New URL for NEMO forge! http://forge.nemo-ocean.eu

Since March 2022 along with NEMO 4.2 release, the code development moved to a self-hosted GitLab.
This present forge is now archived and remained online for history.

nemogcm.F90 in branches/2011/DEV_r2739_STFC_dCSE/NEMOGCM/NEMO/OPA_SRC – NEMO

Context Navigation

source: branches/2011/DEV_r2739_STFC_dCSE/NEMOGCM/NEMO/OPA_SRC/nemogcm.F90 @ 3187

Last change on this file since 3187 was 3187, checked in by spickles2, 12 years ago
Stephen Pickles, 28 Nov 2011. First commit of dCSE NEMO project work, part 1 - index re-ordering, OPA_SRC top level only. Includes fix for sub-optimal auto-partitioning in nemogcm.F90.
Property svn:keywords set to `Id`
File size: 27.7 KB

Line
1	MODULE nemogcm
2	!!======================================================================
3	!! * MODULE nemogcm *
4	!! Ocean system : NEMO GCM (ocean dynamics, on-line tracers, biochemistry and sea-ice)
5	!!======================================================================
6	!! History : OPA ! 1990-10 (C. Levy, G. Madec) Original code
7	!! 7.0 ! 1991-11 (M. Imbard, C. Levy, G. Madec)
8	!! 7.1 ! 1993-03 (M. Imbard, C. Levy, G. Madec, O. Marti, M. Guyon, A. Lazar,
9	!! P. Delecluse, C. Perigaud, G. Caniaux, B. Colot, C. Maes) release 7.1
10	!! - ! 1992-06 (L.Terray) coupling implementation
11	!! - ! 1993-11 (M.A. Filiberti) IGLOO sea-ice
12	!! 8.0 ! 1996-03 (M. Imbard, C. Levy, G. Madec, O. Marti, M. Guyon, A. Lazar,
13	!! P. Delecluse, L.Terray, M.A. Filiberti, J. Vialar, A.M. Treguier, M. Levy) release 8.0
14	!! 8.1 ! 1997-06 (M. Imbard, G. Madec)
15	!! 8.2 ! 1999-11 (M. Imbard, H. Goosse) LIM sea-ice model
16	!! ! 1999-12 (V. Thierry, A-M. Treguier, M. Imbard, M-A. Foujols) OPEN-MP
17	!! ! 2000-07 (J-M Molines, M. Imbard) Open Boundary Conditions (CLIPPER)
18	!! NEMO 1.0 ! 2002-08 (G. Madec) F90: Free form and modules
19	!! - ! 2004-06 (R. Redler, NEC CCRLE, Germany) add OASIS[3/4] coupled interfaces
20	!! - ! 2004-08 (C. Talandier) New trends organization
21	!! - ! 2005-06 (C. Ethe) Add the 1D configuration possibility
22	!! - ! 2005-11 (V. Garnier) Surface pressure gradient organization
23	!! - ! 2006-03 (L. Debreu, C. Mazauric) Agrif implementation
24	!! - ! 2006-04 (G. Madec, R. Benshila) Step reorganization
25	!! - ! 2007-07 (J. Chanut, A. Sellar) Unstructured open boundaries (BDY)
26	!! 3.2 ! 2009-08 (S. Masson) open/write in the listing file in mpp
27	!! 3.3 ! 2010-05 (K. Mogensen, A. Weaver, M. Martin, D. Lea) Assimilation interface
28	!! - ! 2010-10 (C. Ethe, G. Madec) reorganisation of initialisation phase
29	!! 4.0 ! 2011-01 (A. R. Porter, STFC Daresbury) dynamical allocation
30	!!----------------------------------------------------------------------
31
32	!!----------------------------------------------------------------------
33	!! nemo_gcm : solve ocean dynamics, tracer, biogeochemistry and/or sea-ice
34	!! nemo_init : initialization of the NEMO system
35	!! nemo_ctl : initialisation of the contol print
36	!! nemo_closefile : close remaining open files
37	!! nemo_alloc : dynamical allocation
38	!! nemo_partition : calculate MPP domain decomposition
39	!! sqfact : calculate factors of the no. of MPI processes
40	!!----------------------------------------------------------------------
41	USE step_oce ! module used in the ocean time stepping module
42	USE sbc_oce ! surface boundary condition: ocean
43	USE cla ! cross land advection (tra_cla routine)
44	USE domcfg ! domain configuration (dom_cfg routine)
45	USE mppini ! shared/distributed memory setting (mpp_init routine)
46	USE domain ! domain initialization (dom_init routine)
47	USE obcini ! open boundary cond. initialization (obc_ini routine)
48	USE bdyini ! unstructured open boundary cond. initialization (bdy_init routine)
49	USE istate ! initial state setting (istate_init routine)
50	USE ldfdyn ! lateral viscosity setting (ldfdyn_init routine)
51	USE ldftra ! lateral diffusivity setting (ldftra_init routine)
52	USE zdfini ! vertical physics setting (zdf_init routine)
53	USE phycst ! physical constant (par_cst routine)
54	USE trdmod ! momentum/tracers trends (trd_mod_init routine)
55	USE asminc ! assimilation increments (asm_inc_init routine)
56	USE asmtrj ! writing out state trajectory
57	USE sshwzv ! vertical velocity used in asm
58	USE diaptr ! poleward transports (dia_ptr_init routine)
59	USE diaobs ! Observation diagnostics (dia_obs_init routine)
60	USE step ! NEMO time-stepping (stp routine)
61	#if defined key_oasis3
62	USE cpl_oasis3 ! OASIS3 coupling
63	#elif defined key_oasis4
64	USE cpl_oasis4 ! OASIS4 coupling (not working)
65	#endif
66	USE c1d ! 1D configuration
67	USE step_c1d ! Time stepping loop for the 1D configuration
68	#if defined key_top
69	USE trcini ! passive tracer initialisation
70	#endif
71	USE lib_mpp ! distributed memory computing
72	#if defined key_iomput
73	USE mod_ioclient
74	#endif
75
76	IMPLICIT NONE
77	PRIVATE
78
79	PUBLIC nemo_gcm ! called by model.F90
80	PUBLIC nemo_init ! needed by AGRIF
81
82	CHARACTER(lc) :: cform_aaa="( /, 'AAAAAAAA', / ) " ! flag for output listing
83
84	!!----------------------------------------------------------------------
85	!! NEMO/OPA 4.0 , NEMO Consortium (2011)
86	!! $Id$
87	!! Software governed by the CeCILL licence (NEMOGCM/NEMO_CeCILL.txt)
88	!!----------------------------------------------------------------------
89	CONTAINS
90
91	SUBROUTINE nemo_gcm
92	!!----------------------------------------------------------------------
93	!! * ROUTINE nemo_gcm *
94	!!
95	!! ** Purpose : NEMO solves the primitive equations on an orthogonal
96	!! curvilinear mesh on the sphere.
97	!!
98	!! ** Method : - model general initialization
99	!! - launch the time-stepping (stp routine)
100	!! - finalize the run by closing files and communications
101	!!
102	!! References : Madec, Delecluse, Imbard, and Levy, 1997: internal report, IPSL.
103	!! Madec, 2008, internal report, IPSL.
104	!!----------------------------------------------------------------------
105	INTEGER :: istp ! time step index
106	!!----------------------------------------------------------------------
107	!
108	#if defined key_agrif
109	CALL Agrif_Init_Grids() ! AGRIF: set the meshes
110	#endif
111
112	! !-----------------------!
113	CALL nemo_init !== Initialisations ==!
114	! !-----------------------!
115	#if defined key_agrif
116	CALL Agrif_Declare_Var ! AGRIF: set the meshes
117	# if defined key_top
118	CALL Agrif_Declare_Var_Top ! AGRIF: set the meshes
119	# endif
120	#endif
121	! check that all process are still there... If some process have an error,
122	! they will never enter in step and other processes will wait until the end of the cpu time!
123	IF( lk_mpp ) CALL mpp_max( nstop )
124
125	IF(lwp) WRITE(numout,cform_aaa) ! Flag AAAAAAA
126
127	! !-----------------------!
128	! !== time stepping ==!
129	! !-----------------------!
130	istp = nit000
131	#if defined key_c1d
132	DO WHILE ( istp <= nitend .AND. nstop == 0 )
133	CALL stp_c1d( istp )
134	istp = istp + 1
135	END DO
136	#else
137	IF( lk_asminc ) THEN
138	IF( ln_bkgwri ) CALL asm_bkg_wri( nit000 - 1 ) ! Output background fields
139	IF( ln_trjwri ) CALL asm_trj_wri( nit000 - 1 ) ! Output trajectory fields
140	IF( ln_asmdin ) THEN ! Direct initialization
141	IF( ln_trainc ) CALL tra_asm_inc( nit000 - 1 ) ! Tracers
142	IF( ln_dyninc ) THEN
143	CALL dyn_asm_inc( nit000 - 1 ) ! Dynamics
144	IF ( ln_asmdin ) CALL ssh_wzv ( nit000 - 1 ) ! update vertical velocity
145	ENDIF
146	IF( ln_sshinc ) CALL ssh_asm_inc( nit000 - 1 ) ! SSH
147	ENDIF
148	ENDIF
149
150	DO WHILE ( istp <= nitend .AND. nstop == 0 )
151	#if defined key_agrif
152	CALL Agrif_Step( stp ) ! AGRIF: time stepping
153	#else
154	CALL stp( istp ) ! standard time stepping
155	#endif
156	istp = istp + 1
157	IF( lk_mpp ) CALL mpp_max( nstop )
158	END DO
159	#endif
160
161	IF( lk_diaobs ) CALL dia_obs_wri
162
163	! !------------------------!
164	! !== finalize the run ==!
165	! !------------------------!
166	IF(lwp) WRITE(numout,cform_aaa) ! Flag AAAAAAA
167	!
168	IF( nstop /= 0 .AND. lwp ) THEN ! error print
169	WRITE(numout,cform_err)
170	WRITE(numout,*) nstop, ' error have been found'
171	ENDIF
172	!
173	CALL nemo_closefile
174	#if defined key_oasis3 \|\| defined key_oasis4
175	CALL cpl_prism_finalize ! end coupling and mpp communications with OASIS
176	#else
177	IF( lk_mpp ) CALL mppstop ! end mpp communications
178	#endif
179	!
180	END SUBROUTINE nemo_gcm
181
182
183	SUBROUTINE nemo_init
184	!!----------------------------------------------------------------------
185	!! * ROUTINE nemo_init *
186	!!
187	!! ** Purpose : initialization of the NEMO GCM
188	!!----------------------------------------------------------------------
189	INTEGER :: ji ! dummy loop indices
190	INTEGER :: ilocal_comm ! local integer
191	CHARACTER(len=80), DIMENSION(16) :: cltxt
192	!!
193	NAMELIST/namctl/ ln_ctl , nn_print, nn_ictls, nn_ictle, &
194	& nn_isplt, nn_jsplt, nn_jctls, nn_jctle, nn_bench
195	!!----------------------------------------------------------------------
196	!
197	cltxt = ''
198	!
199	! ! open Namelist file
200	CALL ctl_opn( numnam, 'namelist', 'OLD', 'FORMATTED', 'SEQUENTIAL', -1, 6, .FALSE. )
201	!
202	READ( numnam, namctl ) ! Namelist namctl : Control prints & Benchmark
203	!
204	! !--------------------------------------------!
205	! ! set communicator & select the local node !
206	! !--------------------------------------------!
207	#if defined key_iomput
208	IF( Agrif_Root() ) THEN
209	# if defined key_oasis3 \|\| defined key_oasis4
210	CALL cpl_prism_init( ilocal_comm ) ! nemo local communicator given by oasis
211	# endif
212	CALL init_ioclient( ilocal_comm ) ! exchange io_server nemo local communicator with the io_server
213	ENDIF
214	narea = mynode( cltxt, numnam, nstop, ilocal_comm ) ! Nodes selection
215	#else
216	# if defined key_oasis3 \|\| defined key_oasis4
217	IF( Agrif_Root() ) THEN
218	CALL cpl_prism_init( ilocal_comm ) ! nemo local communicator given by oasis
219	ENDIF
220	narea = mynode( cltxt, numnam, nstop, ilocal_comm ) ! Nodes selection (control print return in cltxt)
221	# else
222	ilocal_comm = 0
223	narea = mynode( cltxt, numnam, nstop ) ! Nodes selection (control print return in cltxt)
224	# endif
225	#endif
226	narea = narea + 1 ! mynode return the rank of proc (0 --> jpnij -1 )
227
228	lwp = (narea == 1) .OR. ln_ctl ! control of all listing output print
229
230	! If dimensions of processor grid weren't specified in the namelist file
231	! then we calculate them here now that we have our communicator size
232	IF( (jpni < 1) .OR. (jpnj < 1) )THEN
233	#if defined key_mpp_mpi
234	IF( Agrif_Root() ) CALL nemo_partition(mppsize)
235	#else
236	jpni = 1
237	jpnj = 1
238	jpnij = jpni*jpnj
239	#endif
240	END IF
241
242	! Calculate domain dimensions given calculated jpni and jpnj
243	! This used to be done in par_oce.F90 when they were parameters rather
244	! than variables
245	IF( Agrif_Root() ) THEN
246	jpi = ( jpiglo-2jpreci + (jpni-1) ) / jpni + 2jpreci ! first dim.
247	jpj = ( jpjglo-2jprecj + (jpnj-1) ) / jpnj + 2jprecj ! second dim.
248	jpk = jpkdta ! third dim
249	jpim1 = jpi-1 ! inner domain indices
250	jpjm1 = jpj-1 ! " "
251	jpkm1 = jpk-1 ! " "
252	jpij = jpi*jpj ! jpi x j
253	ENDIF
254
255	IF(lwp) THEN ! open listing units
256	!
257	CALL ctl_opn( numout, 'ocean.output', 'REPLACE', 'FORMATTED', 'SEQUENTIAL', -1, 6, .FALSE., narea )
258	!
259	WRITE(numout,*)
260	WRITE(numout,*) ' CNRS - NERC - Met OFFICE - MERCATOR-ocean'
261	WRITE(numout,*) ' NEMO team'
262	WRITE(numout,*) ' Ocean General Circulation Model'
263	WRITE(numout,*) ' version 3.3 (2010) '
264	WRITE(numout,*)
265	WRITE(numout,*)
266	DO ji = 1, SIZE(cltxt)
267	IF( TRIM(cltxt(ji)) /= '' ) WRITE(numout,*) cltxt(ji) ! control print of mynode
268	END DO
269	WRITE(numout,cform_aaa) ! Flag AAAAAAA
270	!
271	ENDIF
272
273	! Now we know the dimensions of the grid and numout has been set we can
274	! allocate arrays
275	CALL nemo_alloc()
276
277	! !-------------------------------!
278	! ! NEMO general initialization !
279	! !-------------------------------!
280
281	CALL nemo_ctl ! Control prints & Benchmark
282
283	! ! Domain decomposition
284	IF( jpni*jpnj == jpnij ) THEN ; CALL mpp_init ! standard cutting out
285	ELSE ; CALL mpp_init2 ! eliminate land processors
286	ENDIF
287	!
288	! ! General initialization
289	CALL phy_cst ! Physical constants
290	CALL eos_init ! Equation of state
291	CALL dom_cfg ! Domain configuration
292	CALL dom_init ! Domain
293
294	IF( ln_ctl ) CALL prt_ctl_init ! Print control
295
296	IF( lk_obc ) CALL obc_init ! Open boundaries
297	IF( lk_bdy ) CALL bdy_init ! Unstructured open boundaries
298
299	CALL istate_init ! ocean initial state (Dynamics and tracers)
300
301	! ! Ocean physics
302	CALL sbc_init ! Forcings : surface module
303	! ! Vertical physics
304	CALL zdf_init ! namelist read
305	CALL zdf_bfr_init ! bottom friction
306	IF( lk_zdfric ) CALL zdf_ric_init ! Richardson number dependent Kz
307	IF( lk_zdftke ) CALL zdf_tke_init ! TKE closure scheme
308	IF( lk_zdfgls ) CALL zdf_gls_init ! GLS closure scheme
309	IF( lk_zdfkpp ) CALL zdf_kpp_init ! KPP closure scheme
310	IF( lk_zdftmx ) CALL zdf_tmx_init ! tidal vertical mixing
311	IF( lk_zdfddm .AND. .NOT. lk_zdfkpp ) &
312	& CALL zdf_ddm_init ! double diffusive mixing
313	! ! Lateral physics
314	CALL ldf_tra_init ! Lateral ocean tracer physics
315	CALL ldf_dyn_init ! Lateral ocean momentum physics
316	IF( lk_ldfslp ) CALL ldf_slp_init ! slope of lateral mixing
317
318	! ! Active tracers
319	CALL tra_qsr_init ! penetrative solar radiation qsr
320	CALL tra_bbc_init ! bottom heat flux
321	IF( lk_trabbl ) CALL tra_bbl_init ! advective (and/or diffusive) bottom boundary layer scheme
322	IF( lk_tradmp ) CALL tra_dmp_init ! internal damping trends
323	CALL tra_adv_init ! horizontal & vertical advection
324	CALL tra_ldf_init ! lateral mixing
325	CALL tra_zdf_init ! vertical mixing and after tracer fields
326
327	! ! Dynamics
328	CALL dyn_adv_init ! advection (vector or flux form)
329	CALL dyn_vor_init ! vorticity term including Coriolis
330	CALL dyn_ldf_init ! lateral mixing
331	CALL dyn_hpg_init ! horizontal gradient of Hydrostatic pressure
332	CALL dyn_zdf_init ! vertical diffusion
333	CALL dyn_spg_init ! surface pressure gradient
334
335	! ! Misc. options
336	IF( nn_cla == 1 ) CALL cla_init ! Cross Land Advection
337
338	#if defined key_top
339	! ! Passive tracers
340	CALL trc_init
341	#endif
342	! ! Diagnostics
343	CALL iom_init ! iom_put initialization
344	IF( lk_floats ) CALL flo_init ! drifting Floats
345	IF( lk_diaar5 ) CALL dia_ar5_init ! ar5 diag
346	CALL dia_ptr_init ! Poleward TRansports initialization
347	CALL dia_hsb_init ! heat content, salt content and volume budgets
348	CALL trd_mod_init ! Mixed-layer/Vorticity/Integral constraints trends
349	IF( lk_diaobs ) THEN ! Observation & model comparison
350	CALL dia_obs_init ! Initialize observational data
351	CALL dia_obs( nit000 - 1 ) ! Observation operator for restart
352	ENDIF
353	! ! Assimilation increments
354	IF( lk_asminc ) CALL asm_inc_init ! Initialize assimilation increments
355	IF(lwp) WRITE(numout,*) 'Euler time step switch is ', neuler
356	!
357	END SUBROUTINE nemo_init
358
359
360	SUBROUTINE nemo_ctl
361	!!----------------------------------------------------------------------
362	!! * ROUTINE nemo_ctl *
363	!!
364	!! ** Purpose : control print setting
365	!!
366	!! ** Method : - print namctl information and check some consistencies
367	!!----------------------------------------------------------------------
368	!
369	IF(lwp) THEN ! control print
370	WRITE(numout,*)
371	WRITE(numout,*) 'nemo_ctl: Control prints & Benchmark'
372	WRITE(numout,*) '~~~~~~~ '
373	WRITE(numout,*) ' Namelist namctl'
374	WRITE(numout,*) ' run control (for debugging) ln_ctl = ', ln_ctl
375	WRITE(numout,*) ' level of print nn_print = ', nn_print
376	WRITE(numout,*) ' Start i indice for SUM control nn_ictls = ', nn_ictls
377	WRITE(numout,*) ' End i indice for SUM control nn_ictle = ', nn_ictle
378	WRITE(numout,*) ' Start j indice for SUM control nn_jctls = ', nn_jctls
379	WRITE(numout,*) ' End j indice for SUM control nn_jctle = ', nn_jctle
380	WRITE(numout,*) ' number of proc. following i nn_isplt = ', nn_isplt
381	WRITE(numout,*) ' number of proc. following j nn_jsplt = ', nn_jsplt
382	WRITE(numout,*) ' benchmark parameter (0/1) nn_bench = ', nn_bench
383	ENDIF
384	!
385	nprint = nn_print ! convert DOCTOR namelist names into OLD names
386	nictls = nn_ictls
387	nictle = nn_ictle
388	njctls = nn_jctls
389	njctle = nn_jctle
390	isplt = nn_isplt
391	jsplt = nn_jsplt
392	nbench = nn_bench
393	! ! Parameter control
394	!
395	IF( ln_ctl ) THEN ! sub-domain area indices for the control prints
396	IF( lk_mpp ) THEN
397	isplt = jpni ; jsplt = jpnj ; ijsplt = jpni*jpnj ! the domain is forced to the real split domain
398	ELSE
399	IF( isplt == 1 .AND. jsplt == 1 ) THEN
400	CALL ctl_warn( ' - isplt & jsplt are equal to 1', &
401	& ' - the print control will be done over the whole domain' )
402	ENDIF
403	ijsplt = isplt * jsplt ! total number of processors ijsplt
404	ENDIF
405	IF(lwp) WRITE(numout,*)' - The total number of processors over which the'
406	IF(lwp) WRITE(numout,*)' print control will be done is ijsplt : ', ijsplt
407	!
408	! ! indices used for the SUM control
409	IF( nictls+nictle+njctls+njctle == 0 ) THEN ! print control done over the default area
410	lsp_area = .FALSE.
411	ELSE ! print control done over a specific area
412	lsp_area = .TRUE.
413	IF( nictls < 1 .OR. nictls > jpiglo ) THEN
414	CALL ctl_warn( ' - nictls must be 1<=nictls>=jpiglo, it is forced to 1' )
415	nictls = 1
416	ENDIF
417	IF( nictle < 1 .OR. nictle > jpiglo ) THEN
418	CALL ctl_warn( ' - nictle must be 1<=nictle>=jpiglo, it is forced to jpiglo' )
419	nictle = jpiglo
420	ENDIF
421	IF( njctls < 1 .OR. njctls > jpjglo ) THEN
422	CALL ctl_warn( ' - njctls must be 1<=njctls>=jpjglo, it is forced to 1' )
423	njctls = 1
424	ENDIF
425	IF( njctle < 1 .OR. njctle > jpjglo ) THEN
426	CALL ctl_warn( ' - njctle must be 1<=njctle>=jpjglo, it is forced to jpjglo' )
427	njctle = jpjglo
428	ENDIF
429	ENDIF
430	ENDIF
431	!
432	IF( nbench == 1 ) THEN ! Benchmark
433	SELECT CASE ( cp_cfg )
434	CASE ( 'gyre' ) ; CALL ctl_warn( ' The Benchmark is activated ' )
435	CASE DEFAULT ; CALL ctl_stop( ' The Benchmark is based on the GYRE configuration:', &
436	& ' key_gyre must be used or set nbench = 0' )
437	END SELECT
438	ENDIF
439	!
440	IF( lk_c1d .AND. .NOT.lk_iomput ) CALL ctl_stop( 'nemo_ctl: The 1D configuration must be used ', &
441	& 'with the IOM Input/Output manager. ' , &
442	& 'Compile with key_iomput enabled' )
443	!
444	END SUBROUTINE nemo_ctl
445
446
447	SUBROUTINE nemo_closefile
448	!!----------------------------------------------------------------------
449	!! * ROUTINE nemo_closefile *
450	!!
451	!! ** Purpose : Close the files
452	!!----------------------------------------------------------------------
453	!
454	IF( lk_mpp ) CALL mppsync
455	!
456	CALL iom_close ! close all input/output files managed by iom_*
457	!
458	IF( numstp /= -1 ) CLOSE( numstp ) ! time-step file
459	IF( numsol /= -1 ) CLOSE( numsol ) ! solver file
460	IF( numnam /= -1 ) CLOSE( numnam ) ! oce namelist
461	IF( numnam_ice /= -1 ) CLOSE( numnam_ice ) ! ice namelist
462	IF( numevo_ice /= -1 ) CLOSE( numevo_ice ) ! ice variables (temp. evolution)
463	IF( numout /= 6 ) CLOSE( numout ) ! standard model output file
464	!
465	numout = 6 ! redefine numout in case it is used after this point...
466	!
467	END SUBROUTINE nemo_closefile
468
469
470	SUBROUTINE nemo_alloc
471	!!----------------------------------------------------------------------
472	!! * ROUTINE nemo_alloc *
473	!!
474	!! ** Purpose : Allocate all the dynamic arrays of the OPA modules
475	!!
476	!! ** Method :
477	!!----------------------------------------------------------------------
478	USE diawri , ONLY: dia_wri_alloc
479	USE dom_oce , ONLY: dom_oce_alloc
480	USE ldfdyn_oce, ONLY: ldfdyn_oce_alloc
481	USE ldftra_oce, ONLY: ldftra_oce_alloc
482	USE trc_oce , ONLY: trc_oce_alloc
483	USE wrk_nemo , ONLY: wrk_alloc
484	!
485	INTEGER :: ierr
486	!!----------------------------------------------------------------------
487	!
488	ierr = oce_alloc () ! ocean
489	ierr = ierr + dia_wri_alloc ()
490	ierr = ierr + dom_oce_alloc () ! ocean domain
491	ierr = ierr + ldfdyn_oce_alloc() ! ocean lateral physics : dynamics
492	ierr = ierr + ldftra_oce_alloc() ! ocean lateral physics : tracers
493	ierr = ierr + zdf_oce_alloc () ! ocean vertical physics
494	!
495	ierr = ierr + lib_mpp_alloc (numout) ! mpp exchanges
496	ierr = ierr + trc_oce_alloc () ! shared TRC / TRA arrays
497	!
498	ierr = ierr + wrk_alloc(numout, lwp) ! workspace
499	!
500	IF( lk_mpp ) CALL mpp_sum( ierr )
501	IF( ierr /= 0 ) CALL ctl_stop( 'STOP', 'nemo_alloc : unable to allocate standard ocean arrays' )
502	!
503	END SUBROUTINE nemo_alloc
504
505
506	SUBROUTINE nemo_partition( num_pes )
507	!!----------------------------------------------------------------------
508	!! * ROUTINE nemo_partition *
509	!!
510	!! ** Purpose : Work out a sensible factorisation of the number of
511	!! processors for the x and y dimensions.
512	!! ** Method :
513	!!----------------------------------------------------------------------
514	INTEGER, INTENT(in) :: num_pes ! The number of MPI processes we have
515	!
516	INTEGER :: ifact1, ifact2 ! factors of num_pes, ifact1 <= ifact2
517	!!----------------------------------------------------------------------
518
519	! Factorise the number of processors into ifact1*ifact2, such that
520	! ifact1 and ifact2 are as nearly equal as possible.
521
522	CALL sqfact( num_pes, ifact1, ifact2 )
523
524	! Make sure that the smaller dimension of the processor grid
525	! is given the smaller dimension of the global domain
526	IF( jpiglo <= jpjglo) THEN
527	jpni = ifact1
528	jpnj = ifact2
529	ELSE
530	jpni = ifact2
531	jpnj = ifact1
532	ENDIF
533
534	! This should never happen
535	IF( (jpni*jpnj) /= num_pes) THEN
536	WRITE (numout, *) 'WARNING: internal error - factorisation of number of PEs failed'
537	ENDIF
538
539	! This should only happen if num_pes is prime
540	IF( ifact1 <= 1 ) THEN
541	WRITE (numout, *) 'WARNING: factorisation of number of PEs failed'
542	WRITE (numout, *) ' : using grid of ',jpni,' x ',jpnj
543	ENDIF
544	!
545	jpnij = jpni*jpnj
546	!
547	END SUBROUTINE nemo_partition
548
549	SUBROUTINE sqfact ( kn, kna, knb )
550	!!----------------------------------------------------------------------
551	!! * ROUTINE sqfact *
552	!!
553	!! ** Purpose : return factors (kna, knb) of kn, such that
554	!! (1) kna*knb=kn
555	!! (2) kna and knb are as near equal as possible
556	!! (3) kna < knb
557	!! ** Method : Search backwards from the square root of kn,
558	!! until we find an integer that cleanly divides kn
559	!! ** Preconditions : kn must be positive
560	!!----------------------------------------------------------------------
561	INTEGER, INTENT(in ) :: kn
562	INTEGER, INTENT( out) :: kna, knb
563
564	! Search backwards from the square root of n.
565
566	fact_loop: DO kna=SQRT(REAL(kn)),1,-1
567	IF ( kn/kna*kna == kn ) THEN
568	EXIT fact_loop
569	ENDIF
570	END DO fact_loop
571
572	IF( kna < 1 ) kna = 1
573
574	! kna divides kn cleanly. Work out the other factor.
575	knb = kn/kna
576
577	END SUBROUTINE sqfact
578
579	!!======================================================================
580	END MODULE nemogcm

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

Download in other formats: