Home

Context Navigation

source: palm/trunk/SOURCE/advec_ws.f90 @ 3551

Last change on this file since 3551 was 3551, checked in by suehring, 6 years ago
optimization of advec_ws
Property svn:keywords set to `Id`
File size: 302.9 KB

Line
1	!> @file advec_ws.f90
2	!------------------------------------------------------------------------------!
3	! This file is part of the PALM model system.
4	!
5	! PALM is free software: you can redistribute it and/or modify it under the
6	! terms of the GNU General Public License as published by the Free Software
7	! Foundation, either version 3 of the License, or (at your option) any later
8	! version.
9	!
10	! PALM is distributed in the hope that it will be useful, but WITHOUT ANY
11	! WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR
12	! A PARTICULAR PURPOSE. See the GNU General Public License for more details.
13	!
14	! You should have received a copy of the GNU General Public License along with
15	! PALM. If not, see <http://www.gnu.org/licenses/>.
16	!
17	! Copyright 1997-2018 Leibniz Universitaet Hannover
18	!------------------------------------------------------------------------------!
19	!
20	! Current revisions:
21	! ------------------
22	!
23	!
24	! Former revisions:
25	! -----------------
26	! $Id: advec_ws.f90 3551 2018-11-21 17:05:46Z suehring $
27	! - Computation of vertical fluxes separated from computation of horizontal
28	! fluxes. Loops are splitted in order to avoid indirect indexing and allow
29	! for better vectorization.
30	! - Accelerate code by remove type-conversions of ibits
31	! - replace pointer definition in scalar routine by simple explicit definition
32	! of the passed array
33	!
34	! 3547 2018-11-21 13:21:24Z suehring
35	! variables documented
36	!
37	! 3467 2018-10-30 19:05:21Z suehring
38	! Implementation of a new aerosol module salsa.
39	!
40	! 3302 2018-10-03 02:39:40Z raasch
41	! Stokes drift velocity added in scalar advection
42	!
43	! 3298 2018-10-02 12:21:11Z kanani
44	! Add formatted note from ketelsen about statistics for chemistry
45	!
46	! 3294 2018-10-01 02:37:10Z raasch
47	! ocean renamed ocean_mode
48	!
49	! 3274 2018-09-24 15:42:55Z knoop
50	! Modularization of all bulk cloud physics code components
51	!
52	! 3241 2018-09-12 15:02:00Z raasch
53	! unused variables removed
54	!
55	! 3183 2018-07-27 14:25:55Z suehring
56	! Rename variables for boundary-condition flags and for offline nesting mode
57	!
58	! 3182 2018-07-27 13:36:03Z suehring
59	! Bugfix in calculation of left-sided fluxes for u-component in OpenMP case.
60	!
61	! 2731 2018-01-09 17:44:02Z suehring
62	! Enable loop vectorization by splitting the k-loop
63	!
64	! 2718 2018-01-02 08:49:38Z maronga
65	! Corrected "Former revisions" section
66	!
67	! 2696 2017-12-14 17:12:51Z kanani
68	! Change in file header (GPL part)
69	! Implement advection for TKE-dissipation in case of RANS-TKE-e closure (TG)
70	! Allocate advc_flags_1/2 within ws_init_flags instead of init_grid
71	! Change argument list for exchange_horiz_2d_int (MS)
72	!
73	! 2329 2017-08-03 14:24:56Z knoop
74	! Bugfix concerning density in divergence correction close to buildings
75	!
76	! 2292 2017-06-20 09:51:42Z schwenkel
77	! Implementation of new microphysic scheme: cloud_scheme = 'morrison'
78	! includes two more prognostic equations for cloud drop concentration (nc)
79	! and cloud water content (qc).
80	!
81	! 2233 2017-05-30 18:08:54Z suehring
82	!
83	! 2232 2017-05-30 17:47:52Z suehring
84	! Rename wall_flags_0 and wall_flags_00 into advc_flags_1 and advc_flags_2,
85	! respectively.
86	! Set advc_flags_1/2 on basis of wall_flags_0/00 instead of nzb_s/u/v/w_inner.
87	! Setting advc_flags_1/2 also for downward-facing walls
88	!
89	! 2200 2017-04-11 11:37:51Z suehring
90	! monotonic_adjustment removed
91	!
92	! 2118 2017-01-17 16:38:49Z raasch
93	! OpenACC version of subroutines removed
94	!
95	! 2037 2016-10-26 11:15:40Z knoop
96	! Anelastic approximation implemented
97	!
98	! 2000 2016-08-20 18:09:15Z knoop
99	! Forced header and separation lines into 80 columns
100	!
101	! 1996 2016-08-18 11:42:29Z suehring
102	! Bugfix concerning calculation of turbulent of turbulent fluxes
103	!
104	! 1960 2016-07-12 16:34:24Z suehring
105	! Separate humidity and passive scalar
106	!
107	! 1942 2016-06-14 12:18:18Z suehring
108	! Initialization of flags for ws-scheme moved from init_grid.
109	!
110	! 1873 2016-04-18 14:50:06Z maronga
111	! Module renamed (removed _mod)
112	!
113	!
114	! 1850 2016-04-08 13:29:27Z maronga
115	! Module renamed
116	!
117	!
118	! 1822 2016-04-07 07:49:42Z hoffmann
119	! icloud_scheme removed, microphysics_seifert added
120	!
121	! 1682 2015-10-07 23:56:08Z knoop
122	! Code annotations made doxygen readable
123	!
124	! 1630 2015-08-26 16:57:23Z suehring
125	!
126	!
127	! 1629 2015-08-26 16:56:11Z suehring
128	! Bugfix concerning wall_flags at left and south PE boundaries
129	!
130	! 1581 2015-04-10 13:45:59Z suehring
131	!
132	!
133	! 1580 2015-04-10 13:43:49Z suehring
134	! Bugfix: statistical evaluation of scalar fluxes in case of monotonic limiter
135	!
136	! 1567 2015-03-10 17:57:55Z suehring
137	! Bugfixes in monotonic limiter.
138	!
139	! 2015-03-09 13:10:37Z heinze
140	! Bugfix: REAL constants provided with KIND-attribute in call of
141	! intrinsic functions like MAX and MIN
142	!
143	! 1557 2015-03-05 16:43:04Z suehring
144	! Enable monotone advection for scalars using monotonic limiter
145	!
146	! 1374 2014-04-25 12:55:07Z raasch
147	! missing variables added to ONLY list
148	!
149	! 1361 2014-04-16 15:17:48Z hoffmann
150	! accelerator and vector version for qr and nr added
151	!
152	! 1353 2014-04-08 15:21:23Z heinze
153	! REAL constants provided with KIND-attribute,
154	! module kinds added
155	! some formatting adjustments
156	!
157	! 1322 2014-03-20 16:38:49Z raasch
158	! REAL constants defined as wp-kind
159	!
160	! 1320 2014-03-20 08:40:49Z raasch
161	! ONLY-attribute added to USE-statements,
162	! kind-parameters added to all INTEGER and REAL declaration statements,
163	! kinds are defined in new module kinds,
164	! old module precision_kind is removed,
165	! revision history before 2012 removed,
166	! comment fields (!:) to be used for variable explanations added to
167	! all variable declaration statements
168	!
169	! 1257 2013-11-08 15:18:40Z raasch
170	! accelerator loop directives removed
171	!
172	! 1221 2013-09-10 08:59:13Z raasch
173	! wall_flags_00 introduced, which holds bits 32-...
174	!
175	! 1128 2013-04-12 06:19:32Z raasch
176	! loop index bounds in accelerator version replaced by i_left, i_right, j_south,
177	! j_north
178	!
179	! 1115 2013-03-26 18:16:16Z hoffmann
180	! calculation of qr and nr is restricted to precipitation
181	!
182	! 1053 2012-11-13 17:11:03Z hoffmann
183	! necessary expansions according to the two new prognostic equations (nr, qr)
184	! of the two-moment cloud physics scheme:
185	! +flux_l_, flux_s_, diss_l_, diss_s_, sums_ws*s_ws_l
186	!
187	! 1036 2012-10-22 13:43:42Z raasch
188	! code put under GPL (PALM 3.9)
189	!
190	! 1027 2012-10-15 17:18:39Z suehring
191	! Bugfix in calculation indices k_mm, k_pp in accelerator version
192	!
193	! 1019 2012-09-28 06:46:45Z raasch
194	! small change in comment lines
195	!
196	! 1015 2012-09-27 09:23:24Z raasch
197	! accelerator versions (*_acc) added
198	!
199	! 1010 2012-09-20 07:59:54Z raasch
200	! cpp switch __nopointer added for pointer free version
201	!
202	! 888 2012-04-20 15:03:46Z suehring
203	! Number of REAL( IBITS() calls with identical arguments is reduced.
204	!
205	! 862 2012-03-26 14:21:38Z suehring
206	! ws-scheme also work with topography in combination with vector version.
207	! ws-scheme also work with outflow boundaries in combination with
208	! vector version.
209	! Degradation of the applied order of scheme is now steered by multiplying with
210	! Integer advc_flags_1. 2nd order scheme, WS3 and WS5 are calculated on each
211	! grid point and mulitplied with the appropriate flag.
212	! 2nd order numerical dissipation term changed. Now the appropriate 2nd order
213	! term derived according to the 4th and 6th order terms is applied. It turns
214	! out that diss_2nd does not provide sufficient dissipation near walls.
215	! Therefore, the function diss_2nd is removed.
216	! Near walls a divergence correction is necessary to overcome numerical
217	! instabilities due to too less divergence reduction of the velocity field.
218	! boundary_flags and logicals steering the degradation are removed.
219	! Empty SUBROUTINE local_diss removed.
220	! Further formatting adjustments.
221	!
222	! 801 2012-01-10 17:30:36Z suehring
223	! Bugfix concerning OpenMP parallelization. Summation of sums_wsus_ws_l,
224	! sums_wsvs_ws_l, sums_us2_ws_l, sums_vs2_ws_l, sums_ws2_ws_l, sums_wspts_ws_l,
225	! sums_wsqs_ws_l, sums_wssas_ws_l is now thread-safe by adding an additional
226	! dimension.
227	!
228	! Initial revision
229	!
230	! 411 2009-12-11 12:31:43 Z suehring
231	!
232	! Description:
233	! ------------
234	!> Advection scheme for scalars and momentum using the flux formulation of
235	!> Wicker and Skamarock 5th order. Additionally the module contains of a
236	!> routine using for initialisation and steering of the statical evaluation.
237	!> The computation of turbulent fluxes takes place inside the advection
238	!> routines.
239	!> Near non-cyclic boundaries the order of the applied advection scheme is
240	!> degraded.
241	!> A divergence correction is applied. It is necessary for topography, since
242	!> the divergence is not sufficiently reduced, resulting in erroneous fluxes and
243	!> partly numerical instabilities.
244	!-----------------------------------------------------------------------------!
245	MODULE advec_ws
246
247	USE kinds
248
249	IMPLICIT NONE
250
251	REAL(wp) :: adv_mom_1 !< 1/4 - constant used in 5th-order advection scheme for momentum advection (1st-order part)
252	REAL(wp) :: adv_mom_3 !< 1/24 - constant used in 5th-order advection scheme for momentum advection (3rd-order part)
253	REAL(wp) :: adv_mom_5 !< 1/120 - constant used in 5th-order advection scheme for momentum advection (5th-order part)
254	REAL(wp) :: adv_sca_1 !< 1/2 - constant used in 5th-order advection scheme for scalar advection (1st-order part)
255	REAL(wp) :: adv_sca_3 !< 1/12 - constant used in 5th-order advection scheme for scalar advection (3rd-order part)
256	REAL(wp) :: adv_sca_5 !< 1/60 - constant used in 5th-order advection scheme for scalar advection (5th-order part)
257
258	PRIVATE
259	PUBLIC advec_s_ws, advec_u_ws, advec_v_ws, advec_w_ws, ws_init, &
260	ws_init_flags, ws_statistics
261
262	INTERFACE ws_init
263	MODULE PROCEDURE ws_init
264	END INTERFACE ws_init
265
266	INTERFACE ws_init_flags
267	MODULE PROCEDURE ws_init_flags
268	END INTERFACE ws_init_flags
269
270	INTERFACE ws_statistics
271	MODULE PROCEDURE ws_statistics
272	END INTERFACE ws_statistics
273
274	INTERFACE advec_s_ws
275	MODULE PROCEDURE advec_s_ws
276	MODULE PROCEDURE advec_s_ws_ij
277	END INTERFACE advec_s_ws
278
279	INTERFACE advec_u_ws
280	MODULE PROCEDURE advec_u_ws
281	MODULE PROCEDURE advec_u_ws_ij
282	END INTERFACE advec_u_ws
283
284	INTERFACE advec_v_ws
285	MODULE PROCEDURE advec_v_ws
286	MODULE PROCEDURE advec_v_ws_ij
287	END INTERFACE advec_v_ws
288
289	INTERFACE advec_w_ws
290	MODULE PROCEDURE advec_w_ws
291	MODULE PROCEDURE advec_w_ws_ij
292	END INTERFACE advec_w_ws
293
294	CONTAINS
295
296
297	!------------------------------------------------------------------------------!
298	! Description:
299	! ------------
300	!> Initialization of WS-scheme
301	!------------------------------------------------------------------------------!
302	SUBROUTINE ws_init
303
304	USE arrays_3d, &
305	ONLY: diss_l_diss, diss_l_e, diss_l_nc, diss_l_nr, diss_l_pt, &
306	diss_l_q, diss_l_qc, diss_l_qr, diss_l_s, diss_l_sa, &
307	diss_l_u, diss_l_v, diss_l_w, flux_l_diss, flux_l_e, &
308	flux_l_nc, flux_l_nr, flux_l_pt, flux_l_q, flux_l_qc, &
309	flux_l_qr, flux_l_s, flux_l_sa, flux_l_u, flux_l_v, &
310	flux_l_w, diss_s_diss, diss_s_e, diss_s_nc, diss_s_nr, &
311	diss_s_pt, diss_s_q, diss_s_qc, diss_s_qr, diss_s_s, &
312	diss_s_sa, diss_s_u, diss_s_v, diss_s_w, flux_s_diss, &
313	flux_s_e, flux_s_nc, flux_s_nr, flux_s_pt, flux_s_q, &
314	flux_s_qc, flux_s_qr, flux_s_s, flux_s_sa, flux_s_u, &
315	flux_s_v, flux_s_w
316
317	USE control_parameters, &
318	ONLY: humidity, loop_optimization, passive_scalar, ocean_mode, &
319	rans_tke_e, ws_scheme_mom, ws_scheme_sca
320
321	USE indices, &
322	ONLY: nyn, nys, nzb, nzt
323
324	USE kinds
325
326	USE bulk_cloud_model_mod, &
327	ONLY: bulk_cloud_model, microphysics_morrison, microphysics_seifert
328
329	USE pegrid
330
331	USE salsa_util_mod, &
332	ONLY: sums_salsa_ws_l
333
334	USE statistics, &
335	ONLY: sums_us2_ws_l, sums_vs2_ws_l, sums_ws2_ws_l, sums_wsncs_ws_l,&
336	sums_wsnrs_ws_l,sums_wspts_ws_l, sums_wsqcs_ws_l, &
337	sums_wsqrs_ws_l, sums_wsqs_ws_l, sums_wsss_ws_l, &
338	sums_wssas_ws_l, sums_wsss_ws_l, sums_wsus_ws_l, &
339	sums_wsvs_ws_l
340
341
342	!
343	!-- Set the appropriate factors for scalar and momentum advection.
344	adv_sca_5 = 1.0_wp / 60.0_wp
345	adv_sca_3 = 1.0_wp / 12.0_wp
346	adv_sca_1 = 1.0_wp / 2.0_wp
347	adv_mom_5 = 1.0_wp / 120.0_wp
348	adv_mom_3 = 1.0_wp / 24.0_wp
349	adv_mom_1 = 1.0_wp / 4.0_wp
350	!
351	!-- Arrays needed for statical evaluation of fluxes.
352	IF ( ws_scheme_mom ) THEN
353
354	ALLOCATE( sums_wsus_ws_l(nzb:nzt+1,0:threads_per_task-1), &
355	sums_wsvs_ws_l(nzb:nzt+1,0:threads_per_task-1), &
356	sums_us2_ws_l(nzb:nzt+1,0:threads_per_task-1), &
357	sums_vs2_ws_l(nzb:nzt+1,0:threads_per_task-1), &
358	sums_ws2_ws_l(nzb:nzt+1,0:threads_per_task-1) )
359
360	sums_wsus_ws_l = 0.0_wp
361	sums_wsvs_ws_l = 0.0_wp
362	sums_us2_ws_l = 0.0_wp
363	sums_vs2_ws_l = 0.0_wp
364	sums_ws2_ws_l = 0.0_wp
365
366	ENDIF
367
368	IF ( ws_scheme_sca ) THEN
369
370	ALLOCATE( sums_wspts_ws_l(nzb:nzt+1,0:threads_per_task-1) )
371	sums_wspts_ws_l = 0.0_wp
372
373	IF ( humidity ) THEN
374	ALLOCATE( sums_wsqs_ws_l(nzb:nzt+1,0:threads_per_task-1) )
375	sums_wsqs_ws_l = 0.0_wp
376	ENDIF
377
378	IF ( passive_scalar ) THEN
379	ALLOCATE( sums_wsss_ws_l(nzb:nzt+1,0:threads_per_task-1) )
380	sums_wsss_ws_l = 0.0_wp
381	ENDIF
382
383	IF ( bulk_cloud_model .AND. microphysics_morrison ) THEN
384	ALLOCATE( sums_wsqcs_ws_l(nzb:nzt+1,0:threads_per_task-1) )
385	ALLOCATE( sums_wsncs_ws_l(nzb:nzt+1,0:threads_per_task-1) )
386	sums_wsqcs_ws_l = 0.0_wp
387	sums_wsncs_ws_l = 0.0_wp
388	ENDIF
389
390	IF ( bulk_cloud_model .AND. microphysics_seifert ) THEN
391	ALLOCATE( sums_wsqrs_ws_l(nzb:nzt+1,0:threads_per_task-1) )
392	ALLOCATE( sums_wsnrs_ws_l(nzb:nzt+1,0:threads_per_task-1) )
393	sums_wsqrs_ws_l = 0.0_wp
394	sums_wsnrs_ws_l = 0.0_wp
395	ENDIF
396
397	IF ( ocean_mode ) THEN
398	ALLOCATE( sums_wssas_ws_l(nzb:nzt+1,0:threads_per_task-1) )
399	sums_wssas_ws_l = 0.0_wp
400	ENDIF
401	!
402	!-- Mona: Now always allocated
403	ALLOCATE( sums_salsa_ws_l(nzb:nzt+1,0:threads_per_task-1) )
404	sums_salsa_ws_l = 0.0_wp
405
406	ENDIF
407
408	!
409	!-- Arrays needed for reasons of speed optimization for cache version.
410	!-- For the vector version the buffer arrays are not necessary,
411	!-- because the the fluxes can swapped directly inside the loops of the
412	!-- advection routines.
413	IF ( loop_optimization /= 'vector' ) THEN
414
415	IF ( ws_scheme_mom ) THEN
416
417	ALLOCATE( flux_s_u(nzb+1:nzt,0:threads_per_task-1), &
418	flux_s_v(nzb+1:nzt,0:threads_per_task-1), &
419	flux_s_w(nzb+1:nzt,0:threads_per_task-1), &
420	diss_s_u(nzb+1:nzt,0:threads_per_task-1), &
421	diss_s_v(nzb+1:nzt,0:threads_per_task-1), &
422	diss_s_w(nzb+1:nzt,0:threads_per_task-1) )
423	ALLOCATE( flux_l_u(nzb+1:nzt,nys:nyn,0:threads_per_task-1), &
424	flux_l_v(nzb+1:nzt,nys:nyn,0:threads_per_task-1), &
425	flux_l_w(nzb+1:nzt,nys:nyn,0:threads_per_task-1), &
426	diss_l_u(nzb+1:nzt,nys:nyn,0:threads_per_task-1), &
427	diss_l_v(nzb+1:nzt,nys:nyn,0:threads_per_task-1), &
428	diss_l_w(nzb+1:nzt,nys:nyn,0:threads_per_task-1) )
429
430	ENDIF
431
432	IF ( ws_scheme_sca ) THEN
433
434	ALLOCATE( flux_s_pt(nzb+1:nzt,0:threads_per_task-1), &
435	flux_s_e(nzb+1:nzt,0:threads_per_task-1), &
436	diss_s_pt(nzb+1:nzt,0:threads_per_task-1), &
437	diss_s_e(nzb+1:nzt,0:threads_per_task-1) )
438	ALLOCATE( flux_l_pt(nzb+1:nzt,nys:nyn,0:threads_per_task-1), &
439	flux_l_e(nzb+1:nzt,nys:nyn,0:threads_per_task-1), &
440	diss_l_pt(nzb+1:nzt,nys:nyn,0:threads_per_task-1), &
441	diss_l_e(nzb+1:nzt,nys:nyn,0:threads_per_task-1) )
442
443	IF ( rans_tke_e ) THEN
444	ALLOCATE( flux_s_diss(nzb+1:nzt,0:threads_per_task-1), &
445	diss_s_diss(nzb+1:nzt,0:threads_per_task-1) )
446	ALLOCATE( flux_l_diss(nzb+1:nzt,nys:nyn,0:threads_per_task-1), &
447	diss_l_diss(nzb+1:nzt,nys:nyn,0:threads_per_task-1) )
448	ENDIF
449
450	IF ( humidity ) THEN
451	ALLOCATE( flux_s_q(nzb+1:nzt,0:threads_per_task-1), &
452	diss_s_q(nzb+1:nzt,0:threads_per_task-1) )
453	ALLOCATE( flux_l_q(nzb+1:nzt,nys:nyn,0:threads_per_task-1), &
454	diss_l_q(nzb+1:nzt,nys:nyn,0:threads_per_task-1) )
455	ENDIF
456
457	IF ( passive_scalar ) THEN
458	ALLOCATE( flux_s_s(nzb+1:nzt,0:threads_per_task-1), &
459	diss_s_s(nzb+1:nzt,0:threads_per_task-1) )
460	ALLOCATE( flux_l_s(nzb+1:nzt,nys:nyn,0:threads_per_task-1), &
461	diss_l_s(nzb+1:nzt,nys:nyn,0:threads_per_task-1) )
462	ENDIF
463
464	IF ( bulk_cloud_model .AND. microphysics_morrison ) THEN
465	ALLOCATE( flux_s_qc(nzb+1:nzt,0:threads_per_task-1), &
466	diss_s_qc(nzb+1:nzt,0:threads_per_task-1), &
467	flux_s_nc(nzb+1:nzt,0:threads_per_task-1), &
468	diss_s_nc(nzb+1:nzt,0:threads_per_task-1) )
469	ALLOCATE( flux_l_qc(nzb+1:nzt,nys:nyn,0:threads_per_task-1), &
470	diss_l_qc(nzb+1:nzt,nys:nyn,0:threads_per_task-1), &
471	flux_l_nc(nzb+1:nzt,nys:nyn,0:threads_per_task-1), &
472	diss_l_nc(nzb+1:nzt,nys:nyn,0:threads_per_task-1) )
473	ENDIF
474
475	IF ( bulk_cloud_model .AND. microphysics_seifert ) THEN
476	ALLOCATE( flux_s_qr(nzb+1:nzt,0:threads_per_task-1), &
477	diss_s_qr(nzb+1:nzt,0:threads_per_task-1), &
478	flux_s_nr(nzb+1:nzt,0:threads_per_task-1), &
479	diss_s_nr(nzb+1:nzt,0:threads_per_task-1) )
480	ALLOCATE( flux_l_qr(nzb+1:nzt,nys:nyn,0:threads_per_task-1), &
481	diss_l_qr(nzb+1:nzt,nys:nyn,0:threads_per_task-1), &
482	flux_l_nr(nzb+1:nzt,nys:nyn,0:threads_per_task-1), &
483	diss_l_nr(nzb+1:nzt,nys:nyn,0:threads_per_task-1) )
484	ENDIF
485
486	IF ( ocean_mode ) THEN
487	ALLOCATE( flux_s_sa(nzb+1:nzt,0:threads_per_task-1), &
488	diss_s_sa(nzb+1:nzt,0:threads_per_task-1) )
489	ALLOCATE( flux_l_sa(nzb+1:nzt,nys:nyn,0:threads_per_task-1), &
490	diss_l_sa(nzb+1:nzt,nys:nyn,0:threads_per_task-1) )
491	ENDIF
492
493	ENDIF
494
495	ENDIF
496
497	END SUBROUTINE ws_init
498
499	!------------------------------------------------------------------------------!
500	! Description:
501	! ------------
502	!> Initialization of flags for WS-scheme used to degrade the order of the scheme
503	!> near walls.
504	!------------------------------------------------------------------------------!
505	SUBROUTINE ws_init_flags
506
507	USE control_parameters, &
508	ONLY: bc_dirichlet_l, bc_dirichlet_n, bc_dirichlet_r, &
509	bc_dirichlet_s, bc_radiation_l, bc_radiation_n, &
510	bc_radiation_r, bc_radiation_s, momentum_advec, scalar_advec
511
512	USE indices, &
513	ONLY: advc_flags_1, advc_flags_2, nbgp, nxl, nxlg, nxlu, nxr, nxrg,&
514	nyn, nyng, nys, nysg, nysv, nzb, nzt, wall_flags_0
515
516	USE kinds
517
518	IMPLICIT NONE
519
520	INTEGER(iwp) :: i !< index variable along x
521	INTEGER(iwp) :: j !< index variable along y
522	INTEGER(iwp) :: k !< index variable along z
523	INTEGER(iwp) :: k_mm !< dummy index along z
524	INTEGER(iwp) :: k_pp !< dummy index along z
525	INTEGER(iwp) :: k_ppp !< dummy index along z
526
527	LOGICAL :: flag_set !< steering variable for advection flags
528
529	ALLOCATE( advc_flags_1(nzb:nzt+1,nysg:nyng,nxlg:nxrg), &
530	advc_flags_2(nzb:nzt+1,nysg:nyng,nxlg:nxrg) )
531	advc_flags_1 = 0
532	advc_flags_2 = 0
533	IF ( scalar_advec == 'ws-scheme' ) THEN
534	!
535	!-- Set flags to steer the degradation of the advection scheme in advec_ws
536	!-- near topography, inflow- and outflow boundaries as well as bottom and
537	!-- top of model domain. advc_flags_1 remains zero for all non-prognostic
538	!-- grid points.
539	DO i = nxl, nxr
540	DO j = nys, nyn
541	DO k = nzb+1, nzt
542	!
543	!-- scalar - x-direction
544	!-- WS1 (0), WS3 (1), WS5 (2)
545	IF ( ( .NOT. BTEST(wall_flags_0(k,j,i+1),0) &
546	.OR. .NOT. BTEST(wall_flags_0(k,j,i+2),0) &
547	.OR. .NOT. BTEST(wall_flags_0(k,j,i-1),0) ) &
548	.OR. ( ( bc_dirichlet_l .OR. bc_radiation_l ) &
549	.AND. i == nxl ) &
550	.OR. ( ( bc_dirichlet_r .OR. bc_radiation_r ) &
551	.AND. i == nxr ) ) &
552	THEN
553	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 0 )
554	ELSEIF ( ( .NOT. BTEST(wall_flags_0(k,j,i+3),0) &
555	.AND. BTEST(wall_flags_0(k,j,i+1),0) &
556	.AND. BTEST(wall_flags_0(k,j,i+2),0) &
557	.AND. BTEST(wall_flags_0(k,j,i-1),0) &
558	) .OR. &
559	( .NOT. BTEST(wall_flags_0(k,j,i-2),0) &
560	.AND. BTEST(wall_flags_0(k,j,i+1),0) &
561	.AND. BTEST(wall_flags_0(k,j,i+2),0) &
562	.AND. BTEST(wall_flags_0(k,j,i-1),0) &
563	) &
564	.OR. &
565	( ( bc_dirichlet_r .OR. bc_radiation_r ) &
566	.AND. i == nxr-1 ) .OR. &
567	( ( bc_dirichlet_l .OR. bc_radiation_l ) &
568	.AND. i == nxlu ) ) & ! why not nxl+1
569	THEN
570	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 1 )
571	ELSEIF ( BTEST(wall_flags_0(k,j,i+1),0) &
572	.AND. BTEST(wall_flags_0(k,j,i+2),0) &
573	.AND. BTEST(wall_flags_0(k,j,i+3),0) &
574	.AND. BTEST(wall_flags_0(k,j,i-1),0) &
575	.AND. BTEST(wall_flags_0(k,j,i-2),0) ) &
576	THEN
577	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 2 )
578	ENDIF
579	!
580	!-- scalar - y-direction
581	!-- WS1 (3), WS3 (4), WS5 (5)
582	IF ( ( .NOT. BTEST(wall_flags_0(k,j+1,i),0) &
583	.OR. .NOT. BTEST(wall_flags_0(k,j+2,i),0) &
584	.OR. .NOT. BTEST(wall_flags_0(k,j-1,i),0)) &
585	.OR. ( ( bc_dirichlet_s .OR. bc_radiation_s ) &
586	.AND. j == nys ) &
587	.OR. ( ( bc_dirichlet_n .OR. bc_radiation_n ) &
588	.AND. j == nyn ) ) &
589	THEN
590	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 3 )
591	!
592	!-- WS3
593	ELSEIF ( ( .NOT. BTEST(wall_flags_0(k,j+3,i),0) &
594	.AND. BTEST(wall_flags_0(k,j+1,i),0) &
595	.AND. BTEST(wall_flags_0(k,j+2,i),0) &
596	.AND. BTEST(wall_flags_0(k,j-1,i),0) &
597	) .OR. &
598	( .NOT. BTEST(wall_flags_0(k,j-2,i),0) &
599	.AND. BTEST(wall_flags_0(k,j+1,i),0) &
600	.AND. BTEST(wall_flags_0(k,j+2,i),0) &
601	.AND. BTEST(wall_flags_0(k,j-1,i),0) &
602	) &
603	.OR. &
604	( ( bc_dirichlet_s .OR. bc_radiation_s ) &
605	.AND. j == nysv ) .OR. & ! why not nys+1
606	( ( bc_dirichlet_n .OR. bc_radiation_n ) &
607	.AND. j == nyn-1 ) ) &
608	THEN
609	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 4 )
610	!
611	!-- WS5
612	ELSEIF ( BTEST(wall_flags_0(k,j+1,i),0) &
613	.AND. BTEST(wall_flags_0(k,j+2,i),0) &
614	.AND. BTEST(wall_flags_0(k,j+3,i),0) &
615	.AND. BTEST(wall_flags_0(k,j-1,i),0) &
616	.AND. BTEST(wall_flags_0(k,j-2,i),0) ) &
617	THEN
618	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 5 )
619	ENDIF
620	!
621	!-- scalar - z-direction
622	!-- WS1 (6), WS3 (7), WS5 (8)
623	IF ( k == nzb+1 ) THEN
624	k_mm = nzb
625	ELSE
626	k_mm = k - 2
627	ENDIF
628	IF ( k > nzt-1 ) THEN
629	k_pp = nzt+1
630	ELSE
631	k_pp = k + 2
632	ENDIF
633	IF ( k > nzt-2 ) THEN
634	k_ppp = nzt+1
635	ELSE
636	k_ppp = k + 3
637	ENDIF
638
639	flag_set = .FALSE.
640	IF ( .NOT. BTEST(wall_flags_0(k-1,j,i),0) .AND. &
641	BTEST(wall_flags_0(k,j,i),0) .OR. &
642	.NOT. BTEST(wall_flags_0(k_pp,j,i),0) .AND. &
643	BTEST(wall_flags_0(k,j,i),0) .OR. &
644	k == nzt ) &
645	THEN
646	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 6 )
647	flag_set = .TRUE.
648	ELSEIF ( ( .NOT. BTEST(wall_flags_0(k_mm,j,i),0) .OR. &
649	.NOT. BTEST(wall_flags_0(k_ppp,j,i),0) ) .AND. &
650	BTEST(wall_flags_0(k-1,j,i),0) .AND. &
651	BTEST(wall_flags_0(k,j,i),0) .AND. &
652	BTEST(wall_flags_0(k+1,j,i),0) .AND. &
653	BTEST(wall_flags_0(k_pp,j,i),0) .OR. &
654	k == nzt - 1 ) &
655	THEN
656	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 7 )
657	flag_set = .TRUE.
658	ELSEIF ( BTEST(wall_flags_0(k_mm,j,i),0) &
659	.AND. BTEST(wall_flags_0(k-1,j,i),0) &
660	.AND. BTEST(wall_flags_0(k,j,i),0) &
661	.AND. BTEST(wall_flags_0(k+1,j,i),0) &
662	.AND. BTEST(wall_flags_0(k_pp,j,i),0) &
663	.AND. BTEST(wall_flags_0(k_ppp,j,i),0) &
664	.AND. .NOT. flag_set ) &
665	THEN
666	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 8 )
667	ENDIF
668
669	ENDDO
670	ENDDO
671	ENDDO
672	ENDIF
673
674	IF ( momentum_advec == 'ws-scheme' ) THEN
675	!
676	!-- Set advc_flags_1 to steer the degradation of the advection scheme in advec_ws
677	!-- near topography, inflow- and outflow boundaries as well as bottom and
678	!-- top of model domain. advc_flags_1 remains zero for all non-prognostic
679	!-- grid points.
680	DO i = nxl, nxr
681	DO j = nys, nyn
682	DO k = nzb+1, nzt
683	!
684	!-- At first, set flags to WS1.
685	!-- Since fluxes are swapped in advec_ws.f90, this is necessary to
686	!-- in order to handle the left/south flux.
687	!-- near vertical walls.
688	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 9 )
689	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 12 )
690	!
691	!-- u component - x-direction
692	!-- WS1 (9), WS3 (10), WS5 (11)
693	IF ( .NOT. BTEST(wall_flags_0(k,j,i+1),1) .OR. &
694	( ( bc_dirichlet_l .OR. bc_radiation_l ) &
695	.AND. i <= nxlu ) .OR. &
696	( ( bc_dirichlet_r .OR. bc_radiation_r ) &
697	.AND. i == nxr ) ) &
698	THEN
699	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 9 )
700	ELSEIF ( ( .NOT. BTEST(wall_flags_0(k,j,i+2),1) .AND. &
701	BTEST(wall_flags_0(k,j,i+1),1) .OR. &
702	.NOT. BTEST(wall_flags_0(k,j,i-1),1) ) &
703	.OR. &
704	( ( bc_dirichlet_r .OR. bc_radiation_r ) &
705	.AND. i == nxr-1 ) .OR. &
706	( ( bc_dirichlet_l .OR. bc_radiation_l ) &
707	.AND. i == nxlu+1) ) &
708	THEN
709	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 10 )
710	!
711	!-- Clear flag for WS1
712	advc_flags_1(k,j,i) = IBCLR( advc_flags_1(k,j,i), 9 )
713	ELSEIF ( BTEST(wall_flags_0(k,j,i+1),1) .AND. &
714	BTEST(wall_flags_0(k,j,i+2),1) .AND. &
715	BTEST(wall_flags_0(k,j,i-1),1) ) &
716	THEN
717	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 11 )
718	!
719	!-- Clear flag for WS1
720	advc_flags_1(k,j,i) = IBCLR( advc_flags_1(k,j,i), 9 )
721	ENDIF
722	!
723	!-- u component - y-direction
724	!-- WS1 (12), WS3 (13), WS5 (14)
725	IF ( .NOT. BTEST(wall_flags_0(k,j+1,i),1) .OR. &
726	( ( bc_dirichlet_s .OR. bc_radiation_s ) &
727	.AND. j == nys ) .OR. &
728	( ( bc_dirichlet_n .OR. bc_radiation_n ) &
729	.AND. j == nyn ) ) &
730	THEN
731	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 12 )
732	ELSEIF ( ( .NOT. BTEST(wall_flags_0(k,j+2,i),1) .AND. &
733	BTEST(wall_flags_0(k,j+1,i),1) .OR. &
734	.NOT. BTEST(wall_flags_0(k,j-1,i),1) ) &
735	.OR. &
736	( ( bc_dirichlet_s .OR. bc_radiation_s ) &
737	.AND. j == nysv ) .OR. &
738	( ( bc_dirichlet_n .OR. bc_radiation_n ) &
739	.AND. j == nyn-1 ) ) &
740	THEN
741	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 13 )
742	!
743	!-- Clear flag for WS1
744	advc_flags_1(k,j,i) = IBCLR( advc_flags_1(k,j,i), 12 )
745	ELSEIF ( BTEST(wall_flags_0(k,j+1,i),1) .AND. &
746	BTEST(wall_flags_0(k,j+2,i),1) .AND. &
747	BTEST(wall_flags_0(k,j-1,i),1) ) &
748	THEN
749	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 14 )
750	!
751	!-- Clear flag for WS1
752	advc_flags_1(k,j,i) = IBCLR( advc_flags_1(k,j,i), 12 )
753	ENDIF
754	!
755	!-- u component - z-direction
756	!-- WS1 (15), WS3 (16), WS5 (17)
757	IF ( k == nzb+1 ) THEN
758	k_mm = nzb
759	ELSE
760	k_mm = k - 2
761	ENDIF
762	IF ( k > nzt-1 ) THEN
763	k_pp = nzt+1
764	ELSE
765	k_pp = k + 2
766	ENDIF
767	IF ( k > nzt-2 ) THEN
768	k_ppp = nzt+1
769	ELSE
770	k_ppp = k + 3
771	ENDIF
772
773	flag_set = .FALSE.
774	IF ( .NOT. BTEST(wall_flags_0(k-1,j,i),1) .AND. &
775	BTEST(wall_flags_0(k,j,i),1) .OR. &
776	.NOT. BTEST(wall_flags_0(k_pp,j,i),1) .AND. &
777	BTEST(wall_flags_0(k,j,i),1) .OR. &
778	k == nzt ) &
779	THEN
780	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 15 )
781	flag_set = .TRUE.
782	ELSEIF ( ( .NOT. BTEST(wall_flags_0(k_mm,j,i),1) .OR. &
783	.NOT. BTEST(wall_flags_0(k_ppp,j,i),1) ) .AND. &
784	BTEST(wall_flags_0(k-1,j,i),1) .AND. &
785	BTEST(wall_flags_0(k,j,i),1) .AND. &
786	BTEST(wall_flags_0(k+1,j,i),1) .AND. &
787	BTEST(wall_flags_0(k_pp,j,i),1) .OR. &
788	k == nzt - 1 ) &
789	THEN
790	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 16 )
791	flag_set = .TRUE.
792	ELSEIF ( BTEST(wall_flags_0(k_mm,j,i),1) .AND. &
793	BTEST(wall_flags_0(k-1,j,i),1) .AND. &
794	BTEST(wall_flags_0(k,j,i),1) .AND. &
795	BTEST(wall_flags_0(k+1,j,i),1) .AND. &
796	BTEST(wall_flags_0(k_pp,j,i),1) .AND. &
797	BTEST(wall_flags_0(k_ppp,j,i),1) .AND. &
798	.NOT. flag_set ) &
799	THEN
800	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 17 )
801	ENDIF
802
803	ENDDO
804	ENDDO
805	ENDDO
806
807	DO i = nxl, nxr
808	DO j = nys, nyn
809	DO k = nzb+1, nzt
810	!
811	!-- At first, set flags to WS1.
812	!-- Since fluxes are swapped in advec_ws.f90, this is necessary to
813	!-- in order to handle the left/south flux.
814	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 18 )
815	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 21 )
816	!
817	!-- v component - x-direction
818	!-- WS1 (18), WS3 (19), WS5 (20)
819	IF ( .NOT. BTEST(wall_flags_0(k,j,i+1),2) .OR. &
820	( ( bc_dirichlet_l .OR. bc_radiation_l ) &
821	.AND. i == nxl ) .OR. &
822	( ( bc_dirichlet_r .OR. bc_radiation_r ) &
823	.AND. i == nxr ) ) &
824	THEN
825	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 18 )
826	!
827	!-- WS3
828	ELSEIF ( ( .NOT. BTEST(wall_flags_0(k,j,i+2),2) .AND. &
829	BTEST(wall_flags_0(k,j,i+1),2) ) .OR. &
830	.NOT. BTEST(wall_flags_0(k,j,i-1),2) &
831	.OR. &
832	( ( bc_dirichlet_r .OR. bc_radiation_r ) &
833	.AND. i == nxr-1 ) .OR. &
834	( ( bc_dirichlet_l .OR. bc_radiation_l ) &
835	.AND. i == nxlu ) ) &
836	THEN
837	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 19 )
838	!
839	!-- Clear flag for WS1
840	advc_flags_1(k,j,i) = IBCLR( advc_flags_1(k,j,i), 18 )
841	ELSEIF ( BTEST(wall_flags_0(k,j,i+1),2) .AND. &
842	BTEST(wall_flags_0(k,j,i+2),2) .AND. &
843	BTEST(wall_flags_0(k,j,i-1),2) ) &
844	THEN
845	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 20 )
846	!
847	!-- Clear flag for WS1
848	advc_flags_1(k,j,i) = IBCLR( advc_flags_1(k,j,i), 18 )
849	ENDIF
850	!
851	!-- v component - y-direction
852	!-- WS1 (21), WS3 (22), WS5 (23)
853	IF ( .NOT. BTEST(wall_flags_0(k,j+1,i),2) .OR. &
854	( ( bc_dirichlet_s .OR. bc_radiation_s ) &
855	.AND. j <= nysv ) .OR. &
856	( ( bc_dirichlet_n .OR. bc_radiation_n ) &
857	.AND. j == nyn ) ) &
858	THEN
859	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 21 )
860	ELSEIF ( ( .NOT. BTEST(wall_flags_0(k,j+2,i),2) .AND. &
861	BTEST(wall_flags_0(k,j+1,i),2) .OR. &
862	.NOT. BTEST(wall_flags_0(k,j-1,i),2) ) &
863	.OR. &
864	( ( bc_dirichlet_s .OR. bc_radiation_s ) &
865	.AND. j == nysv+1) .OR. &
866	( ( bc_dirichlet_n .OR. bc_radiation_n ) &
867	.AND. j == nyn-1 ) ) &
868	THEN
869	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 22 )
870	!
871	!-- Clear flag for WS1
872	advc_flags_1(k,j,i) = IBCLR( advc_flags_1(k,j,i), 21 )
873	ELSEIF ( BTEST(wall_flags_0(k,j+1,i),2) .AND. &
874	BTEST(wall_flags_0(k,j+2,i),2) .AND. &
875	BTEST(wall_flags_0(k,j-1,i),2) ) &
876	THEN
877	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 23 )
878	!
879	!-- Clear flag for WS1
880	advc_flags_1(k,j,i) = IBCLR( advc_flags_1(k,j,i), 21 )
881	ENDIF
882	!
883	!-- v component - z-direction
884	!-- WS1 (24), WS3 (25), WS5 (26)
885	IF ( k == nzb+1 ) THEN
886	k_mm = nzb
887	ELSE
888	k_mm = k - 2
889	ENDIF
890	IF ( k > nzt-1 ) THEN
891	k_pp = nzt+1
892	ELSE
893	k_pp = k + 2
894	ENDIF
895	IF ( k > nzt-2 ) THEN
896	k_ppp = nzt+1
897	ELSE
898	k_ppp = k + 3
899	ENDIF
900
901	flag_set = .FALSE.
902	IF ( .NOT. BTEST(wall_flags_0(k-1,j,i),2) .AND. &
903	BTEST(wall_flags_0(k,j,i),2) .OR. &
904	.NOT. BTEST(wall_flags_0(k_pp,j,i),2) .AND. &
905	BTEST(wall_flags_0(k,j,i),2) .OR. &
906	k == nzt ) &
907	THEN
908	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 24 )
909	flag_set = .TRUE.
910	ELSEIF ( ( .NOT. BTEST(wall_flags_0(k_mm,j,i),2) .OR. &
911	.NOT. BTEST(wall_flags_0(k_ppp,j,i),2) ) .AND. &
912	BTEST(wall_flags_0(k-1,j,i),2) .AND. &
913	BTEST(wall_flags_0(k,j,i),2) .AND. &
914	BTEST(wall_flags_0(k+1,j,i),2) .AND. &
915	BTEST(wall_flags_0(k_pp,j,i),2) .OR. &
916	k == nzt - 1 ) &
917	THEN
918	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 25 )
919	flag_set = .TRUE.
920	ELSEIF ( BTEST(wall_flags_0(k_mm,j,i),2) .AND. &
921	BTEST(wall_flags_0(k-1,j,i),2) .AND. &
922	BTEST(wall_flags_0(k,j,i),2) .AND. &
923	BTEST(wall_flags_0(k+1,j,i),2) .AND. &
924	BTEST(wall_flags_0(k_pp,j,i),2) .AND. &
925	BTEST(wall_flags_0(k_ppp,j,i),2) .AND. &
926	.NOT. flag_set ) &
927	THEN
928	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 26 )
929	ENDIF
930
931	ENDDO
932	ENDDO
933	ENDDO
934	DO i = nxl, nxr
935	DO j = nys, nyn
936	DO k = nzb+1, nzt
937	!
938	!-- At first, set flags to WS1.
939	!-- Since fluxes are swapped in advec_ws.f90, this is necessary to
940	!-- in order to handle the left/south flux.
941	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 27 )
942	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 30 )
943	!
944	!-- w component - x-direction
945	!-- WS1 (27), WS3 (28), WS5 (29)
946	IF ( .NOT. BTEST(wall_flags_0(k,j,i+1),3) .OR. &
947	( ( bc_dirichlet_l .OR. bc_radiation_l ) &
948	.AND. i == nxl ) .OR. &
949	( ( bc_dirichlet_r .OR. bc_radiation_r ) &
950	.AND. i == nxr ) ) &
951	THEN
952	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 27 )
953	ELSEIF ( ( .NOT. BTEST(wall_flags_0(k,j,i+2),3) .AND. &
954	BTEST(wall_flags_0(k,j,i+1),3) .OR. &
955	.NOT. BTEST(wall_flags_0(k,j,i-1),3) ) &
956	.OR. &
957	( ( bc_dirichlet_r .OR. bc_radiation_r ) &
958	.AND. i == nxr-1 ) .OR. &
959	( ( bc_dirichlet_l .OR. bc_radiation_l ) &
960	.AND. i == nxlu ) ) &
961	THEN
962	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 28 )
963	!
964	!-- Clear flag for WS1
965	advc_flags_1(k,j,i) = IBCLR( advc_flags_1(k,j,i), 27 )
966	ELSEIF ( BTEST(wall_flags_0(k,j,i+1),3) .AND. &
967	BTEST(wall_flags_0(k,j,i+2),3) .AND. &
968	BTEST(wall_flags_0(k,j,i-1),3) ) &
969	THEN
970	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i),29 )
971	!
972	!-- Clear flag for WS1
973	advc_flags_1(k,j,i) = IBCLR( advc_flags_1(k,j,i), 27 )
974	ENDIF
975	!
976	!-- w component - y-direction
977	!-- WS1 (30), WS3 (31), WS5 (32)
978	IF ( .NOT. BTEST(wall_flags_0(k,j+1,i),3) .OR. &
979	( ( bc_dirichlet_s .OR. bc_radiation_s ) &
980	.AND. j == nys ) .OR. &
981	( ( bc_dirichlet_n .OR. bc_radiation_n ) &
982	.AND. j == nyn ) ) &
983	THEN
984	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 30 )
985	ELSEIF ( ( .NOT. BTEST(wall_flags_0(k,j+2,i),3) .AND. &
986	BTEST(wall_flags_0(k,j+1,i),3) .OR. &
987	.NOT. BTEST(wall_flags_0(k,j-1,i),3) ) &
988	.OR. &
989	( ( bc_dirichlet_s .OR. bc_radiation_s ) &
990	.AND. j == nysv ) .OR. &
991	( ( bc_dirichlet_n .OR. bc_radiation_n ) &
992	.AND. j == nyn-1 ) ) &
993	THEN
994	advc_flags_1(k,j,i) = IBSET( advc_flags_1(k,j,i), 31 )
995	!
996	!-- Clear flag for WS1
997	advc_flags_1(k,j,i) = IBCLR( advc_flags_1(k,j,i), 30 )
998	ELSEIF ( BTEST(wall_flags_0(k,j+1,i),3) .AND. &
999	BTEST(wall_flags_0(k,j+2,i),3) .AND. &
1000	BTEST(wall_flags_0(k,j-1,i),3) ) &
1001	THEN
1002	advc_flags_2(k,j,i) = IBSET( advc_flags_2(k,j,i), 0 )
1003	!
1004	!-- Clear flag for WS1
1005	advc_flags_1(k,j,i) = IBCLR( advc_flags_1(k,j,i), 30 )
1006	ENDIF
1007	!
1008	!-- w component - z-direction
1009	!-- WS1 (33), WS3 (34), WS5 (35)
1010	flag_set = .FALSE.
1011	IF ( k == nzb+1 ) THEN
1012	k_mm = nzb
1013	ELSE
1014	k_mm = k - 2
1015	ENDIF
1016	IF ( k > nzt-1 ) THEN
1017	k_pp = nzt+1
1018	ELSE
1019	k_pp = k + 2
1020	ENDIF
1021	IF ( k > nzt-2 ) THEN
1022	k_ppp = nzt+1
1023	ELSE
1024	k_ppp = k + 3
1025	ENDIF
1026
1027	IF ( ( .NOT. BTEST(wall_flags_0(k-1,j,i),3) .AND. &
1028	.NOT. BTEST(wall_flags_0(k,j,i),3) .AND. &
1029	BTEST(wall_flags_0(k+1,j,i),3) ) .OR. &
1030	( .NOT. BTEST(wall_flags_0(k-1,j,i),3) .AND. &
1031	BTEST(wall_flags_0(k,j,i),3) ) .OR. &
1032	( .NOT. BTEST(wall_flags_0(k+1,j,i),3) .AND. &
1033	BTEST(wall_flags_0(k,j,i),3) ) .OR. &
1034	k == nzt ) &
1035	THEN
1036	!
1037	!-- Please note, at k == nzb_w_inner(j,i) a flag is explictely
1038	!-- set, although this is not a prognostic level. However,
1039	!-- contrary to the advection of u,v and s this is necessary
1040	!-- because flux_t(nzb_w_inner(j,i)) is used for the tendency
1041	!-- at k == nzb_w_inner(j,i)+1.
1042	advc_flags_2(k,j,i) = IBSET( advc_flags_2(k,j,i), 1 )
1043	flag_set = .TRUE.
1044	ELSEIF ( ( .NOT. BTEST(wall_flags_0(k_mm,j,i),3) .OR. &
1045	.NOT. BTEST(wall_flags_0(k_ppp,j,i),3) ) .AND. &
1046	BTEST(wall_flags_0(k-1,j,i),3) .AND. &
1047	BTEST(wall_flags_0(k,j,i),3) .AND. &
1048	BTEST(wall_flags_0(k+1,j,i),3) .OR. &
1049	k == nzt - 1 ) &
1050	THEN
1051	advc_flags_2(k,j,i) = IBSET( advc_flags_2(k,j,i), 2 )
1052	flag_set = .TRUE.
1053	ELSEIF ( BTEST(wall_flags_0(k_mm,j,i),3) .AND. &
1054	BTEST(wall_flags_0(k-1,j,i),3) .AND. &
1055	BTEST(wall_flags_0(k,j,i),3) .AND. &
1056	BTEST(wall_flags_0(k+1,j,i),3) .AND. &
1057	BTEST(wall_flags_0(k_pp,j,i),3) .AND. &
1058	BTEST(wall_flags_0(k_ppp,j,i),3) .AND. &
1059	.NOT. flag_set ) &
1060	THEN
1061	advc_flags_2(k,j,i) = IBSET( advc_flags_2(k,j,i), 3 )
1062	ENDIF
1063
1064	ENDDO
1065	ENDDO
1066	ENDDO
1067
1068	ENDIF
1069
1070
1071	!
1072	!-- Exchange 3D integer wall_flags.
1073	IF ( momentum_advec == 'ws-scheme' .OR. scalar_advec == 'ws-scheme' &
1074	) THEN
1075	!
1076	!-- Exchange ghost points for advection flags
1077	CALL exchange_horiz_int( advc_flags_1, nys, nyn, nxl, nxr, nzt, nbgp )
1078	CALL exchange_horiz_int( advc_flags_2, nys, nyn, nxl, nxr, nzt, nbgp )
1079	!
1080	!-- Set boundary flags at inflow and outflow boundary in case of
1081	!-- non-cyclic boundary conditions.
1082	IF ( bc_dirichlet_l .OR. bc_radiation_l ) THEN
1083	advc_flags_1(:,:,nxl-1) = advc_flags_1(:,:,nxl)
1084	advc_flags_2(:,:,nxl-1) = advc_flags_2(:,:,nxl)
1085	ENDIF
1086
1087	IF ( bc_dirichlet_r .OR. bc_radiation_r ) THEN
1088	advc_flags_1(:,:,nxr+1) = advc_flags_1(:,:,nxr)
1089	advc_flags_2(:,:,nxr+1) = advc_flags_2(:,:,nxr)
1090	ENDIF
1091
1092	IF ( bc_dirichlet_n .OR. bc_radiation_n ) THEN
1093	advc_flags_1(:,nyn+1,:) = advc_flags_1(:,nyn,:)
1094	advc_flags_2(:,nyn+1,:) = advc_flags_2(:,nyn,:)
1095	ENDIF
1096
1097	IF ( bc_dirichlet_s .OR. bc_radiation_s ) THEN
1098	advc_flags_1(:,nys-1,:) = advc_flags_1(:,nys,:)
1099	advc_flags_2(:,nys-1,:) = advc_flags_2(:,nys,:)
1100	ENDIF
1101
1102	ENDIF
1103
1104
1105	END SUBROUTINE ws_init_flags
1106
1107
1108	!------------------------------------------------------------------------------!
1109	! Description:
1110	! ------------
1111	!> Initialize variables used for storing statistic quantities (fluxes, variances)
1112	!------------------------------------------------------------------------------!
1113	SUBROUTINE ws_statistics
1114
1115	USE control_parameters, &
1116	ONLY: humidity, passive_scalar, ocean_mode, ws_scheme_mom, &
1117	ws_scheme_sca
1118
1119	USE kinds
1120
1121	USE bulk_cloud_model_mod, &
1122	ONLY: bulk_cloud_model, microphysics_morrison, microphysics_seifert
1123
1124	USE salsa_util_mod, &
1125	ONLY: sums_salsa_ws_l
1126
1127	USE statistics, &
1128	ONLY: sums_us2_ws_l, sums_vs2_ws_l, sums_ws2_ws_l, sums_wsncs_ws_l,&
1129	sums_wsnrs_ws_l, sums_wspts_ws_l, sums_wsqcs_ws_l, &
1130	sums_wsqrs_ws_l, sums_wsqs_ws_l, sums_wsss_ws_l, &
1131	sums_wssas_ws_l, sums_wsus_ws_l, sums_wsvs_ws_l
1132
1133
1134	IMPLICIT NONE
1135
1136	!
1137	!-- The arrays needed for statistical evaluation are set to to 0 at the
1138	!-- beginning of prognostic_equations.
1139	IF ( ws_scheme_mom ) THEN
1140	sums_wsus_ws_l = 0.0_wp
1141	sums_wsvs_ws_l = 0.0_wp
1142	sums_us2_ws_l = 0.0_wp
1143	sums_vs2_ws_l = 0.0_wp
1144	sums_ws2_ws_l = 0.0_wp
1145	ENDIF
1146
1147	IF ( ws_scheme_sca ) THEN
1148	sums_wspts_ws_l = 0.0_wp
1149	IF ( humidity ) sums_wsqs_ws_l = 0.0_wp
1150	IF ( passive_scalar ) sums_wsss_ws_l = 0.0_wp
1151	IF ( bulk_cloud_model .AND. microphysics_morrison ) THEN
1152	sums_wsqcs_ws_l = 0.0_wp
1153	sums_wsncs_ws_l = 0.0_wp
1154	ENDIF
1155	IF ( bulk_cloud_model .AND. microphysics_seifert ) THEN
1156	sums_wsqrs_ws_l = 0.0_wp
1157	sums_wsnrs_ws_l = 0.0_wp
1158	ENDIF
1159	IF ( ocean_mode ) sums_wssas_ws_l = 0.0_wp
1160
1161	sums_salsa_ws_l = 0.0_wp
1162
1163
1164	ENDIF
1165
1166	END SUBROUTINE ws_statistics
1167
1168
1169	!------------------------------------------------------------------------------!
1170	! Description:
1171	! ------------
1172	!> Scalar advection - Call for grid point i,j
1173	!------------------------------------------------------------------------------!
1174	SUBROUTINE advec_s_ws_ij( i, j, sk, sk_char, swap_flux_y_local, &
1175	swap_diss_y_local, swap_flux_x_local, &
1176	swap_diss_x_local, i_omp, tn )
1177
1178	USE arrays_3d, &
1179	ONLY: ddzw, drho_air, rho_air_zw, tend, u, u_stokes_zu, v, &
1180	v_stokes_zu, w
1181
1182	USE control_parameters, &
1183	ONLY: intermediate_timestep_count, u_gtrans, v_gtrans
1184
1185	USE grid_variables, &
1186	ONLY: ddx, ddy
1187
1188	USE indices, &
1189	ONLY: nxl, nxlg, nxr, nxrg, nyn, nyng, nys, nysg, nzb, nzb_max, &
1190	nzt, advc_flags_1
1191
1192	USE kinds
1193
1194	USE pegrid
1195
1196	USE salsa_util_mod, &
1197	ONLY: sums_salsa_ws_l
1198
1199	USE statistics, &
1200	ONLY: hom, sums_wsncs_ws_l, sums_wsnrs_ws_l, sums_wspts_ws_l, &
1201	sums_wsqcs_ws_l, sums_wsqrs_ws_l, sums_wsqs_ws_l, &
1202	sums_wssas_ws_l, sums_wsss_ws_l, weight_substep
1203
1204
1205	IMPLICIT NONE
1206
1207	CHARACTER (LEN = *), INTENT(IN) :: sk_char !< string identifier, used for assign fluxes to the correct dimension in the analysis array
1208
1209	INTEGER(iwp) :: i !< grid index along x-direction
1210	INTEGER(iwp) :: i_omp !< leftmost index on subdomain, or in case of OpenMP, on thread
1211	INTEGER(iwp) :: j !< grid index along y-direction
1212	INTEGER(iwp) :: k !< grid index along z-direction
1213	INTEGER(iwp) :: k_mm !< k-2 index in disretization, can be modified to avoid segmentation faults
1214	INTEGER(iwp) :: k_pp !< k+2 index in disretization, can be modified to avoid segmentation faults
1215	INTEGER(iwp) :: k_ppp !< k+3 index in disretization, can be modified to avoid segmentation faults
1216	INTEGER(iwp) :: tn !< number of OpenMP thread
1217
1218	REAL(wp) :: ibit0 !< flag indicating 1st-order scheme along x-direction
1219	REAL(wp) :: ibit1 !< flag indicating 3rd-order scheme along x-direction
1220	REAL(wp) :: ibit2 !< flag indicating 5th-order scheme along x-direction
1221	REAL(wp) :: ibit3 !< flag indicating 1st-order scheme along y-direction
1222	REAL(wp) :: ibit4 !< flag indicating 3rd-order scheme along y-direction
1223	REAL(wp) :: ibit5 !< flag indicating 5th-order scheme along y-direction
1224	REAL(wp) :: ibit6 !< flag indicating 1st-order scheme along z-direction
1225	REAL(wp) :: ibit7 !< flag indicating 3rd-order scheme along z-direction
1226	REAL(wp) :: ibit8 !< flag indicating 5th-order scheme along z-direction
1227	REAL(wp) :: diss_d !< artificial dissipation term at grid box bottom
1228	REAL(wp) :: div !< diverence on scalar grid
1229	REAL(wp) :: flux_d !< 6th-order flux at grid box bottom
1230	REAL(wp) :: u_comp !< advection velocity along x-direction
1231	REAL(wp) :: v_comp !< advection velocity along y-direction
1232	!
1233	!-- sk is an array from parameter list. It should not be a pointer, because
1234	!-- in that case the compiler can not assume a stride 1 and cannot perform
1235	!-- a strided one vector load. Adding the CONTIGUOUS keyword makes things
1236	!-- even worse, because the compiler cannot assume strided one in the
1237	!-- caller side.
1238	REAL(wp), INTENT(IN),DIMENSION(nzb:nzt+1,nysg:nyng,nxlg:nxrg) :: sk !< advected scalar
1239
1240	REAL(wp), DIMENSION(nzb:nzt+1) :: diss_n !< discretized artificial dissipation at northward-side of the grid box
1241	REAL(wp), DIMENSION(nzb:nzt+1) :: diss_r !< discretized artificial dissipation at rightward-side of the grid box
1242	REAL(wp), DIMENSION(nzb:nzt+1) :: diss_t !< discretized artificial dissipation at top of the grid box
1243	REAL(wp), DIMENSION(nzb:nzt+1) :: flux_n !< discretized 6th-order flux at northward-side of the grid box
1244	REAL(wp), DIMENSION(nzb:nzt+1) :: flux_r !< discretized 6th-order flux at rightward-side of the grid box
1245	REAL(wp), DIMENSION(nzb:nzt+1) :: flux_t !< discretized 6th-order flux at top of the grid box
1246
1247	REAL(wp), DIMENSION(nzb+1:nzt,0:threads_per_task-1) :: swap_diss_y_local !< discretized artificial dissipation at southward-side of the grid box
1248	REAL(wp), DIMENSION(nzb+1:nzt,0:threads_per_task-1) :: swap_flux_y_local !< discretized 6th-order flux at northward-side of the grid box
1249
1250	REAL(wp), DIMENSION(nzb+1:nzt,nys:nyn,0:threads_per_task-1) :: swap_diss_x_local !< discretized artificial dissipation at leftward-side of the grid box
1251	REAL(wp), DIMENSION(nzb+1:nzt,nys:nyn,0:threads_per_task-1) :: swap_flux_x_local !< discretized 6th-order flux at leftward-side of the grid box
1252
1253
1254	!
1255	!-- Compute southside fluxes of the respective PE bounds.
1256	IF ( j == nys ) THEN
1257	!
1258	!-- Up to the top of the highest topography.
1259	DO k = nzb+1, nzb_max
1260
1261	ibit5 = REAL( IBITS(advc_flags_1(k,j-1,i),5,1), KIND = wp )
1262	ibit4 = REAL( IBITS(advc_flags_1(k,j-1,i),4,1), KIND = wp )
1263	ibit3 = REAL( IBITS(advc_flags_1(k,j-1,i),3,1), KIND = wp )
1264
1265	v_comp = v(k,j,i) - v_gtrans + v_stokes_zu(k)
1266	swap_flux_y_local(k,tn) = v_comp * ( &
1267	( 37.0_wp * ibit5 * adv_sca_5 &
1268	+ 7.0_wp * ibit4 * adv_sca_3 &
1269	+ ibit3 * adv_sca_1 &
1270	) * &
1271	( sk(k,j,i) + sk(k,j-1,i) ) &
1272	- ( 8.0_wp * ibit5 * adv_sca_5 &
1273	+ ibit4 * adv_sca_3 &
1274	) * &
1275	( sk(k,j+1,i) + sk(k,j-2,i) ) &
1276	+ ( ibit5 * adv_sca_5 &
1277	) * &
1278	( sk(k,j+2,i) + sk(k,j-3,i) ) &
1279	)
1280
1281	swap_diss_y_local(k,tn) = -ABS( v_comp ) * ( &
1282	( 10.0_wp * ibit5 * adv_sca_5 &
1283	+ 3.0_wp * ibit4 * adv_sca_3 &
1284	+ ibit3 * adv_sca_1 &
1285	) * &
1286	( sk(k,j,i) - sk(k,j-1,i) ) &
1287	- ( 5.0_wp * ibit5 * adv_sca_5 &
1288	+ ibit4 * adv_sca_3 &
1289	) * &
1290	( sk(k,j+1,i) - sk(k,j-2,i) ) &
1291	+ ( ibit5 * adv_sca_5 &
1292	) * &
1293	( sk(k,j+2,i) - sk(k,j-3,i) ) &
1294	)
1295
1296	ENDDO
1297	!
1298	!-- Above to the top of the highest topography. No degradation necessary.
1299	DO k = nzb_max+1, nzt
1300
1301	v_comp = v(k,j,i) - v_gtrans + v_stokes_zu(k)
1302	swap_flux_y_local(k,tn) = v_comp * ( &
1303	37.0_wp * ( sk(k,j,i) + sk(k,j-1,i) ) &
1304	- 8.0_wp * ( sk(k,j+1,i) + sk(k,j-2,i) ) &
1305	+ ( sk(k,j+2,i) + sk(k,j-3,i) ) &
1306	) * adv_sca_5
1307	swap_diss_y_local(k,tn) = -ABS( v_comp ) * ( &
1308	10.0_wp * ( sk(k,j,i) - sk(k,j-1,i) ) &
1309	- 5.0_wp * ( sk(k,j+1,i) - sk(k,j-2,i) ) &
1310	+ sk(k,j+2,i) - sk(k,j-3,i) &
1311	) * adv_sca_5
1312
1313	ENDDO
1314
1315	ENDIF
1316	!
1317	!-- Compute leftside fluxes of the respective PE bounds.
1318	IF ( i == i_omp ) THEN
1319
1320	DO k = nzb+1, nzb_max
1321
1322	ibit2 = REAL( IBITS(advc_flags_1(k,j,i-1),2,1), KIND = wp )
1323	ibit1 = REAL( IBITS(advc_flags_1(k,j,i-1),1,1), KIND = wp )
1324	ibit0 = REAL( IBITS(advc_flags_1(k,j,i-1),0,1), KIND = wp )
1325
1326	u_comp = u(k,j,i) - u_gtrans + u_stokes_zu(k)
1327	swap_flux_x_local(k,j,tn) = u_comp * ( &
1328	( 37.0_wp * ibit2 * adv_sca_5 &
1329	+ 7.0_wp * ibit1 * adv_sca_3 &
1330	+ ibit0 * adv_sca_1 &
1331	) * &
1332	( sk(k,j,i) + sk(k,j,i-1) ) &
1333	- ( 8.0_wp * ibit2 * adv_sca_5 &
1334	+ ibit1 * adv_sca_3 &
1335	) * &
1336	( sk(k,j,i+1) + sk(k,j,i-2) ) &
1337	+ ( ibit2 * adv_sca_5 &
1338	) * &
1339	( sk(k,j,i+2) + sk(k,j,i-3) ) &
1340	)
1341
1342	swap_diss_x_local(k,j,tn) = -ABS( u_comp ) * ( &
1343	( 10.0_wp * ibit2 * adv_sca_5 &
1344	+ 3.0_wp * ibit1 * adv_sca_3 &
1345	+ ibit0 * adv_sca_1 &
1346	) * &
1347	( sk(k,j,i) - sk(k,j,i-1) ) &
1348	- ( 5.0_wp * ibit2 * adv_sca_5 &
1349	+ ibit1 * adv_sca_3 &
1350	) * &
1351	( sk(k,j,i+1) - sk(k,j,i-2) ) &
1352	+ ( ibit2 * adv_sca_5 &
1353	) * &
1354	( sk(k,j,i+2) - sk(k,j,i-3) ) &
1355	)
1356
1357	ENDDO
1358
1359	DO k = nzb_max+1, nzt
1360
1361	u_comp = u(k,j,i) - u_gtrans + u_stokes_zu(k)
1362	swap_flux_x_local(k,j,tn) = u_comp * ( &
1363	37.0_wp * ( sk(k,j,i) + sk(k,j,i-1) ) &
1364	- 8.0_wp * ( sk(k,j,i+1) + sk(k,j,i-2) ) &
1365	+ ( sk(k,j,i+2) + sk(k,j,i-3) ) &
1366	) * adv_sca_5
1367
1368	swap_diss_x_local(k,j,tn) = -ABS( u_comp ) * ( &
1369	10.0_wp * ( sk(k,j,i) - sk(k,j,i-1) ) &
1370	- 5.0_wp * ( sk(k,j,i+1) - sk(k,j,i-2) ) &
1371	+ ( sk(k,j,i+2) - sk(k,j,i-3) ) &
1372	) * adv_sca_5
1373
1374	ENDDO
1375
1376	ENDIF
1377	!
1378	!-- Now compute the fluxes and tendency terms for the horizontal and
1379	!-- vertical parts up to the top of the highest topography.
1380	DO k = nzb+1, nzb_max
1381	!
1382	!-- Note: It is faster to conduct all multiplications explicitly, e.g.
1383	!-- * adv_sca_5 ... than to determine a factor and multiplicate the
1384	!-- flux at the end.
1385
1386	ibit2 = REAL( IBITS(advc_flags_1(k,j,i),2,1), KIND = wp )
1387	ibit1 = REAL( IBITS(advc_flags_1(k,j,i),1,1), KIND = wp )
1388	ibit0 = REAL( IBITS(advc_flags_1(k,j,i),0,1), KIND = wp )
1389
1390	u_comp = u(k,j,i+1) - u_gtrans + u_stokes_zu(k)
1391	flux_r(k) = u_comp * ( &
1392	( 37.0_wp * ibit2 * adv_sca_5 &
1393	+ 7.0_wp * ibit1 * adv_sca_3 &
1394	+ ibit0 * adv_sca_1 &
1395	) * &
1396	( sk(k,j,i+1) + sk(k,j,i) ) &
1397	- ( 8.0_wp * ibit2 * adv_sca_5 &
1398	+ ibit1 * adv_sca_3 &
1399	) * &
1400	( sk(k,j,i+2) + sk(k,j,i-1) ) &
1401	+ ( ibit2 * adv_sca_5 &
1402	) * &
1403	( sk(k,j,i+3) + sk(k,j,i-2) ) &
1404	)
1405
1406	diss_r(k) = -ABS( u_comp ) * ( &
1407	( 10.0_wp * ibit2 * adv_sca_5 &
1408	+ 3.0_wp * ibit1 * adv_sca_3 &
1409	+ ibit0 * adv_sca_1 &
1410	) * &
1411	( sk(k,j,i+1) - sk(k,j,i) ) &
1412	- ( 5.0_wp * ibit2 * adv_sca_5 &
1413	+ ibit1 * adv_sca_3 &
1414	) * &
1415	( sk(k,j,i+2) - sk(k,j,i-1) ) &
1416	+ ( ibit2 * adv_sca_5 &
1417	) * &
1418	( sk(k,j,i+3) - sk(k,j,i-2) ) &
1419	)
1420
1421	ibit5 = REAL( IBITS(advc_flags_1(k,j,i),5,1), KIND = wp )
1422	ibit4 = REAL( IBITS(advc_flags_1(k,j,i),4,1), KIND = wp )
1423	ibit3 = REAL( IBITS(advc_flags_1(k,j,i),3,1), KIND = wp )
1424
1425	v_comp = v(k,j+1,i) - v_gtrans + v_stokes_zu(k)
1426	flux_n(k) = v_comp * ( &
1427	( 37.0_wp * ibit5 * adv_sca_5 &
1428	+ 7.0_wp * ibit4 * adv_sca_3 &
1429	+ ibit3 * adv_sca_1 &
1430	) * &
1431	( sk(k,j+1,i) + sk(k,j,i) ) &
1432	- ( 8.0_wp * ibit5 * adv_sca_5 &
1433	+ ibit4 * adv_sca_3 &
1434	) * &
1435	( sk(k,j+2,i) + sk(k,j-1,i) ) &
1436	+ ( ibit5 * adv_sca_5 &
1437	) * &
1438	( sk(k,j+3,i) + sk(k,j-2,i) ) &
1439	)
1440
1441	diss_n(k) = -ABS( v_comp ) * ( &
1442	( 10.0_wp * ibit5 * adv_sca_5 &
1443	+ 3.0_wp * ibit4 * adv_sca_3 &
1444	+ ibit3 * adv_sca_1 &
1445	) * &
1446	( sk(k,j+1,i) - sk(k,j,i) ) &
1447	- ( 5.0_wp * ibit5 * adv_sca_5 &
1448	+ ibit4 * adv_sca_3 &
1449	) * &
1450	( sk(k,j+2,i) - sk(k,j-1,i) ) &
1451	+ ( ibit5 * adv_sca_5 &
1452	) * &
1453	( sk(k,j+3,i) - sk(k,j-2,i) ) &
1454	)
1455	ENDDO
1456	!
1457	!-- Now compute the fluxes and tendency terms for the horizontal and
1458	!-- vertical parts above the top of the highest topography. No degradation
1459	!-- for the horizontal parts, but for the vertical it is stell needed.
1460	DO k = nzb_max+1, nzt
1461
1462	u_comp = u(k,j,i+1) - u_gtrans + u_stokes_zu(k)
1463	flux_r(k) = u_comp * ( &
1464	37.0_wp * ( sk(k,j,i+1) + sk(k,j,i) ) &
1465	- 8.0_wp * ( sk(k,j,i+2) + sk(k,j,i-1) ) &
1466	+ ( sk(k,j,i+3) + sk(k,j,i-2) ) ) * adv_sca_5
1467	diss_r(k) = -ABS( u_comp ) * ( &
1468	10.0_wp * ( sk(k,j,i+1) - sk(k,j,i) ) &
1469	- 5.0_wp * ( sk(k,j,i+2) - sk(k,j,i-1) ) &
1470	+ ( sk(k,j,i+3) - sk(k,j,i-2) ) ) * adv_sca_5
1471
1472	v_comp = v(k,j+1,i) - v_gtrans + v_stokes_zu(k)
1473	flux_n(k) = v_comp * ( &
1474	37.0_wp * ( sk(k,j+1,i) + sk(k,j,i) ) &
1475	- 8.0_wp * ( sk(k,j+2,i) + sk(k,j-1,i) ) &
1476	+ ( sk(k,j+3,i) + sk(k,j-2,i) ) ) * adv_sca_5
1477	diss_n(k) = -ABS( v_comp ) * ( &
1478	10.0_wp * ( sk(k,j+1,i) - sk(k,j,i) ) &
1479	- 5.0_wp * ( sk(k,j+2,i) - sk(k,j-1,i) ) &
1480	+ ( sk(k,j+3,i) - sk(k,j-2,i) ) ) * adv_sca_5
1481
1482	ENDDO
1483	!
1484	!-- Now, compute vertical fluxes. Split loop into a part treating the
1485	!-- lowest 2 grid points with indirect indexing, a main loop without
1486	!-- indirect indexing, and a loop for the uppermost 2 grip points with
1487	!-- indirect indexing. This allows better vectorization for the main loop.
1488	!-- First, compute the flux at model surface, which need has to be
1489	!-- calculated explicetely for the tendency at
1490	!-- the first w-level. For topography wall this is done implicitely by
1491	!-- advc_flags_1.
1492	flux_t(nzb) = 0.0_wp
1493	diss_t(nzb) = 0.0_wp
1494	DO k = nzb+1, nzb+2
1495	ibit8 = REAL( IBITS(advc_flags_1(k,j,i),8,1), KIND = wp )
1496	ibit7 = REAL( IBITS(advc_flags_1(k,j,i),7,1), KIND = wp )
1497	ibit6 = REAL( IBITS(advc_flags_1(k,j,i),6,1), KIND = wp )
1498	!
1499	!-- k index has to be modified near bottom and top, else array
1500	!-- subscripts will be exceeded.
1501	k_ppp = k + 3 * ibit8
1502	k_pp = k + 2 * ( 1 - ibit6 )
1503	k_mm = k - 2 * ibit8
1504
1505
1506	flux_t(k) = w(k,j,i) * rho_air_zw(k) * ( &
1507	( 37.0_wp * ibit8 * adv_sca_5 &
1508	+ 7.0_wp * ibit7 * adv_sca_3 &
1509	+ ibit6 * adv_sca_1 &
1510	) * &
1511	( sk(k+1,j,i) + sk(k,j,i) ) &
1512	- ( 8.0_wp * ibit8 * adv_sca_5 &
1513	+ ibit7 * adv_sca_3 &
1514	) * &
1515	( sk(k_pp,j,i) + sk(k-1,j,i) ) &
1516	+ ( ibit8 * adv_sca_5 &
1517	) * ( sk(k_ppp,j,i)+ sk(k_mm,j,i) ) &
1518	)
1519
1520	diss_t(k) = -ABS( w(k,j,i) ) * rho_air_zw(k) * ( &
1521	( 10.0_wp * ibit8 * adv_sca_5 &
1522	+ 3.0_wp * ibit7 * adv_sca_3 &
1523	+ ibit6 * adv_sca_1 &
1524	) * &
1525	( sk(k+1,j,i) - sk(k,j,i) ) &
1526	- ( 5.0_wp * ibit8 * adv_sca_5 &
1527	+ ibit7 * adv_sca_3 &
1528	) * &
1529	( sk(k_pp,j,i) - sk(k-1,j,i) ) &
1530	+ ( ibit8 * adv_sca_5 &
1531	) * &
1532	( sk(k_ppp,j,i) - sk(k_mm,j,i) ) &
1533	)
1534	ENDDO
1535
1536	DO k = nzb+3, nzt-2
1537	ibit8 = REAL( IBITS(advc_flags_1(k,j,i),8,1), KIND = wp )
1538	ibit7 = REAL( IBITS(advc_flags_1(k,j,i),7,1), KIND = wp )
1539	ibit6 = REAL( IBITS(advc_flags_1(k,j,i),6,1), KIND = wp )
1540
1541	flux_t(k) = w(k,j,i) * rho_air_zw(k) * ( &
1542	( 37.0_wp * ibit8 * adv_sca_5 &
1543	+ 7.0_wp * ibit7 * adv_sca_3 &
1544	+ ibit6 * adv_sca_1 &
1545	) * &
1546	( sk(k+1,j,i) + sk(k,j,i) ) &
1547	- ( 8.0_wp * ibit8 * adv_sca_5 &
1548	+ ibit7 * adv_sca_3 &
1549	) * &
1550	( sk(k+2,j,i) + sk(k-1,j,i) ) &
1551	+ ( ibit8 * adv_sca_5 &
1552	) * ( sk(k+3,j,i)+ sk(k-2,j,i) ) &
1553	)
1554
1555	diss_t(k) = -ABS( w(k,j,i) ) * rho_air_zw(k) * ( &
1556	( 10.0_wp * ibit8 * adv_sca_5 &
1557	+ 3.0_wp * ibit7 * adv_sca_3 &
1558	+ ibit6 * adv_sca_1 &
1559	) * &
1560	( sk(k+1,j,i) - sk(k,j,i) ) &
1561	- ( 5.0_wp * ibit8 * adv_sca_5 &
1562	+ ibit7 * adv_sca_3 &
1563	) * &
1564	( sk(k+2,j,i) - sk(k-1,j,i) ) &
1565	+ ( ibit8 * adv_sca_5 &
1566	) * &
1567	( sk(k+3,j,i) - sk(k-2,j,i) ) &
1568	)
1569	ENDDO
1570
1571	DO k = nzt-1, nzt
1572	ibit8 = REAL( IBITS(advc_flags_1(k,j,i),8,1), KIND = wp )
1573	ibit7 = REAL( IBITS(advc_flags_1(k,j,i),7,1), KIND = wp )
1574	ibit6 = REAL( IBITS(advc_flags_1(k,j,i),6,1), KIND = wp )
1575	!
1576	!-- k index has to be modified near bottom and top, else array
1577	!-- subscripts will be exceeded.
1578	k_ppp = k + 3 * ibit8
1579	k_pp = k + 2 * ( 1 - ibit6 )
1580	k_mm = k - 2 * ibit8
1581
1582
1583	flux_t(k) = w(k,j,i) * rho_air_zw(k) * ( &
1584	( 37.0_wp * ibit8 * adv_sca_5 &
1585	+ 7.0_wp * ibit7 * adv_sca_3 &
1586	+ ibit6 * adv_sca_1 &
1587	) * &
1588	( sk(k+1,j,i) + sk(k,j,i) ) &
1589	- ( 8.0_wp * ibit8 * adv_sca_5 &
1590	+ ibit7 * adv_sca_3 &
1591	) * &
1592	( sk(k_pp,j,i) + sk(k-1,j,i) ) &
1593	+ ( ibit8 * adv_sca_5 &
1594	) * ( sk(k_ppp,j,i)+ sk(k_mm,j,i) ) &
1595	)
1596
1597	diss_t(k) = -ABS( w(k,j,i) ) * rho_air_zw(k) * ( &
1598	( 10.0_wp * ibit8 * adv_sca_5 &
1599	+ 3.0_wp * ibit7 * adv_sca_3 &
1600	+ ibit6 * adv_sca_1 &
1601	) * &
1602	( sk(k+1,j,i) - sk(k,j,i) ) &
1603	- ( 5.0_wp * ibit8 * adv_sca_5 &
1604	+ ibit7 * adv_sca_3 &
1605	) * &
1606	( sk(k_pp,j,i) - sk(k-1,j,i) ) &
1607	+ ( ibit8 * adv_sca_5 &
1608	) * &
1609	( sk(k_ppp,j,i) - sk(k_mm,j,i) ) &
1610	)
1611	ENDDO
1612
1613	DO k = nzb+1, nzt
1614
1615	flux_d = flux_t(k-1)
1616	diss_d = diss_t(k-1)
1617
1618	!
1619	!-- Calculate the divergence of the velocity field. A respective
1620	!-- correction is needed to overcome numerical instabilities introduced
1621	!-- by a not sufficient reduction of divergences near topography.
1622	div = ( u(k,j,i+1) * ( ibit0 + ibit1 + ibit2 ) &
1623	- u(k,j,i) * ( &
1624	REAL( IBITS(advc_flags_1(k,j,i-1),0,1), KIND = wp ) &
1625	+ REAL( IBITS(advc_flags_1(k,j,i-1),1,1), KIND = wp ) &
1626	+ REAL( IBITS(advc_flags_1(k,j,i-1),2,1), KIND = wp ) &
1627	) &
1628	) * ddx &
1629	+ ( v(k,j+1,i) * ( ibit3 + ibit4 + ibit5 ) &
1630	- v(k,j,i) * ( &
1631	REAL( IBITS(advc_flags_1(k,j-1,i),3,1), KIND = wp ) &
1632	+ REAL( IBITS(advc_flags_1(k,j-1,i),4,1), KIND = wp ) &
1633	+ REAL( IBITS(advc_flags_1(k,j-1,i),5,1), KIND = wp ) &
1634	) &
1635	) * ddy &
1636	+ ( w(k,j,i) * rho_air_zw(k) * &
1637	( ibit6 + ibit7 + ibit8 ) &
1638	- w(k-1,j,i) * rho_air_zw(k-1) * &
1639	( &
1640	REAL( IBITS(advc_flags_1(k-1,j,i),6,1), KIND = wp ) &
1641	+ REAL( IBITS(advc_flags_1(k-1,j,i),7,1), KIND = wp ) &
1642	+ REAL( IBITS(advc_flags_1(k-1,j,i),8,1), KIND = wp ) &
1643	) &
1644	) * drho_air(k) * ddzw(k)
1645
1646	tend(k,j,i) = tend(k,j,i) - ( &
1647	( flux_r(k) + diss_r(k) - swap_flux_x_local(k,j,tn) - &
1648	swap_diss_x_local(k,j,tn) ) * ddx &
1649	+ ( flux_n(k) + diss_n(k) - swap_flux_y_local(k,tn) - &
1650	swap_diss_y_local(k,tn) ) * ddy &
1651	+ ( ( flux_t(k) + diss_t(k) ) - &
1652	( flux_d + diss_d ) &
1653	) * drho_air(k) * ddzw(k) &
1654	) + sk(k,j,i) * div
1655
1656
1657	swap_flux_y_local(k,tn) = flux_n(k)
1658	swap_diss_y_local(k,tn) = diss_n(k)
1659	swap_flux_x_local(k,j,tn) = flux_r(k)
1660	swap_diss_x_local(k,j,tn) = diss_r(k)
1661
1662	ENDDO
1663
1664	!
1665	!-- Evaluation of statistics.
1666	SELECT CASE ( sk_char )
1667
1668	CASE ( 'pt' )
1669
1670	DO k = nzb, nzt
1671	sums_wspts_ws_l(k,tn) = sums_wspts_ws_l(k,tn) + &
1672	( flux_t(k) / ( w(k,j,i) + SIGN( 1.0E-20_wp, w(k,j,i) ) ) &
1673	* ( w(k,j,i) - hom(k,1,3,0) ) &
1674	+ diss_t(k) / ( ABS(w(k,j,i)) + 1.0E-20_wp ) &
1675	* ABS( w(k,j,i) - hom(k,1,3,0) ) &
1676	) * weight_substep(intermediate_timestep_count)
1677	ENDDO
1678
1679	CASE ( 'sa' )
1680
1681	DO k = nzb, nzt
1682	sums_wssas_ws_l(k,tn) = sums_wssas_ws_l(k,tn) + &
1683	( flux_t(k) / ( w(k,j,i) + SIGN( 1.0E-20_wp, w(k,j,i) ) ) &
1684	* ( w(k,j,i) - hom(k,1,3,0) ) &
1685	+ diss_t(k) / ( ABS(w(k,j,i)) + 1.0E-20_wp ) &
1686	* ABS( w(k,j,i) - hom(k,1,3,0) ) &
1687	) * weight_substep(intermediate_timestep_count)
1688	ENDDO
1689
1690	CASE ( 'q' )
1691
1692	DO k = nzb, nzt
1693	sums_wsqs_ws_l(k,tn) = sums_wsqs_ws_l(k,tn) + &
1694	( flux_t(k) / ( w(k,j,i) + SIGN( 1.0E-20_wp, w(k,j,i) ) ) &
1695	* ( w(k,j,i) - hom(k,1,3,0) ) &
1696	+ diss_t(k) / ( ABS(w(k,j,i)) + 1.0E-20_wp ) &
1697	* ABS( w(k,j,i) - hom(k,1,3,0) ) &
1698	) * weight_substep(intermediate_timestep_count)
1699	ENDDO
1700
1701	CASE ( 'qc' )
1702
1703	DO k = nzb, nzt
1704	sums_wsqcs_ws_l(k,tn) = sums_wsqcs_ws_l(k,tn) + &
1705	( flux_t(k) / ( w(k,j,i) + SIGN( 1.0E-20_wp, w(k,j,i) ) ) &
1706	* ( w(k,j,i) - hom(k,1,3,0) ) &
1707	+ diss_t(k) / ( ABS(w(k,j,i)) + 1.0E-20_wp ) &
1708	* ABS( w(k,j,i) - hom(k,1,3,0) ) &
1709	) * weight_substep(intermediate_timestep_count)
1710	ENDDO
1711
1712
1713	CASE ( 'qr' )
1714
1715	DO k = nzb, nzt
1716	sums_wsqrs_ws_l(k,tn) = sums_wsqrs_ws_l(k,tn) + &
1717	( flux_t(k) / ( w(k,j,i) + SIGN( 1.0E-20_wp, w(k,j,i) ) ) &
1718	* ( w(k,j,i) - hom(k,1,3,0) ) &
1719	+ diss_t(k) / ( ABS(w(k,j,i)) + 1.0E-20_wp ) &
1720	* ABS( w(k,j,i) - hom(k,1,3,0) ) &
1721	) * weight_substep(intermediate_timestep_count)
1722	ENDDO
1723
1724	CASE ( 'nc' )
1725
1726	DO k = nzb, nzt
1727	sums_wsncs_ws_l(k,tn) = sums_wsncs_ws_l(k,tn) + &
1728	( flux_t(k) / ( w(k,j,i) + SIGN( 1.0E-20_wp, w(k,j,i) ) ) &
1729	* ( w(k,j,i) - hom(k,1,3,0) ) &
1730	+ diss_t(k) / ( ABS(w(k,j,i)) + 1.0E-20_wp ) &
1731	* ABS( w(k,j,i) - hom(k,1,3,0) ) &
1732	) * weight_substep(intermediate_timestep_count)
1733	ENDDO
1734
1735	CASE ( 'nr' )
1736
1737	DO k = nzb, nzt
1738	sums_wsnrs_ws_l(k,tn) = sums_wsnrs_ws_l(k,tn) + &
1739	( flux_t(k) / ( w(k,j,i) + SIGN( 1.0E-20_wp, w(k,j,i) ) ) &
1740	* ( w(k,j,i) - hom(k,1,3,0) ) &
1741	+ diss_t(k) / ( ABS(w(k,j,i)) + 1.0E-20_wp ) &
1742	* ABS( w(k,j,i) - hom(k,1,3,0) ) &
1743	) * weight_substep(intermediate_timestep_count)
1744	ENDDO
1745
1746	CASE ( 's' )
1747
1748	DO k = nzb, nzt
1749	sums_wsss_ws_l(k,tn) = sums_wsss_ws_l(k,tn) + &
1750	( flux_t(k) / ( w(k,j,i) + SIGN( 1.0E-20_wp, w(k,j,i) ) ) &
1751	* ( w(k,j,i) - hom(k,1,3,0) ) &
1752	+ diss_t(k) / ( ABS(w(k,j,i)) + 1.0E-20_wp ) &
1753	* ABS( w(k,j,i) - hom(k,1,3,0) ) &
1754	) * weight_substep(intermediate_timestep_count)
1755	ENDDO
1756
1757	CASE ( 'aerosol_mass', 'aerosol_number', 'salsa_gas' )
1758
1759	DO k = nzb, nzt
1760	sums_salsa_ws_l(k,tn) = sums_salsa_ws_l(k,tn) + &
1761	( flux_t(k) / ( w(k,j,i) + SIGN( 1.0E-20_wp, w(k,j,i) ) ) &
1762	* ( w(k,j,i) - hom(k,1,3,0) ) &
1763	+ diss_t(k) / ( ABS(w(k,j,i)) + 1.0E-20_wp ) &
1764	* ABS( w(k,j,i) - hom(k,1,3,0) ) &
1765	) * weight_substep(intermediate_timestep_count)
1766	ENDDO
1767
1768	! CASE ( 'kc' )
1769	!kk Has to be implemented for kpp chemistry
1770
1771
1772	END SELECT
1773
1774	END SUBROUTINE advec_s_ws_ij
1775
1776
1777
1778
1779	!------------------------------------------------------------------------------!
1780	! Description:
1781	! ------------
1782	!> Advection of u-component - Call for grid point i,j
1783	!------------------------------------------------------------------------------!
1784	SUBROUTINE advec_u_ws_ij( i, j, i_omp, tn )
1785
1786	USE arrays_3d, &
1787	ONLY: ddzw, diss_l_u, diss_s_u, flux_l_u, flux_s_u, tend, u, v, w,&
1788	drho_air, rho_air_zw
1789
1790	USE control_parameters, &
1791	ONLY: intermediate_timestep_count, u_gtrans, v_gtrans
1792
1793	USE grid_variables, &
1794	ONLY: ddx, ddy
1795
1796	USE indices, &
1797	ONLY: nxlu, nys, nzb, nzb_max, nzt, advc_flags_1
1798
1799	USE kinds
1800
1801	USE statistics, &
1802	ONLY: hom, sums_us2_ws_l, sums_wsus_ws_l, weight_substep
1803
1804	IMPLICIT NONE
1805
1806	INTEGER(iwp) :: i !< grid index along x-direction
1807	INTEGER(iwp) :: i_omp !< leftmost index on subdomain, or in case of OpenMP, on thread
1808	INTEGER(iwp) :: j !< grid index along y-direction
1809	INTEGER(iwp) :: k !< grid index along z-direction
1810	INTEGER(iwp) :: k_mm !< k-2 index in disretization, can be modified to avoid segmentation faults
1811	INTEGER(iwp) :: k_pp !< k+2 index in disretization, can be modified to avoid segmentation faults
1812	INTEGER(iwp) :: k_ppp !< k+3 index in disretization, can be modified to avoid segmentation faults
1813	INTEGER(iwp) :: tn !< number of OpenMP thread
1814
1815	REAL(wp) :: ibit9 !< flag indicating 1st-order scheme along x-direction
1816	REAL(wp) :: ibit10 !< flag indicating 3rd-order scheme along x-direction
1817	REAL(wp) :: ibit11 !< flag indicating 5th-order scheme along x-direction
1818	REAL(wp) :: ibit12 !< flag indicating 1st-order scheme along y-direction
1819	REAL(wp) :: ibit13 !< flag indicating 3rd-order scheme along y-direction
1820	REAL(wp) :: ibit14 !< flag indicating 5th-order scheme along y-direction
1821	REAL(wp) :: ibit15 !< flag indicating 1st-order scheme along z-direction
1822	REAL(wp) :: ibit16 !< flag indicating 3rd-order scheme along z-direction
1823	REAL(wp) :: ibit17 !< flag indicating 5th-order scheme along z-direction
1824	REAL(wp) :: diss_d !< artificial dissipation term at grid box bottom
1825	REAL(wp) :: div !< diverence on u-grid
1826	REAL(wp) :: flux_d !< 6th-order flux at grid box bottom
1827	REAL(wp) :: gu !< Galilei-transformation velocity along x
1828	REAL(wp) :: gv !< Galilei-transformation velocity along y
1829	REAL(wp) :: u_comp_l !< advection velocity along x at leftmost grid point on subdomain
1830
1831	REAL(wp), DIMENSION(nzb:nzt+1) :: diss_n !< discretized artificial dissipation at northward-side of the grid box
1832	REAL(wp), DIMENSION(nzb:nzt+1) :: diss_r !< discretized artificial dissipation at rightward-side of the grid box
1833	REAL(wp), DIMENSION(nzb:nzt+1) :: diss_t !< discretized artificial dissipation at top of the grid box
1834	REAL(wp), DIMENSION(nzb:nzt+1) :: flux_n !< discretized 6th-order flux at northward-side of the grid box
1835	REAL(wp), DIMENSION(nzb:nzt+1) :: flux_r !< discretized 6th-order flux at rightward-side of the grid box
1836	REAL(wp), DIMENSION(nzb:nzt+1) :: flux_t !< discretized 6th-order flux at top of the grid box
1837	REAL(wp), DIMENSION(nzb:nzt+1) :: u_comp !< advection velocity along x
1838	REAL(wp), DIMENSION(nzb:nzt+1) :: v_comp !< advection velocity along y
1839	REAL(wp), DIMENSION(nzb:nzt+1) :: w_comp !< advection velocity along z
1840
1841	gu = 2.0_wp * u_gtrans
1842	gv = 2.0_wp * v_gtrans
1843	!
1844	!-- Compute southside fluxes for the respective boundary of PE
1845	IF ( j == nys ) THEN
1846
1847	DO k = nzb+1, nzb_max
1848
1849	ibit14 = REAL( IBITS(advc_flags_1(k,j-1,i),14,1), KIND = wp )
1850	ibit13 = REAL( IBITS(advc_flags_1(k,j-1,i),13,1), KIND = wp )
1851	ibit12 = REAL( IBITS(advc_flags_1(k,j-1,i),12,1), KIND = wp )
1852
1853	v_comp(k) = v(k,j,i) + v(k,j,i-1) - gv
1854	flux_s_u(k,tn) = v_comp(k) * ( &
1855	( 37.0_wp * ibit14 * adv_mom_5 &
1856	+ 7.0_wp * ibit13 * adv_mom_3 &
1857	+ ibit12 * adv_mom_1 &
1858	) * &
1859	( u(k,j,i) + u(k,j-1,i) ) &
1860	- ( 8.0_wp * ibit14 * adv_mom_5 &
1861	+ ibit13 * adv_mom_3 &
1862	) * &
1863	( u(k,j+1,i) + u(k,j-2,i) ) &
1864	+ ( ibit14 * adv_mom_5 &
1865	) * &
1866	( u(k,j+2,i) + u(k,j-3,i) ) &
1867	)
1868
1869	diss_s_u(k,tn) = - ABS ( v_comp(k) ) * ( &
1870	( 10.0_wp * ibit14 * adv_mom_5 &
1871	+ 3.0_wp * ibit13 * adv_mom_3 &
1872	+ ibit12 * adv_mom_1 &
1873	) * &
1874	( u(k,j,i) - u(k,j-1,i) ) &
1875	- ( 5.0_wp * ibit14 * adv_mom_5 &
1876	+ ibit13 * adv_mom_3 &
1877	) * &
1878	( u(k,j+1,i) - u(k,j-2,i) ) &
1879	+ ( ibit14 * adv_mom_5 &
1880	) * &
1881	( u(k,j+2,i) - u(k,j-3,i) ) &
1882	)
1883
1884	ENDDO
1885
1886	DO k = nzb_max+1, nzt
1887
1888	v_comp(k) = v(k,j,i) + v(k,j,i-1) - gv
1889	flux_s_u(k,tn) = v_comp(k) * ( &
1890	37.0_wp * ( u(k,j,i) + u(k,j-1,i) ) &
1891	- 8.0_wp * ( u(k,j+1,i) + u(k,j-2,i) ) &
1892	+ ( u(k,j+2,i) + u(k,j-3,i) ) ) * adv_mom_5
1893	diss_s_u(k,tn) = - ABS(v_comp(k)) * ( &
1894	10.0_wp * ( u(k,j,i) - u(k,j-1,i) ) &
1895	- 5.0_wp * ( u(k,j+1,i) - u(k,j-2,i) ) &
1896	+ ( u(k,j+2,i) - u(k,j-3,i) ) ) * adv_mom_5
1897
1898	ENDDO
1899
1900	ENDIF
1901	!
1902	!-- Compute leftside fluxes for the respective boundary of PE
1903	IF ( i == i_omp .OR. i == nxlu ) THEN
1904
1905	DO k = nzb+1, nzb_max
1906
1907	ibit11 = REAL( IBITS(advc_flags_1(k,j,i-1),11,1), KIND = wp )
1908	ibit10 = REAL( IBITS(advc_flags_1(k,j,i-1),10,1), KIND = wp )
1909	ibit9 = REAL( IBITS(advc_flags_1(k,j,i-1),9,1), KIND = wp )
1910
1911	u_comp_l = u(k,j,i) + u(k,j,i-1) - gu
1912	flux_l_u(k,j,tn) = u_comp_l * ( &
1913	( 37.0_wp * ibit11 * adv_mom_5 &
1914	+ 7.0_wp * ibit10 * adv_mom_3 &
1915	+ ibit9 * adv_mom_1 &
1916	) * &
1917	( u(k,j,i) + u(k,j,i-1) ) &
1918	- ( 8.0_wp * ibit11 * adv_mom_5 &
1919	+ ibit10 * adv_mom_3 &
1920	) * &
1921	( u(k,j,i+1) + u(k,j,i-2) ) &
1922	+ ( ibit11 * adv_mom_5 &
1923	) * &
1924	( u(k,j,i+2) + u(k,j,i-3) ) &
1925	)
1926
1927	diss_l_u(k,j,tn) = - ABS( u_comp_l ) * ( &
1928	( 10.0_wp * ibit11 * adv_mom_5 &
1929	+ 3.0_wp * ibit10 * adv_mom_3 &
1930	+ ibit9 * adv_mom_1 &
1931	) * &
1932	( u(k,j,i) - u(k,j,i-1) ) &
1933	- ( 5.0_wp * ibit11 * adv_mom_5 &
1934	+ ibit10 * adv_mom_3 &
1935	) * &
1936	( u(k,j,i+1) - u(k,j,i-2) ) &
1937	+ ( ibit11 * adv_mom_5 &
1938	) * &
1939	( u(k,j,i+2) - u(k,j,i-3) ) &
1940	)
1941
1942	ENDDO
1943
1944	DO k = nzb_max+1, nzt
1945
1946	u_comp_l = u(k,j,i) + u(k,j,i-1) - gu
1947	flux_l_u(k,j,tn) = u_comp_l * ( &
1948	37.0_wp * ( u(k,j,i) + u(k,j,i-1) ) &
1949	- 8.0_wp * ( u(k,j,i+1) + u(k,j,i-2) ) &
1950	+ ( u(k,j,i+2) + u(k,j,i-3) ) ) * adv_mom_5
1951	diss_l_u(k,j,tn) = - ABS(u_comp_l) * ( &
1952	10.0_wp * ( u(k,j,i) - u(k,j,i-1) ) &
1953	- 5.0_wp * ( u(k,j,i+1) - u(k,j,i-2) ) &
1954	+ ( u(k,j,i+2) - u(k,j,i-3) ) ) * adv_mom_5
1955
1956	ENDDO
1957
1958	ENDIF
1959	!
1960	!-- Now compute the fluxes tendency terms for the horizontal and
1961	!-- vertical parts.
1962	DO k = nzb+1, nzb_max
1963
1964	ibit11 = REAL( IBITS(advc_flags_1(k,j,i),11,1), KIND = wp )
1965	ibit10 = REAL( IBITS(advc_flags_1(k,j,i),10,1), KIND = wp )
1966	ibit9 = REAL( IBITS(advc_flags_1(k,j,i),9,1), KIND = wp )
1967
1968	u_comp(k) = u(k,j,i+1) + u(k,j,i)
1969	flux_r(k) = ( u_comp(k) - gu ) * ( &
1970	( 37.0_wp * ibit11 * adv_mom_5 &
1971	+ 7.0_wp * ibit10 * adv_mom_3 &
1972	+ ibit9 * adv_mom_1 &
1973	) * &
1974	( u(k,j,i+1) + u(k,j,i) ) &
1975	- ( 8.0_wp * ibit11 * adv_mom_5 &
1976	+ ibit10 * adv_mom_3 &
1977	) * &
1978	( u(k,j,i+2) + u(k,j,i-1) ) &
1979	+ ( ibit11 * adv_mom_5 &
1980	) * &
1981	( u(k,j,i+3) + u(k,j,i-2) ) &
1982	)
1983
1984	diss_r(k) = - ABS( u_comp(k) - gu ) * ( &
1985	( 10.0_wp * ibit11 * adv_mom_5 &
1986	+ 3.0_wp * ibit10 * adv_mom_3 &
1987	+ ibit9 * adv_mom_1 &
1988	) * &
1989	( u(k,j,i+1) - u(k,j,i) ) &
1990	- ( 5.0_wp * ibit11 * adv_mom_5 &
1991	+ ibit10 * adv_mom_3 &
1992	) * &
1993	( u(k,j,i+2) - u(k,j,i-1) ) &
1994	+ ( ibit11 * adv_mom_5 &
1995	) * &
1996	( u(k,j,i+3) - u(k,j,i-2) ) &
1997	)
1998
1999	ibit14 = REAL( IBITS(advc_flags_1(k,j,i),14,1), KIND = wp )
2000	ibit13 = REAL( IBITS(advc_flags_1(k,j,i),13,1), KIND = wp )
2001	ibit12 = REAL( IBITS(advc_flags_1(k,j,i),12,1), KIND = wp )
2002
2003	v_comp(k) = v(k,j+1,i) + v(k,j+1,i-1) - gv
2004	flux_n(k) = v_comp(k) * ( &
2005	( 37.0_wp * ibit14 * adv_mom_5 &
2006	+ 7.0_wp * ibit13 * adv_mom_3 &
2007	+ ibit12 * adv_mom_1 &
2008	) * &
2009	( u(k,j+1,i) + u(k,j,i) ) &
2010	- ( 8.0_wp * ibit14 * adv_mom_5 &
2011	+ ibit13 * adv_mom_3 &
2012	) * &
2013	( u(k,j+2,i) + u(k,j-1,i) ) &
2014	+ ( ibit14 * adv_mom_5 &
2015	) * &
2016	( u(k,j+3,i) + u(k,j-2,i) ) &
2017	)
2018
2019	diss_n(k) = - ABS ( v_comp(k) ) * ( &
2020	( 10.0_wp * ibit14 * adv_mom_5 &
2021	+ 3.0_wp * ibit13 * adv_mom_3 &
2022	+ ibit12 * adv_mom_1 &
2023	) * &
2024	( u(k,j+1,i) - u(k,j,i) ) &
2025	- ( 5.0_wp * ibit14 * adv_mom_5 &
2026	+ ibit13 * adv_mom_3 &
2027	) * &
2028	( u(k,j+2,i) - u(k,j-1,i) ) &
2029	+ ( ibit14 * adv_mom_5 &
2030	) * &
2031	( u(k,j+3,i) - u(k,j-2,i) ) &
2032	)
2033	ENDDO
2034
2035	DO k = nzb_max+1, nzt
2036
2037	u_comp(k) = u(k,j,i+1) + u(k,j,i)
2038	flux_r(k) = ( u_comp(k) - gu ) * ( &
2039	37.0_wp * ( u(k,j,i+1) + u(k,j,i) ) &
2040	- 8.0_wp * ( u(k,j,i+2) + u(k,j,i-1) ) &
2041	+ ( u(k,j,i+3) + u(k,j,i-2) ) ) * adv_mom_5
2042	diss_r(k) = - ABS( u_comp(k) - gu ) * ( &
2043	10.0_wp * ( u(k,j,i+1) - u(k,j,i) ) &
2044	- 5.0_wp * ( u(k,j,i+2) - u(k,j,i-1) ) &
2045	+ ( u(k,j,i+3) - u(k,j,i-2) ) ) * adv_mom_5
2046
2047	v_comp(k) = v(k,j+1,i) + v(k,j+1,i-1) - gv
2048	flux_n(k) = v_comp(k) * ( &
2049	37.0_wp * ( u(k,j+1,i) + u(k,j,i) ) &
2050	- 8.0_wp * ( u(k,j+2,i) + u(k,j-1,i) ) &
2051	+ ( u(k,j+3,i) + u(k,j-2,i) ) ) * adv_mom_5
2052	diss_n(k) = - ABS( v_comp(k) ) * ( &
2053	10.0_wp * ( u(k,j+1,i) - u(k,j,i) ) &
2054	- 5.0_wp * ( u(k,j+2,i) - u(k,j-1,i) ) &
2055	+ ( u(k,j+3,i) - u(k,j-2,i) ) ) * adv_mom_5
2056
2057	ENDDO
2058	!
2059	!-- Now, compute vertical fluxes. Split loop into a part treating the
2060	!-- lowest 2 grid points with indirect indexing, a main loop without
2061	!-- indirect indexing, and a loop for the uppermost 2 grip points with
2062	!-- indirect indexing. This allows better vectorization for the main loop.
2063	!-- First, compute the flux at model surface, which need has to be
2064	!-- calculated explicetely for the tendency at
2065	!-- the first w-level. For topography wall this is done implicitely by
2066	!-- advc_flags_1.
2067	flux_t(nzb) = 0.0_wp
2068	diss_t(nzb) = 0.0_wp
2069	w_comp(nzb) = 0.0_wp
2070	DO k = nzb+1, nzb+2
2071	!
2072	!-- k index has to be modified near bottom and top, else array
2073	!-- subscripts will be exceeded.
2074	ibit17 = REAL( IBITS(advc_flags_1(k,j,i),17,1), KIND = wp )
2075	ibit16 = REAL( IBITS(advc_flags_1(k,j,i),16,1), KIND = wp )
2076	ibit15 = REAL( IBITS(advc_flags_1(k,j,i),15,1), KIND = wp )
2077
2078	k_ppp = k + 3 * ibit17
2079	k_pp = k + 2 * ( 1 - ibit15 )
2080	k_mm = k - 2 * ibit17
2081
2082	w_comp(k) = w(k,j,i) + w(k,j,i-1)
2083	flux_t(k) = w_comp(k) * rho_air_zw(k) * ( &
2084	( 37.0_wp * ibit17 * adv_mom_5 &
2085	+ 7.0_wp * ibit16 * adv_mom_3 &
2086	+ ibit15 * adv_mom_1 &
2087	) * &
2088	( u(k+1,j,i) + u(k,j,i) ) &
2089	- ( 8.0_wp * ibit17 * adv_mom_5 &
2090	+ ibit16 * adv_mom_3 &
2091	) * &
2092	( u(k_pp,j,i) + u(k-1,j,i) ) &
2093	+ ( ibit17 * adv_mom_5 &
2094	) * &
2095	( u(k_ppp,j,i) + u(k_mm,j,i) ) &
2096	)
2097
2098	diss_t(k) = - ABS( w_comp(k) ) * rho_air_zw(k) * ( &
2099	( 10.0_wp * ibit17 * adv_mom_5 &
2100	+ 3.0_wp * ibit16 * adv_mom_3 &
2101	+ ibit15 * adv_mom_1 &
2102	) * &
2103	( u(k+1,j,i) - u(k,j,i) ) &
2104	- ( 5.0_wp * ibit17 * adv_mom_5 &
2105	+ ibit16 * adv_mom_3 &
2106	) * &
2107	( u(k_pp,j,i) - u(k-1,j,i) ) &
2108	+ ( ibit17 * adv_mom_5 &
2109	) * &
2110	( u(k_ppp,j,i) - u(k_mm,j,i) ) &
2111	)
2112	ENDDO
2113
2114	DO k = nzb+3, nzt-2
2115
2116	ibit17 = REAL( IBITS(advc_flags_1(k,j,i),17,1), KIND = wp )
2117	ibit16 = REAL( IBITS(advc_flags_1(k,j,i),16,1), KIND = wp )
2118	ibit15 = REAL( IBITS(advc_flags_1(k,j,i),15,1), KIND = wp )
2119
2120	w_comp(k) = w(k,j,i) + w(k,j,i-1)
2121	flux_t(k) = w_comp(k) * rho_air_zw(k) * ( &
2122	( 37.0_wp * ibit17 * adv_mom_5 &
2123	+ 7.0_wp * ibit16 * adv_mom_3 &
2124	+ ibit15 * adv_mom_1 &
2125	) * &
2126	( u(k+1,j,i) + u(k,j,i) ) &
2127	- ( 8.0_wp * ibit17 * adv_mom_5 &
2128	+ ibit16 * adv_mom_3 &
2129	) * &
2130	( u(k+2,j,i) + u(k-1,j,i) ) &
2131	+ ( ibit17 * adv_mom_5 &
2132	) * &
2133	( u(k+3,j,i) + u(k-2,j,i) ) &
2134	)
2135
2136	diss_t(k) = - ABS( w_comp(k) ) * rho_air_zw(k) * ( &
2137	( 10.0_wp * ibit17 * adv_mom_5 &
2138	+ 3.0_wp * ibit16 * adv_mom_3 &
2139	+ ibit15 * adv_mom_1 &
2140	) * &
2141	( u(k+1,j,i) - u(k,j,i) ) &
2142	- ( 5.0_wp * ibit17 * adv_mom_5 &
2143	+ ibit16 * adv_mom_3 &
2144	) * &
2145	( u(k+2,j,i) - u(k-1,j,i) ) &
2146	+ ( ibit17 * adv_mom_5 &
2147	) * &
2148	( u(k+3,j,i) - u(k-2,j,i) ) &
2149	)
2150	ENDDO
2151
2152	DO k = nzt-1, nzt
2153	!
2154	!-- k index has to be modified near bottom and top, else array
2155	!-- subscripts will be exceeded.
2156	ibit17 = REAL( IBITS(advc_flags_1(k,j,i),17,1), KIND = wp )
2157	ibit16 = REAL( IBITS(advc_flags_1(k,j,i),16,1), KIND = wp )
2158	ibit15 = REAL( IBITS(advc_flags_1(k,j,i),15,1), KIND = wp )
2159
2160	k_ppp = k + 3 * ibit17
2161	k_pp = k + 2 * ( 1 - ibit15 )
2162	k_mm = k - 2 * ibit17
2163
2164	w_comp(k) = w(k,j,i) + w(k,j,i-1)
2165	flux_t(k) = w_comp(k) * rho_air_zw(k) * ( &
2166	( 37.0_wp * ibit17 * adv_mom_5 &
2167	+ 7.0_wp * ibit16 * adv_mom_3 &
2168	+ ibit15 * adv_mom_1 &
2169	) * &
2170	( u(k+1,j,i) + u(k,j,i) ) &
2171	- ( 8.0_wp * ibit17 * adv_mom_5 &
2172	+ ibit16 * adv_mom_3 &
2173	) * &
2174	( u(k_pp,j,i) + u(k-1,j,i) ) &
2175	+ ( ibit17 * adv_mom_5 &
2176	) * &
2177	( u(k_ppp,j,i) + u(k_mm,j,i) ) &
2178	)
2179
2180	diss_t(k) = - ABS( w_comp(k) ) * rho_air_zw(k) * ( &
2181	( 10.0_wp * ibit17 * adv_mom_5 &
2182	+ 3.0_wp * ibit16 * adv_mom_3 &
2183	+ ibit15 * adv_mom_1 &
2184	) * &
2185	( u(k+1,j,i) - u(k,j,i) ) &
2186	- ( 5.0_wp * ibit17 * adv_mom_5 &
2187	+ ibit16 * adv_mom_3 &
2188	) * &
2189	( u(k_pp,j,i) - u(k-1,j,i) ) &
2190	+ ( ibit17 * adv_mom_5 &
2191	) * &
2192	( u(k_ppp,j,i) - u(k_mm,j,i) ) &
2193	)
2194	ENDDO
2195
2196	DO k = nzb+1, nzt
2197
2198	flux_d = flux_t(k-1)
2199	diss_d = diss_t(k-1)
2200	!
2201	!-- Calculate the divergence of the velocity field. A respective
2202	!-- correction is needed to overcome numerical instabilities introduced
2203	!-- by a not sufficient reduction of divergences near topography.
2204	div = ( ( u_comp(k) * ( ibit9 + ibit10 + ibit11 ) &
2205	- ( u(k,j,i) + u(k,j,i-1) ) &
2206	* ( &
2207	REAL( IBITS(advc_flags_1(k,j,i-1),9,1), KIND = wp ) &
2208	+ REAL( IBITS(advc_flags_1(k,j,i-1),10,1), KIND = wp ) &
2209	+ REAL( IBITS(advc_flags_1(k,j,i-1),11,1), KIND = wp ) &
2210	) &
2211	) * ddx &
2212	+ ( ( v_comp(k) + gv ) * ( ibit12 + ibit13 + ibit14 ) &
2213	- ( v(k,j,i) + v(k,j,i-1 ) ) &
2214	* ( &
2215	REAL( IBITS(advc_flags_1(k,j-1,i),12,1), KIND = wp ) &
2216	+ REAL( IBITS(advc_flags_1(k,j-1,i),13,1), KIND = wp ) &
2217	+ REAL( IBITS(advc_flags_1(k,j-1,i),14,1), KIND = wp ) &
2218	) &
2219	) * ddy &
2220	+ ( w_comp(k) * rho_air_zw(k) * ( ibit15 + ibit16 + ibit17 )&
2221	- w_comp(k-1) * rho_air_zw(k-1) &
2222	* ( &
2223	REAL( IBITS(advc_flags_1(k-1,j,i),15,1), KIND = wp ) &
2224	+ REAL( IBITS(advc_flags_1(k-1,j,i),16,1), KIND = wp ) &
2225	+ REAL( IBITS(advc_flags_1(k-1,j,i),17,1), KIND = wp ) &
2226	) &
2227	) * drho_air(k) * ddzw(k) &
2228	) * 0.5_wp
2229
2230	tend(k,j,i) = tend(k,j,i) - ( &
2231	( flux_r(k) + diss_r(k) &
2232	- flux_l_u(k,j,tn) - diss_l_u(k,j,tn) ) * ddx &
2233	+ ( flux_n(k) + diss_n(k) &
2234	- flux_s_u(k,tn) - diss_s_u(k,tn) ) * ddy &
2235	+ ( ( flux_t(k) + diss_t(k) ) &
2236	- ( flux_d + diss_d ) &
2237	) * drho_air(k) * ddzw(k) &
2238	) + div * u(k,j,i)
2239
2240	flux_l_u(k,j,tn) = flux_r(k)
2241	diss_l_u(k,j,tn) = diss_r(k)
2242	flux_s_u(k,tn) = flux_n(k)
2243	diss_s_u(k,tn) = diss_n(k)
2244	!
2245	!-- Statistical Evaluation of u'u'. The factor has to be applied for
2246	!-- right evaluation when gallilei_trans = .T. .
2247	sums_us2_ws_l(k,tn) = sums_us2_ws_l(k,tn) &
2248	+ ( flux_r(k) &
2249	* ( u_comp(k) - 2.0_wp * hom(k,1,1,0) ) &
2250	/ ( u_comp(k) - gu + SIGN( 1.0E-20_wp, u_comp(k) - gu ) ) &
2251	+ diss_r(k) &
2252	* ABS( u_comp(k) - 2.0_wp * hom(k,1,1,0) ) &
2253	/ ( ABS( u_comp(k) - gu ) + 1.0E-20_wp ) &
2254	) * weight_substep(intermediate_timestep_count)
2255	!
2256	!-- Statistical Evaluation of w'u'.
2257	sums_wsus_ws_l(k,tn) = sums_wsus_ws_l(k,tn) &
2258	+ ( flux_t(k) &
2259	* ( w_comp(k) - 2.0_wp * hom(k,1,3,0) ) &
2260	/ ( w_comp(k) + SIGN( 1.0E-20_wp, w_comp(k) ) ) &
2261	+ diss_t(k) &
2262	* ABS( w_comp(k) - 2.0_wp * hom(k,1,3,0) ) &
2263	/ ( ABS( w_comp(k) ) + 1.0E-20_wp ) &
2264	) * weight_substep(intermediate_timestep_count)
2265	ENDDO
2266
2267	sums_us2_ws_l(nzb,tn) = sums_us2_ws_l(nzb+1,tn)
2268
2269
2270
2271	END SUBROUTINE advec_u_ws_ij
2272
2273
2274
2275	!-----------------------------------------------------------------------------!
2276	! Description:
2277	! ------------
2278	!> Advection of v-component - Call for grid point i,j
2279	!-----------------------------------------------------------------------------!
2280	SUBROUTINE advec_v_ws_ij( i, j, i_omp, tn )
2281
2282	USE arrays_3d, &
2283	ONLY: ddzw, diss_l_v, diss_s_v, flux_l_v, flux_s_v, tend, u, v, w, &
2284	drho_air, rho_air_zw
2285
2286	USE control_parameters, &
2287	ONLY: intermediate_timestep_count, u_gtrans, v_gtrans
2288
2289	USE grid_variables, &
2290	ONLY: ddx, ddy
2291
2292	USE indices, &
2293	ONLY: nysv, nzb, nzb_max, nzt, advc_flags_1
2294
2295	USE kinds
2296
2297	USE statistics, &
2298	ONLY: hom, sums_vs2_ws_l, sums_wsvs_ws_l, weight_substep
2299
2300	IMPLICIT NONE
2301
2302	INTEGER(iwp) :: i !< grid index along x-direction
2303	INTEGER(iwp) :: i_omp !< leftmost index on subdomain, or in case of OpenMP, on thread
2304	INTEGER(iwp) :: j !< grid index along y-direction
2305	INTEGER(iwp) :: k !< grid index along z-direction
2306	INTEGER(iwp) :: k_mm !< k-2 index in disretization, can be modified to avoid segmentation faults
2307	INTEGER(iwp) :: k_pp !< k+2 index in disretization, can be modified to avoid segmentation faults
2308	INTEGER(iwp) :: k_ppp !< k+3 index in disretization, can be modified to avoid segmentation faults
2309	INTEGER(iwp) :: tn !< number of OpenMP thread
2310
2311	REAL(wp) :: ibit18 !< flag indicating 1st-order scheme along x-direction
2312	REAL(wp) :: ibit19 !< flag indicating 3rd-order scheme along x-direction
2313	REAL(wp) :: ibit20 !< flag indicating 5th-order scheme along x-direction
2314	REAL(wp) :: ibit21 !< flag indicating 1st-order scheme along y-direction
2315	REAL(wp) :: ibit22 !< flag indicating 3rd-order scheme along y-direction
2316	REAL(wp) :: ibit23 !< flag indicating 3rd-order scheme along y-direction
2317	REAL(wp) :: ibit24 !< flag indicating 1st-order scheme along z-direction
2318	REAL(wp) :: ibit25 !< flag indicating 3rd-order scheme along z-direction
2319	REAL(wp) :: ibit26 !< flag indicating 3rd-order scheme along z-direction
2320	REAL(wp) :: diss_d !< artificial dissipation term at grid box bottom
2321	REAL(wp) :: div !< divergence on v-grid
2322	REAL(wp) :: flux_d !< 6th-order flux at grid box bottom
2323	REAL(wp) :: gu !< Galilei-transformation velocity along x
2324	REAL(wp) :: gv !< Galilei-transformation velocity along y
2325	REAL(wp) :: v_comp_l !< advection velocity along y on leftmost grid point on subdomain
2326
2327	REAL(wp), DIMENSION(nzb:nzt+1) :: diss_n !< discretized artificial dissipation at northward-side of the grid box
2328	REAL(wp), DIMENSION(nzb:nzt+1) :: diss_r !< discretized artificial dissipation at rightward-side of the grid box
2329	REAL(wp), DIMENSION(nzb:nzt+1) :: diss_t !< discretized artificial dissipation at top of the grid box
2330	REAL(wp), DIMENSION(nzb:nzt+1) :: flux_n !< discretized 6th-order flux at northward-side of the grid box
2331	REAL(wp), DIMENSION(nzb:nzt+1) :: flux_r !< discretized 6th-order flux at rightward-side of the grid box
2332	REAL(wp), DIMENSION(nzb:nzt+1) :: flux_t !< discretized 6th-order flux at top of the grid box
2333	REAL(wp), DIMENSION(nzb:nzt+1) :: u_comp !< advection velocity along x
2334	REAL(wp), DIMENSION(nzb:nzt+1) :: v_comp !< advection velocity along y
2335	REAL(wp), DIMENSION(nzb:nzt+1) :: w_comp !< advection velocity along z
2336
2337	gu = 2.0_wp * u_gtrans
2338	gv = 2.0_wp * v_gtrans
2339
2340	!
2341	!-- Compute leftside fluxes for the respective boundary.
2342	IF ( i == i_omp ) THEN
2343
2344	DO k = nzb+1, nzb_max
2345
2346	ibit20 = REAL( IBITS(advc_flags_1(k,j,i-1),20,1), KIND = wp )
2347	ibit19 = REAL( IBITS(advc_flags_1(k,j,i-1),19,1), KIND = wp )
2348	ibit18 = REAL( IBITS(advc_flags_1(k,j,i-1),18,1), KIND = wp )
2349
2350	u_comp(k) = u(k,j-1,i) + u(k,j,i) - gu
2351	flux_l_v(k,j,tn) = u_comp(k) * ( &
2352	( 37.0_wp * ibit20 * adv_mom_5 &
2353	+ 7.0_wp * ibit19 * adv_mom_3 &
2354	+ ibit18 * adv_mom_1 &
2355	) * &
2356	( v(k,j,i) + v(k,j,i-1) ) &
2357	- ( 8.0_wp * ibit20 * adv_mom_5 &
2358	+ ibit19 * adv_mom_3 &
2359	) * &
2360	( v(k,j,i+1) + v(k,j,i-2) ) &
2361	+ ( ibit20 * adv_mom_5 &
2362	) * &
2363	( v(k,j,i+2) + v(k,j,i-3) ) &
2364	)
2365
2366	diss_l_v(k,j,tn) = - ABS( u_comp(k) ) * ( &
2367	( 10.0_wp * ibit20 * adv_mom_5 &
2368	+ 3.0_wp * ibit19 * adv_mom_3 &
2369	+ ibit18 * adv_mom_1 &
2370	) * &
2371	( v(k,j,i) - v(k,j,i-1) ) &
2372	- ( 5.0_wp * ibit20 * adv_mom_5 &
2373	+ ibit19 * adv_mom_3 &
2374	) * &
2375	( v(k,j,i+1) - v(k,j,i-2) ) &
2376	+ ( ibit20 * adv_mom_5 &
2377	) * &
2378	( v(k,j,i+2) - v(k,j,i-3) ) &
2379	)
2380
2381	ENDDO
2382
2383	DO k = nzb_max+1, nzt
2384
2385	u_comp(k) = u(k,j-1,i) + u(k,j,i) - gu
2386	flux_l_v(k,j,tn) = u_comp(k) * ( &
2387	37.0_wp * ( v(k,j,i) + v(k,j,i-1) ) &
2388	- 8.0_wp * ( v(k,j,i+1) + v(k,j,i-2) ) &
2389	+ ( v(k,j,i+2) + v(k,j,i-3) ) ) * adv_mom_5
2390	diss_l_v(k,j,tn) = - ABS( u_comp(k) ) * ( &
2391	10.0_wp * ( v(k,j,i) - v(k,j,i-1) ) &
2392	- 5.0_wp * ( v(k,j,i+1) - v(k,j,i-2) ) &
2393	+ ( v(k,j,i+2) - v(k,j,i-3) ) ) * adv_mom_5
2394
2395	ENDDO
2396
2397	ENDIF
2398	!
2399	!-- Compute southside fluxes for the respective boundary.
2400	IF ( j == nysv ) THEN
2401
2402	DO k = nzb+1, nzb_max
2403
2404	ibit23 = REAL( IBITS(advc_flags_1(k,j-1,i),23,1), KIND = wp )
2405	ibit22 = REAL( IBITS(advc_flags_1(k,j-1,i),22,1), KIND = wp )
2406	ibit21 = REAL( IBITS(advc_flags_1(k,j-1,i),21,1), KIND = wp )
2407
2408	v_comp_l = v(k,j,i) + v(k,j-1,i) - gv
2409	flux_s_v(k,tn) = v_comp_l * ( &
2410	( 37.0_wp * ibit23 * adv_mom_5 &
2411	+ 7.0_wp * ibit22 * adv_mom_3 &
2412	+ ibit21 * adv_mom_1 &
2413	) * &
2414	( v(k,j,i) + v(k,j-1,i) ) &
2415	- ( 8.0_wp * ibit23 * adv_mom_5 &
2416	+ ibit22 * adv_mom_3 &
2417	) * &
2418	( v(k,j+1,i) + v(k,j-2,i) ) &
2419	+ ( ibit23 * adv_mom_5 &
2420	) * &
2421	( v(k,j+2,i) + v(k,j-3,i) ) &
2422	)
2423
2424	diss_s_v(k,tn) = - ABS( v_comp_l ) * ( &
2425	( 10.0_wp * ibit23 * adv_mom_5 &
2426	+ 3.0_wp * ibit22 * adv_mom_3 &
2427	+ ibit21 * adv_mom_1 &
2428	) * &
2429	( v(k,j,i) - v(k,j-1,i) ) &
2430	- ( 5.0_wp * ibit23 * adv_mom_5 &
2431	+ ibit22 * adv_mom_3 &
2432	) * &
2433	( v(k,j+1,i) - v(k,j-2,i) ) &
2434	+ ( ibit23 * adv_mom_5 &
2435	) * &
2436	( v(k,j+2,i) - v(k,j-3,i) ) &
2437	)
2438
2439	ENDDO
2440
2441	DO k = nzb_max+1, nzt
2442
2443	v_comp_l = v(k,j,i) + v(k,j-1,i) - gv
2444	flux_s_v(k,tn) = v_comp_l * ( &
2445	37.0_wp * ( v(k,j,i) + v(k,j-1,i) ) &
2446	- 8.0_wp * ( v(k,j+1,i) + v(k,j-2,i) ) &
2447	+ ( v(k,j+2,i) + v(k,j-3,i) ) ) * adv_mom_5
2448	diss_s_v(k,tn) = - ABS( v_comp_l ) * ( &
2449	10.0_wp * ( v(k,j,i) - v(k,j-1,i) ) &
2450	- 5.0_wp * ( v(k,j+1,i) - v(k,j-2,i) ) &
2451	+ ( v(k,j+2,i) - v(k,j-3,i) ) ) * adv_mom_5
2452
2453	ENDDO
2454
2455	ENDIF
2456	!
2457	!-- Now compute the fluxes and tendency terms for the horizontal and
2458	!-- verical parts.
2459	DO k = nzb+1, nzb_max
2460
2461	ibit20 = REAL( IBITS(advc_flags_1(k,j,i),20,1), KIND = wp )
2462	ibit19 = REAL( IBITS(advc_flags_1(k,j,i),19,1), KIND = wp )
2463	ibit18 = REAL( IBITS(advc_flags_1(k,j,i),18,1), KIND = wp )
2464
2465	u_comp(k) = u(k,j-1,i+1) + u(k,j,i+1) - gu
2466	flux_r(k) = u_comp(k) * ( &
2467	( 37.0_wp * ibit20 * adv_mom_5 &
2468	+ 7.0_wp * ibit19 * adv_mom_3 &
2469	+ ibit18 * adv_mom_1 &
2470	) * &
2471	( v(k,j,i+1) + v(k,j,i) ) &
2472	- ( 8.0_wp * ibit20 * adv_mom_5 &
2473	+ ibit19 * adv_mom_3 &
2474	) * &
2475	( v(k,j,i+2) + v(k,j,i-1) ) &
2476	+ ( ibit20 * adv_mom_5 &
2477	) * &
2478	( v(k,j,i+3) + v(k,j,i-2) ) &
2479	)
2480
2481	diss_r(k) = - ABS( u_comp(k) ) * ( &
2482	( 10.0_wp * ibit20 * adv_mom_5 &
2483	+ 3.0_wp * ibit19 * adv_mom_3 &
2484	+ ibit18 * adv_mom_1 &
2485	) * &
2486	( v(k,j,i+1) - v(k,j,i) ) &
2487	- ( 5.0_wp * ibit20 * adv_mom_5 &
2488	+ ibit19 * adv_mom_3 &
2489	) * &
2490	( v(k,j,i+2) - v(k,j,i-1) ) &
2491	+ ( ibit20 * adv_mom_5 &
2492	) * &
2493	( v(k,j,i+3) - v(k,j,i-2) ) &
2494	)
2495
2496	ibit23 = REAL( IBITS(advc_flags_1(k,j,i),23,1), KIND = wp )
2497	ibit22 = REAL( IBITS(advc_flags_1(k,j,i),22,1), KIND = wp )
2498	ibit21 = REAL( IBITS(advc_flags_1(k,j,i),21,1), KIND = wp )
2499
2500
2501	v_comp(k) = v(k,j+1,i) + v(k,j,i)
2502	flux_n(k) = ( v_comp(k) - gv ) * ( &
2503	( 37.0_wp * ibit23 * adv_mom_5 &
2504	+ 7.0_wp * ibit22 * adv_mom_3 &
2505	+ ibit21 * adv_mom_1 &
2506	) * &
2507	( v(k,j+1,i) + v(k,j,i) ) &
2508	- ( 8.0_wp * ibit23 * adv_mom_5 &
2509	+ ibit22 * adv_mom_3 &
2510	) * &
2511	( v(k,j+2,i) + v(k,j-1,i) ) &
2512	+ ( ibit23 * adv_mom_5 &
2513	) * &
2514	( v(k,j+3,i) + v(k,j-2,i) ) &
2515	)
2516
2517	diss_n(k) = - ABS( v_comp(k) - gv ) * ( &
2518	( 10.0_wp * ibit23 * adv_mom_5 &
2519	+ 3.0_wp * ibit22 * adv_mom_3 &
2520	+ ibit21 * adv_mom_1 &
2521	) * &
2522	( v(k,j+1,i) - v(k,j,i) ) &
2523	- ( 5.0_wp * ibit23 * adv_mom_5 &
2524	+ ibit22 * adv_mom_3 &
2525	) * &
2526	( v(k,j+2,i) - v(k,j-1,i) ) &
2527	+ ( ibit23 * adv_mom_5 &
2528	) * &
2529	( v(k,j+3,i) - v(k,j-2,i) ) &
2530	)
2531	ENDDO
2532
2533	DO k = nzb_max+1, nzt
2534
2535	u_comp(k) = u(k,j-1,i+1) + u(k,j,i+1) - gu
2536	flux_r(k) = u_comp(k) * ( &
2537	37.0_wp * ( v(k,j,i+1) + v(k,j,i) ) &
2538	- 8.0_wp * ( v(k,j,i+2) + v(k,j,i-1) ) &
2539	+ ( v(k,j,i+3) + v(k,j,i-2) ) ) * adv_mom_5
2540
2541	diss_r(k) = - ABS( u_comp(k) ) * ( &
2542	10.0_wp * ( v(k,j,i+1) - v(k,j,i) ) &
2543	- 5.0_wp * ( v(k,j,i+2) - v(k,j,i-1) ) &
2544	+ ( v(k,j,i+3) - v(k,j,i-2) ) ) * adv_mom_5
2545
2546
2547	v_comp(k) = v(k,j+1,i) + v(k,j,i)
2548	flux_n(k) = ( v_comp(k) - gv ) * ( &
2549	37.0_wp * ( v(k,j+1,i) + v(k,j,i) ) &
2550	- 8.0_wp * ( v(k,j+2,i) + v(k,j-1,i) ) &
2551	+ ( v(k,j+3,i) + v(k,j-2,i) ) ) * adv_mom_5
2552
2553	diss_n(k) = - ABS( v_comp(k) - gv ) * ( &
2554	10.0_wp * ( v(k,j+1,i) - v(k,j,i) ) &
2555	- 5.0_wp * ( v(k,j+2,i) - v(k,j-1,i) ) &
2556	+ ( v(k,j+3,i) - v(k,j-2,i) ) ) * adv_mom_5
2557	ENDDO
2558	!
2559	!-- Now, compute vertical fluxes. Split loop into a part treating the
2560	!-- lowest 2 grid points with indirect indexing, a main loop without
2561	!-- indirect indexing, and a loop for the uppermost 2 grip points with
2562	!-- indirect indexing. This allows better vectorization for the main loop.
2563	!-- First, compute the flux at model surface, which need has to be
2564	!-- calculated explicetely for the tendency at
2565	!-- the first w-level. For topography wall this is done implicitely by
2566	!-- advc_flags_1.
2567	flux_t(nzb) = 0.0_wp
2568	diss_t(nzb) = 0.0_wp
2569	w_comp(nzb) = 0.0_wp
2570
2571	DO k = nzb+1, nzb+2
2572	!
2573	!-- k index has to be modified near bottom and top, else array
2574	!-- subscripts will be exceeded.
2575	ibit26 = REAL( IBITS(advc_flags_1(k,j,i),26,1), KIND = wp )
2576	ibit25 = REAL( IBITS(advc_flags_1(k,j,i),25,1), KIND = wp )
2577	ibit24 = REAL( IBITS(advc_flags_1(k,j,i),24,1), KIND = wp )
2578
2579	k_ppp = k + 3 * ibit26
2580	k_pp = k + 2 * ( 1 - ibit24 )
2581	k_mm = k - 2 * ibit26
2582
2583	w_comp(k) = w(k,j-1,i) + w(k,j,i)
2584	flux_t(k) = w_comp(k) * rho_air_zw(k) * ( &
2585	( 37.0_wp * ibit26 * adv_mom_5 &
2586	+ 7.0_wp * ibit25 * adv_mom_3 &
2587	+ ibit24 * adv_mom_1 &
2588	) * &
2589	( v(k+1,j,i) + v(k,j,i) ) &
2590	- ( 8.0_wp * ibit26 * adv_mom_5 &
2591	+ ibit25 * adv_mom_3 &
2592	) * &
2593	( v(k_pp,j,i) + v(k-1,j,i) ) &
2594	+ ( ibit26 * adv_mom_5 &
2595	) * &
2596	( v(k_ppp,j,i) + v(k_mm,j,i) ) &
2597	)
2598
2599	diss_t(k) = - ABS( w_comp(k) ) * rho_air_zw(k) * ( &
2600	( 10.0_wp * ibit26 * adv_mom_5 &
2601	+ 3.0_wp * ibit25 * adv_mom_3 &
2602	+ ibit24 * adv_mom_1 &
2603	) * &
2604	( v(k+1,j,i) - v(k,j,i) ) &
2605	- ( 5.0_wp * ibit26 * adv_mom_5 &
2606	+ ibit25 * adv_mom_3 &
2607	) * &
2608	( v(k_pp,j,i) - v(k-1,j,i) ) &
2609	+ ( ibit26 * adv_mom_5 &
2610	) * &
2611	( v(k_ppp,j,i) - v(k_mm,j,i) ) &
2612	)
2613	ENDDO
2614
2615	DO k = nzb+3, nzt-2
2616
2617	ibit26 = REAL( IBITS(advc_flags_1(k,j,i),26,1), KIND = wp )
2618	ibit25 = REAL( IBITS(advc_flags_1(k,j,i),25,1), KIND = wp )
2619	ibit24 = REAL( IBITS(advc_flags_1(k,j,i),24,1), KIND = wp )
2620
2621	w_comp(k) = w(k,j-1,i) + w(k,j,i)
2622	flux_t(k) = w_comp(k) * rho_air_zw(k) * ( &
2623	( 37.0_wp * ibit26 * adv_mom_5 &
2624	+ 7.0_wp * ibit25 * adv_mom_3 &
2625	+ ibit24 * adv_mom_1 &
2626	) * &
2627	( v(k+1,j,i) + v(k,j,i) ) &
2628	- ( 8.0_wp * ibit26 * adv_mom_5 &
2629	+ ibit25 * adv_mom_3 &
2630	) * &
2631	( v(k+2,j,i) + v(k-1,j,i) ) &
2632	+ ( ibit26 * adv_mom_5 &
2633	) * &
2634	( v(k+3,j,i) + v(k-2,j,i) ) &
2635	)
2636
2637	diss_t(k) = - ABS( w_comp(k) ) * rho_air_zw(k) * ( &
2638	( 10.0_wp * ibit26 * adv_mom_5 &
2639	+ 3.0_wp * ibit25 * adv_mom_3 &
2640	+ ibit24 * adv_mom_1 &
2641	) * &
2642	( v(k+1,j,i) - v(k,j,i) ) &
2643	- ( 5.0_wp * ibit26 * adv_mom_5 &
2644	+ ibit25 * adv_mom_3 &
2645	) * &
2646	( v(k+2,j,i) - v(k-1,j,i) ) &
2647	+ ( ibit26 * adv_mom_5 &
2648	) * &
2649	( v(k+3,j,i) - v(k-2,j,i) ) &
2650	)
2651	ENDDO
2652
2653	DO k = nzt-1, nzt
2654	!
2655	!-- k index has to be modified near bottom and top, else array
2656	!-- subscripts will be exceeded.
2657	ibit26 = REAL( IBITS(advc_flags_1(k,j,i),26,1), KIND = wp )
2658	ibit25 = REAL( IBITS(advc_flags_1(k,j,i),25,1), KIND = wp )
2659	ibit24 = REAL( IBITS(advc_flags_1(k,j,i),24,1), KIND = wp )
2660
2661	k_ppp = k + 3 * ibit26
2662	k_pp = k + 2 * ( 1 - ibit24 )
2663	k_mm = k - 2 * ibit26
2664
2665	w_comp(k) = w(k,j-1,i) + w(k,j,i)
2666	flux_t(k) = w_comp(k) * rho_air_zw(k) * ( &
2667	( 37.0_wp * ibit26 * adv_mom_5 &
2668	+ 7.0_wp * ibit25 * adv_mom_3 &
2669	+ ibit24 * adv_mom_1 &
2670	) * &
2671	( v(k+1,j,i) + v(k,j,i) ) &
2672	- ( 8.0_wp * ibit26 * adv_mom_5 &
2673	+ ibit25 * adv_mom_3 &
2674	) * &
2675	( v(k_pp,j,i) + v(k-1,j,i) ) &
2676	+ ( ibit26 * adv_mom_5 &
2677	) * &
2678	( v(k_ppp,j,i) + v(k_mm,j,i) ) &
2679	)
2680
2681	diss_t(k) = - ABS( w_comp(k) ) * rho_air_zw(k) * ( &
2682	( 10.0_wp * ibit26 * adv_mom_5 &
2683	+ 3.0_wp * ibit25 * adv_mom_3 &
2684	+ ibit24 * adv_mom_1 &
2685	) * &
2686	( v(k+1,j,i) - v(k,j,i) ) &
2687	- ( 5.0_wp * ibit26 * adv_mom_5 &
2688	+ ibit25 * adv_mom_3 &
2689	) * &
2690	( v(k_pp,j,i) - v(k-1,j,i) ) &
2691	+ ( ibit26 * adv_mom_5 &
2692	) * &
2693	( v(k_ppp,j,i) - v(k_mm,j,i) ) &
2694	)
2695	ENDDO
2696
2697	DO k = nzb+1, nzt
2698
2699	flux_d = flux_t(k-1)
2700	diss_d = diss_t(k-1)
2701	!
2702	!-- Calculate the divergence of the velocity field. A respective
2703	!-- correction is needed to overcome numerical instabilities introduced
2704	!-- by a not sufficient reduction of divergences near topography.
2705	div = ( ( ( u_comp(k) + gu ) &
2706	* ( ibit18 + ibit19 + ibit20 ) &
2707	- ( u(k,j-1,i) + u(k,j,i) ) &
2708	* ( &
2709	REAL( IBITS(advc_flags_1(k,j,i-1),18,1), KIND = wp ) &
2710	+ REAL( IBITS(advc_flags_1(k,j,i-1),19,1), KIND = wp ) &
2711	+ REAL( IBITS(advc_flags_1(k,j,i-1),20,1), KIND = wp ) &
2712	) &
2713	) * ddx &
2714	+ ( v_comp(k) &
2715	* ( ibit21 + ibit22 + ibit23 ) &
2716	- ( v(k,j,i) + v(k,j-1,i) ) &
2717	* ( &
2718	REAL( IBITS(advc_flags_1(k,j-1,i),21,1), KIND = wp ) &
2719	+ REAL( IBITS(advc_flags_1(k,j-1,i),22,1), KIND = wp ) &
2720	+ REAL( IBITS(advc_flags_1(k,j-1,i),23,1), KIND = wp ) &
2721	) &
2722	) * ddy &
2723	+ ( w_comp(k) * rho_air_zw(k) * ( ibit24 + ibit25 + ibit26 )&
2724	- w_comp(k-1) * rho_air_zw(k-1) &
2725	* ( &
2726	REAL( IBITS(advc_flags_1(k-1,j,i),24,1), KIND = wp ) &
2727	+ REAL( IBITS(advc_flags_1(k-1,j,i),25,1), KIND = wp ) &
2728	+ REAL( IBITS(advc_flags_1(k-1,j,i),26,1), KIND = wp ) &
2729	) &
2730	) * drho_air(k) * ddzw(k) &
2731	) * 0.5_wp
2732
2733	tend(k,j,i) = tend(k,j,i) - ( &
2734	( flux_r(k) + diss_r(k) &
2735	- flux_l_v(k,j,tn) - diss_l_v(k,j,tn) ) * ddx &
2736	+ ( flux_n(k) + diss_n(k) &
2737	- flux_s_v(k,tn) - diss_s_v(k,tn) ) * ddy &
2738	+ ( ( flux_t(k) + diss_t(k) ) &
2739	- ( flux_d + diss_d ) &
2740	) * drho_air(k) * ddzw(k) &
2741	) + v(k,j,i) * div
2742
2743	flux_l_v(k,j,tn) = flux_r(k)
2744	diss_l_v(k,j,tn) = diss_r(k)
2745	flux_s_v(k,tn) = flux_n(k)
2746	diss_s_v(k,tn) = diss_n(k)
2747	!
2748	!-- Statistical Evaluation of v'v'. The factor has to be applied for
2749	!-- right evaluation when gallilei_trans = .T. .
2750	sums_vs2_ws_l(k,tn) = sums_vs2_ws_l(k,tn) &
2751	+ ( flux_n(k) &
2752	* ( v_comp(k) - 2.0_wp * hom(k,1,2,0) ) &
2753	/ ( v_comp(k) - gv + SIGN( 1.0E-20_wp, v_comp(k) - gv ) ) &
2754	+ diss_n(k) &
2755	* ABS( v_comp(k) - 2.0_wp * hom(k,1,2,0) ) &
2756	/ ( ABS( v_comp(k) - gv ) + 1.0E-20_wp ) &
2757	) * weight_substep(intermediate_timestep_count)
2758	!
2759	!-- Statistical Evaluation of w'u'.
2760	sums_wsvs_ws_l(k,tn) = sums_wsvs_ws_l(k,tn) &
2761	+ ( flux_t(k) &
2762	* ( w_comp(k) - 2.0_wp * hom(k,1,3,0) ) &
2763	/ ( w_comp(k) + SIGN( 1.0E-20_wp, w_comp(k) ) ) &
2764	+ diss_t(k) &
2765	* ABS( w_comp(k) - 2.0_wp * hom(k,1,3,0) ) &
2766	/ ( ABS( w_comp(k) ) + 1.0E-20_wp ) &
2767	) * weight_substep(intermediate_timestep_count)
2768
2769	ENDDO
2770	sums_vs2_ws_l(nzb,tn) = sums_vs2_ws_l(nzb+1,tn)
2771
2772
2773	END SUBROUTINE advec_v_ws_ij
2774
2775
2776
2777	!------------------------------------------------------------------------------!
2778	! Description:
2779	! ------------
2780	!> Advection of w-component - Call for grid point i,j
2781	!------------------------------------------------------------------------------!
2782	SUBROUTINE advec_w_ws_ij( i, j, i_omp, tn )
2783
2784	USE arrays_3d, &
2785	ONLY: ddzu, diss_l_w, diss_s_w, flux_l_w, flux_s_w, tend, u, v, w,&
2786	drho_air_zw, rho_air
2787
2788	USE control_parameters, &
2789	ONLY: intermediate_timestep_count, u_gtrans, v_gtrans
2790
2791	USE grid_variables, &
2792	ONLY: ddx, ddy
2793
2794	USE indices, &
2795	ONLY: nys, nzb, nzb_max, nzt, advc_flags_1, advc_flags_2
2796
2797	USE kinds
2798
2799	USE statistics, &
2800	ONLY: hom, sums_ws2_ws_l, weight_substep
2801
2802	IMPLICIT NONE
2803
2804	INTEGER(iwp) :: i !< grid index along x-direction
2805	INTEGER(iwp) :: i_omp !< leftmost index on subdomain, or in case of OpenMP, on thread
2806	INTEGER(iwp) :: j !< grid index along y-direction
2807	INTEGER(iwp) :: k !< grid index along z-direction
2808	INTEGER(iwp) :: k_mm !< k-2 index in disretization, can be modified to avoid segmentation faults
2809	INTEGER(iwp) :: k_pp !< k+2 index in disretization, can be modified to avoid segmentation faults
2810	INTEGER(iwp) :: k_ppp !< k+3 index in disretization, can be modified to avoid segmentation faults
2811	INTEGER(iwp) :: tn !< number of OpenMP thread
2812
2813	REAL(wp) :: ibit27 !< flag indicating 1st-order scheme along x-direction
2814	REAL(wp) :: ibit28 !< flag indicating 3rd-order scheme along x-direction
2815	REAL(wp) :: ibit29 !< flag indicating 5th-order scheme along x-direction
2816	REAL(wp) :: ibit30 !< flag indicating 1st-order scheme along y-direction
2817	REAL(wp) :: ibit31 !< flag indicating 3rd-order scheme along y-direction
2818	REAL(wp) :: ibit32 !< flag indicating 5th-order scheme along y-direction
2819	REAL(wp) :: ibit33 !< flag indicating 1st-order scheme along z-direction
2820	REAL(wp) :: ibit34 !< flag indicating 3rd-order scheme along z-direction
2821	REAL(wp) :: ibit35 !< flag indicating 5th-order scheme along z-direction
2822	REAL(wp) :: diss_d !< discretized artificial dissipation at top of the grid box
2823	REAL(wp) :: div !< divergence on w-grid
2824	REAL(wp) :: flux_d !< discretized 6th-order flux at top of the grid box
2825	REAL(wp) :: gu !< Galilei-transformation velocity along x
2826	REAL(wp) :: gv !< Galilei-transformation velocity along y
2827
2828	REAL(wp), DIMENSION(nzb:nzt+1) :: diss_n !< discretized artificial dissipation at northward-side of the grid box
2829	REAL(wp), DIMENSION(nzb:nzt+1) :: diss_r !< discretized artificial dissipation at rightward-side of the grid box
2830	REAL(wp), DIMENSION(nzb:nzt+1) :: diss_t !< discretized artificial dissipation at top of the grid box
2831	REAL(wp), DIMENSION(nzb:nzt+1) :: flux_n !< discretized 6th-order flux at northward-side of the grid box
2832	REAL(wp), DIMENSION(nzb:nzt+1) :: flux_r !< discretized 6th-order flux at rightward-side of the grid box
2833	REAL(wp), DIMENSION(nzb:nzt+1) :: flux_t !< discretized 6th-order flux at top of the grid box
2834	REAL(wp), DIMENSION(nzb:nzt+1) :: u_comp !< advection velocity along x
2835	REAL(wp), DIMENSION(nzb:nzt+1) :: v_comp !< advection velocity along y
2836	REAL(wp), DIMENSION(nzb:nzt+1) :: w_comp !< advection velocity along z
2837
2838	gu = 2.0_wp * u_gtrans
2839	gv = 2.0_wp * v_gtrans
2840
2841	!
2842	!-- Compute southside fluxes for the respective boundary.
2843	IF ( j == nys ) THEN
2844
2845	DO k = nzb+1, nzb_max
2846	ibit32 = REAL( IBITS(advc_flags_2(k,j-1,i),0,1), KIND = wp )
2847	ibit31 = REAL( IBITS(advc_flags_1(k,j-1,i),31,1), KIND = wp )
2848	ibit30 = REAL( IBITS(advc_flags_1(k,j-1,i),30,1), KIND = wp )
2849
2850	v_comp(k) = v(k+1,j,i) + v(k,j,i) - gv
2851	flux_s_w(k,tn) = v_comp(k) * ( &
2852	( 37.0_wp * ibit32 * adv_mom_5 &
2853	+ 7.0_wp * ibit31 * adv_mom_3 &
2854	+ ibit30 * adv_mom_1 &
2855	) * &
2856	( w(k,j,i) + w(k,j-1,i) ) &
2857	- ( 8.0_wp * ibit32 * adv_mom_5 &
2858	+ ibit31 * adv_mom_3 &
2859	) * &
2860	( w(k,j+1,i) + w(k,j-2,i) ) &
2861	+ ( ibit32 * adv_mom_5 &
2862	) * &
2863	( w(k,j+2,i) + w(k,j-3,i) ) &
2864	)
2865
2866	diss_s_w(k,tn) = - ABS( v_comp(k) ) * ( &
2867	( 10.0_wp * ibit32 * adv_mom_5 &
2868	+ 3.0_wp * ibit31 * adv_mom_3 &
2869	+ ibit30 * adv_mom_1 &
2870	) * &
2871	( w(k,j,i) - w(k,j-1,i) ) &
2872	- ( 5.0_wp * ibit32 * adv_mom_5 &
2873	+ ibit31 * adv_mom_3 &
2874	) * &
2875	( w(k,j+1,i) - w(k,j-2,i) ) &
2876	+ ( ibit32 * adv_mom_5 &
2877	) * &
2878	( w(k,j+2,i) - w(k,j-3,i) ) &
2879	)
2880
2881	ENDDO
2882
2883	DO k = nzb_max+1, nzt
2884
2885	v_comp(k) = v(k+1,j,i) + v(k,j,i) - gv
2886	flux_s_w(k,tn) = v_comp(k) * ( &
2887	37.0_wp * ( w(k,j,i) + w(k,j-1,i) ) &
2888	- 8.0_wp * ( w(k,j+1,i) +w(k,j-2,i) ) &
2889	+ ( w(k,j+2,i) + w(k,j-3,i) ) ) * adv_mom_5
2890	diss_s_w(k,tn) = - ABS( v_comp(k) ) * ( &
2891	10.0_wp * ( w(k,j,i) - w(k,j-1,i) ) &
2892	- 5.0_wp * ( w(k,j+1,i) - w(k,j-2,i) ) &
2893	+ ( w(k,j+2,i) - w(k,j-3,i) ) ) * adv_mom_5
2894
2895	ENDDO
2896
2897	ENDIF
2898	!
2899	!-- Compute leftside fluxes for the respective boundary.
2900	IF ( i == i_omp ) THEN
2901
2902	DO k = nzb+1, nzb_max
2903
2904	ibit29 = REAL( IBITS(advc_flags_1(k,j,i-1),29,1), KIND = wp )
2905	ibit28 = REAL( IBITS(advc_flags_1(k,j,i-1),28,1), KIND = wp )
2906	ibit27 = REAL( IBITS(advc_flags_1(k,j,i-1),27,1), KIND = wp )
2907
2908	u_comp(k) = u(k+1,j,i) + u(k,j,i) - gu
2909	flux_l_w(k,j,tn) = u_comp(k) * ( &
2910	( 37.0_wp * ibit29 * adv_mom_5 &
2911	+ 7.0_wp * ibit28 * adv_mom_3 &
2912	+ ibit27 * adv_mom_1 &
2913	) * &
2914	( w(k,j,i) + w(k,j,i-1) ) &
2915	- ( 8.0_wp * ibit29 * adv_mom_5 &
2916	+ ibit28 * adv_mom_3 &
2917	) * &
2918	( w(k,j,i+1) + w(k,j,i-2) ) &
2919	+ ( ibit29 * adv_mom_5 &
2920	) * &
2921	( w(k,j,i+2) + w(k,j,i-3) ) &
2922	)
2923
2924	diss_l_w(k,j,tn) = - ABS( u_comp(k) ) * ( &
2925	( 10.0_wp * ibit29 * adv_mom_5 &
2926	+ 3.0_wp * ibit28 * adv_mom_3 &
2927	+ ibit27 * adv_mom_1 &
2928	) * &
2929	( w(k,j,i) - w(k,j,i-1) ) &
2930	- ( 5.0_wp * ibit29 * adv_mom_5 &
2931	+ ibit28 * adv_mom_3 &
2932	) * &
2933	( w(k,j,i+1) - w(k,j,i-2) ) &
2934	+ ( ibit29 * adv_mom_5 &
2935	) * &
2936	( w(k,j,i+2) - w(k,j,i-3) ) &
2937	)
2938
2939	ENDDO
2940
2941	DO k = nzb_max+1, nzt
2942
2943	u_comp(k) = u(k+1,j,i) + u(k,j,i) - gu
2944	flux_l_w(k,j,tn) = u_comp(k) * ( &
2945	37.0_wp * ( w(k,j,i) + w(k,j,i-1) ) &
2946	- 8.0_wp * ( w(k,j,i+1) + w(k,j,i-2) ) &
2947	+ ( w(k,j,i+2) + w(k,j,i-3) ) ) * adv_mom_5
2948	diss_l_w(k,j,tn) = - ABS( u_comp(k) ) * ( &
2949	10.0_wp * ( w(k,j,i) - w(k,j,i-1) ) &
2950	- 5.0_wp * ( w(k,j,i+1) - w(k,j,i-2) ) &
2951	+ ( w(k,j,i+2) - w(k,j,i-3) ) ) * adv_mom_5
2952
2953	ENDDO
2954
2955	ENDIF
2956	!
2957	!-- Now compute the fluxes and tendency terms for the horizontal
2958	!-- and vertical parts.
2959	DO k = nzb+1, nzb_max
2960
2961	ibit29 = REAL( IBITS(advc_flags_1(k,j,i),29,1), KIND = wp )
2962	ibit28 = REAL( IBITS(advc_flags_1(k,j,i),28,1), KIND = wp )
2963	ibit27 = REAL( IBITS(advc_flags_1(k,j,i),27,1), KIND = wp )
2964
2965	u_comp(k) = u(k+1,j,i+1) + u(k,j,i+1) - gu
2966	flux_r(k) = u_comp(k) * ( &
2967	( 37.0_wp * ibit29 * adv_mom_5 &
2968	+ 7.0_wp * ibit28 * adv_mom_3 &
2969	+ ibit27 * adv_mom_1 &
2970	) * &
2971	( w(k,j,i+1) + w(k,j,i) ) &
2972	- ( 8.0_wp * ibit29 * adv_mom_5 &
2973	+ ibit28 * adv_mom_3 &
2974	) * &
2975	( w(k,j,i+2) + w(k,j,i-1) ) &
2976	+ ( ibit29 * adv_mom_5 &
2977	) * &
2978	( w(k,j,i+3) + w(k,j,i-2) ) &
2979	)
2980
2981	diss_r(k) = - ABS( u_comp(k) ) * ( &
2982	( 10.0_wp * ibit29 * adv_mom_5 &
2983	+ 3.0_wp * ibit28 * adv_mom_3 &
2984	+ ibit27 * adv_mom_1 &
2985	) * &
2986	( w(k,j,i+1) - w(k,j,i) ) &
2987	- ( 5.0_wp * ibit29 * adv_mom_5 &
2988	+ ibit28 * adv_mom_3 &
2989	) * &
2990	( w(k,j,i+2) - w(k,j,i-1) ) &
2991	+ ( ibit29 * adv_mom_5 &
2992	) * &
2993	( w(k,j,i+3) - w(k,j,i-2) ) &
2994	)
2995
2996	ibit32 = REAL( IBITS(advc_flags_2(k,j,i),0,1), KIND = wp )
2997	ibit31 = REAL( IBITS(advc_flags_1(k,j,i),31,1), KIND = wp )
2998	ibit30 = REAL( IBITS(advc_flags_1(k,j,i),30,1), KIND = wp )
2999
3000	v_comp(k) = v(k+1,j+1,i) + v(k,j+1,i) - gv
3001	flux_n(k) = v_comp(k) * ( &
3002	( 37.0_wp * ibit32 * adv_mom_5 &
3003	+ 7.0_wp * ibit31 * adv_mom_3 &
3004	+ ibit30 * adv_mom_1 &
3005	) * &
3006	( w(k,j+1,i) + w(k,j,i) ) &
3007	- ( 8.0_wp * ibit32 * adv_mom_5 &
3008	+ ibit31 * adv_mom_3 &
3009	) * &
3010	( w(k,j+2,i) + w(k,j-1,i) ) &
3011	+ ( ibit32 * adv_mom_5 &
3012	) * &
3013	( w(k,j+3,i) + w(k,j-2,i) ) &
3014	)
3015
3016	diss_n(k) = - ABS( v_comp(k) ) * ( &
3017	( 10.0_wp * ibit32 * adv_mom_5 &
3018	+ 3.0_wp * ibit31 * adv_mom_3 &
3019	+ ibit30 * adv_mom_1 &
3020	) * &
3021	( w(k,j+1,i) - w(k,j,i) ) &
3022	- ( 5.0_wp * ibit32 * adv_mom_5 &
3023	+ ibit31 * adv_mom_3 &
3024	) * &
3025	( w(k,j+2,i) - w(k,j-1,i) ) &
3026	+ ( ibit32 * adv_mom_5 &
3027	) * &
3028	( w(k,j+3,i) - w(k,j-2,i) ) &
3029	)
3030	ENDDO
3031
3032	DO k = nzb_max+1, nzt
3033
3034	u_comp(k) = u(k+1,j,i+1) + u(k,j,i+1) - gu
3035	flux_r(k) = u_comp(k) * ( &
3036	37.0_wp * ( w(k,j,i+1) + w(k,j,i) ) &
3037	- 8.0_wp * ( w(k,j,i+2) + w(k,j,i-1) ) &
3038	+ ( w(k,j,i+3) + w(k,j,i-2) ) ) * adv_mom_5
3039
3040	diss_r(k) = - ABS( u_comp(k) ) * ( &
3041	10.0_wp * ( w(k,j,i+1) - w(k,j,i) ) &
3042	- 5.0_wp * ( w(k,j,i+2) - w(k,j,i-1) ) &
3043	+ ( w(k,j,i+3) - w(k,j,i-2) ) ) * adv_mom_5
3044
3045	v_comp(k) = v(k+1,j+1,i) + v(k,j+1,i) - gv
3046	flux_n(k) = v_comp(k) * ( &
3047	37.0_wp * ( w(k,j+1,i) + w(k,j,i) ) &
3048	- 8.0_wp * ( w(k,j+2,i) + w(k,j-1,i) ) &
3049	+ ( w(k,j+3,i) + w(k,j-2,i) ) ) * adv_mom_5
3050
3051	diss_n(k) = - ABS( v_comp(k) ) * ( &
3052	10.0_wp * ( w(k,j+1,i) - w(k,j,i) ) &
3053	- 5.0_wp * ( w(k,j+2,i) - w(k,j-1,i) ) &
3054	+ ( w(k,j+3,i) - w(k,j-2,i) ) ) * adv_mom_5
3055	ENDDO
3056
3057	!
3058	!-- Now, compute vertical fluxes. Split loop into a part treating the
3059	!-- lowest 2 grid points with indirect indexing, a main loop without
3060	!-- indirect indexing, and a loop for the uppermost 2 grip points with
3061	!-- indirect indexing. This allows better vectorization for the main loop.
3062	!-- First, compute the flux at model surface, which need has to be
3063	!-- calculated explicetely for the tendency at
3064	!-- the first w-level. For topography wall this is done implicitely by
3065	!-- advc_flags_1.
3066	k = nzb + 1
3067	w_comp(k) = w(k,j,i) + w(k-1,j,i)
3068	flux_t(0) = w_comp(k) * ( w(k,j,i) + w(k-1,j,i) ) * adv_mom_1
3069	diss_t(0) = -ABS(w_comp(k)) * ( w(k,j,i) - w(k-1,j,i) ) * adv_mom_1
3070
3071	DO k = nzb+1, nzb+2
3072	!
3073	!-- k index has to be modified near bottom and top, else array
3074	!-- subscripts will be exceeded.
3075	ibit35 = REAL( IBITS(advc_flags_2(k,j,i),3,1), KIND = wp )
3076	ibit34 = REAL( IBITS(advc_flags_2(k,j,i),2,1), KIND = wp )
3077	ibit33 = REAL( IBITS(advc_flags_2(k,j,i),1,1), KIND = wp )
3078
3079	k_ppp = k + 3 * ibit35
3080	k_pp = k + 2 * ( 1 - ibit33 )
3081	k_mm = k - 2 * ibit35
3082
3083	w_comp(k) = w(k+1,j,i) + w(k,j,i)
3084	flux_t(k) = w_comp(k) * rho_air(k+1) * ( &
3085	( 37.0_wp * ibit35 * adv_mom_5 &
3086	+ 7.0_wp * ibit34 * adv_mom_3 &
3087	+ ibit33 * adv_mom_1 &
3088	) * &
3089	( w(k+1,j,i) + w(k,j,i) ) &
3090	- ( 8.0_wp * ibit35 * adv_mom_5 &
3091	+ ibit34 * adv_mom_3 &
3092	) * &
3093	( w(k_pp,j,i) + w(k-1,j,i) ) &
3094	+ ( ibit35 * adv_mom_5 &
3095	) * &
3096	( w(k_ppp,j,i) + w(k_mm,j,i) ) &
3097	)
3098
3099	diss_t(k) = - ABS( w_comp(k) ) * rho_air(k+1) * ( &
3100	( 10.0_wp * ibit35 * adv_mom_5 &
3101	+ 3.0_wp * ibit34 * adv_mom_3 &
3102	+ ibit33 * adv_mom_1 &
3103	) * &
3104	( w(k+1,j,i) - w(k,j,i) ) &
3105	- ( 5.0_wp * ibit35 * adv_mom_5 &
3106	+ ibit34 * adv_mom_3 &
3107	) * &
3108	( w(k_pp,j,i) - w(k-1,j,i) ) &
3109	+ ( ibit35 * adv_mom_5 &
3110	) * &
3111	( w(k_ppp,j,i) - w(k_mm,j,i) ) &
3112	)
3113	ENDDO
3114
3115	DO k = nzb+3, nzt-2
3116
3117	ibit35 = REAL( IBITS(advc_flags_2(k,j,i),3,1), KIND = wp )
3118	ibit34 = REAL( IBITS(advc_flags_2(k,j,i),2,1), KIND = wp )
3119	ibit33 = REAL( IBITS(advc_flags_2(k,j,i),1,1), KIND = wp )
3120
3121	w_comp(k) = w(k+1,j,i) + w(k,j,i)
3122	flux_t(k) = w_comp(k) * rho_air(k+1) * ( &
3123	( 37.0_wp * ibit35 * adv_mom_5 &
3124	+ 7.0_wp * ibit34 * adv_mom_3 &
3125	+ ibit33 * adv_mom_1 &
3126	) * &
3127	( w(k+1,j,i) + w(k,j,i) ) &
3128	- ( 8.0_wp * ibit35 * adv_mom_5 &
3129	+ ibit34 * adv_mom_3 &
3130	) * &
3131	( w(k+2,j,i) + w(k-1,j,i) ) &
3132	+ ( ibit35 * adv_mom_5 &
3133	) * &
3134	( w(k+3,j,i) + w(k-2,j,i) ) &
3135	)
3136
3137	diss_t(k) = - ABS( w_comp(k) ) * rho_air(k+1) * ( &
3138	( 10.0_wp * ibit35 * adv_mom_5 &
3139	+ 3.0_wp * ibit34 * adv_mom_3 &
3140	+ ibit33 * adv_mom_1 &
3141	) * &
3142	( w(k+1,j,i) - w(k,j,i) ) &
3143	- ( 5.0_wp * ibit35 * adv_mom_5 &
3144	+ ibit34 * adv_mom_3 &
3145	) * &
3146	( w(k+2,j,i) - w(k-1,j,i) ) &
3147	+ ( ibit35 * adv_mom_5 &
3148	) * &
3149	( w(k+3,j,i) - w(k-2,j,i) ) &
3150	)
3151	ENDDO
3152
3153	DO k = nzt-1, nzt
3154	!
3155	!-- k index has to be modified near bottom and top, else array
3156	!-- subscripts will be exceeded.
3157	ibit35 = REAL( IBITS(advc_flags_2(k,j,i),3,1), KIND = wp )
3158	ibit34 = REAL( IBITS(advc_flags_2(k,j,i),2,1), KIND = wp )
3159	ibit33 = REAL( IBITS(advc_flags_2(k,j,i),1,1), KIND = wp )
3160
3161	k_ppp = k + 3 * ibit35
3162	k_pp = k + 2 * ( 1 - ibit33 )
3163	k_mm = k - 2 * ibit35
3164
3165	w_comp(k) = w(k+1,j,i) + w(k,j,i)
3166	flux_t(k) = w_comp(k) * rho_air(k+1) * ( &
3167	( 37.0_wp * ibit35 * adv_mom_5 &
3168	+ 7.0_wp * ibit34 * adv_mom_3 &
3169	+ ibit33 * adv_mom_1 &
3170	) * &
3171	( w(k+1,j,i) + w(k,j,i) ) &
3172	- ( 8.0_wp * ibit35 * adv_mom_5 &
3173	+ ibit34 * adv_mom_3 &
3174	) * &
3175	( w(k_pp,j,i) + w(k-1,j,i) ) &
3176	+ ( ibit35 * adv_mom_5 &
3177	) * &
3178	( w(k_ppp,j,i) + w(k_mm,j,i) ) &
3179	)
3180
3181	diss_t(k) = - ABS( w_comp(k) ) * rho_air(k+1) * ( &
3182	( 10.0_wp * ibit35 * adv_mom_5 &
3183	+ 3.0_wp * ibit34 * adv_mom_3 &
3184	+ ibit33 * adv_mom_1 &
3185	) * &
3186	( w(k+1,j,i) - w(k,j,i) ) &
3187	- ( 5.0_wp * ibit35 * adv_mom_5 &
3188	+ ibit34 * adv_mom_3 &
3189	) * &
3190	( w(k_pp,j,i) - w(k-1,j,i) ) &
3191	+ ( ibit35 * adv_mom_5 &
3192	) * &
3193	( w(k_ppp,j,i) - w(k_mm,j,i) ) &
3194	)
3195	ENDDO
3196
3197	DO k = nzb+1, nzt
3198
3199	flux_d = flux_t(k-1)
3200	diss_d = diss_t(k-1)
3201	!
3202	!-- Calculate the divergence of the velocity field. A respective
3203	!-- correction is needed to overcome numerical instabilities introduced
3204	!-- by a not sufficient reduction of divergences near topography.
3205	div = ( ( ( u_comp(k) + gu ) * ( ibit27 + ibit28 + ibit29 ) &
3206	- ( u(k+1,j,i) + u(k,j,i) ) &
3207	* ( &
3208	REAL( IBITS(advc_flags_1(k,j,i-1),27,1), KIND = wp ) &
3209	+ REAL( IBITS(advc_flags_1(k,j,i-1),28,1), KIND = wp ) &
3210	+ REAL( IBITS(advc_flags_1(k,j,i-1),29,1), KIND = wp ) &
3211	) &
3212	) * ddx &
3213	+ ( ( v_comp(k) + gv ) * ( ibit30 + ibit31 + ibit32 ) &
3214	- ( v(k+1,j,i) + v(k,j,i) ) &
3215	* ( &
3216	REAL( IBITS(advc_flags_1(k,j-1,i),30,1), KIND = wp ) &
3217	+ REAL( IBITS(advc_flags_1(k,j-1,i),31,1), KIND = wp ) &
3218	+ REAL( IBITS(advc_flags_2(k,j-1,i),0,1), KIND = wp ) &
3219	) &
3220	) * ddy &
3221	+ ( w_comp(k) * rho_air(k+1) &
3222	* ( ibit33 + ibit34 + ibit35 ) &
3223	- ( w(k,j,i) + w(k-1,j,i) ) * rho_air(k) &
3224	* ( &
3225	REAL( IBITS(advc_flags_2(k-1,j,i),1,1), KIND = wp ) &
3226	+ REAL( IBITS(advc_flags_2(k-1,j,i),2,1), KIND = wp ) &
3227	+ REAL( IBITS(advc_flags_2(k-1,j,i),3,1), KIND = wp ) &
3228	) &
3229	) * drho_air_zw(k) * ddzu(k+1) &
3230	) * 0.5_wp
3231
3232	tend(k,j,i) = tend(k,j,i) - ( &
3233	( flux_r(k) + diss_r(k) &
3234	- flux_l_w(k,j,tn) - diss_l_w(k,j,tn) ) * ddx &
3235	+ ( flux_n(k) + diss_n(k) &
3236	- flux_s_w(k,tn) - diss_s_w(k,tn) ) * ddy &
3237	+ ( ( flux_t(k) + diss_t(k) ) &
3238	- ( flux_d + diss_d ) &
3239	) * drho_air_zw(k) * ddzu(k+1) &
3240	) + div * w(k,j,i)
3241
3242	flux_l_w(k,j,tn) = flux_r(k)
3243	diss_l_w(k,j,tn) = diss_r(k)
3244	flux_s_w(k,tn) = flux_n(k)
3245	diss_s_w(k,tn) = diss_n(k)
3246	!
3247	!-- Statistical Evaluation of w'w'.
3248	sums_ws2_ws_l(k,tn) = sums_ws2_ws_l(k,tn) &
3249	+ ( flux_t(k) &
3250	* ( w_comp(k) - 2.0_wp * hom(k,1,3,0) ) &
3251	/ ( w_comp(k) + SIGN( 1.0E-20_wp, w_comp(k) ) ) &
3252	+ diss_t(k) &
3253	* ABS( w_comp(k) - 2.0_wp * hom(k,1,3,0) ) &
3254	/ ( ABS( w_comp(k) ) + 1.0E-20_wp ) &
3255	) * weight_substep(intermediate_timestep_count)
3256
3257	ENDDO
3258
3259
3260	END SUBROUTINE advec_w_ws_ij
3261
3262
3263	!------------------------------------------------------------------------------!
3264	! Description:
3265	! ------------
3266	!> Scalar advection - Call for all grid points
3267	!------------------------------------------------------------------------------!
3268	SUBROUTINE advec_s_ws( sk, sk_char )
3269
3270	USE arrays_3d, &
3271	ONLY: ddzw, drho_air, rho_air_zw, tend, u, u_stokes_zu, v, &
3272	v_stokes_zu, w
3273
3274	USE control_parameters, &
3275	ONLY: intermediate_timestep_count, u_gtrans, v_gtrans
3276
3277	USE grid_variables, &
3278	ONLY: ddx, ddy
3279
3280	USE indices, &
3281	ONLY: nxl, nxlg, nxr, nxrg, nyn, nyng, nys, nysg, nzb, nzb_max, &
3282	nzt, advc_flags_1
3283
3284	USE kinds
3285
3286	USE salsa_util_mod, &
3287	ONLY: sums_salsa_ws_l
3288
3289	USE statistics, &
3290	ONLY: hom, sums_wspts_ws_l, sums_wsqs_ws_l, sums_wssas_ws_l, &
3291	sums_wsqcs_ws_l, sums_wsqrs_ws_l, sums_wsncs_ws_l, &
3292	sums_wsnrs_ws_l, sums_wsss_ws_l, weight_substep
3293
3294
3295
3296	IMPLICIT NONE
3297
3298	CHARACTER (LEN = *), INTENT(IN) :: sk_char !< string identifier, used for assign fluxes to the correct dimension in the analysis array
3299
3300	INTEGER(iwp) :: i !< grid index along x-direction
3301	INTEGER(iwp) :: j !< grid index along y-direction
3302	INTEGER(iwp) :: k !< grid index along z-direction
3303	INTEGER(iwp) :: k_mm !< k-2 index in disretization, can be modified to avoid segmentation faults
3304	INTEGER(iwp) :: k_pp !< k+2 index in disretization, can be modified to avoid segmentation faults
3305	INTEGER(iwp) :: k_ppp !< k+3 index in disretization, can be modified to avoid segmentation faults
3306	INTEGER(iwp) :: tn = 0 !< number of OpenMP thread
3307
3308	!
3309	!-- sk is an array from parameter list. It should not be a pointer, because
3310	!-- in that case the compiler can not assume a stride 1 and cannot perform
3311	!-- a strided one vector load. Adding the CONTIGUOUS keyword makes things
3312	!-- even worse, because the compiler cannot assume strided one in the
3313	!-- caller side.
3314	REAL(wp), INTENT(IN),DIMENSION(nzb:nzt+1,nysg:nyng,nxlg:nxrg) :: sk !< advected scalar
3315
3316	REAL(wp) :: ibit0 !< flag indicating 1st-order scheme along x-direction
3317	REAL(wp) :: ibit1 !< flag indicating 3rd-order scheme along x-direction
3318	REAL(wp) :: ibit2 !< flag indicating 5th-order scheme along x-direction
3319	REAL(wp) :: ibit3 !< flag indicating 1st-order scheme along y-direction
3320	REAL(wp) :: ibit4 !< flag indicating 3rd-order scheme along y-direction
3321	REAL(wp) :: ibit5 !< flag indicating 5th-order scheme along y-direction
3322	REAL(wp) :: ibit6 !< flag indicating 1st-order scheme along z-direction
3323	REAL(wp) :: ibit7 !< flag indicating 3rd-order scheme along z-direction
3324	REAL(wp) :: ibit8 !< flag indicating 5th-order scheme along z-direction
3325	REAL(wp) :: diss_d !< artificial dissipation term at grid box bottom
3326	REAL(wp) :: div !< diverence on scalar grid
3327	REAL(wp) :: flux_d !< 6th-order flux at grid box bottom
3328	REAL(wp) :: u_comp !< advection velocity along x-direction
3329	REAL(wp) :: v_comp !< advection velocity along y-direction
3330
3331	REAL(wp), DIMENSION(nzb:nzt) :: diss_n !< discretized artificial dissipation at northward-side of the grid box
3332	REAL(wp), DIMENSION(nzb:nzt) :: diss_r !< discretized artificial dissipation at rightward-side of the grid box
3333	REAL(wp), DIMENSION(nzb:nzt) :: diss_t !< discretized artificial dissipation at rightward-side of the grid box
3334	REAL(wp), DIMENSION(nzb:nzt) :: flux_n !< discretized 6th-order flux at northward-side of the grid box
3335	REAL(wp), DIMENSION(nzb:nzt) :: flux_r !< discretized 6th-order flux at rightward-side of the grid box
3336	REAL(wp), DIMENSION(nzb:nzt) :: flux_t !< discretized 6th-order flux at rightward-side of the grid box
3337
3338	REAL(wp), DIMENSION(nzb+1:nzt) :: swap_diss_y_local !< discretized artificial dissipation term at southward-side of the grid box
3339	REAL(wp), DIMENSION(nzb+1:nzt) :: swap_flux_y_local !< discretized 6th-order flux at northward-side of the grid box
3340
3341	REAL(wp), DIMENSION(nzb+1:nzt,nys:nyn) :: swap_diss_x_local !< discretized artificial dissipation term at leftward-side of the grid box
3342	REAL(wp), DIMENSION(nzb+1:nzt,nys:nyn) :: swap_flux_x_local !< discretized 6th-order flux at leftward-side of the grid box
3343
3344
3345	!
3346	!-- Compute the fluxes for the whole left boundary of the processor domain.
3347	i = nxl
3348	DO j = nys, nyn
3349
3350	DO k = nzb+1, nzb_max
3351
3352	ibit2 = REAL( IBITS(advc_flags_1(k,j,i-1),2,1), KIND = wp )
3353	ibit1 = REAL( IBITS(advc_flags_1(k,j,i-1),1,1), KIND = wp )
3354	ibit0 = REAL( IBITS(advc_flags_1(k,j,i-1),0,1), KIND = wp )
3355
3356	u_comp = u(k,j,i) - u_gtrans + u_stokes_zu(k)
3357	swap_flux_x_local(k,j) = u_comp * ( &
3358	( 37.0_wp * ibit2 * adv_sca_5 &
3359	+ 7.0_wp * ibit1 * adv_sca_3 &
3360	+ ibit0 * adv_sca_1 &
3361	) * &
3362	( sk(k,j,i) + sk(k,j,i-1) ) &
3363	- ( 8.0_wp * ibit2 * adv_sca_5 &
3364	+ ibit1 * adv_sca_3 &
3365	) * &
3366	( sk(k,j,i+1) + sk(k,j,i-2) ) &
3367	+ ( ibit2 * adv_sca_5 &
3368	) * &
3369	( sk(k,j,i+2) + sk(k,j,i-3) ) &
3370	)
3371
3372	swap_diss_x_local(k,j) = -ABS( u_comp ) * ( &
3373	( 10.0_wp * ibit2 * adv_sca_5 &
3374	+ 3.0_wp * ibit1 * adv_sca_3 &
3375	+ ibit0 * adv_sca_1 &
3376	) * &
3377	( sk(k,j,i) - sk(k,j,i-1) ) &
3378	- ( 5.0_wp * ibit2 * adv_sca_5 &
3379	+ ibit1 * adv_sca_3 &
3380	) * &
3381	( sk(k,j,i+1) - sk(k,j,i-2) ) &
3382	+ ( ibit2 * adv_sca_5 &
3383	) * &
3384	( sk(k,j,i+2) - sk(k,j,i-3) ) &
3385	)
3386
3387	ENDDO
3388
3389	DO k = nzb_max+1, nzt
3390
3391	u_comp = u(k,j,i) - u_gtrans + u_stokes_zu(k)
3392	swap_flux_x_local(k,j) = u_comp * ( &
3393	37.0_wp * ( sk(k,j,i) + sk(k,j,i-1) ) &
3394	- 8.0_wp * ( sk(k,j,i+1) + sk(k,j,i-2) ) &
3395	+ ( sk(k,j,i+2) + sk(k,j,i-3) ) &
3396	) * adv_sca_5
3397
3398	swap_diss_x_local(k,j) = -ABS( u_comp ) * ( &
3399	10.0_wp * ( sk(k,j,i) - sk(k,j,i-1) ) &
3400	- 5.0_wp * ( sk(k,j,i+1) - sk(k,j,i-2) ) &
3401	+ ( sk(k,j,i+2) - sk(k,j,i-3) ) &
3402	) * adv_sca_5
3403
3404	ENDDO
3405
3406	ENDDO
3407
3408	DO i = nxl, nxr
3409
3410	j = nys
3411	DO k = nzb+1, nzb_max
3412
3413	ibit5 = REAL( IBITS(advc_flags_1(k,j-1,i),5,1), KIND = wp )
3414	ibit4 = REAL( IBITS(advc_flags_1(k,j-1,i),4,1), KIND = wp )
3415	ibit3 = REAL( IBITS(advc_flags_1(k,j-1,i),3,1), KIND = wp )
3416
3417	v_comp = v(k,j,i) - v_gtrans + v_stokes_zu(k)
3418	swap_flux_y_local(k) = v_comp * ( &
3419	( 37.0_wp * ibit5 * adv_sca_5 &
3420	+ 7.0_wp * ibit4 * adv_sca_3 &
3421	+ ibit3 * adv_sca_1 &
3422	) * &
3423	( sk(k,j,i) + sk(k,j-1,i) ) &
3424	- ( 8.0_wp * ibit5 * adv_sca_5 &
3425	+ ibit4 * adv_sca_3 &
3426	) * &
3427	( sk(k,j+1,i) + sk(k,j-2,i) ) &
3428	+ ( ibit5 * adv_sca_5 &
3429	) * &
3430	( sk(k,j+2,i) + sk(k,j-3,i) ) &
3431	)
3432
3433	swap_diss_y_local(k) = -ABS( v_comp ) * ( &
3434	( 10.0_wp * ibit5 * adv_sca_5 &
3435	+ 3.0_wp * ibit4 * adv_sca_3 &
3436	+ ibit3 * adv_sca_1 &
3437	) * &
3438	( sk(k,j,i) - sk(k,j-1,i) ) &
3439	- ( 5.0_wp * ibit5 * adv_sca_5 &
3440	+ ibit4 * adv_sca_3 &
3441	) * &
3442	( sk(k,j+1,i) - sk(k,j-2,i) ) &
3443	+ ( ibit5 * adv_sca_5 &
3444	) * &
3445	( sk(k,j+2,i) - sk(k,j-3,i) ) &
3446	)
3447
3448	ENDDO
3449	!
3450	!-- Above to the top of the highest topography. No degradation necessary.
3451	DO k = nzb_max+1, nzt
3452
3453	v_comp = v(k,j,i) - v_gtrans + v_stokes_zu(k)
3454	swap_flux_y_local(k) = v_comp * ( &
3455	37.0_wp * ( sk(k,j,i) + sk(k,j-1,i) ) &
3456	- 8.0_wp * ( sk(k,j+1,i) + sk(k,j-2,i) ) &
3457	+ ( sk(k,j+2,i) + sk(k,j-3,i) ) &
3458	) * adv_sca_5
3459	swap_diss_y_local(k) = -ABS( v_comp ) * ( &
3460	10.0_wp * ( sk(k,j,i) - sk(k,j-1,i) ) &
3461	- 5.0_wp * ( sk(k,j+1,i) - sk(k,j-2,i) ) &
3462	+ sk(k,j+2,i) - sk(k,j-3,i) &
3463	) * adv_sca_5
3464
3465	ENDDO
3466
3467	DO j = nys, nyn
3468
3469	flux_t(0) = 0.0_wp
3470	diss_t(0) = 0.0_wp
3471	flux_d = 0.0_wp
3472	diss_d = 0.0_wp
3473
3474	DO k = nzb+1, nzb_max
3475
3476	ibit2 = REAL( IBITS(advc_flags_1(k,j,i),2,1), KIND = wp )
3477	ibit1 = REAL( IBITS(advc_flags_1(k,j,i),1,1), KIND = wp )
3478	ibit0 = REAL( IBITS(advc_flags_1(k,j,i),0,1), KIND = wp )
3479
3480	u_comp = u(k,j,i+1) - u_gtrans + u_stokes_zu(k)
3481	flux_r(k) = u_comp * ( &
3482	( 37.0_wp * ibit2 * adv_sca_5 &
3483	+ 7.0_wp * ibit1 * adv_sca_3 &
3484	+ ibit0 * adv_sca_1 &
3485	) * &
3486	( sk(k,j,i+1) + sk(k,j,i) ) &
3487	- ( 8.0_wp * ibit2 * adv_sca_5 &
3488	+ ibit1 * adv_sca_3 &
3489	) * &
3490	( sk(k,j,i+2) + sk(k,j,i-1) ) &
3491	+ ( ibit2 * adv_sca_5 &
3492	) * &
3493	( sk(k,j,i+3) + sk(k,j,i-2) ) &
3494	)
3495
3496	diss_r(k) = -ABS( u_comp ) * ( &
3497	( 10.0_wp * ibit2 * adv_sca_5 &
3498	+ 3.0_wp * ibit1 * adv_sca_3 &
3499	+ ibit0 * adv_sca_1 &
3500	) * &
3501	( sk(k,j,i+1) - sk(k,j,i) ) &
3502	- ( 5.0_wp * ibit2 * adv_sca_5 &
3503	+ ibit1 * adv_sca_3 &
3504	) * &
3505	( sk(k,j,i+2) - sk(k,j,i-1) ) &
3506	+ ( ibit2 * adv_sca_5 &
3507	) * &
3508	( sk(k,j,i+3) - sk(k,j,i-2) ) &
3509	)
3510
3511	ibit5 = REAL( IBITS(advc_flags_1(k,j,i),5,1), KIND = wp )
3512	ibit4 = REAL( IBITS(advc_flags_1(k,j,i),4,1), KIND = wp )
3513	ibit3 = REAL( IBITS(advc_flags_1(k,j,i),3,1), KIND = wp )
3514
3515	v_comp = v(k,j+1,i) - v_gtrans + v_stokes_zu(k)
3516	flux_n(k) = v_comp * ( &
3517	( 37.0_wp * ibit5 * adv_sca_5 &
3518	+ 7.0_wp * ibit4 * adv_sca_3 &
3519	+ ibit3 * adv_sca_1 &
3520	) * &
3521	( sk(k,j+1,i) + sk(k,j,i) ) &
3522	- ( 8.0_wp * ibit5 * adv_sca_5 &
3523	+ ibit4 * adv_sca_3 &
3524	) * &
3525	( sk(k,j+2,i) + sk(k,j-1,i) ) &
3526	+ ( ibit5 * adv_sca_5 &
3527	) * &
3528	( sk(k,j+3,i) + sk(k,j-2,i) ) &
3529	)
3530
3531	diss_n(k) = -ABS( v_comp ) * ( &
3532	( 10.0_wp * ibit5 * adv_sca_5 &
3533	+ 3.0_wp * ibit4 * adv_sca_3 &
3534	+ ibit3 * adv_sca_1 &
3535	) * &
3536	( sk(k,j+1,i) - sk(k,j,i) ) &
3537	- ( 5.0_wp * ibit5 * adv_sca_5 &
3538	+ ibit4 * adv_sca_3 &
3539	) * &
3540	( sk(k,j+2,i) - sk(k,j-1,i) ) &
3541	+ ( ibit5 * adv_sca_5 &
3542	) * &
3543	( sk(k,j+3,i) - sk(k,j-2,i) ) &
3544	)
3545	!
3546	!-- k index has to be modified near bottom and top, else array
3547	!-- subscripts will be exceeded.
3548	ibit8 = REAL( IBITS(advc_flags_1(k,j,i),8,1), KIND = wp )
3549	ibit7 = REAL( IBITS(advc_flags_1(k,j,i),7,1), KIND = wp )
3550	ibit6 = REAL( IBITS(advc_flags_1(k,j,i),6,1), KIND = wp )
3551
3552	k_ppp = k + 3 * ibit8
3553	k_pp = k + 2 * ( 1 - ibit6 )
3554	k_mm = k - 2 * ibit8
3555
3556
3557	flux_t(k) = w(k,j,i) * rho_air_zw(k) * ( &
3558	( 37.0_wp * ibit8 * adv_sca_5 &
3559	+ 7.0_wp * ibit7 * adv_sca_3 &
3560	+ ibit6 * adv_sca_1 &
3561	) * &
3562	( sk(k+1,j,i) + sk(k,j,i) ) &
3563	- ( 8.0_wp * ibit8 * adv_sca_5 &
3564	+ ibit7 * adv_sca_3 &
3565	) * &
3566	( sk(k_pp,j,i) + sk(k-1,j,i) ) &
3567	+ ( ibit8 * adv_sca_5 &
3568	) * ( sk(k_ppp,j,i)+ sk(k_mm,j,i) ) &
3569	)
3570
3571	diss_t(k) = -ABS( w(k,j,i) ) * rho_air_zw(k) * ( &
3572	( 10.0_wp * ibit8 * adv_sca_5 &
3573	+ 3.0_wp * ibit7 * adv_sca_3 &
3574	+ ibit6 * adv_sca_1 &
3575	) * &
3576	( sk(k+1,j,i) - sk(k,j,i) ) &
3577	- ( 5.0_wp * ibit8 * adv_sca_5 &
3578	+ ibit7 * adv_sca_3 &
3579	) * &
3580	( sk(k_pp,j,i) - sk(k-1,j,i) ) &
3581	+ ( ibit8 * adv_sca_5 &
3582	) * &
3583	( sk(k_ppp,j,i) - sk(k_mm,j,i) ) &
3584	)
3585	!
3586	!-- Calculate the divergence of the velocity field. A respective
3587	!-- correction is needed to overcome numerical instabilities caused
3588	!-- by a not sufficient reduction of divergences near topography.
3589	div = ( u(k,j,i+1) * ( ibit0 + ibit1 + ibit2 ) &
3590	- u(k,j,i) * ( &
3591	REAL( IBITS(advc_flags_1(k,j,i-1),0,1), KIND = wp ) &
3592	+ REAL( IBITS(advc_flags_1(k,j,i-1),1,1), KIND = wp ) &
3593	+ REAL( IBITS(advc_flags_1(k,j,i-1),2,1), KIND = wp ) &
3594	) &
3595	) * ddx &
3596	+ ( v(k,j+1,i) * ( ibit3 + ibit4 + ibit5 ) &
3597	- v(k,j,i) * ( &
3598	REAL( IBITS(advc_flags_1(k,j-1,i),3,1), KIND = wp ) &
3599	+ REAL( IBITS(advc_flags_1(k,j-1,i),4,1), KIND = wp ) &
3600	+ REAL( IBITS(advc_flags_1(k,j-1,i),5,1), KIND = wp ) &
3601	) &
3602	) * ddy &
3603	+ ( w(k,j,i) * rho_air_zw(k) * &
3604	( ibit6 + ibit7 + ibit8 ) &
3605	- w(k-1,j,i) * rho_air_zw(k-1) * &
3606	( &
3607	REAL( IBITS(advc_flags_1(k-1,j,i),6,1), KIND = wp ) &
3608	+ REAL( IBITS(advc_flags_1(k-1,j,i),7,1), KIND = wp ) &
3609	+ REAL( IBITS(advc_flags_1(k-1,j,i),8,1), KIND = wp ) &
3610	) &
3611	) * drho_air(k) * ddzw(k)
3612
3613
3614	tend(k,j,i) = tend(k,j,i) - ( &
3615	( flux_r(k) + diss_r(k) - swap_flux_x_local(k,j) - &
3616	swap_diss_x_local(k,j) ) * ddx &
3617	+ ( flux_n(k) + diss_n(k) - swap_flux_y_local(k) - &
3618	swap_diss_y_local(k) ) * ddy &
3619	+ ( ( flux_t(k) + diss_t(k) ) - &
3620	( flux_d + diss_d ) &
3621	) * drho_air(k) * ddzw(k) &
3622	) + sk(k,j,i) * div
3623
3624	swap_flux_y_local(k) = flux_n(k)
3625	swap_diss_y_local(k) = diss_n(k)
3626	swap_flux_x_local(k,j) = flux_r(k)
3627	swap_diss_x_local(k,j) = diss_r(k)
3628	flux_d = flux_t(k)
3629	diss_d = diss_t(k)
3630
3631	ENDDO
3632
3633	DO k = nzb_max+1, nzt
3634
3635	u_comp = u(k,j,i+1) - u_gtrans + u_stokes_zu(k)
3636	flux_r(k) = u_comp * ( &
3637	37.0_wp * ( sk(k,j,i+1) + sk(k,j,i) ) &
3638	- 8.0_wp * ( sk(k,j,i+2) + sk(k,j,i-1) ) &
3639	+ ( sk(k,j,i+3) + sk(k,j,i-2) ) ) * adv_sca_5
3640	diss_r(k) = -ABS( u_comp ) * ( &
3641	10.0_wp * ( sk(k,j,i+1) - sk(k,j,i) ) &
3642	- 5.0_wp * ( sk(k,j,i+2) - sk(k,j,i-1) ) &
3643	+ ( sk(k,j,i+3) - sk(k,j,i-2) ) ) * adv_sca_5
3644
3645	v_comp = v(k,j+1,i) - v_gtrans + v_stokes_zu(k)
3646	flux_n(k) = v_comp * ( &
3647	37.0_wp * ( sk(k,j+1,i) + sk(k,j,i) ) &
3648	- 8.0_wp * ( sk(k,j+2,i) + sk(k,j-1,i) ) &
3649	+ ( sk(k,j+3,i) + sk(k,j-2,i) ) ) * adv_sca_5
3650	diss_n(k) = -ABS( v_comp ) * ( &
3651	10.0_wp * ( sk(k,j+1,i) - sk(k,j,i) ) &
3652	- 5.0_wp * ( sk(k,j+2,i) - sk(k,j-1,i) ) &
3653	+ ( sk(k,j+3,i) - sk(k,j-2,i) ) ) * adv_sca_5
3654	!
3655	!-- k index has to be modified near bottom and top, else array
3656	!-- subscripts will be exceeded.
3657	ibit8 = REAL( IBITS(advc_flags_1(k,j,i),8,1), KIND = wp )
3658	ibit7 = REAL( IBITS(advc_flags_1(k,j,i),7,1), KIND = wp )
3659	ibit6 = REAL( IBITS(advc_flags_1(k,j,i),6,1), KIND = wp )
3660
3661	k_ppp = k + 3 * ibit8
3662	k_pp = k + 2 * ( 1 - ibit6 )
3663	k_mm = k - 2 * ibit8
3664
3665
3666	flux_t(k) = w(k,j,i) * rho_air_zw(k) * ( &
3667	( 37.0_wp * ibit8 * adv_sca_5 &
3668	+ 7.0_wp * ibit7 * adv_sca_3 &
3669	+ ibit6 * adv_sca_1 &
3670	) * &
3671	( sk(k+1,j,i) + sk(k,j,i) ) &
3672	- ( 8.0_wp * ibit8 * adv_sca_5 &
3673	+ ibit7 * adv_sca_3 &
3674	) * &
3675	( sk(k_pp,j,i) + sk(k-1,j,i) ) &
3676	+ ( ibit8 * adv_sca_5 &
3677	) * ( sk(k_ppp,j,i)+ sk(k_mm,j,i) ) &
3678	)
3679
3680	diss_t(k) = -ABS( w(k,j,i) ) * rho_air_zw(k) * ( &
3681	( 10.0_wp * ibit8 * adv_sca_5 &
3682	+ 3.0_wp * ibit7 * adv_sca_3 &
3683	+ ibit6 * adv_sca_1 &
3684	) * &
3685	( sk(k+1,j,i) - sk(k,j,i) ) &
3686	- ( 5.0_wp * ibit8 * adv_sca_5 &
3687	+ ibit7 * adv_sca_3 &
3688	) * &
3689	( sk(k_pp,j,i) - sk(k-1,j,i) ) &
3690	+ ( ibit8 * adv_sca_5 &
3691	) * &
3692	( sk(k_ppp,j,i) - sk(k_mm,j,i) ) &
3693	)
3694	!
3695	!-- Calculate the divergence of the velocity field. A respective
3696	!-- correction is needed to overcome numerical instabilities introduced
3697	!-- by a not sufficient reduction of divergences near topography.
3698	div = ( u(k,j,i+1) - u(k,j,i) ) * ddx &
3699	+ ( v(k,j+1,i) - v(k,j,i) ) * ddy &
3700	+ ( w(k,j,i) * rho_air_zw(k) - &
3701	w(k-1,j,i) * rho_air_zw(k-1) &
3702	) * drho_air(k) * ddzw(k)
3703
3704	tend(k,j,i) = tend(k,j,i) - ( &
3705	( flux_r(k) + diss_r(k) - swap_flux_x_local(k,j) - &
3706	swap_diss_x_local(k,j) ) * ddx &
3707	+ ( flux_n(k) + diss_n(k) - swap_flux_y_local(k) - &
3708	swap_diss_y_local(k) ) * ddy &
3709	+ ( ( flux_t(k) + diss_t(k) ) - &
3710	( flux_d + diss_d ) &
3711	) * drho_air(k) * ddzw(k) &
3712	) + sk(k,j,i) * div
3713
3714	swap_flux_y_local(k) = flux_n(k)
3715	swap_diss_y_local(k) = diss_n(k)
3716	swap_flux_x_local(k,j) = flux_r(k)
3717	swap_diss_x_local(k,j) = diss_r(k)
3718	flux_d = flux_t(k)
3719	diss_d = diss_t(k)
3720
3721	ENDDO
3722	!
3723	!-- Evaluation of statistics.
3724	SELECT CASE ( sk_char )
3725
3726	CASE ( 'pt' )
3727	DO k = nzb, nzt
3728	sums_wspts_ws_l(k,tn) = sums_wspts_ws_l(k,tn) &
3729	+ ( flux_t(k) &
3730	/ ( w(k,j,i) + SIGN( 1.0E-20_wp, w(k,j,i) ) ) &
3731	* ( w(k,j,i) - hom(k,1,3,0) ) &
3732	+ diss_t(k) &
3733	/ ( ABS(w(k,j,i)) + 1.0E-20_wp ) &
3734	* ABS(w(k,j,i) - hom(k,1,3,0) ) &
3735	) * weight_substep(intermediate_timestep_count)
3736	ENDDO
3737	CASE ( 'sa' )
3738	DO k = nzb, nzt
3739	sums_wssas_ws_l(k,tn) = sums_wssas_ws_l(k,tn) &
3740	+ ( flux_t(k) &
3741	/ ( w(k,j,i) + SIGN( 1.0E-20_wp, w(k,j,i) ) ) &
3742	* ( w(k,j,i) - hom(k,1,3,0) ) &
3743	+ diss_t(k) &
3744	/ ( ABS(w(k,j,i)) + 1.0E-20_wp ) &
3745	* ABS(w(k,j,i) - hom(k,1,3,0) ) &
3746	) * weight_substep(intermediate_timestep_count)
3747	ENDDO
3748	CASE ( 'q' )
3749	DO k = nzb, nzt
3750	sums_wsqs_ws_l(k,tn) = sums_wsqs_ws_l(k,tn) &
3751	+ ( flux_t(k) &
3752	/ ( w(k,j,i) + SIGN( 1.0E-20_wp, w(k,j,i) ) ) &
3753	* ( w(k,j,i) - hom(k,1,3,0) ) &
3754	+ diss_t(k) &
3755	/ ( ABS(w(k,j,i)) + 1.0E-20_wp ) &
3756	* ABS(w(k,j,i) - hom(k,1,3,0) ) &
3757	) * weight_substep(intermediate_timestep_count)
3758	ENDDO
3759	CASE ( 'qc' )
3760	DO k = nzb, nzt
3761	sums_wsqcs_ws_l(k,tn) = sums_wsqcs_ws_l(k,tn) &
3762	+ ( flux_t(k) &
3763	/ ( w(k,j,i) + SIGN( 1.0E-20_wp, w(k,j,i) ) ) &
3764	* ( w(k,j,i) - hom(k,1,3,0) ) &
3765	+ diss_t(k) &
3766	/ ( ABS(w(k,j,i)) + 1.0E-20_wp ) &
3767	* ABS(w(k,j,i) - hom(k,1,3,0) ) &
3768	) * weight_substep(intermediate_timestep_count)
3769	ENDDO
3770	CASE ( 'qr' )
3771	DO k = nzb, nzt
3772	sums_wsqrs_ws_l(k,tn) = sums_wsqrs_ws_l(k,tn) &
3773	+ ( flux_t(k) &
3774	/ ( w(k,j,i) + SIGN( 1.0E-20_wp, w(k,j,i) ) ) &
3775	* ( w(k,j,i) - hom(k,1,3,0) ) &
3776	+ diss_t(k) &
3777	/ ( ABS(w(k,j,i)) + 1.0E-20_wp ) &
3778	* ABS(w(k,j,i) - hom(k,1,3,0) ) &
3779	) * weight_substep(intermediate_timestep_count)
3780	ENDDO
3781	CASE ( 'nc' )
3782	DO k = nzb, nzt
3783	sums_wsncs_ws_l(k,tn) = sums_wsncs_ws_l(k,tn) &
3784	+ ( flux_t(k) &
3785	/ ( w(k,j,i) + SIGN( 1.0E-20_wp, w(k,j,i) ) ) &
3786	* ( w(k,j,i) - hom(k,1,3,0) ) &
3787	+ diss_t(k) &
3788	/ ( ABS(w(k,j,i)) + 1.0E-20_wp ) &
3789	* ABS(w(k,j,i) - hom(k,1,3,0) ) &
3790	) * weight_substep(intermediate_timestep_count)
3791	ENDDO
3792	CASE ( 'nr' )
3793	DO k = nzb, nzt
3794	sums_wsnrs_ws_l(k,tn) = sums_wsnrs_ws_l(k,tn) &
3795	+ ( flux_t(k) &
3796	/ ( w(k,j,i) + SIGN( 1.0E-20_wp, w(k,j,i) ) ) &
3797	* ( w(k,j,i) - hom(k,1,3,0) ) &
3798	+ diss_t(k) &
3799	/ ( ABS(w(k,j,i)) + 1.0E-20_wp ) &
3800	* ABS(w(k,j,i) - hom(k,1,3,0) ) &
3801	) * weight_substep(intermediate_timestep_count)
3802	ENDDO
3803	CASE ( 's' )
3804	DO k = nzb, nzt
3805	sums_wsss_ws_l(k,tn) = sums_wsss_ws_l(k,tn) &
3806	+ ( flux_t(k) &
3807	/ ( w(k,j,i) + SIGN( 1.0E-20_wp, w(k,j,i) ) ) &
3808	* ( w(k,j,i) - hom(k,1,3,0) ) &
3809	+ diss_t(k) &
3810	/ ( ABS(w(k,j,i)) + 1.0E-20_wp ) &
3811	* ABS(w(k,j,i) - hom(k,1,3,0) ) &
3812	) * weight_substep(intermediate_timestep_count)
3813	ENDDO
3814
3815	CASE ( 'aerosol_mass', 'aerosol_number', 'salsa_gas' )
3816	DO k = nzb, nzt
3817	sums_salsa_ws_l(k,tn) = sums_salsa_ws_l(k,tn) &
3818	+ ( flux_t(k) &
3819	/ ( w(k,j,i) + SIGN( 1.0E-20_wp, w(k,j,i) ) ) &
3820	* ( w(k,j,i) - hom(k,1,3,0) ) &
3821	+ diss_t(k) &
3822	/ ( ABS(w(k,j,i)) + 1.0E-20_wp ) &
3823	* ABS(w(k,j,i) - hom(k,1,3,0) ) &
3824	) * weight_substep(intermediate_timestep_count)
3825	ENDDO
3826
3827
3828	END SELECT
3829
3830	ENDDO
3831	ENDDO
3832
3833	END SUBROUTINE advec_s_ws
3834
3835
3836	!------------------------------------------------------------------------------!
3837	! Description:
3838	! ------------
3839	!> Advection of u - Call for all grid points
3840	!------------------------------------------------------------------------------!
3841	SUBROUTINE advec_u_ws
3842
3843	USE arrays_3d, &
3844	ONLY: ddzw, drho_air, tend, u, v, w, rho_air_zw
3845
3846	USE control_parameters, &
3847	ONLY: intermediate_timestep_count, u_gtrans, v_gtrans
3848
3849	USE grid_variables, &
3850	ONLY: ddx, ddy
3851
3852	USE indices, &
3853	ONLY: nxlu, nxr, nyn, nys, nzb, nzb_max, nzt, advc_flags_1
3854
3855	USE kinds
3856
3857	USE statistics, &
3858	ONLY: hom, sums_us2_ws_l, sums_wsus_ws_l, weight_substep
3859
3860	IMPLICIT NONE
3861
3862	INTEGER(iwp) :: i !< grid index along x-direction
3863	INTEGER(iwp) :: j !< grid index along y-direction
3864	INTEGER(iwp) :: k !< grid index along z-direction
3865	INTEGER(iwp) :: k_mm !< k-2 index in disretization, can be modified to avoid segmentation faults
3866	INTEGER(iwp) :: k_pp !< k+2 index in disretization, can be modified to avoid segmentation faults
3867	INTEGER(iwp) :: k_ppp !< k+3 index in disretization, can be modified to avoid segmentation faults
3868	INTEGER(iwp) :: tn = 0 !< number of OpenMP thread
3869
3870	REAL(wp) :: ibit9 !< flag indicating 1st-order scheme along x-direction
3871	REAL(wp) :: ibit10 !< flag indicating 3rd-order scheme along x-direction
3872	REAL(wp) :: ibit11 !< flag indicating 5th-order scheme along x-direction
3873	REAL(wp) :: ibit12 !< flag indicating 1st-order scheme along y-direction
3874	REAL(wp) :: ibit13 !< flag indicating 3rd-order scheme along y-direction
3875	REAL(wp) :: ibit14 !< flag indicating 5th-order scheme along y-direction
3876	REAL(wp) :: ibit15 !< flag indicating 1st-order scheme along z-direction
3877	REAL(wp) :: ibit16 !< flag indicating 3rd-order scheme along z-direction
3878	REAL(wp) :: ibit17 !< flag indicating 5th-order scheme along z-direction
3879	REAL(wp) :: diss_d !< artificial dissipation term at grid box bottom
3880	REAL(wp) :: div !< diverence on u-grid
3881	REAL(wp) :: flux_d !< 6th-order flux at grid box bottom
3882	REAL(wp) :: gu !< Galilei-transformation velocity along x
3883	REAL(wp) :: gv !< Galilei-transformation velocity along y
3884	REAL(wp) :: v_comp !< advection velocity along y
3885	REAL(wp) :: w_comp !< advection velocity along z
3886
3887	REAL(wp), DIMENSION(nzb+1:nzt) :: swap_diss_y_local_u !< discretized artificial dissipation at southward-side of the grid box
3888	REAL(wp), DIMENSION(nzb+1:nzt) :: swap_flux_y_local_u !< discretized 6th-order flux at southward-side of the grid box
3889
3890	REAL(wp), DIMENSION(nzb+1:nzt,nys:nyn) :: swap_diss_x_local_u !< discretized artificial dissipation at leftward-side of the grid box
3891	REAL(wp), DIMENSION(nzb+1:nzt,nys:nyn) :: swap_flux_x_local_u !< discretized 6th-order flux at leftward-side of the grid box
3892
3893	REAL(wp), DIMENSION(nzb:nzt) :: diss_n !< discretized artificial dissipation at northward-side of the grid box
3894	REAL(wp), DIMENSION(nzb:nzt) :: diss_r !< discretized artificial dissipation at leftward-side of the grid box
3895	REAL(wp), DIMENSION(nzb:nzt) :: diss_t !< discretized artificial dissipation at top of the grid box
3896	REAL(wp), DIMENSION(nzb:nzt) :: flux_n !< discretized 6th-order flux at northward-side of the grid box
3897	REAL(wp), DIMENSION(nzb:nzt) :: flux_r !< discretized 6th-order flux at rightward-side of the grid box
3898	REAL(wp), DIMENSION(nzb:nzt) :: flux_t !< discretized 6th-order flux at top of the grid box
3899	REAL(wp), DIMENSION(nzb:nzt) :: u_comp !< advection velocity along x
3900
3901	gu = 2.0_wp * u_gtrans
3902	gv = 2.0_wp * v_gtrans
3903
3904	!
3905	!-- Compute the fluxes for the whole left boundary of the processor domain.
3906	i = nxlu
3907	DO j = nys, nyn
3908	DO k = nzb+1, nzb_max
3909
3910	ibit11 = REAL( IBITS(advc_flags_1(k,j,i-1),11,1), KIND = wp )
3911	ibit10 = REAL( IBITS(advc_flags_1(k,j,i-1),10,1), KIND = wp )
3912	ibit9 = REAL( IBITS(advc_flags_1(k,j,i-1),9,1), KIND = wp )
3913
3914	u_comp(k) = u(k,j,i) + u(k,j,i-1) - gu
3915	swap_flux_x_local_u(k,j) = u_comp(k) * ( &
3916	( 37.0_wp * ibit11 * adv_mom_5 &
3917	+ 7.0_wp * ibit10 * adv_mom_3 &
3918	+ ibit9 * adv_mom_1 &
3919	) * &
3920	( u(k,j,i) + u(k,j,i-1) ) &
3921	- ( 8.0_wp * ibit11 * adv_mom_5 &
3922	+ ibit10 * adv_mom_3 &
3923	) * &
3924	( u(k,j,i+1) + u(k,j,i-2) ) &
3925	+ ( ibit11 * adv_mom_5 &
3926	) * &
3927	( u(k,j,i+2) + u(k,j,i-3) ) &
3928	)
3929
3930	swap_diss_x_local_u(k,j) = - ABS( u_comp(k) ) * ( &
3931	( 10.0_wp * ibit11 * adv_mom_5 &
3932	+ 3.0_wp * ibit10 * adv_mom_3 &
3933	+ ibit9 * adv_mom_1 &
3934	) * &
3935	( u(k,j,i) - u(k,j,i-1) ) &
3936	- ( 5.0_wp * ibit11 * adv_mom_5 &
3937	+ ibit10 * adv_mom_3 &
3938	) * &
3939	( u(k,j,i+1) - u(k,j,i-2) ) &
3940	+ ( ibit11 * adv_mom_5 &
3941	) * &
3942	( u(k,j,i+2) - u(k,j,i-3) ) &
3943	)
3944
3945	ENDDO
3946
3947	DO k = nzb_max+1, nzt
3948
3949	u_comp(k) = u(k,j,i) + u(k,j,i-1) - gu
3950	swap_flux_x_local_u(k,j) = u_comp(k) * ( &
3951	37.0_wp * ( u(k,j,i) + u(k,j,i-1) ) &
3952	- 8.0_wp * ( u(k,j,i+1) + u(k,j,i-2) ) &
3953	+ ( u(k,j,i+2) + u(k,j,i-3) ) ) * adv_mom_5
3954	swap_diss_x_local_u(k,j) = - ABS(u_comp(k)) * ( &
3955	10.0_wp * ( u(k,j,i) - u(k,j,i-1) ) &
3956	- 5.0_wp * ( u(k,j,i+1) - u(k,j,i-2) ) &
3957	+ ( u(k,j,i+2) - u(k,j,i-3) ) ) * adv_mom_5
3958
3959	ENDDO
3960	ENDDO
3961
3962	DO i = nxlu, nxr
3963	!
3964	!-- The following loop computes the fluxes for the south boundary points
3965	j = nys
3966	DO k = nzb+1, nzb_max
3967
3968	ibit14 = REAL( IBITS(advc_flags_1(k,j-1,i),14,1), KIND = wp )
3969	ibit13 = REAL( IBITS(advc_flags_1(k,j-1,i),13,1), KIND = wp )
3970	ibit12 = REAL( IBITS(advc_flags_1(k,j-1,i),12,1), KIND = wp )
3971
3972	v_comp = v(k,j,i) + v(k,j,i-1) - gv
3973	swap_flux_y_local_u(k) = v_comp * ( &
3974	( 37.0_wp * ibit14 * adv_mom_5 &
3975	+ 7.0_wp * ibit13 * adv_mom_3 &
3976	+ ibit12 * adv_mom_1 &
3977	) * &
3978	( u(k,j,i) + u(k,j-1,i) ) &
3979	- ( 8.0_wp * ibit14 * adv_mom_5 &
3980	+ ibit13 * adv_mom_3 &
3981	) * &
3982	( u(k,j+1,i) + u(k,j-2,i) ) &
3983	+ ( ibit14 * adv_mom_5 &
3984	) * &
3985	( u(k,j+2,i) + u(k,j-3,i) ) &
3986	)
3987
3988	swap_diss_y_local_u(k) = - ABS ( v_comp ) * ( &
3989	( 10.0_wp * ibit14 * adv_mom_5 &
3990	+ 3.0_wp * ibit13 * adv_mom_3 &
3991	+ ibit12 * adv_mom_1 &
3992	) * &
3993	( u(k,j,i) - u(k,j-1,i) ) &
3994	- ( 5.0_wp * ibit14 * adv_mom_5 &
3995	+ ibit13 * adv_mom_3 &
3996	) * &
3997	( u(k,j+1,i) - u(k,j-2,i) ) &
3998	+ ( ibit14 * adv_mom_5 &
3999	) * &
4000	( u(k,j+2,i) - u(k,j-3,i) ) &
4001	)
4002
4003	ENDDO
4004
4005	DO k = nzb_max+1, nzt
4006
4007	v_comp = v(k,j,i) + v(k,j,i-1) - gv
4008	swap_flux_y_local_u(k) = v_comp * ( &
4009	37.0_wp * ( u(k,j,i) + u(k,j-1,i) ) &
4010	- 8.0_wp * ( u(k,j+1,i) + u(k,j-2,i) ) &
4011	+ ( u(k,j+2,i) + u(k,j-3,i) ) ) * adv_mom_5
4012	swap_diss_y_local_u(k) = - ABS(v_comp) * ( &
4013	10.0_wp * ( u(k,j,i) - u(k,j-1,i) ) &
4014	- 5.0_wp * ( u(k,j+1,i) - u(k,j-2,i) ) &
4015	+ ( u(k,j+2,i) - u(k,j-3,i) ) ) * adv_mom_5
4016
4017	ENDDO
4018	!
4019	!-- Computation of interior fluxes and tendency terms
4020	DO j = nys, nyn
4021
4022	flux_t(0) = 0.0_wp
4023	diss_t(0) = 0.0_wp
4024	flux_d = 0.0_wp
4025	diss_d = 0.0_wp
4026
4027	DO k = nzb+1, nzb_max
4028
4029	ibit11 = REAL( IBITS(advc_flags_1(k,j,i),11,1), KIND = wp )
4030	ibit10 = REAL( IBITS(advc_flags_1(k,j,i),10,1), KIND = wp )
4031	ibit9 = REAL( IBITS(advc_flags_1(k,j,i),9,1), KIND = wp )
4032
4033	u_comp(k) = u(k,j,i+1) + u(k,j,i)
4034	flux_r(k) = ( u_comp(k) - gu ) * ( &
4035	( 37.0_wp * ibit11 * adv_mom_5 &
4036	+ 7.0_wp * ibit10 * adv_mom_3 &
4037	+ ibit9 * adv_mom_1 &
4038	) * &
4039	( u(k,j,i+1) + u(k,j,i) ) &
4040	- ( 8.0_wp * ibit11 * adv_mom_5 &
4041	+ ibit10 * adv_mom_3 &
4042	) * &
4043	( u(k,j,i+2) + u(k,j,i-1) ) &
4044	+ ( ibit11 * adv_mom_5 &
4045	) * &
4046	( u(k,j,i+3) + u(k,j,i-2) ) &
4047	)
4048
4049	diss_r(k) = - ABS( u_comp(k) - gu ) * ( &
4050	( 10.0_wp * ibit11 * adv_mom_5 &
4051	+ 3.0_wp * ibit10 * adv_mom_3 &
4052	+ ibit9 * adv_mom_1 &
4053	) * &
4054	( u(k,j,i+1) - u(k,j,i) ) &
4055	- ( 5.0_wp * ibit11 * adv_mom_5 &
4056	+ ibit10 * adv_mom_3 &
4057	) * &
4058	( u(k,j,i+2) - u(k,j,i-1) ) &
4059	+ ( ibit11 * adv_mom_5 &
4060	) * &
4061	( u(k,j,i+3) - u(k,j,i-2) ) &
4062	)
4063
4064	ibit14 = REAL( IBITS(advc_flags_1(k,j,i),14,1), KIND = wp )
4065	ibit13 = REAL( IBITS(advc_flags_1(k,j,i),13,1), KIND = wp )
4066	ibit12 = REAL( IBITS(advc_flags_1(k,j,i),12,1), KIND = wp )
4067
4068	v_comp = v(k,j+1,i) + v(k,j+1,i-1) - gv
4069	flux_n(k) = v_comp * ( &
4070	( 37.0_wp * ibit14 * adv_mom_5 &
4071	+ 7.0_wp * ibit13 * adv_mom_3 &
4072	+ ibit12 * adv_mom_1 &
4073	) * &
4074	( u(k,j+1,i) + u(k,j,i) ) &
4075	- ( 8.0_wp * ibit14 * adv_mom_5 &
4076	+ ibit13 * adv_mom_3 &
4077	) * &
4078	( u(k,j+2,i) + u(k,j-1,i) ) &
4079	+ ( ibit14 * adv_mom_5 &
4080	) * &
4081	( u(k,j+3,i) + u(k,j-2,i) ) &
4082	)
4083
4084	diss_n(k) = - ABS ( v_comp ) * ( &
4085	( 10.0_wp * ibit14 * adv_mom_5 &
4086	+ 3.0_wp * ibit13 * adv_mom_3 &
4087	+ ibit12 * adv_mom_1 &
4088	) * &
4089	( u(k,j+1,i) - u(k,j,i) ) &
4090	- ( 5.0_wp * ibit14 * adv_mom_5 &
4091	+ ibit13 * adv_mom_3 &
4092	) * &
4093	( u(k,j+2,i) - u(k,j-1,i) ) &
4094	+ ( ibit14 * adv_mom_5 &
4095	) * &
4096	( u(k,j+3,i) - u(k,j-2,i) ) &
4097	)
4098	!
4099	!-- k index has to be modified near bottom and top, else array
4100	!-- subscripts will be exceeded.
4101	ibit17 = REAL( IBITS(advc_flags_1(k,j,i),17,1), KIND = wp )
4102	ibit16 = REAL( IBITS(advc_flags_1(k,j,i),16,1), KIND = wp )
4103	ibit15 = REAL( IBITS(advc_flags_1(k,j,i),15,1), KIND = wp )
4104
4105	k_ppp = k + 3 * ibit17
4106	k_pp = k + 2 * ( 1 - ibit15 )
4107	k_mm = k - 2 * ibit17
4108
4109	w_comp = w(k,j,i) + w(k,j,i-1)
4110	flux_t(k) = w_comp * rho_air_zw(k) * ( &
4111	( 37.0_wp * ibit17 * adv_mom_5 &
4112	+ 7.0_wp * ibit16 * adv_mom_3 &
4113	+ ibit15 * adv_mom_1 &
4114	) * &
4115	( u(k+1,j,i) + u(k,j,i) ) &
4116	- ( 8.0_wp * ibit17 * adv_mom_5 &
4117	+ ibit16 * adv_mom_3 &
4118	) * &
4119	( u(k_pp,j,i) + u(k-1,j,i) ) &
4120	+ ( ibit17 * adv_mom_5 &
4121	) * &
4122	( u(k_ppp,j,i) + u(k_mm,j,i) ) &
4123	)
4124
4125	diss_t(k) = - ABS( w_comp ) * rho_air_zw(k) * ( &
4126	( 10.0_wp * ibit17 * adv_mom_5 &
4127	+ 3.0_wp * ibit16 * adv_mom_3 &
4128	+ ibit15 * adv_mom_1 &
4129	) * &
4130	( u(k+1,j,i) - u(k,j,i) ) &
4131	- ( 5.0_wp * ibit17 * adv_mom_5 &
4132	+ ibit16 * adv_mom_3 &
4133	) * &
4134	( u(k_pp,j,i) - u(k-1,j,i) ) &
4135	+ ( ibit17 * adv_mom_5 &
4136	) * &
4137	( u(k_ppp,j,i) - u(k_mm,j,i) ) &
4138	)
4139	!
4140	!-- Calculate the divergence of the velocity field. A respective
4141	!-- correction is needed to overcome numerical instabilities caused
4142	!-- by a not sufficient reduction of divergences near topography.
4143	div = ( ( u_comp(k) * ( ibit9 + ibit10 + ibit11 ) &
4144	- ( u(k,j,i) + u(k,j,i-1) ) &
4145	* ( &
4146	REAL( IBITS(advc_flags_1(k,j,i-1),9,1), KIND = wp ) &
4147	+ REAL( IBITS(advc_flags_1(k,j,i-1),10,1), KIND = wp ) &
4148	+ REAL( IBITS(advc_flags_1(k,j,i-1),11,1), KIND = wp ) &
4149	) &
4150	) * ddx &
4151	+ ( ( v_comp + gv ) * ( ibit12 + ibit13 + ibit14 ) &
4152	- ( v(k,j,i) + v(k,j,i-1 ) ) &
4153	* ( &
4154	REAL( IBITS(advc_flags_1(k,j-1,i),12,1), KIND = wp ) &
4155	+ REAL( IBITS(advc_flags_1(k,j-1,i),13,1), KIND = wp ) &
4156	+ REAL( IBITS(advc_flags_1(k,j-1,i),14,1), KIND = wp ) &
4157	) &
4158	) * ddy &
4159	+ ( w_comp * rho_air_zw(k) * ( ibit15 + ibit16 + ibit17 ) &
4160	- ( w(k-1,j,i) + w(k-1,j,i-1) ) * rho_air_zw(k-1) &
4161	* ( &
4162	REAL( IBITS(advc_flags_1(k-1,j,i),15,1), KIND = wp ) &
4163	+ REAL( IBITS(advc_flags_1(k-1,j,i),16,1), KIND = wp ) &
4164	+ REAL( IBITS(advc_flags_1(k-1,j,i),17,1), KIND = wp ) &
4165	) &
4166	) * drho_air(k) * ddzw(k) &
4167	) * 0.5_wp
4168
4169
4170
4171	tend(k,j,i) = tend(k,j,i) - ( &
4172	( flux_r(k) + diss_r(k) &
4173	- swap_flux_x_local_u(k,j) - swap_diss_x_local_u(k,j) ) * ddx &
4174	+ ( flux_n(k) + diss_n(k) &
4175	- swap_flux_y_local_u(k) - swap_diss_y_local_u(k) ) * ddy &
4176	+ ( ( flux_t(k) + diss_t(k) ) &
4177	- ( flux_d + diss_d ) &
4178	) * drho_air(k) * ddzw(k) &
4179	) + div * u(k,j,i)
4180
4181	swap_flux_x_local_u(k,j) = flux_r(k)
4182	swap_diss_x_local_u(k,j) = diss_r(k)
4183	swap_flux_y_local_u(k) = flux_n(k)
4184	swap_diss_y_local_u(k) = diss_n(k)
4185	flux_d = flux_t(k)
4186	diss_d = diss_t(k)
4187	!
4188	!-- Statistical Evaluation of u'u'. The factor has to be applied
4189	!-- for right evaluation when gallilei_trans = .T. .
4190	sums_us2_ws_l(k,tn) = sums_us2_ws_l(k,tn) &
4191	+ ( flux_r(k) &
4192	* ( u_comp(k) - 2.0_wp * hom(k,1,1,0) ) &
4193	/ ( u_comp(k) - gu + SIGN( 1.0E-20_wp, u_comp(k) - gu ) ) &
4194	+ diss_r(k) &
4195	* ABS( u_comp(k) - 2.0_wp * hom(k,1,1,0) ) &
4196	/ ( ABS( u_comp(k) - gu ) + 1.0E-20_wp ) &
4197	) * weight_substep(intermediate_timestep_count)
4198	!
4199	!-- Statistical Evaluation of w'u'.
4200	sums_wsus_ws_l(k,tn) = sums_wsus_ws_l(k,tn) &
4201	+ ( flux_t(k) &
4202	* ( w_comp - 2.0_wp * hom(k,1,3,0) ) &
4203	/ ( w_comp + SIGN( 1.0E-20_wp, w_comp ) ) &
4204	+ diss_t(k) &
4205	* ABS( w_comp - 2.0_wp * hom(k,1,3,0) ) &
4206	/ ( ABS( w_comp ) + 1.0E-20_wp ) &
4207	) * weight_substep(intermediate_timestep_count)
4208
4209	ENDDO
4210
4211	DO k = nzb_max+1, nzt
4212
4213	u_comp(k) = u(k,j,i+1) + u(k,j,i)
4214	flux_r(k) = ( u_comp(k) - gu ) * ( &
4215	37.0_wp * ( u(k,j,i+1) + u(k,j,i) ) &
4216	- 8.0_wp * ( u(k,j,i+2) + u(k,j,i-1) ) &
4217	+ ( u(k,j,i+3) + u(k,j,i-2) ) ) * adv_mom_5
4218	diss_r(k) = - ABS( u_comp(k) - gu ) * ( &
4219	10.0_wp * ( u(k,j,i+1) - u(k,j,i) ) &
4220	- 5.0_wp * ( u(k,j,i+2) - u(k,j,i-1) ) &
4221	+ ( u(k,j,i+3) - u(k,j,i-2) ) ) * adv_mom_5
4222
4223	v_comp = v(k,j+1,i) + v(k,j+1,i-1) - gv
4224	flux_n(k) = v_comp * ( &
4225	37.0_wp * ( u(k,j+1,i) + u(k,j,i) ) &
4226	- 8.0_wp * ( u(k,j+2,i) + u(k,j-1,i) ) &
4227	+ ( u(k,j+3,i) + u(k,j-2,i) ) ) * adv_mom_5
4228	diss_n(k) = - ABS( v_comp ) * ( &
4229	10.0_wp * ( u(k,j+1,i) - u(k,j,i) ) &
4230	- 5.0_wp * ( u(k,j+2,i) - u(k,j-1,i) ) &
4231	+ ( u(k,j+3,i) - u(k,j-2,i) ) ) * adv_mom_5
4232	!
4233	!-- k index has to be modified near bottom and top, else array
4234	!-- subscripts will be exceeded.
4235	ibit17 = REAL( IBITS(advc_flags_1(k,j,i),17,1), KIND = wp )
4236	ibit16 = REAL( IBITS(advc_flags_1(k,j,i),16,1), KIND = wp )
4237	ibit15 = REAL( IBITS(advc_flags_1(k,j,i),15,1), KIND = wp )
4238
4239	k_ppp = k + 3 * ibit17
4240	k_pp = k + 2 * ( 1 - ibit15 )
4241	k_mm = k - 2 * ibit17
4242
4243	w_comp = w(k,j,i) + w(k,j,i-1)
4244	flux_t(k) = w_comp * rho_air_zw(k) * ( &
4245	( 37.0_wp * ibit17 * adv_mom_5 &
4246	+ 7.0_wp * ibit16 * adv_mom_3 &
4247	+ ibit15 * adv_mom_1 &
4248	) * &
4249	( u(k+1,j,i) + u(k,j,i) ) &
4250	- ( 8.0_wp * ibit17 * adv_mom_5 &
4251	+ ibit16 * adv_mom_3 &
4252	) * &
4253	( u(k_pp,j,i) + u(k-1,j,i) ) &
4254	+ ( ibit17 * adv_mom_5 &
4255	) * &
4256	( u(k_ppp,j,i) + u(k_mm,j,i) ) &
4257	)
4258
4259	diss_t(k) = - ABS( w_comp ) * rho_air_zw(k) * ( &
4260	( 10.0_wp * ibit17 * adv_mom_5 &
4261	+ 3.0_wp * ibit16 * adv_mom_3 &
4262	+ ibit15 * adv_mom_1 &
4263	) * &
4264	( u(k+1,j,i) - u(k,j,i) ) &
4265	- ( 5.0_wp * ibit17 * adv_mom_5 &
4266	+ ibit16 * adv_mom_3 &
4267	) * &
4268	( u(k_pp,j,i) - u(k-1,j,i) ) &
4269	+ ( ibit17 * adv_mom_5 &
4270	) * &
4271	( u(k_ppp,j,i) - u(k_mm,j,i) ) &
4272	)
4273	!
4274	!-- Calculate the divergence of the velocity field. A respective
4275	!-- correction is needed to overcome numerical instabilities caused
4276	!-- by a not sufficient reduction of divergences near topography.
4277	div = ( ( u_comp(k) - ( u(k,j,i) + u(k,j,i-1) ) ) * ddx &
4278	+ ( v_comp + gv - ( v(k,j,i) + v(k,j,i-1 ) ) ) * ddy &
4279	+ ( w_comp * rho_air_zw(k) - &
4280	( w(k-1,j,i) + w(k-1,j,i-1) ) * rho_air_zw(k-1) &
4281	) * drho_air(k) * ddzw(k) &
4282	) * 0.5_wp
4283
4284	tend(k,j,i) = tend(k,j,i) - ( &
4285	( flux_r(k) + diss_r(k) &
4286	- swap_flux_x_local_u(k,j) - swap_diss_x_local_u(k,j) ) * ddx &
4287	+ ( flux_n(k) + diss_n(k) &
4288	- swap_flux_y_local_u(k) - swap_diss_y_local_u(k) ) * ddy &
4289	+ ( ( flux_t(k) + diss_t(k) ) &
4290	- ( flux_d + diss_d ) &
4291	) * drho_air(k) * ddzw(k) &
4292	) + div * u(k,j,i)
4293
4294	swap_flux_x_local_u(k,j) = flux_r(k)
4295	swap_diss_x_local_u(k,j) = diss_r(k)
4296	swap_flux_y_local_u(k) = flux_n(k)
4297	swap_diss_y_local_u(k) = diss_n(k)
4298	flux_d = flux_t(k)
4299	diss_d = diss_t(k)
4300	!
4301	!-- Statistical Evaluation of u'u'. The factor has to be applied
4302	!-- for right evaluation when gallilei_trans = .T. .
4303	sums_us2_ws_l(k,tn) = sums_us2_ws_l(k,tn) &
4304	+ ( flux_r(k) &
4305	* ( u_comp(k) - 2.0_wp * hom(k,1,1,0) ) &
4306	/ ( u_comp(k) - gu + SIGN( 1.0E-20_wp, u_comp(k) - gu ) ) &
4307	+ diss_r(k) &
4308	* ABS( u_comp(k) - 2.0_wp * hom(k,1,1,0) ) &
4309	/ ( ABS( u_comp(k) - gu ) + 1.0E-20_wp ) &
4310	) * weight_substep(intermediate_timestep_count)
4311	!
4312	!-- Statistical Evaluation of w'u'.
4313	sums_wsus_ws_l(k,tn) = sums_wsus_ws_l(k,tn) &
4314	+ ( flux_t(k) &
4315	* ( w_comp - 2.0_wp * hom(k,1,3,0) ) &
4316	/ ( w_comp + SIGN( 1.0E-20_wp, w_comp ) ) &
4317	+ diss_t(k) &
4318	* ABS( w_comp - 2.0_wp * hom(k,1,3,0) ) &
4319	/ ( ABS( w_comp ) + 1.0E-20_wp ) &
4320	) * weight_substep(intermediate_timestep_count)
4321	ENDDO
4322	ENDDO
4323	ENDDO
4324	sums_us2_ws_l(nzb,tn) = sums_us2_ws_l(nzb+1,tn)
4325
4326
4327	END SUBROUTINE advec_u_ws
4328
4329
4330	!------------------------------------------------------------------------------!
4331	! Description:
4332	! ------------
4333	!> Advection of v - Call for all grid points
4334	!------------------------------------------------------------------------------!
4335	SUBROUTINE advec_v_ws
4336
4337	USE arrays_3d, &
4338	ONLY: ddzw, drho_air, tend, u, v, w, rho_air_zw
4339
4340	USE control_parameters, &
4341	ONLY: intermediate_timestep_count, u_gtrans, v_gtrans
4342
4343	USE grid_variables, &
4344	ONLY: ddx, ddy
4345
4346	USE indices, &
4347	ONLY: nxl, nxr, nyn, nys, nysv, nzb, nzb_max, nzt, advc_flags_1
4348
4349	USE kinds
4350
4351	USE statistics, &
4352	ONLY: hom, sums_vs2_ws_l, sums_wsvs_ws_l, weight_substep
4353
4354	IMPLICIT NONE
4355
4356
4357	INTEGER(iwp) :: i !< grid index along x-direction
4358	INTEGER(iwp) :: j !< grid index along y-direction
4359	INTEGER(iwp) :: k !< grid index along z-direction
4360	INTEGER(iwp) :: k_mm !< k-2 index in disretization, can be modified to avoid segmentation faults
4361	INTEGER(iwp) :: k_pp !< k+2 index in disretization, can be modified to avoid segmentation faults
4362	INTEGER(iwp) :: k_ppp !< k+3 index in disretization, can be modified to avoid segmentation faults
4363	INTEGER(iwp) :: tn = 0 !< number of OpenMP thread
4364
4365	REAL(wp) :: ibit18 !< flag indicating 1st-order scheme along x-direction
4366	REAL(wp) :: ibit19 !< flag indicating 3rd-order scheme along x-direction
4367	REAL(wp) :: ibit20 !< flag indicating 5th-order scheme along x-direction
4368	REAL(wp) :: ibit21 !< flag indicating 1st-order scheme along y-direction
4369	REAL(wp) :: ibit22 !< flag indicating 3rd-order scheme along y-direction
4370	REAL(wp) :: ibit23 !< flag indicating 5th-order scheme along y-direction
4371	REAL(wp) :: ibit24 !< flag indicating 1st-order scheme along z-direction
4372	REAL(wp) :: ibit25 !< flag indicating 3rd-order scheme along z-direction
4373	REAL(wp) :: ibit26 !< flag indicating 5th-order scheme along z-direction
4374	REAL(wp) :: diss_d !< artificial dissipation term at grid box bottom
4375	REAL(wp) :: div !< diverence on v-grid
4376	REAL(wp) :: flux_d !< artificial 6th-order flux at grid box bottom
4377	REAL(wp) :: gu !< Galilei-transformation velocity along x
4378	REAL(wp) :: gv !< Galilei-transformation velocity along y
4379	REAL(wp) :: u_comp !< advection velocity along x
4380	REAL(wp) :: w_comp !< advection velocity along z
4381
4382	REAL(wp), DIMENSION(nzb+1:nzt) :: swap_diss_y_local_v !< discretized artificial dissipation at southward-side of the grid box
4383	REAL(wp), DIMENSION(nzb+1:nzt) :: swap_flux_y_local_v !< discretized 6th-order flux at southward-side of the grid box
4384
4385	REAL(wp), DIMENSION(nzb+1:nzt,nys:nyn) :: swap_diss_x_local_v !< discretized artificial dissipation at leftward-side of the grid box
4386	REAL(wp), DIMENSION(nzb+1:nzt,nys:nyn) :: swap_flux_x_local_v !< discretized 6th-order flux at leftward-side of the grid box
4387
4388	REAL(wp), DIMENSION(nzb:nzt) :: diss_n !< discretized artificial dissipation at northward-side of the grid box
4389	REAL(wp), DIMENSION(nzb:nzt) :: diss_r !< discretized artificial dissipation at rightward-side of the grid box
4390	REAL(wp), DIMENSION(nzb:nzt) :: diss_t !< discretized artificial dissipation at top of the grid box
4391	REAL(wp), DIMENSION(nzb:nzt) :: flux_n !< discretized 6th-order flux at northward-side of the grid box
4392	REAL(wp), DIMENSION(nzb:nzt) :: flux_r !< discretized 6th-order flux at rightward-side of the grid box
4393	REAL(wp), DIMENSION(nzb:nzt) :: flux_t !< discretized 6th-order flux at top of the grid box
4394	REAL(wp), DIMENSION(nzb:nzt) :: v_comp !< advection velocity along y
4395
4396	gu = 2.0_wp * u_gtrans
4397	gv = 2.0_wp * v_gtrans
4398	!
4399	!-- First compute the whole left boundary of the processor domain
4400	i = nxl
4401	DO j = nysv, nyn
4402	DO k = nzb+1, nzb_max
4403
4404	ibit20 = REAL( IBITS(advc_flags_1(k,j,i-1),20,1), KIND = wp )
4405	ibit19 = REAL( IBITS(advc_flags_1(k,j,i-1),19,1), KIND = wp )
4406	ibit18 = REAL( IBITS(advc_flags_1(k,j,i-1),18,1), KIND = wp )
4407
4408	u_comp = u(k,j-1,i) + u(k,j,i) - gu
4409	swap_flux_x_local_v(k,j) = u_comp * ( &
4410	( 37.0_wp * ibit20 * adv_mom_5 &
4411	+ 7.0_wp * ibit19 * adv_mom_3 &
4412	+ ibit18 * adv_mom_1 &
4413	) * &
4414	( v(k,j,i) + v(k,j,i-1) ) &
4415	- ( 8.0_wp * ibit20 * adv_mom_5 &
4416	+ ibit19 * adv_mom_3 &
4417	) * &
4418	( v(k,j,i+1) + v(k,j,i-2) ) &
4419	+ ( ibit20 * adv_mom_5 &
4420	) * &
4421	( v(k,j,i+2) + v(k,j,i-3) ) &
4422	)
4423
4424	swap_diss_x_local_v(k,j) = - ABS( u_comp ) * ( &
4425	( 10.0_wp * ibit20 * adv_mom_5 &
4426	+ 3.0_wp * ibit19 * adv_mom_3 &
4427	+ ibit18 * adv_mom_1 &
4428	) * &
4429	( v(k,j,i) - v(k,j,i-1) ) &
4430	- ( 5.0_wp * ibit20 * adv_mom_5 &
4431	+ ibit19 * adv_mom_3 &
4432	) * &
4433	( v(k,j,i+1) - v(k,j,i-2) ) &
4434	+ ( ibit20 * adv_mom_5 &
4435	) * &
4436	( v(k,j,i+2) - v(k,j,i-3) ) &
4437	)
4438
4439	ENDDO
4440
4441	DO k = nzb_max+1, nzt
4442
4443	u_comp = u(k,j-1,i) + u(k,j,i) - gu
4444	swap_flux_x_local_v(k,j) = u_comp * ( &
4445	37.0_wp * ( v(k,j,i) + v(k,j,i-1) ) &
4446	- 8.0_wp * ( v(k,j,i+1) + v(k,j,i-2) ) &
4447	+ ( v(k,j,i+2) + v(k,j,i-3) ) ) * adv_mom_5
4448	swap_diss_x_local_v(k,j) = - ABS( u_comp ) * ( &
4449	10.0_wp * ( v(k,j,i) - v(k,j,i-1) ) &
4450	- 5.0_wp * ( v(k,j,i+1) - v(k,j,i-2) ) &
4451	+ ( v(k,j,i+2) - v(k,j,i-3) ) ) * adv_mom_5
4452
4453	ENDDO
4454
4455	ENDDO
4456
4457	DO i = nxl, nxr
4458
4459	j = nysv
4460	DO k = nzb+1, nzb_max
4461
4462	ibit23 = REAL( IBITS(advc_flags_1(k,j-1,i),23,1), KIND = wp )
4463	ibit22 = REAL( IBITS(advc_flags_1(k,j-1,i),22,1), KIND = wp )
4464	ibit21 = REAL( IBITS(advc_flags_1(k,j-1,i),21,1), KIND = wp )
4465
4466	v_comp(k) = v(k,j,i) + v(k,j-1,i) - gv
4467	swap_flux_y_local_v(k) = v_comp(k) * ( &
4468	( 37.0_wp * ibit23 * adv_mom_5 &
4469	+ 7.0_wp * ibit22 * adv_mom_3 &
4470	+ ibit21 * adv_mom_1 &
4471	) * &
4472	( v(k,j,i) + v(k,j-1,i) ) &
4473	- ( 8.0_wp * ibit23 * adv_mom_5 &
4474	+ ibit22 * adv_mom_3 &
4475	) * &
4476	( v(k,j+1,i) + v(k,j-2,i) ) &
4477	+ ( ibit23 * adv_mom_5 &
4478	) * &
4479	( v(k,j+2,i) + v(k,j-3,i) ) &
4480	)
4481
4482	swap_diss_y_local_v(k) = - ABS( v_comp(k) ) * ( &
4483	( 10.0_wp * ibit23 * adv_mom_5 &
4484	+ 3.0_wp * ibit22 * adv_mom_3 &
4485	+ ibit21 * adv_mom_1 &
4486	) * &
4487	( v(k,j,i) - v(k,j-1,i) ) &
4488	- ( 5.0_wp * ibit23 * adv_mom_5 &
4489	+ ibit22 * adv_mom_3 &
4490	) * &
4491	( v(k,j+1,i) - v(k,j-2,i) ) &
4492	+ ( ibit23 * adv_mom_5 &
4493	) * &
4494	( v(k,j+2,i) - v(k,j-3,i) ) &
4495	)
4496
4497	ENDDO
4498
4499	DO k = nzb_max+1, nzt
4500
4501	v_comp(k) = v(k,j,i) + v(k,j-1,i) - gv
4502	swap_flux_y_local_v(k) = v_comp(k) * ( &
4503	37.0_wp * ( v(k,j,i) + v(k,j-1,i) ) &
4504	- 8.0_wp * ( v(k,j+1,i) + v(k,j-2,i) ) &
4505	+ ( v(k,j+2,i) + v(k,j-3,i) ) ) * adv_mom_5
4506	swap_diss_y_local_v(k) = - ABS( v_comp(k) ) * ( &
4507	10.0_wp * ( v(k,j,i) - v(k,j-1,i) ) &
4508	- 5.0_wp * ( v(k,j+1,i) - v(k,j-2,i) ) &
4509	+ ( v(k,j+2,i) - v(k,j-3,i) ) ) * adv_mom_5
4510
4511	ENDDO
4512
4513	DO j = nysv, nyn
4514
4515	flux_t(0) = 0.0_wp
4516	diss_t(0) = 0.0_wp
4517	flux_d = 0.0_wp
4518	diss_d = 0.0_wp
4519
4520	DO k = nzb+1, nzb_max
4521
4522	ibit20 = REAL( IBITS(advc_flags_1(k,j,i),20,1), KIND = wp )
4523	ibit19 = REAL( IBITS(advc_flags_1(k,j,i),19,1), KIND = wp )
4524	ibit18 = REAL( IBITS(advc_flags_1(k,j,i),18,1), KIND = wp )
4525
4526	u_comp = u(k,j-1,i+1) + u(k,j,i+1) - gu
4527	flux_r(k) = u_comp * ( &
4528	( 37.0_wp * ibit20 * adv_mom_5 &
4529	+ 7.0_wp * ibit19 * adv_mom_3 &
4530	+ ibit18 * adv_mom_1 &
4531	) * &
4532	( v(k,j,i+1) + v(k,j,i) ) &
4533	- ( 8.0_wp * ibit20 * adv_mom_5 &
4534	+ ibit19 * adv_mom_3 &
4535	) * &
4536	( v(k,j,i+2) + v(k,j,i-1) ) &
4537	+ ( ibit20 * adv_mom_5 &
4538	) * &
4539	( v(k,j,i+3) + v(k,j,i-2) ) &
4540	)
4541
4542	diss_r(k) = - ABS( u_comp ) * ( &
4543	( 10.0_wp * ibit20 * adv_mom_5 &
4544	+ 3.0_wp * ibit19 * adv_mom_3 &
4545	+ ibit18 * adv_mom_1 &
4546	) * &
4547	( v(k,j,i+1) - v(k,j,i) ) &
4548	- ( 5.0_wp * ibit20 * adv_mom_5 &
4549	+ ibit19 * adv_mom_3 &
4550	) * &
4551	( v(k,j,i+2) - v(k,j,i-1) ) &
4552	+ ( ibit20 * adv_mom_5 &
4553	) * &
4554	( v(k,j,i+3) - v(k,j,i-2) ) &
4555	)
4556
4557	ibit23 = REAL( IBITS(advc_flags_1(k,j,i),23,1), KIND = wp )
4558	ibit22 = REAL( IBITS(advc_flags_1(k,j,i),22,1), KIND = wp )
4559	ibit21 = REAL( IBITS(advc_flags_1(k,j,i),21,1), KIND = wp )
4560
4561	v_comp(k) = v(k,j+1,i) + v(k,j,i)
4562	flux_n(k) = ( v_comp(k) - gv ) * ( &
4563	( 37.0_wp * ibit23 * adv_mom_5 &
4564	+ 7.0_wp * ibit22 * adv_mom_3 &
4565	+ ibit21 * adv_mom_1 &
4566	) * &
4567	( v(k,j+1,i) + v(k,j,i) ) &
4568	- ( 8.0_wp * ibit23 * adv_mom_5 &
4569	+ ibit22 * adv_mom_3 &
4570	) * &
4571	( v(k,j+2,i) + v(k,j-1,i) ) &
4572	+ ( ibit23 * adv_mom_5 &
4573	) * &
4574	( v(k,j+3,i) + v(k,j-2,i) ) &
4575	)
4576
4577	diss_n(k) = - ABS( v_comp(k) - gv ) * ( &
4578	( 10.0_wp * ibit23 * adv_mom_5 &
4579	+ 3.0_wp * ibit22 * adv_mom_3 &
4580	+ ibit21 * adv_mom_1 &
4581	) * &
4582	( v(k,j+1,i) - v(k,j,i) ) &
4583	- ( 5.0_wp * ibit23 * adv_mom_5 &
4584	+ ibit22 * adv_mom_3 &
4585	) * &
4586	( v(k,j+2,i) - v(k,j-1,i) ) &
4587	+ ( ibit23 * adv_mom_5 &
4588	) * &
4589	( v(k,j+3,i) - v(k,j-2,i) ) &
4590	)
4591	!
4592	!-- k index has to be modified near bottom and top, else array
4593	!-- subscripts will be exceeded.
4594	ibit26 = REAL( IBITS(advc_flags_1(k,j,i),26,1), KIND = wp )
4595	ibit25 = REAL( IBITS(advc_flags_1(k,j,i),25,1), KIND = wp )
4596	ibit24 = REAL( IBITS(advc_flags_1(k,j,i),24,1), KIND = wp )
4597
4598	k_ppp = k + 3 * ibit26
4599	k_pp = k + 2 * ( 1 - ibit24 )
4600	k_mm = k - 2 * ibit26
4601
4602	w_comp = w(k,j-1,i) + w(k,j,i)
4603	flux_t(k) = w_comp * rho_air_zw(k) * ( &
4604	( 37.0_wp * ibit26 * adv_mom_5 &
4605	+ 7.0_wp * ibit25 * adv_mom_3 &
4606	+ ibit24 * adv_mom_1 &
4607	) * &
4608	( v(k+1,j,i) + v(k,j,i) ) &
4609	- ( 8.0_wp * ibit26 * adv_mom_5 &
4610	+ ibit25 * adv_mom_3 &
4611	) * &
4612	( v(k_pp,j,i) + v(k-1,j,i) ) &
4613	+ ( ibit26 * adv_mom_5 &
4614	) * &
4615	( v(k_ppp,j,i) + v(k_mm,j,i) ) &
4616	)
4617
4618	diss_t(k) = - ABS( w_comp ) * rho_air_zw(k) * ( &
4619	( 10.0_wp * ibit26 * adv_mom_5 &
4620	+ 3.0_wp * ibit25 * adv_mom_3 &
4621	+ ibit24 * adv_mom_1 &
4622	) * &
4623	( v(k+1,j,i) - v(k,j,i) ) &
4624	- ( 5.0_wp * ibit26 * adv_mom_5 &
4625	+ ibit25 * adv_mom_3 &
4626	) * &
4627	( v(k_pp,j,i) - v(k-1,j,i) ) &
4628	+ ( ibit26 * adv_mom_5 &
4629	) * &
4630	( v(k_ppp,j,i) - v(k_mm,j,i) ) &
4631	)
4632	!
4633	!-- Calculate the divergence of the velocity field. A respective
4634	!-- correction is needed to overcome numerical instabilities caused
4635	!-- by a not sufficient reduction of divergences near topography.
4636	div = ( ( ( u_comp + gu ) &
4637	* ( ibit18 + ibit19 + ibit20 ) &
4638	- ( u(k,j-1,i) + u(k,j,i) ) &
4639	* ( &
4640	REAL( IBITS(advc_flags_1(k,j,i-1),18,1), KIND = wp ) &
4641	+ REAL( IBITS(advc_flags_1(k,j,i-1),19,1), KIND = wp ) &
4642	+ REAL( IBITS(advc_flags_1(k,j,i-1),20,1), KIND = wp ) &
4643	) &
4644	) * ddx &
4645	+ ( v_comp(k) &
4646	* ( ibit21 + ibit22 + ibit23 ) &
4647	- ( v(k,j,i) + v(k,j-1,i) ) &
4648	* ( &
4649	REAL( IBITS(advc_flags_1(k,j-1,i),21,1), KIND = wp ) &
4650	+ REAL( IBITS(advc_flags_1(k,j-1,i),22,1), KIND = wp ) &
4651	+ REAL( IBITS(advc_flags_1(k,j-1,i),23,1), KIND = wp ) &
4652	) &
4653	) * ddy &
4654	+ ( w_comp * rho_air_zw(k) &
4655	* ( ibit24 + ibit25 + ibit26 ) &
4656	- ( w(k-1,j-1,i) + w(k-1,j,i) ) * rho_air_zw(k-1) &
4657	* ( &
4658	REAL( IBITS(advc_flags_1(k-1,j,i),24,1), KIND = wp ) &
4659	+ REAL( IBITS(advc_flags_1(k-1,j,i),25,1), KIND = wp ) &
4660	+ REAL( IBITS(advc_flags_1(k-1,j,i),26,1), KIND = wp ) &
4661	) &
4662	) * drho_air(k) * ddzw(k) &
4663	) * 0.5_wp
4664
4665
4666	tend(k,j,i) = tend(k,j,i) - ( &
4667	( flux_r(k) + diss_r(k) &
4668	- swap_flux_x_local_v(k,j) - swap_diss_x_local_v(k,j) &
4669	) * ddx &
4670	+ ( flux_n(k) + diss_n(k) &
4671	- swap_flux_y_local_v(k) - swap_diss_y_local_v(k) &
4672	) * ddy &
4673	+ ( ( flux_t(k) + diss_t(k) ) &
4674	- ( flux_d + diss_d ) &
4675	) * drho_air(k) * ddzw(k) &
4676	) + v(k,j,i) * div
4677
4678	swap_flux_x_local_v(k,j) = flux_r(k)
4679	swap_diss_x_local_v(k,j) = diss_r(k)
4680	swap_flux_y_local_v(k) = flux_n(k)
4681	swap_diss_y_local_v(k) = diss_n(k)
4682	flux_d = flux_t(k)
4683	diss_d = diss_t(k)
4684
4685	!
4686	!-- Statistical Evaluation of v'v'. The factor has to be applied
4687	!-- for right evaluation when gallilei_trans = .T. .
4688	sums_vs2_ws_l(k,tn) = sums_vs2_ws_l(k,tn) &
4689	+ ( flux_n(k) &
4690	* ( v_comp(k) - 2.0_wp * hom(k,1,2,0) ) &
4691	/ ( v_comp(k) - gv + SIGN( 1.0E-20_wp, v_comp(k) - gv ) ) &
4692	+ diss_n(k) &
4693	* ABS( v_comp(k) - 2.0_wp * hom(k,1,2,0) ) &
4694	/ ( ABS( v_comp(k) - gv ) + 1.0E-20_wp ) &
4695	) * weight_substep(intermediate_timestep_count)
4696	!
4697	!-- Statistical Evaluation of w'u'.
4698	sums_wsvs_ws_l(k,tn) = sums_wsvs_ws_l(k,tn) &
4699	+ ( flux_t(k) &
4700	* ( w_comp - 2.0_wp * hom(k,1,3,0) ) &
4701	/ ( w_comp + SIGN( 1.0E-20_wp, w_comp ) ) &
4702	+ diss_t(k) &
4703	* ABS( w_comp - 2.0_wp * hom(k,1,3,0) ) &
4704	/ ( ABS( w_comp ) + 1.0E-20_wp ) &
4705	) * weight_substep(intermediate_timestep_count)
4706
4707	ENDDO
4708
4709	DO k = nzb_max+1, nzt
4710
4711	u_comp = u(k,j-1,i+1) + u(k,j,i+1) - gu
4712	flux_r(k) = u_comp * ( &
4713	37.0_wp * ( v(k,j,i+1) + v(k,j,i) ) &
4714	- 8.0_wp * ( v(k,j,i+2) + v(k,j,i-1) ) &
4715	+ ( v(k,j,i+3) + v(k,j,i-2) ) ) * adv_mom_5
4716
4717	diss_r(k) = - ABS( u_comp ) * ( &
4718	10.0_wp * ( v(k,j,i+1) - v(k,j,i) ) &
4719	- 5.0_wp * ( v(k,j,i+2) - v(k,j,i-1) ) &
4720	+ ( v(k,j,i+3) - v(k,j,i-2) ) ) * adv_mom_5
4721
4722
4723	v_comp(k) = v(k,j+1,i) + v(k,j,i)
4724	flux_n(k) = ( v_comp(k) - gv ) * ( &
4725	37.0_wp * ( v(k,j+1,i) + v(k,j,i) ) &
4726	- 8.0_wp * ( v(k,j+2,i) + v(k,j-1,i) ) &
4727	+ ( v(k,j+3,i) + v(k,j-2,i) ) ) * adv_mom_5
4728
4729	diss_n(k) = - ABS( v_comp(k) - gv ) * ( &
4730	10.0_wp * ( v(k,j+1,i) - v(k,j,i) ) &
4731	- 5.0_wp * ( v(k,j+2,i) - v(k,j-1,i) ) &
4732	+ ( v(k,j+3,i) - v(k,j-2,i) ) ) * adv_mom_5
4733	!
4734	!-- k index has to be modified near bottom and top, else array
4735	!-- subscripts will be exceeded.
4736	ibit26 = REAL( IBITS(advc_flags_1(k,j,i),26,1), KIND = wp )
4737	ibit25 = REAL( IBITS(advc_flags_1(k,j,i),25,1), KIND = wp )
4738	ibit24 = REAL( IBITS(advc_flags_1(k,j,i),24,1), KIND = wp )
4739
4740	k_ppp = k + 3 * ibit26
4741	k_pp = k + 2 * ( 1 - ibit24 )
4742	k_mm = k - 2 * ibit26
4743
4744	w_comp = w(k,j-1,i) + w(k,j,i)
4745	flux_t(k) = w_comp * rho_air_zw(k) * ( &
4746	( 37.0_wp * ibit26 * adv_mom_5 &
4747	+ 7.0_wp * ibit25 * adv_mom_3 &
4748	+ ibit24 * adv_mom_1 &
4749	) * &
4750	( v(k+1,j,i) + v(k,j,i) ) &
4751	- ( 8.0_wp * ibit26 * adv_mom_5 &
4752	+ ibit25 * adv_mom_3 &
4753	) * &
4754	( v(k_pp,j,i) + v(k-1,j,i) ) &
4755	+ ( ibit26 * adv_mom_5 &
4756	) * &
4757	( v(k_ppp,j,i) + v(k_mm,j,i) ) &
4758	)
4759
4760	diss_t(k) = - ABS( w_comp ) * rho_air_zw(k) * ( &
4761	( 10.0_wp * ibit26 * adv_mom_5 &
4762	+ 3.0_wp * ibit25 * adv_mom_3 &
4763	+ ibit24 * adv_mom_1 &
4764	) * &
4765	( v(k+1,j,i) - v(k,j,i) ) &
4766	- ( 5.0_wp * ibit26 * adv_mom_5 &
4767	+ ibit25 * adv_mom_3 &
4768	) * &
4769	( v(k_pp,j,i) - v(k-1,j,i) ) &
4770	+ ( ibit26 * adv_mom_5 &
4771	) * &
4772	( v(k_ppp,j,i) - v(k_mm,j,i) ) &
4773	)
4774	!
4775	!-- Calculate the divergence of the velocity field. A respective
4776	!-- correction is needed to overcome numerical instabilities caused
4777	!-- by a not sufficient reduction of divergences near topography.
4778	div = ( ( u_comp + gu - ( u(k,j-1,i) + u(k,j,i) ) ) * ddx &
4779	+ ( v_comp(k) - ( v(k,j,i) + v(k,j-1,i) ) ) * ddy &
4780	+ ( w_comp * rho_air_zw(k) - &
4781	( w(k-1,j-1,i) + w(k-1,j,i) ) * rho_air_zw(k-1) &
4782	) * drho_air(k) * ddzw(k) &
4783	) * 0.5_wp
4784
4785	tend(k,j,i) = tend(k,j,i) - ( &
4786	( flux_r(k) + diss_r(k) &
4787	- swap_flux_x_local_v(k,j) - swap_diss_x_local_v(k,j) &
4788	) * ddx &
4789	+ ( flux_n(k) + diss_n(k) &
4790	- swap_flux_y_local_v(k) - swap_diss_y_local_v(k) &
4791	) * ddy &
4792	+ ( ( flux_t(k) + diss_t(k) ) &
4793	- ( flux_d + diss_d ) &
4794	) * drho_air(k) * ddzw(k) &
4795	) + v(k,j,i) * div
4796
4797	swap_flux_x_local_v(k,j) = flux_r(k)
4798	swap_diss_x_local_v(k,j) = diss_r(k)
4799	swap_flux_y_local_v(k) = flux_n(k)
4800	swap_diss_y_local_v(k) = diss_n(k)
4801	flux_d = flux_t(k)
4802	diss_d = diss_t(k)
4803
4804	!
4805	!-- Statistical Evaluation of v'v'. The factor has to be applied
4806	!-- for right evaluation when gallilei_trans = .T. .
4807	sums_vs2_ws_l(k,tn) = sums_vs2_ws_l(k,tn) &
4808	+ ( flux_n(k) &
4809	* ( v_comp(k) - 2.0_wp * hom(k,1,2,0) ) &
4810	/ ( v_comp(k) - gv + SIGN( 1.0E-20_wp, v_comp(k) - gv ) ) &
4811	+ diss_n(k) &
4812	* ABS( v_comp(k) - 2.0_wp * hom(k,1,2,0) ) &
4813	/ ( ABS( v_comp(k) - gv ) + 1.0E-20_wp ) &
4814	) * weight_substep(intermediate_timestep_count)
4815	!
4816	!-- Statistical Evaluation of w'u'.
4817	sums_wsvs_ws_l(k,tn) = sums_wsvs_ws_l(k,tn) &
4818	+ ( flux_t(k) &
4819	* ( w_comp - 2.0_wp * hom(k,1,3,0) ) &
4820	/ ( w_comp + SIGN( 1.0E-20_wp, w_comp ) ) &
4821	+ diss_t(k) &
4822	* ABS( w_comp - 2.0_wp * hom(k,1,3,0) ) &
4823	/ ( ABS( w_comp ) + 1.0E-20_wp ) &
4824	) * weight_substep(intermediate_timestep_count)
4825
4826	ENDDO
4827	ENDDO
4828	ENDDO
4829	sums_vs2_ws_l(nzb,tn) = sums_vs2_ws_l(nzb+1,tn)
4830
4831
4832	END SUBROUTINE advec_v_ws
4833
4834
4835	!------------------------------------------------------------------------------!
4836	! Description:
4837	! ------------
4838	!> Advection of w - Call for all grid points
4839	!------------------------------------------------------------------------------!
4840	SUBROUTINE advec_w_ws
4841
4842	USE arrays_3d, &
4843	ONLY: ddzu, drho_air_zw, tend, u, v, w, rho_air
4844
4845	USE control_parameters, &
4846	ONLY: intermediate_timestep_count, u_gtrans, v_gtrans
4847
4848	USE grid_variables, &
4849	ONLY: ddx, ddy
4850
4851	USE indices, &
4852	ONLY: nxl, nxr, nyn, nys, nzb, nzb_max, nzt, advc_flags_1, &
4853	advc_flags_2
4854
4855	USE kinds
4856
4857	USE statistics, &
4858	ONLY: hom, sums_ws2_ws_l, weight_substep
4859
4860	IMPLICIT NONE
4861
4862	INTEGER(iwp) :: i !< grid index along x-direction
4863	INTEGER(iwp) :: j !< grid index along y-direction
4864	INTEGER(iwp) :: k !< grid index along z-direction
4865	INTEGER(iwp) :: k_mm !< k-2 index in disretization, can be modified to avoid segmentation faults
4866	INTEGER(iwp) :: k_pp !< k+2 index in disretization, can be modified to avoid segmentation faults
4867	INTEGER(iwp) :: k_ppp !< k+3 index in disretization, can be modified to avoid segmentation faults
4868	INTEGER(iwp) :: tn = 0 !< number of OpenMP thread
4869
4870	REAL(wp) :: ibit27 !< flag indicating 1st-order scheme along x-direction
4871	REAL(wp) :: ibit28 !< flag indicating 3rd-order scheme along x-direction
4872	REAL(wp) :: ibit29 !< flag indicating 5th-order scheme along x-direction
4873	REAL(wp) :: ibit30 !< flag indicating 1st-order scheme along y-direction
4874	REAL(wp) :: ibit31 !< flag indicating 3rd-order scheme along y-direction
4875	REAL(wp) :: ibit32 !< flag indicating 5th-order scheme along y-direction
4876	REAL(wp) :: ibit33 !< flag indicating 1st-order scheme along z-direction
4877	REAL(wp) :: ibit34 !< flag indicating 3rd-order scheme along z-direction
4878	REAL(wp) :: ibit35 !< flag indicating 5th-order scheme along z-direction
4879	REAL(wp) :: diss_d !< artificial dissipation term at grid box bottom
4880	REAL(wp) :: div !< divergence on w-grid
4881	REAL(wp) :: flux_d !< 6th-order flux at grid box bottom
4882	REAL(wp) :: gu !< Galilei-transformation velocity along x
4883	REAL(wp) :: gv !< Galilei-transformation velocity along y
4884	REAL(wp) :: u_comp !< advection velocity along x
4885	REAL(wp) :: v_comp !< advection velocity along y
4886	REAL(wp) :: w_comp !< advection velocity along z
4887
4888	REAL(wp), DIMENSION(nzb:nzt) :: diss_t !< discretized artificial dissipation at top of the grid box
4889	REAL(wp), DIMENSION(nzb:nzt) :: flux_t !< discretized 6th-order flux at top of the grid box
4890
4891	REAL(wp), DIMENSION(nzb+1:nzt) :: diss_n !< discretized artificial dissipation at northward-side of the grid box
4892	REAL(wp), DIMENSION(nzb+1:nzt) :: diss_r !< discretized artificial dissipation at rightward-side of the grid box
4893	REAL(wp), DIMENSION(nzb+1:nzt) :: flux_n !< discretized 6th-order flux at northward-side of the grid box
4894	REAL(wp), DIMENSION(nzb+1:nzt) :: flux_r !< discretized 6th-order flux at rightward-side of the grid box
4895	REAL(wp), DIMENSION(nzb+1:nzt) :: swap_diss_y_local_w !< discretized artificial dissipation at southward-side of the grid box
4896	REAL(wp), DIMENSION(nzb+1:nzt) :: swap_flux_y_local_w !< discretized 6th-order flux at southward-side of the grid box
4897
4898	REAL(wp), DIMENSION(nzb+1:nzt,nys:nyn) :: swap_diss_x_local_w !< discretized artificial dissipation at leftward-side of the grid box
4899	REAL(wp), DIMENSION(nzb+1:nzt,nys:nyn) :: swap_flux_x_local_w !< discretized 6th-order flux at leftward-side of the grid box
4900
4901	gu = 2.0_wp * u_gtrans
4902	gv = 2.0_wp * v_gtrans
4903	!
4904	!-- compute the whole left boundary of the processor domain
4905	i = nxl
4906	DO j = nys, nyn
4907	DO k = nzb+1, nzb_max
4908
4909	ibit29 = REAL( IBITS(advc_flags_1(k,j,i-1),29,1), KIND = wp )
4910	ibit28 = REAL( IBITS(advc_flags_1(k,j,i-1),28,1), KIND = wp )
4911	ibit27 = REAL( IBITS(advc_flags_1(k,j,i-1),27,1), KIND = wp )
4912
4913	u_comp = u(k+1,j,i) + u(k,j,i) - gu
4914	swap_flux_x_local_w(k,j) = u_comp * ( &
4915	( 37.0_wp * ibit29 * adv_mom_5 &
4916	+ 7.0_wp * ibit28 * adv_mom_3 &
4917	+ ibit27 * adv_mom_1 &
4918	) * &
4919	( w(k,j,i) + w(k,j,i-1) ) &
4920	- ( 8.0_wp * ibit29 * adv_mom_5 &
4921	+ ibit28 * adv_mom_3 &
4922	) * &
4923	( w(k,j,i+1) + w(k,j,i-2) ) &
4924	+ ( ibit29 * adv_mom_5 &
4925	) * &
4926	( w(k,j,i+2) + w(k,j,i-3) ) &
4927	)
4928
4929	swap_diss_x_local_w(k,j) = - ABS( u_comp ) * ( &
4930	( 10.0_wp * ibit29 * adv_mom_5 &
4931	+ 3.0_wp * ibit28 * adv_mom_3 &
4932	+ ibit27 * adv_mom_1 &
4933	) * &
4934	( w(k,j,i) - w(k,j,i-1) ) &
4935	- ( 5.0_wp * ibit29 * adv_mom_5 &
4936	+ ibit28 * adv_mom_3 &
4937	) * &
4938	( w(k,j,i+1) - w(k,j,i-2) ) &
4939	+ ( ibit29 * adv_mom_5 &
4940	) * &
4941	( w(k,j,i+2) - w(k,j,i-3) ) &
4942	)
4943
4944	ENDDO
4945
4946	DO k = nzb_max+1, nzt
4947
4948	u_comp = u(k+1,j,i) + u(k,j,i) - gu
4949	swap_flux_x_local_w(k,j) = u_comp * ( &
4950	37.0_wp * ( w(k,j,i) + w(k,j,i-1) ) &
4951	- 8.0_wp * ( w(k,j,i+1) + w(k,j,i-2) ) &
4952	+ ( w(k,j,i+2) + w(k,j,i-3) ) ) * adv_mom_5
4953	swap_diss_x_local_w(k,j) = - ABS( u_comp ) * ( &
4954	10.0_wp * ( w(k,j,i) - w(k,j,i-1) ) &
4955	- 5.0_wp * ( w(k,j,i+1) - w(k,j,i-2) ) &
4956	+ ( w(k,j,i+2) - w(k,j,i-3) ) ) * adv_mom_5
4957
4958	ENDDO
4959
4960	ENDDO
4961
4962	DO i = nxl, nxr
4963
4964	j = nys
4965	DO k = nzb+1, nzb_max
4966
4967	ibit32 = REAL( IBITS(advc_flags_2(k,j-1,i),0,1), KIND = wp )
4968	ibit31 = REAL( IBITS(advc_flags_1(k,j-1,i),31,1), KIND = wp )
4969	ibit30 = REAL( IBITS(advc_flags_1(k,j-1,i),30,1), KIND = wp )
4970
4971	v_comp = v(k+1,j,i) + v(k,j,i) - gv
4972	swap_flux_y_local_w(k) = v_comp * ( &
4973	( 37.0_wp * ibit32 * adv_mom_5 &
4974	+ 7.0_wp * ibit31 * adv_mom_3 &
4975	+ ibit30 * adv_mom_1 &
4976	) * &
4977	( w(k,j,i) + w(k,j-1,i) ) &
4978	- ( 8.0_wp * ibit32 * adv_mom_5 &
4979	+ ibit31 * adv_mom_3 &
4980	) * &
4981	( w(k,j+1,i) + w(k,j-2,i) ) &
4982	+ ( ibit32 * adv_mom_5 &
4983	) * &
4984	( w(k,j+2,i) + w(k,j-3,i) ) &
4985	)
4986
4987	swap_diss_y_local_w(k) = - ABS( v_comp ) * ( &
4988	( 10.0_wp * ibit32 * adv_mom_5 &
4989	+ 3.0_wp * ibit31 * adv_mom_3 &
4990	+ ibit30 * adv_mom_1 &
4991	) * &
4992	( w(k,j,i) - w(k,j-1,i) ) &
4993	- ( 5.0_wp * ibit32 * adv_mom_5 &
4994	+ ibit31 * adv_mom_3 &
4995	) * &
4996	( w(k,j+1,i) - w(k,j-2,i) ) &
4997	+ ( ibit32 * adv_mom_5 &
4998	) * &
4999	( w(k,j+2,i) - w(k,j-3,i) ) &
5000	)
5001
5002	ENDDO
5003
5004	DO k = nzb_max+1, nzt
5005
5006	v_comp = v(k+1,j,i) + v(k,j,i) - gv
5007	swap_flux_y_local_w(k) = v_comp * ( &
5008	37.0_wp * ( w(k,j,i) + w(k,j-1,i) ) &
5009	- 8.0_wp * ( w(k,j+1,i) +w(k,j-2,i) ) &
5010	+ ( w(k,j+2,i) + w(k,j-3,i) ) ) * adv_mom_5
5011	swap_diss_y_local_w(k) = - ABS( v_comp ) * ( &
5012	10.0_wp * ( w(k,j,i) - w(k,j-1,i) ) &
5013	- 5.0_wp * ( w(k,j+1,i) - w(k,j-2,i) ) &
5014	+ ( w(k,j+2,i) - w(k,j-3,i) ) ) * adv_mom_5
5015
5016	ENDDO
5017
5018	DO j = nys, nyn
5019
5020	!
5021	!-- The lower flux has to be calculated explicetely for the tendency
5022	!-- at the first w-level. For topography wall this is done implicitely
5023	!-- by advc_flags_1.
5024	k = nzb + 1
5025	w_comp = w(k,j,i) + w(k-1,j,i)
5026	flux_t(0) = w_comp * ( w(k,j,i) + w(k-1,j,i) ) * adv_mom_1
5027	diss_t(0) = -ABS(w_comp) * ( w(k,j,i) - w(k-1,j,i) ) * adv_mom_1
5028	flux_d = flux_t(0)
5029	diss_d = diss_t(0)
5030
5031	DO k = nzb+1, nzb_max
5032
5033	ibit29 = REAL( IBITS(advc_flags_1(k,j,i),29,1), KIND = wp )
5034	ibit28 = REAL( IBITS(advc_flags_1(k,j,i),28,1), KIND = wp )
5035	ibit27 = REAL( IBITS(advc_flags_1(k,j,i),27,1), KIND = wp )
5036
5037	u_comp = u(k+1,j,i+1) + u(k,j,i+1) - gu
5038	flux_r(k) = u_comp * ( &
5039	( 37.0_wp * ibit29 * adv_mom_5 &
5040	+ 7.0_wp * ibit28 * adv_mom_3 &
5041	+ ibit27 * adv_mom_1 &
5042	) * &
5043	( w(k,j,i+1) + w(k,j,i) ) &
5044	- ( 8.0_wp * ibit29 * adv_mom_5 &
5045	+ ibit28 * adv_mom_3 &
5046	) * &
5047	( w(k,j,i+2) + w(k,j,i-1) ) &
5048	+ ( ibit29 * adv_mom_5 &
5049	) * &
5050	( w(k,j,i+3) + w(k,j,i-2) ) &
5051	)
5052
5053	diss_r(k) = - ABS( u_comp ) * ( &
5054	( 10.0_wp * ibit29 * adv_mom_5 &
5055	+ 3.0_wp * ibit28 * adv_mom_3 &
5056	+ ibit27 * adv_mom_1 &
5057	) * &
5058	( w(k,j,i+1) - w(k,j,i) ) &
5059	- ( 5.0_wp * ibit29 * adv_mom_5 &
5060	+ ibit28 * adv_mom_3 &
5061	) * &
5062	( w(k,j,i+2) - w(k,j,i-1) ) &
5063	+ ( ibit29 * adv_mom_5 &
5064	) * &
5065	( w(k,j,i+3) - w(k,j,i-2) ) &
5066	)
5067
5068	ibit32 = REAL( IBITS(advc_flags_2(k,j,i),0,1), KIND = wp )
5069	ibit31 = REAL( IBITS(advc_flags_1(k,j,i),31,1), KIND = wp )
5070	ibit30 = REAL( IBITS(advc_flags_1(k,j,i),30,1), KIND = wp )
5071
5072	v_comp = v(k+1,j+1,i) + v(k,j+1,i) - gv
5073	flux_n(k) = v_comp * ( &
5074	( 37.0_wp * ibit32 * adv_mom_5 &
5075	+ 7.0_wp * ibit31 * adv_mom_3 &
5076	+ ibit30 * adv_mom_1 &
5077	) * &
5078	( w(k,j+1,i) + w(k,j,i) ) &
5079	- ( 8.0_wp * ibit32 * adv_mom_5 &
5080	+ ibit31 * adv_mom_3 &
5081	) * &
5082	( w(k,j+2,i) + w(k,j-1,i) ) &
5083	+ ( ibit32 * adv_mom_5 &
5084	) * &
5085	( w(k,j+3,i) + w(k,j-2,i) ) &
5086	)
5087
5088	diss_n(k) = - ABS( v_comp ) * ( &
5089	( 10.0_wp * ibit32 * adv_mom_5 &
5090	+ 3.0_wp * ibit31 * adv_mom_3 &
5091	+ ibit30 * adv_mom_1 &
5092	) * &
5093	( w(k,j+1,i) - w(k,j,i) ) &
5094	- ( 5.0_wp * ibit32 * adv_mom_5 &
5095	+ ibit31 * adv_mom_3 &
5096	) * &
5097	( w(k,j+2,i) - w(k,j-1,i) ) &
5098	+ ( ibit32 * adv_mom_5 &
5099	) * &
5100	( w(k,j+3,i) - w(k,j-2,i) ) &
5101	)
5102	!
5103	!-- k index has to be modified near bottom and top, else array
5104	!-- subscripts will be exceeded.
5105	ibit35 = REAL( IBITS(advc_flags_2(k,j,i),3,1), KIND = wp )
5106	ibit34 = REAL( IBITS(advc_flags_2(k,j,i),2,1), KIND = wp )
5107	ibit33 = REAL( IBITS(advc_flags_2(k,j,i),1,1), KIND = wp )
5108
5109	k_ppp = k + 3 * ibit35
5110	k_pp = k + 2 * ( 1 - ibit33 )
5111	k_mm = k - 2 * ibit35
5112
5113	w_comp = w(k+1,j,i) + w(k,j,i)
5114	flux_t(k) = w_comp * rho_air(k+1) * ( &
5115	( 37.0_wp * ibit35 * adv_mom_5 &
5116	+ 7.0_wp * ibit34 * adv_mom_3 &
5117	+ ibit33 * adv_mom_1 &
5118	) * &
5119	( w(k+1,j,i) + w(k,j,i) ) &
5120	- ( 8.0_wp * ibit35 * adv_mom_5 &
5121	+ ibit34 * adv_mom_3 &
5122	) * &
5123	( w(k_pp,j,i) + w(k-1,j,i) ) &
5124	+ ( ibit35 * adv_mom_5 &
5125	) * &
5126	( w(k_ppp,j,i) + w(k_mm,j,i) ) &
5127	)
5128
5129	diss_t(k) = - ABS( w_comp ) * rho_air(k+1) * ( &
5130	( 10.0_wp * ibit35 * adv_mom_5 &
5131	+ 3.0_wp * ibit34 * adv_mom_3 &
5132	+ ibit33 * adv_mom_1 &
5133	) * &
5134	( w(k+1,j,i) - w(k,j,i) ) &
5135	- ( 5.0_wp * ibit35 * adv_mom_5 &
5136	+ ibit34 * adv_mom_3 &
5137	) * &
5138	( w(k_pp,j,i) - w(k-1,j,i) ) &
5139	+ ( ibit35 * adv_mom_5 &
5140	) * &
5141	( w(k_ppp,j,i) - w(k_mm,j,i) ) &
5142	)
5143	!
5144	!-- Calculate the divergence of the velocity field. A respective
5145	!-- correction is needed to overcome numerical instabilities caused
5146	!-- by a not sufficient reduction of divergences near topography.
5147	div = ( ( ( u_comp + gu ) * ( ibit27 + ibit28 + ibit29 ) &
5148	- ( u(k+1,j,i) + u(k,j,i) ) &
5149	* ( &
5150	REAL( IBITS(advc_flags_1(k,j,i-1),27,1), KIND = wp ) &
5151	+ REAL( IBITS(advc_flags_1(k,j,i-1),28,1), KIND = wp ) &
5152	+ REAL( IBITS(advc_flags_1(k,j,i-1),29,1), KIND = wp ) &
5153	) &
5154	) * ddx &
5155	+ ( ( v_comp + gv ) * ( ibit30 + ibit31 + ibit32 ) &
5156	- ( v(k+1,j,i) + v(k,j,i) ) &
5157	* ( &
5158	REAL( IBITS(advc_flags_1(k,j-1,i),30,1), KIND = wp ) &
5159	+ REAL( IBITS(advc_flags_1(k,j-1,i),31,1), KIND = wp ) &
5160	+ REAL( IBITS(advc_flags_2(k,j-1,i),0,1), KIND = wp ) &
5161	) &
5162	) * ddy &
5163	+ ( w_comp * rho_air(k+1) * ( ibit33 + ibit34 + ibit35 ) &
5164	- ( w(k,j,i) + w(k-1,j,i) ) * rho_air(k) &
5165	* ( &
5166	REAL( IBITS(advc_flags_2(k-1,j,i),1,1), KIND = wp ) &
5167	+ REAL( IBITS(advc_flags_2(k-1,j,i),2,1), KIND = wp ) &
5168	+ REAL( IBITS(advc_flags_2(k-1,j,i),3,1), KIND = wp ) &
5169	) &
5170	) * drho_air_zw(k) * ddzu(k+1) &
5171	) * 0.5_wp
5172
5173
5174
5175	tend(k,j,i) = tend(k,j,i) - ( &
5176	( flux_r(k) + diss_r(k) &
5177	- swap_flux_x_local_w(k,j) - swap_diss_x_local_w(k,j) &
5178	) * ddx &
5179	+ ( flux_n(k) + diss_n(k) &
5180	- swap_flux_y_local_w(k) - swap_diss_y_local_w(k) &
5181	) * ddy &
5182	+ ( ( flux_t(k) + diss_t(k) ) &
5183	- ( flux_d + diss_d ) &
5184	) * drho_air_zw(k) * ddzu(k+1) &
5185	) + div * w(k,j,i)
5186
5187	swap_flux_x_local_w(k,j) = flux_r(k)
5188	swap_diss_x_local_w(k,j) = diss_r(k)
5189	swap_flux_y_local_w(k) = flux_n(k)
5190	swap_diss_y_local_w(k) = diss_n(k)
5191	flux_d = flux_t(k)
5192	diss_d = diss_t(k)
5193
5194	sums_ws2_ws_l(k,tn) = sums_ws2_ws_l(k,tn) &
5195	+ ( flux_t(k) &
5196	* ( w_comp - 2.0_wp * hom(k,1,3,0) ) &
5197	/ ( w_comp + SIGN( 1.0E-20_wp, w_comp ) ) &
5198	+ diss_t(k) &
5199	* ABS( w_comp - 2.0_wp * hom(k,1,3,0) ) &
5200	/ ( ABS( w_comp ) + 1.0E-20_wp ) &
5201	) * weight_substep(intermediate_timestep_count)
5202
5203	ENDDO
5204
5205	DO k = nzb_max+1, nzt
5206
5207	u_comp = u(k+1,j,i+1) + u(k,j,i+1) - gu
5208	flux_r(k) = u_comp * ( &
5209	37.0_wp * ( w(k,j,i+1) + w(k,j,i) ) &
5210	- 8.0_wp * ( w(k,j,i+2) + w(k,j,i-1) ) &
5211	+ ( w(k,j,i+3) + w(k,j,i-2) ) ) * adv_mom_5
5212
5213	diss_r(k) = - ABS( u_comp ) * ( &
5214	10.0_wp * ( w(k,j,i+1) - w(k,j,i) ) &
5215	- 5.0_wp * ( w(k,j,i+2) - w(k,j,i-1) ) &
5216	+ ( w(k,j,i+3) - w(k,j,i-2) ) ) * adv_mom_5
5217
5218	v_comp = v(k+1,j+1,i) + v(k,j+1,i) - gv
5219	flux_n(k) = v_comp * ( &
5220	37.0_wp * ( w(k,j+1,i) + w(k,j,i) ) &
5221	- 8.0_wp * ( w(k,j+2,i) + w(k,j-1,i) ) &
5222	+ ( w(k,j+3,i) + w(k,j-2,i) ) ) * adv_mom_5
5223
5224	diss_n(k) = - ABS( v_comp ) * ( &
5225	10.0_wp * ( w(k,j+1,i) - w(k,j,i) ) &
5226	- 5.0_wp * ( w(k,j+2,i) - w(k,j-1,i) ) &
5227	+ ( w(k,j+3,i) - w(k,j-2,i) ) ) * adv_mom_5
5228	!
5229	!-- k index has to be modified near bottom and top, else array
5230	!-- subscripts will be exceeded.
5231	ibit35 = REAL( IBITS(advc_flags_2(k,j,i),3,1), KIND = wp )
5232	ibit34 = REAL( IBITS(advc_flags_2(k,j,i),2,1), KIND = wp )
5233	ibit33 = REAL( IBITS(advc_flags_2(k,j,i),1,1), KIND = wp )
5234
5235	k_ppp = k + 3 * ibit35
5236	k_pp = k + 2 * ( 1 - ibit33 )
5237	k_mm = k - 2 * ibit35
5238
5239	w_comp = w(k+1,j,i) + w(k,j,i)
5240	flux_t(k) = w_comp * rho_air(k+1) * ( &
5241	( 37.0_wp * ibit35 * adv_mom_5 &
5242	+ 7.0_wp * ibit34 * adv_mom_3 &
5243	+ ibit33 * adv_mom_1 &
5244	) * &
5245	( w(k+1,j,i) + w(k,j,i) ) &
5246	- ( 8.0_wp * ibit35 * adv_mom_5 &
5247	+ ibit34 * adv_mom_3 &
5248	) * &
5249	( w(k_pp,j,i) + w(k-1,j,i) ) &
5250	+ ( ibit35 * adv_mom_5 &
5251	) * &
5252	( w(k_ppp,j,i) + w(k_mm,j,i) ) &
5253	)
5254
5255	diss_t(k) = - ABS( w_comp ) * rho_air(k+1) * ( &
5256	( 10.0_wp * ibit35 * adv_mom_5 &
5257	+ 3.0_wp * ibit34 * adv_mom_3 &
5258	+ ibit33 * adv_mom_1 &
5259	) * &
5260	( w(k+1,j,i) - w(k,j,i) ) &
5261	- ( 5.0_wp * ibit35 * adv_mom_5 &
5262	+ ibit34 * adv_mom_3 &
5263	) * &
5264	( w(k_pp,j,i) - w(k-1,j,i) ) &
5265	+ ( ibit35 * adv_mom_5 &
5266	) * &
5267	( w(k_ppp,j,i) - w(k_mm,j,i) ) &
5268	)
5269	!
5270	!-- Calculate the divergence of the velocity field. A respective
5271	!-- correction is needed to overcome numerical instabilities caused
5272	!-- by a not sufficient reduction of divergences near topography.
5273	div = ( ( u_comp + gu - ( u(k+1,j,i) + u(k,j,i) ) ) * ddx &
5274	+ ( v_comp + gv - ( v(k+1,j,i) + v(k,j,i) ) ) * ddy &
5275	+ ( w_comp * rho_air(k+1) - &
5276	( w(k,j,i) + w(k-1,j,i) ) * rho_air(k) &
5277	) * drho_air_zw(k) * ddzu(k+1) &
5278	) * 0.5_wp
5279
5280	tend(k,j,i) = tend(k,j,i) - ( &
5281	( flux_r(k) + diss_r(k) &
5282	- swap_flux_x_local_w(k,j) - swap_diss_x_local_w(k,j) &
5283	) * ddx &
5284	+ ( flux_n(k) + diss_n(k) &
5285	- swap_flux_y_local_w(k) - swap_diss_y_local_w(k) &
5286	) * ddy &
5287	+ ( ( flux_t(k) + diss_t(k) ) &
5288	- ( flux_d + diss_d ) &
5289	) * drho_air_zw(k) * ddzu(k+1) &
5290	) + div * w(k,j,i)
5291
5292	swap_flux_x_local_w(k,j) = flux_r(k)
5293	swap_diss_x_local_w(k,j) = diss_r(k)
5294	swap_flux_y_local_w(k) = flux_n(k)
5295	swap_diss_y_local_w(k) = diss_n(k)
5296	flux_d = flux_t(k)
5297	diss_d = diss_t(k)
5298
5299	sums_ws2_ws_l(k,tn) = sums_ws2_ws_l(k,tn) &
5300	+ ( flux_t(k) &
5301	* ( w_comp - 2.0_wp * hom(k,1,3,0) ) &
5302	/ ( w_comp + SIGN( 1.0E-20_wp, w_comp ) ) &
5303	+ diss_t(k) &
5304	* ABS( w_comp - 2.0_wp * hom(k,1,3,0) ) &
5305	/ ( ABS( w_comp ) + 1.0E-20_wp ) &
5306	) * weight_substep(intermediate_timestep_count)
5307
5308	ENDDO
5309	ENDDO
5310	ENDDO
5311
5312	END SUBROUTINE advec_w_ws
5313
5314	END MODULE advec_ws

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

Download in other formats:

| Impressum | ©Leibniz Universität Hannover |