect horizontal average in hom.
3044	!-- Compute deduced averages (e.g. total heat flux)
3045	hom(:,1,3,sr) = sums(:,3) ! w
3046	hom(:,1,8,sr) = sums(:,8) ! e profiles 5-7 are initial profiles
3047	hom(:,1,9,sr) = sums(:,9) ! km
3048	hom(:,1,10,sr) = sums(:,10) ! kh
3049	hom(:,1,11,sr) = sums(:,11) ! l
3050	hom(:,1,12,sr) = sums(:,12) ! w"u"
3051	hom(:,1,13,sr) = sums(:,13) ! wu
3052	hom(:,1,14,sr) = sums(:,14) ! w"v"
3053	hom(:,1,15,sr) = sums(:,15) ! wv
3054	hom(:,1,16,sr) = sums(:,16) ! w"pt"
3055	hom(:,1,17,sr) = sums(:,17) ! wpt
3056	hom(:,1,18,sr) = sums(:,16) + sums(:,17) ! wpt
3057	hom(:,1,19,sr) = sums(:,12) + sums(:,13) ! wu
3058	hom(:,1,20,sr) = sums(:,14) + sums(:,15) ! wv
3059	hom(:,1,21,sr) = sums(:,21) ! wptBC
3060	hom(:,1,22,sr) = sums(:,16) + sums(:,21) ! wptBC
3061	! profile 24 is initial profile (sa)
3062	! profiles 25-29 left empty for initial
3063	! profiles
3064	hom(:,1,30,sr) = sums(:,30) ! u*2
3065	hom(:,1,31,sr) = sums(:,31) ! v*2
3066	hom(:,1,32,sr) = sums(:,32) ! w*2
3067	hom(:,1,33,sr) = sums(:,33) ! pt*2
3068	hom(:,1,34,sr) = sums(:,34) ! e*
3069	hom(:,1,35,sr) = sums(:,35) ! w2pt
3070	hom(:,1,36,sr) = sums(:,36) ! wpt2
3071	hom(:,1,37,sr) = sums(:,37) ! we
3072	hom(:,1,38,sr) = sums(:,38) ! w*3
3073	hom(:,1,39,sr) = sums(:,38) / ( abs( sums(:,32) ) + 1E-20_wp )**1.5_wp ! Sw
3074	hom(:,1,40,sr) = sums(:,40) ! p
3075	hom(:,1,45,sr) = sums(:,45) ! w"vpt"
3076	hom(:,1,46,sr) = sums(:,46) ! wvpt
3077	hom(:,1,47,sr) = sums(:,45) + sums(:,46) ! wvpt
3078	hom(:,1,48,sr) = sums(:,48) ! w"q" (w"qv")
3079	hom(:,1,49,sr) = sums(:,49) ! wq (wqv)
3080	hom(:,1,50,sr) = sums(:,48) + sums(:,49) ! wq (wqv)
3081	hom(:,1,51,sr) = sums(:,51) ! w"qv"
3082	hom(:,1,52,sr) = sums(:,52) ! wqv
3083	hom(:,1,53,sr) = sums(:,52) + sums(:,51) ! wq (wqv)
3084	hom(:,1,54,sr) = sums(:,54) ! ql
3085	hom(:,1,55,sr) = sums(:,55) ! wuu*/dz
3086	hom(:,1,56,sr) = sums(:,56) ! wp/dz
3087	hom(:,1,57,sr) = sums(:,57) ! ( w"e + w"p"/rho )/dz
3088	hom(:,1,58,sr) = sums(:,58) ! u"pt"
3089	hom(:,1,59,sr) = sums(:,59) ! upt
3090	hom(:,1,60,sr) = sums(:,58) + sums(:,59) ! upt_t
3091	hom(:,1,61,sr) = sums(:,61) ! v"pt"
3092	hom(:,1,62,sr) = sums(:,62) ! vpt
3093	hom(:,1,63,sr) = sums(:,61) + sums(:,62) ! vpt_t
3094	hom(:,1,64,sr) = sums(:,64) ! rho
3095	hom(:,1,65,sr) = sums(:,65) ! w"sa"
3096	hom(:,1,66,sr) = sums(:,66) ! wsa
3097	hom(:,1,67,sr) = sums(:,65) + sums(:,66) ! wsa
3098	hom(:,1,68,sr) = sums(:,68) ! wp
3099	hom(:,1,69,sr) = sums(:,69) ! w"e + w"p"/rho
3100	hom(:,1,70,sr) = sums(:,70) ! q*2
3101	hom(:,1,71,sr) = sums(:,71) ! prho
3102	hom(:,1,72,sr) = hyp * 1E-4_wp ! hyp in dbar
3103	hom(:,1,73,sr) = sums(:,73) ! nr
3104	hom(:,1,74,sr) = sums(:,74) ! qr
3105	hom(:,1,75,sr) = sums(:,75) ! qc
3106	hom(:,1,76,sr) = sums(:,76) ! prr (precipitation rate)
3107	! 77 is initial density profile
3108	hom(:,1,78,sr) = ug ! ug
3109	hom(:,1,79,sr) = vg ! vg
3110	hom(:,1,80,sr) = w_subs ! w_subs
3111
3112	IF ( large_scale_forcing ) THEN
3113	hom(:,1,81,sr) = sums_ls_l(:,0) ! td_lsa_lpt
3114	hom(:,1,82,sr) = sums_ls_l(:,1) ! td_lsa_q
3115	IF ( use_subsidence_tendencies ) THEN
3116	hom(:,1,83,sr) = sums_ls_l(:,2) ! td_sub_lpt
3117	hom(:,1,84,sr) = sums_ls_l(:,3) ! td_sub_q
3118	ELSE
3119	hom(:,1,83,sr) = sums(:,83) ! td_sub_lpt
3120	hom(:,1,84,sr) = sums(:,84) ! td_sub_q
3121	ENDIF
3122	hom(:,1,85,sr) = sums(:,85) ! td_nud_lpt
3123	hom(:,1,86,sr) = sums(:,86) ! td_nud_q
3124	hom(:,1,87,sr) = sums(:,87) ! td_nud_u
3125	hom(:,1,88,sr) = sums(:,88) ! td_nud_v
3126	END IF
3127
3128	hom(:,1,pr_palm-1,sr) = sums(:,pr_palm-1)
3129	! upstream-parts u_x, u_y, u_z, v_x,
3130	! v_y, usw. (in last but one profile)
3131	hom(:,1,pr_palm,sr) = sums(:,pr_palm)
3132	! u, w'u', w'v', t (in last profile)
3133
3134	IF ( max_pr_user > 0 ) THEN ! user-defined profiles
3135	hom(:,1,pr_palm+1:pr_palm+max_pr_user,sr) = &
3136	sums(:,pr_palm+1:pr_palm+max_pr_user)
3137	ENDIF
3138
3139	!
3140	!-- Determine the boundary layer height using two different schemes.
3141	!-- First scheme: Starting from the Earth's (Ocean's) surface, look for the
3142	!-- first relative minimum (maximum) of the total heat flux.
3143	!-- The corresponding height is assumed as the boundary layer height, if it
3144	!-- is less than 1.5 times the height where the heat flux becomes negative
3145	!-- (positive) for the first time.
3146	z_i(1) = 0.0_wp
3147	first = .TRUE.
3148
3149	IF ( ocean ) THEN
3150	DO k = nzt, nzb+1, -1
3151	IF ( first .AND. hom(k,1,18,sr) < 0.0_wp &
3152	.AND. abs(hom(k,1,18,sr)) > 1.0E-8_wp ) THEN
3153	first = .FALSE.
3154	height = zw(k)
3155	ENDIF
3156	IF ( hom(k,1,18,sr) < 0.0_wp .AND. &
3157	abs(hom(k,1,18,sr)) > 1.0E-8_wp .AND. &
3158	hom(k-1,1,18,sr) > hom(k,1,18,sr) ) THEN
3159	IF ( zw(k) < 1.5_wp * height ) THEN
3160	z_i(1) = zw(k)
3161	ELSE
3162	z_i(1) = height
3163	ENDIF
3164	EXIT
3165	ENDIF
3166	ENDDO
3167	ELSE
3168	DO k = nzb, nzt-1
3169	IF ( first .AND. hom(k,1,18,sr) < 0.0_wp &
3170	.AND. abs(hom(k,1,18,sr)) > 1.0E-8_wp ) THEN
3171	first = .FALSE.
3172	height = zw(k)
3173	ENDIF
3174	IF ( hom(k,1,18,sr) < 0.0 .AND. &
3175	abs(hom(k,1,18,sr)) > 1.0E-8_wp .AND. &
3176	hom(k+1,1,18,sr) > hom(k,1,18,sr) ) THEN
3177	IF ( zw(k) < 1.5_wp * height ) THEN
3178	z_i(1) = zw(k)
3179	ELSE
3180	z_i(1) = height
3181	ENDIF
3182	EXIT
3183	ENDIF
3184	ENDDO
3185	ENDIF
3186
3187	!
3188	!-- Second scheme: Gradient scheme from Sullivan et al. (1998), modified
3189	!-- by Uhlenbrock(2006). The boundary layer height is the height with the
3190	!-- maximal local temperature gradient: starting from the second (the last
3191	!-- but one) vertical gridpoint, the local gradient must be at least
3192	!-- 0.2K/100m and greater than the next four gradients.
3193	!-- WARNING: The threshold value of 0.2K/100m must be adjusted for the
3194	!-- ocean case!
3195	z_i(2) = 0.0_wp
3196	DO k = nzb+1, nzt+1
3197	dptdz(k) = ( hom(k,1,4,sr) - hom(k-1,1,4,sr) ) * ddzu(k)
3198	ENDDO
3199	dptdz_threshold = 0.2_wp / 100.0_wp
3200
3201	IF ( ocean ) THEN
3202	DO k = nzt+1, nzb+5, -1
3203	IF ( dptdz(k) > dptdz_threshold .AND. &
3204	dptdz(k) > dptdz(k-1) .AND. dptdz(k) > dptdz(k-2) .AND. &
3205	dptdz(k) > dptdz(k-3) .AND. dptdz(k) > dptdz(k-4) ) THEN
3206	z_i(2) = zw(k-1)
3207	EXIT
3208	ENDIF
3209	ENDDO
3210	ELSE
3211	DO k = nzb+1, nzt-3
3212	IF ( dptdz(k) > dptdz_threshold .AND. &
3213	dptdz(k) > dptdz(k+1) .AND. dptdz(k) > dptdz(k+2) .AND. &
3214	dptdz(k) > dptdz(k+3) .AND. dptdz(k) > dptdz(k+4) ) THEN
3215	z_i(2) = zw(k-1)
3216	EXIT
3217	ENDIF
3218	ENDDO
3219	ENDIF
3220
3221	hom(nzb+6,1,pr_palm,sr) = z_i(1)
3222	hom(nzb+7,1,pr_palm,sr) = z_i(2)
3223
3224	!
3225	!-- Computation of both the characteristic vertical velocity and
3226	!-- the characteristic convective boundary layer temperature.
3227	!-- The horizontal average at nzb+1 is input for the average temperature.
3228	IF ( hom(nzb,1,18,sr) > 0.0_wp .AND. abs(hom(nzb,1,18,sr)) > 1.0E-8_wp &
3229	.AND. z_i(1) /= 0.0_wp ) THEN
3230	hom(nzb+8,1,pr_palm,sr) = ( g / hom(nzb+1,1,4,sr) * &
3231	hom(nzb,1,18,sr) * &
3232	ABS( z_i(1) ) )**0.333333333_wp
3233	!-- so far this only works if Prandtl layer is used
3234	hom(nzb+11,1,pr_palm,sr) = hom(nzb,1,16,sr) / hom(nzb+8,1,pr_palm,sr)
3235	ELSE
3236	hom(nzb+8,1,pr_palm,sr) = 0.0_wp
3237	hom(nzb+11,1,pr_palm,sr) = 0.0_wp
3238	ENDIF
3239
3240	!
3241	!-- Collect the time series quantities
3242	ts_value(1,sr) = hom(nzb+4,1,pr_palm,sr) ! E
3243	ts_value(2,sr) = hom(nzb+5,1,pr_palm,sr) ! E*
3244	ts_value(3,sr) = dt_3d
3245	ts_value(4,sr) = hom(nzb,1,pr_palm,sr) ! u*
3246	ts_value(5,sr) = hom(nzb+3,1,pr_palm,sr) ! th*
3247	ts_value(6,sr) = u_max
3248	ts_value(7,sr) = v_max
3249	ts_value(8,sr) = w_max
3250	ts_value(9,sr) = hom(nzb+10,1,pr_palm,sr) ! new divergence
3251	ts_value(10,sr) = hom(nzb+9,1,pr_palm,sr) ! old Divergence
3252	ts_value(11,sr) = hom(nzb+6,1,pr_palm,sr) ! z_i(1)
3253	ts_value(12,sr) = hom(nzb+7,1,pr_palm,sr) ! z_i(2)
3254	ts_value(13,sr) = hom(nzb+8,1,pr_palm,sr) ! w*
3255	ts_value(14,sr) = hom(nzb,1,16,sr) ! w'pt' at k=0
3256	ts_value(15,sr) = hom(nzb+1,1,16,sr) ! w'pt' at k=1
3257	ts_value(16,sr) = hom(nzb+1,1,18,sr) ! wpt at k=1
3258	ts_value(17,sr) = hom(nzb,1,4,sr) ! pt(0)
3259	ts_value(18,sr) = hom(nzb+1,1,4,sr) ! pt(zp)
3260	ts_value(19,sr) = hom(nzb+1,1,pr_palm,sr) ! u'w' at k=0
3261	ts_value(20,sr) = hom(nzb+2,1,pr_palm,sr) ! v'w' at k=0
3262	ts_value(21,sr) = hom(nzb,1,48,sr) ! w"q" at k=0
3263
3264	IF ( ts_value(5,sr) /= 0.0_wp ) THEN
3265	ts_value(22,sr) = ts_value(4,sr)**2_wp / &
3266	( kappa * g * ts_value(5,sr) / ts_value(18,sr) ) ! L
3267	ELSE
3268	ts_value(22,sr) = 10000.0_wp
3269	ENDIF
3270
3271	ts_value(23,sr) = hom(nzb+12,1,pr_palm,sr) ! q*
3272
3273	!
3274	!-- Collect land surface model timeseries
3275	IF ( land_surface ) THEN
3276	ts_value(dots_soil ,sr) = hom(nzb,1,93,sr) ! ghf_eb
3277	ts_value(dots_soil+1,sr) = hom(nzb,1,94,sr) ! shf_eb
3278	ts_value(dots_soil+2,sr) = hom(nzb,1,95,sr) ! qsws_eb
3279	ts_value(dots_soil+3,sr) = hom(nzb,1,96,sr) ! qsws_liq_eb
3280	ts_value(dots_soil+4,sr) = hom(nzb,1,97,sr) ! qsws_soil_eb
3281	ts_value(dots_soil+5,sr) = hom(nzb,1,98,sr) ! qsws_veg_eb
3282	ts_value(dots_soil+6,sr) = hom(nzb,1,99,sr) ! r_a
3283	ts_value(dots_soil+7,sr) = hom(nzb,1,100,sr) ! r_s
3284	ENDIF
3285	!
3286	!-- Collect radiation model timeseries
3287	IF ( radiation ) THEN
3288	ts_value(dots_rad,sr) = hom(nzb,1,99,sr) ! rad_net
3289	ts_value(dots_rad+1,sr) = hom(nzb,1,100,sr) ! rad_sw_in
3290	ENDIF
3291
3292	!
3293	!-- Calculate additional statistics provided by the user interface
3294	CALL user_statistics( 'time_series', sr, 0 )
3295
3296	ENDDO ! loop of the subregions
3297
3298	!$acc end data
3299
3300	!
3301	!-- If required, sum up horizontal averages for subsequent time averaging
3302	IF ( do_sum ) THEN
3303	IF ( average_count_pr == 0 ) hom_sum = 0.0_wp
3304	hom_sum = hom_sum + hom(:,1,:,:)
3305	average_count_pr = average_count_pr + 1
3306	do_sum = .FALSE.
3307	ENDIF
3308
3309	!
3310	!-- Set flag for other UPs (e.g. output routines, but also buoyancy).
3311	!-- This flag is reset after each time step in time_integration.
3312	flow_statistics_called = .TRUE.
3313
3314	CALL cpu_log( log_point(10), 'flow_statistics', 'stop' )
3315
3316
3317	END SUBROUTINE flow_statistics
3318	#endif

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

Download in other formats:

| Impressum | ©Leibniz Universität Hannover |