Home

Context Navigation

source: palm/trunk/DOC/app/chapter_3.5.4.html @ 2416

Last change on this file since 2416 was 405, checked in by raasch, 15 years ago
some initial data assignments changed due to problems with gfortran compiler; bugfix in docu chapter 3.5.4; mrun test modification for hybrid mpi-openmp
Property svn:keywords set to `Id`
File size: 46.4 KB

Line
1	<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
2	<html><head>
3
4	<meta http-equiv="CONTENT-TYPE" content="text/html; charset=windows-1252"><title>PALM chapter 3.5.4</title> <meta name="GENERATOR" content="StarOffice 7 (Win32)"> <meta name="AUTHOR" content="Siegfried Raasch"> <meta name="CREATED" content="20040802;14050943"> <meta name="CHANGED" content="20041117;12180008"> <meta name="KEYWORDS" content="parallel LES model"> <style>
5	<!--
6	@page { size: 21cm 29.7cm }
7	-->
8	</style></head><body style="direction: ltr;" lang="en-US"><h3 style="line-height: 100%;"><font size="4">3.5.4
9	User-defined output quantities<br>
10	</font></h3>A very typical request of users is the
11	calculation and
12	output of
13	quantities which are not part of PALM's standard output. The basic user
14	interface includes a number of subroutines which allow the calculation
15	of user-defined quantities and output of these quantities as 1.
16	(horizontally averaged) <a href="#vertical_profiles">vertical
17	profiles</a>, 2. <a href="#time_series">time
18	series</a>, 3. <a href="#2D_and_3D">2d cross
19	section</a> or <a href="#2D_and_3D">3d volume data</a>,
20	4. <a href="#dvrp">dvrp</a> objects, and 5. <a href="#spectra">spectra</a>. The respective
21	subroutines
22	contain sample code lines (written as comment lines) for defining,
23	calculating and
24	output of such quantities.<br><br>Output times, averaging
25	intervals, etc. are steered by the same variables as used for the
26	standard PALM output quantities, e.g. <a href="chapter_4.2.html#dt_data_output">dt_data_output</a>.<br><br>The
27	following five parts of this chapter explains step-by-step how
28	to modify/extend the
29	respective default user interface subroutines in order to generate the respective
30	output:<br><ol><li>(horizontally averaged) <a href="chapter_3.5.4.html#vertical_profiles">vertical profiles</a>, </li><li><a href="chapter_3.5.4.html#time_series">time series</a>, </li><li><a href="chapter_3.5.4.html#2D_and_3D">2d cross
31	section</a> or <a href="chapter_3.5.4.html#2D_and_3D">3d
32	volume data</a>, </li><li><a href="chapter_3.5.4.html#dvrp">dvrp</a> objects,
33	and </li><li><a href="chapter_3.5.4.html#spectra">spectra</a>.</li></ol><br><h4><a name="vertical_profiles"></a>1. Output of user-defined
34	vertical profiles</h4>This example shows the output of the
35	quantity "turbulent resolved-scale horizontal momentum flux" (uv). If
36	more than one user-defined
37	quantity shall be output, the following steps have to be carried out in
38	the
39	same way for each of the quantities.<br><br><ol><li>The
40	quantity has to be given a unique string identifier, e.g. <span style="font-style: italic;">'uv'</span>.
41	This identifier must be different from the identifiers used for the
42	PALM standard output (see list in description of parameter <a href="chapter_4.2.html#data_output_pr">data_output_pr</a>).
43	To switch on output of this quantity, the user has to assign the string
44	identifier to the parameter <a href="chapter_4.3.html#data_output_pr_user">data_output_pr_user</a>,
45	eg.:<br><br><span style="font-family: monospace;">
46	data_output_pr_user</span> = <span style="font-style: italic;">'uv'</span>,<span style="font-style: italic;"></span><span style="font-style: italic;"></span><span style="font-style: italic;"></span><span style="font-style: italic;"></span><span style="font-style: italic;"></span><span style="font-style: italic;"></span><br><br></li><li>For
47	the
48	quantity, an identification number, a physical unit, and the vertical
49	grid on which it is defined (u- or w-grid), has to be assigned
50	(subroutine <a href="chapter_3.5.1.html#user_check_data_output_pr"><span style="font-family: monospace;">user_check_data_output_pr</span></a>):<span style="font-family: monospace;"><br><br>
51	CASE (
52	'uv' )<br></span><span style="font-family: monospace;">
53	index = pr_palm + <span style="color: rgb(255, 0, 0);">1<span style="color: rgb(0, 0, 0);">
54
55
56
57
58	! identification number</span></span><br>
59	dopr_index(var_count) = index<br>
60	dopr_unit(var_count)  = '<span style="color: rgb(255, 0, 0);">m2/s2</span>'
61
62
63	!
64	physical unit<br>
65	hom(:,2,index,:)      = SPREAD( <span style="color: rgb(255, 0, 0);">zu</span>, 2,
66	statistic_regions+1 )    ! vertical grid<br></span><br>Here
67	only the those parts in <span style="color: rgb(255, 0, 0);">red</span>
68	color have to be given by the user appropriately.<br>The
69	identification number (<span style="font-family: Courier New,Courier,monospace;">index</span>)
70	must be within the range [<span style="font-family: Courier New,Courier,monospace;">
71	pr_palm+1 , pr_palm+max_pr_user</span> ], where <span style="font-family: Courier New,Courier,monospace;">max_pr_user</span>
72	is the number of user-defined profiles as given by parameter <a href="chapter_4.3.html#data_output_pr_user">data_output_pr_user</a>
73	in the respective PALM run. The physical unit has to be given
74	with respect to the NetCDF conventions. If no unit is given,
75	PALM will abort. The vertical grid has to be either <span style="font-family: Courier New,Courier,monospace;">zu</span>
76	(u-grid) or <span style="font-family: Courier New,Courier,monospace;">zw</span>
77	(w-grid).<br><br></li><li>The quantity has to
78	be calculated for all gridpoints (subroutine <a href="chapter_3.5.1.html#user_statistics">user_statistics</a>):<br><br><span style="font-family: monospace;">    !$OMP
79	DO</span><br style="font-family: monospace;"><span style="font-family: monospace;">
80	DO  i = nxl, nxr</span><br style="font-family: monospace;"><span style="font-family: monospace;">
81	DO  j = nys, nyn</span><br style="font-family: monospace;"><span style="font-family: monospace;">
82	DO  k = nzb_s_inner(j,i)+1, nzt</span><br style="font-family: monospace;"><span style="font-family: monospace;"></span><span style="font-family: monospace;"></span><span style="font-family: monospace;"></span><span style="font-family: monospace;">
83	sums_l(k,pr_palm+<span style="color: rgb(255, 0, 0);">1</span>,tn)
84	= sums_l(k,pr_palm+<span style="color: rgb(255, 0, 0);">1</span>,tn)
85	+
86	&</span><br style="font-family: monospace;"><span style="font-family: monospace;">
87	<span style="color: rgb(255, 0, 0);">( 0.5 * ( u(k,j,i) +
88	u(k,j,i+1) ) - hom(k,1,1,sr) ) * &</span></span><br style="font-family: monospace; color: rgb(255, 0, 0);"><span style="font-family: monospace; color: rgb(255, 0, 0);">
89	( 0.5 * ( v(k,j,i) + v(k,j+1,i) ) - hom(k,1,2,sr) )   &</span><br style="font-family: monospace;"><span style="font-family: monospace;">
90	* rmask(j,i,sr)<br>
91	ENDDO<br>
92	ENDDO<br>    ENDDO<br></span><br style="font-family: monospace;">Once again, only those parts
93	in <span style="color: rgb(255, 0, 0);">red</span>
94	have to be adjusted by the user.<br>The
95	turbulent resolved-scale momentum flux uv is defined as the product
96	of the deviations of the horizontal velocities from their respective
97	horizontally averaged mean values. These mean values are stored in
98	array <span style="font-family: Courier New,Courier,monospace;">hom(..,1,1,sr)</span>
99	and <span style="font-family: Courier New,Courier,monospace;">hom(..,1,2,sr)</span>
100	for the u- and v-component, respectively. Since due to the staggered
101	grid, <span style="font-family: Courier New,Courier,monospace;">u</span>
102	and <span style="font-family: Courier New,Courier,monospace;">v</span>
103	are not defined at the same gridpoints, they have to be interpolated
104	appropriately (here to the center of the gridbox). The result of the
105	calculation is stored in array <span style="font-family: Courier New,Courier,monospace;">sums_l</span>.
106	The second index of this array is the identification number of the
107	profile which must match the one given in the previous step 2.</li></ol><br><h4><a name="time_series"></a>2. Output of user-defined
108	timeseries</h4>This example shows the output of two time series
109	for the absolut extremal values of the horizontal velocities <span style="font-style: italic;">u</span> and <span style="font-style: italic;">v</span>. If more than one
110	user-defined
111	quantity shall be output, the following steps have to be carried out in
112	the
113	same way for each of the quantities.<br><br><ol><li>For
114	each time series quantity you have to give a label and a unit
115	(subroutine <a href="chapter_3.5.1.html#user_init">user_init</a>),
116	which will be used for the NetCDF file. They must not contain more than
117	seven characters. The value of <code>dots_num</code> has
118	to be increased by the number of new time series quantities. Its old
119	value has to be stored in <code>dots_num_palm</code>.<br><br><code>
120	dots_label(dots_num<span style="color: rgb(255, 0, 0);">+1</span>)
121	= '<span style="color: rgb(255, 0, 0);">abs_umx</span>'<br>
122	dots_unit(dots_num<span style="color: rgb(255, 0, 0);">+1</span>)
123	= '<span style="color: rgb(255, 0, 0);">m/s</span>'<br>
124	dots_label(dots_num<span style="color: rgb(255, 0, 0);">+2</span>)
125	= '<span style="color: rgb(255, 0, 0);">abs_vmx</span>'<br>
126	dots_unit(dots_num<span style="color: rgb(255, 0, 0);">+2</span>)
127	= '<span style="color: rgb(255, 0, 0);">m/s</span>'<br><br>
128	dots_num_palm = dots_num<br>
129	dots_num = dots_num <span style="color: rgb(255, 0, 0);">+
130	2</span><br><br></code>Only those parts in <span style="color: rgb(255, 0, 0);">red</span> have to be
131	adjusted by the user.<br><br></li><li>These
132	quantities are calculated and output in subroutine <a href="chapter_3.5.1.html#user_statistics">user_statistics</a>
133	for every statistic region <code>sr</code> defined by the
134	user, but at least for the region <span style="font-style: italic;">"total domain"</span>:<br> <br><code>
135	ts_value(dots_num_palm+1,sr) = ABS( u_max )<br>
136	ts_value(dots_num_palm+2,sr) = ABS( v_max )<br></code></li></ol> <br><h4><span style="font-weight: bold;"><a name="2D_and_3D"></a>3.
137	Output of user-defined 2d cross sections or 3d volume data</span></h4><br>This
138	example shows the output of the
139	quantity "square of the u-component" (<span style="font-weight: bold;">Note:</span> this quantity
140	could of course easily be calculated from the u-component by
141	postprocessing the PALM output so that calculation within PALM is not
142	necessarily required). If more than one user-defined
143	quantity shall be output, the following steps have to be carried out in
144	the
145	same way for each of the quantities.<br><br><ol><li>The
146	quantity has to be given a unique string identifier, e.g. <span style="font-style: italic;">'u2'</span>.
147	This identifier must be different from the identifiers used for the
148	PALM standard output (see list in description of parameter <a href="chapter_4.2.html#data_output">data_output</a>).
149	To switch on output of this quantity, the user has to assign the string
150	identifier to the parameter <a href="chapter_4.3.html#data_output_user">data_output_user</a>,
151	eg.:<br><br><span style="font-family: monospace;">
152	data_output_user</span> = <span style="font-style: italic;">'u2'</span>, <span style="font-style: italic;"> 'u2_xy_av'</span><br><br>The
153	pure string <span style="font-style: italic;">'u2'</span>
154	switches on the output of instantaneous 3d volume data. Output of cross
155	section data and time averaged data is switched on by additionally
156	appending the strings <span style="font-style: italic;">'_xy'</span>,
157	<span style="font-style: italic;">'_xz'</span>, <span style="font-style: italic;">'_yz'</span>, and/or <span style="font-style: italic;">'_av'</span> (for a
158	detailed explanation see parameter <a href="chapter_4.2.html#data_output">data_output</a>).<br><br></li><li>In
159	order to store the quantities' grid point data within PALM, a 3d data
160	array has to be declared in module <a href="chapter_3.5.1.html#user"><span style="font-family: monospace;">user</span></a>:<br><br><span style="font-family: monospace;">    REAL,
161	DIMENSION(:,:,:), ALLOCATABLE ::  u2, u2_av</span><br><br>The
162	second array <span style="font-family: monospace;">u2_av</span>
163	is needed in case that output of time averaged data is requested. It is
164	used to store the sum of the data of the respective time levels over
165	which the average has to be carried out.<br><br><br></li><li>The
166	data array has to be allocated in subroutine <a href="chapter_3.5.1.html#user_init"><span style="font-family: monospace;">user_init</span></a>:<br><br><span style="font-family: monospace;">
167	ALLOCATE( u2(nzb:nzt+1,nys-1:nyn+1,nxl-1:nxr+1) )</span><br><br>In
168	case that output of time averaged data is requested, the array
169	containing the sum has possibly to be read from the restart file (local
170	filename <a href="chapter_3.4.html#BININ">BININ</a>)
171	by executing the following code in <span style="font-family: monospace;">user_init</span>:<br><br><span style="font-family: monospace;">
172	IF ( initializing_actions == 'read_restart_data' )
173	THEN</span><br style="font-family: monospace;"><span style="font-family: monospace;">
174	READ ( 13 )  field_char</span><br style="font-family: monospace;"><span style="font-family: monospace;">
175	DO  WHILE ( TRIM( field_char ) /= '*** end
176	user ***' )</span><br style="font-family: monospace;"><br style="font-family: monospace;"><span style="font-family: monospace;">
177	SELECT CASE ( TRIM( field_char ) )</span><br style="font-family: monospace;"><br style="font-family: monospace;"><span style="font-family: monospace;">
178	CASE ( 'u2_av' )</span><br style="font-family: monospace;"><span style="font-family: monospace;">
179	ALLOCATE(
180	u2_av(nzb:nzt+1,nys-1:nyn+1,nxl-1:nxr+1) )</span><br style="font-family: monospace;"><span style="font-family: monospace;">
181	READ ( 13 )  u2_av</span><br style="font-family: monospace;"><br style="font-family: monospace;"><span style="font-family: monospace;">
182	CASE DEFAULT</span><br style="font-family: monospace;"><span style="font-family: monospace;">
183	PRINT*,
184	'+++ user_init: unknown
185	variable named "', &</span><br style="font-family: monospace;"><span style="font-family: monospace;">
186
187
188	TRIM( field_char ), '" found in'</span><br style="font-family: monospace;"><span style="font-family: monospace;">
189	PRINT*,
190	'
191	data from prior run on PE ', myid</span><br style="font-family: monospace;"><span style="font-family: monospace;">
192
193	CALL local_stop</span><br style="font-family: monospace;"><br style="font-family: monospace;"><span style="font-family: monospace;">
194	END SELECT<br>
195	ENDDO</span><br style="font-family: monospace;"><span style="font-family: monospace;">
196	ENDIF</span><br style="font-family: monospace;"><br><br></li><li>The
197	quantity has to be given a unit (subroutine <a href="chapter_3.5.1.html#user_check_data_output"><span style="font-family: monospace;">user_check_data_output</span></a>):<br><br><span style="font-family: monospace;">    CASE (
198	'u2' )</span><br style="font-family: monospace;"><span style="font-family: monospace;">
199	unit = 'm2/s2'</span><br> <br>Otherwise,
200	PALM will abort.<br><br><br></li><li>The
201	vertical grid on which the quantity is defined (given by the levels
202	'zu' or 'zw', on which the u- or w-component of the velocity are
203	defined) has to be specified for the NetCDF output files in subroutine <a href="chapter_3.5.1.html#user_define_netcdf_grid"><span style="font-family: monospace;">user_define_netcdf_grid</span></a>:<br><br><span style="font-family: monospace;">
204	CASE ( 'u2', 'u2_xy', 'u2_xz', 'u2_yz' )</span><br style="font-family: monospace;"><span style="font-family: monospace;">
205	grid = 'zu'</span><br> <br>As
206	the example shows, this grid has to be defined for the 3d volume data
207	as well as for all of the three cross sections.<br><span style="font-family: monospace;"><br><br></span></li><li>After
208	each timestep, the quantity has to be calculated at all gridpoints and
209	to be stored. This has to be done in subroutine <a href="chapter_3.5.1#user_actions"><span style="font-family: monospace;">user_actions</span></a>
210	at location 'after_integration':<br><br><span style="font-family: monospace;">    CASE
211	( 'after_integration' )</span><br style="font-family: monospace;"><span style="font-family: monospace;">!</span><br style="font-family: monospace;"><span style="font-family: monospace;">!--
212	Enter actions to be done after every time integration (before</span><br style="font-family: monospace;"><span style="font-family: monospace;">!--
213	data output)</span><br style="font-family: monospace;"><span style="font-family: monospace;">!--
214	Sample for user-defined output:</span><br style="font-family: monospace;"><span style="font-family: monospace;">
215	DO  i = nxl-1, nxr+1</span><br style="font-family: monospace;"><span style="font-family: monospace;">
216	DO  j = nys-1, nyn+1</span><br style="font-family: monospace;"><span style="font-family: monospace;">
217	DO  k = nzb, nzt+1</span><br style="font-family: monospace;"><span style="font-family: monospace;">
218	u2(k,j,i) = u(k,j,i)**2</span><br style="font-family: monospace;"><span style="font-family: monospace;">
219	ENDDO</span><br style="font-family: monospace;"><span style="font-family: monospace;">
220	ENDDO</span><br style="font-family: monospace;"><span style="font-family: monospace;">
221	ENDDO</span><br><br> <br></li><li>In
222	case that output of time-averaged data is requested, the sum- and
223	average-operations as well as the allocation of the sum-array have to
224	be carried out in subroutine <a href="chapter_3.5.1.html#user_3d_data_averaging"><span style="font-family: monospace;">user_3d_data_averaging</span></a>:<br><br><span style="font-family: monospace;">
225	IF ( mode == 'allocate' )  THEN<br>
226	...</span><br style="font-family: monospace;"><span style="font-family: monospace;">
227	CASE ( 'u2' )</span><br style="font-family: monospace;"><span style="font-family: monospace;">
228	IF ( .NOT. ALLOCATED( u2_av ) )  THEN</span><br style="font-family: monospace;"><span style="font-family: monospace;">
229	ALLOCATE( u2_av(nzb:nzt+1,nys-1:nyn+1,nxl-1:nxr+1) )</span><br style="font-family: monospace;"><span style="font-family: monospace;">
230	ENDIF</span><br style="font-family: monospace;"><span style="font-family: monospace;">
231	u2_av = 0.0<br>    ...</span><br style="font-family: monospace;"><span style="font-family: monospace;">
232	ELSEIF ( mode == 'sum' )  THEN<br>
233	...</span><br style="font-family: monospace;"><span style="font-family: monospace;">       CASE
234	( 'u2' )</span><br style="font-family: monospace;"><span style="font-family: monospace;">
235	DO  i = nxl-1, nxr+1</span><br style="font-family: monospace;"><span style="font-family: monospace;">
236	DO  j = nys-1, nyn+1</span><br style="font-family: monospace;"><span style="font-family: monospace;">
237	DO  k = nzb, nzt+1</span><br style="font-family: monospace;"><span style="font-family: monospace;">
238	u2_av(k,j,i) = u2_av(k,j,i) + u2(k,j,i)</span><br style="font-family: monospace;"><span style="font-family: monospace;">
239	ENDDO</span><br style="font-family: monospace;"><span style="font-family: monospace;">
240	ENDDO</span><br style="font-family: monospace;"><span style="font-family: monospace;">
241	ENDDO<br>    ...</span><br style="font-family: monospace;"><span style="font-family: monospace;">
242	ELSEIF ( mode == 'average' )  THEN<br>
243	...</span><br style="font-family: monospace;"><span style="font-family: monospace;">       CASE
244	( 'u2' )</span><br style="font-family: monospace;"><span style="font-family: monospace;">
245	DO  i = nxl-1, nxr+1</span><br style="font-family: monospace;"><span style="font-family: monospace;">
246	DO  j = nys-1, nyn+1</span><br style="font-family: monospace;"><span style="font-family: monospace;">
247	DO  k = nzb, nzt+1</span><br style="font-family: monospace;"><span style="font-family: monospace;">
248	u2_av(k,j,i) = u2_av(k,j,i) / REAL( average_count_3d )</span><br style="font-family: monospace;"><span style="font-family: monospace;">
249	ENDDO</span><br style="font-family: monospace;"><span style="font-family: monospace;">
250	ENDDO</span><br style="font-family: monospace;"><span style="font-family: monospace;">
251	ENDDO</span><br><br> </li><li>For
252	output of 2d cross sections, the gridpoint data of the quantity has to
253	be resorted to array <span style="font-family: monospace;">local_pf</span>
254	in subroutine <a href="chapter_3.5.1.html#user_data_output_2d"><span style="font-family: monospace;">user_data_output_2d</span></a>.
255	Also the vertical grid, on which the quantity is defined, has to be set
256	again:<br><br><span style="font-family: monospace;">    CASE
257	( 'u2_xy', 'u2_xz', 'u2_yz' )</span><br style="font-family: monospace;"><span style="font-family: monospace;">
258	IF ( av == 0 )  THEN</span><br style="font-family: monospace;"><span style="font-family: monospace;">
259	DO  i = nxl-1, nxr+1</span><br style="font-family: monospace;"><span style="font-family: monospace;">
260	DO  j = nys-1, nyn+1</span><br style="font-family: monospace;"><span style="font-family: monospace;">
261	DO  k = nzb, nzt+1</span><br style="font-family: monospace;"><span style="font-family: monospace;">
262	local_pf(i,j,k) = u2(k,j,i)</span><br style="font-family: monospace;"><span style="font-family: monospace;">
263	ENDDO</span><br style="font-family: monospace;"><span style="font-family: monospace;">
264	ENDDO</span><br style="font-family: monospace;"><span style="font-family: monospace;">
265	ENDDO</span><br style="font-family: monospace;"><span style="font-family: monospace;">
266	ELSE</span><br style="font-family: monospace;"><span style="font-family: monospace;">
267	DO  i = nxl-1, nxr+1</span><br style="font-family: monospace;"><span style="font-family: monospace;">
268	DO  j = nys-1, nyn+1</span><br style="font-family: monospace;"><span style="font-family: monospace;">
269	DO  k = nzb, nzt+1</span><br style="font-family: monospace;"><span style="font-family: monospace;">
270	local_pf(i,j,k) = u2_av(k,j,i)</span><br style="font-family: monospace;"><span style="font-family: monospace;">
271	ENDDO</span><br style="font-family: monospace;"><span style="font-family: monospace;">
272	ENDDO</span><br style="font-family: monospace;"><span style="font-family: monospace;">
273	ENDDO</span><br style="font-family: monospace;"><span style="font-family: monospace;">
274	ENDIF</span><br style="font-family: monospace;"><br style="font-family: monospace;"><span style="font-family: monospace;">
275	grid = 'zu'</span><br> <br>The <span style="font-family: monospace;">ELSE</span> case is
276	only needed in case that output of time-averaged data is requested.<br><br>
277	As a special case, xy cross section output can also be defined at one single level at height <span style="font-family: monospace;">k=nzb+1</span>
278	on the u-grid. This features is useful for output of surface data (e.g.
279	heat fluxes). In this case, the corresponding 2d data has to be
280	resorted to the array <span style="font-family: monospace;">local_pf(i,j,nzb+1).</span> In addition to this, the grid in <a href="chapter_3.5.1.html#user_define_netcdf_grid"><span style="font-family: monospace;">user_define_netcdf_grid</span></a> as well as in <a href="chapter_3.5.1.html#user_data_output_2d"><span style="font-family: monospace;">user_data_output_2d</span></a> must be set to <span style="font-family: monospace;">grid = 'zu1'</span>.<br><br>
281	<span style="font-family: monospace;">    CASE
282	( 'u2_xy' )</span><br style="font-family: monospace;"><span style="font-family: monospace;">
283	IF ( av == 0 )  THEN</span><br style="font-family: monospace;"><span style="font-family: monospace;">
284	DO  i = nxl-1, nxr+1</span><br style="font-family: monospace;"><span style="font-family: monospace;">
285	DO  j = nys-1, nyn+1</span><br style="font-family: monospace;"><span style="font-family: monospace;">
286	local_pf(i,j,nzb+1) = u2(j,i)</span><br style="font-family: monospace;"><span style="font-family: monospace;">
287	ENDDO</span><br style="font-family: monospace;"><span style="font-family: monospace;">
288	ENDDO</span><br style="font-family: monospace;"><span style="font-family: monospace;">
289	ELSE</span><br style="font-family: monospace;"><span style="font-family: monospace;">
290	DO  i = nxl-1, nxr+1</span><br style="font-family: monospace;"><span style="font-family: monospace;">
291	DO  j = nys-1, nyn+1</span><br style="font-family: monospace;"><span style="font-family: monospace;">
292	local_pf(i,j,nzb+1) = u2_av(j,i)</span><br style="font-family: monospace;"><span style="font-family: monospace;">
293	ENDDO</span><br style="font-family: monospace;"><span style="font-family: monospace;">
294	ENDDO</span><br style="font-family: monospace;"><span style="font-family: monospace;">
295	ENDIF</span><br style="font-family: monospace;"><br style="font-family: monospace;"><span style="font-family: monospace;">
296	grid = 'zu1'</span><br><span style="font-family: monospace;">
297	two_d = .TRUE.</span><br><br>
298	Note that <span style="font-family: monospace;">two_d = .TRUE.</span> is necessary for output of a 2d data slice.<br><br>
299	<br></li><li>For
300	output of 3d volume data, the gridpoint data of the quantity has to be
301	resorted to array <span style="font-family: monospace;">local_pf</span>
302	in subroutine <a href="chapter_3.5.1.html#user_data_output_3d"><span style="font-family: monospace;">user_data_output_3d</span></a>.:<br><br><span style="font-family: monospace;">    CASE
303	( 'u2' )</span><br style="font-family: monospace;"><span style="font-family: monospace;">
304	IF ( av == 0 )  THEN</span><br style="font-family: monospace;"><span style="font-family: monospace;">
305	DO  i = nxl-1, nxr+1</span><br style="font-family: monospace;"><span style="font-family: monospace;">
306	DO  j = nys-1, nyn+1</span><br style="font-family: monospace;"><span style="font-family: monospace;">
307	DO  k = nzb, nz_do</span><br style="font-family: monospace;"><span style="font-family: monospace;">
308	local_pf(i,j,k) = u2(k,j,i)</span><br style="font-family: monospace;"><span style="font-family: monospace;">
309	ENDDO</span><br style="font-family: monospace;"><span style="font-family: monospace;">
310	ENDDO</span><br style="font-family: monospace;"><span style="font-family: monospace;">
311	ENDDO</span><br style="font-family: monospace;"><span style="font-family: monospace;">
312	ELSE</span><br style="font-family: monospace;"><span style="font-family: monospace;">
313	DO  i = nxl-1, nxr+1</span><br style="font-family: monospace;"><span style="font-family: monospace;">
314	DO  j = nys-1, nyn+1</span><br style="font-family: monospace;"><span style="font-family: monospace;">
315	DO  k = nzb, nz_do</span><br style="font-family: monospace;"><span style="font-family: monospace;">
316	local_pf(i,j,k) = u2_av(k,j,i)</span><br style="font-family: monospace;"><span style="font-family: monospace;">
317	ENDDO</span><br style="font-family: monospace;"><span style="font-family: monospace;">
318	ENDDO</span><br style="font-family: monospace;"><span style="font-family: monospace;">
319	ENDDO</span><br style="font-family: monospace;"><span style="font-family: monospace;">
320	ENDIF</span><br><br>The <span style="font-family: monospace;">ELSE</span> case is
321	only needed in case that output of time-averaged data is requested.<br><br><br></li><li>In
322	case of job chains, the sum array has to be written to the (binary)
323	restart file (local filename <a href="chapter_3.4.html#BINOUT">BINOUT</a>)
324	in subroutine <a href="chapter_3.5.1.html#user_last_actions"><span style="font-family: monospace;">user_last_actions</span></a>:<br><br><span style="font-family: monospace;">
325	IF ( ALLOCATED( u2_av ) )  THEN</span><br style="font-family: monospace;"><span style="font-family: monospace;">
326	WRITE ( 14 )
327	'u2_av
328	';  WRITE ( 14 )  u2_av</span><br style="font-family: monospace;"><span style="font-family: monospace;">
329	ENDIF</span><br><br>Otherwise,
330	the calculated
331	time-average may be wrong. In the restart run, this quantity has to be
332	read from the restart file by including the following code in
333	subroutine <a href="chapter_3.5.1.html#user_read_restart_data"><span style="font-family: monospace;">user_read_restart_data</span></a>:<br><br><span style="font-family: monospace;">    IF ( initializing_actions == 'read_restart_data' )  THEN<br>       READ ( 13 )  field_char<br>       DO  WHILE ( TRIM( field_char ) /= '* end user *' )<br><br>          SELECT CASE ( TRIM( field_char ) )<br><br>             CASE ( 'u2_av' )<br>                IF ( .NOT. ALLOCATED( u2_av ) ) THEN<br>
334	ALLOCATE( u2_av(nzb:nzt+1,nys-1:nyn+1,nxl-1:nxr+1) )<br>                ENDIF<br>                READ ( 13 )  tmp_3d<br>
335	u2_av(:,nysc-1:nync+1,nxlc-1:nxrc+1) = &<br>
336	tmp_3d(:,nysf-1:nynf+1,nxlf-1:nxrf+1)<br><br>             CASE DEFAULT<br>
337	PRINT*, '+++ user_init: unknown variable named "', &<br>
338	TRIM( field_char ), '" found in'<br>
339	PRINT*,
340	'
341	data from prior run on PE ', myid<br>                CALL local_stop<br><br>          END SELECT<br><br>          READ ( 13 )  field_char<br><br>       ENDDO<br>    ENDIF<br></span></li></ol><h4><a name="dvrp"></a>4. Output of user-defined DVRP objects</h4>This
342	example shows the output of the
343	quantity "square of the u-component", <span style="font-style: italic;">u<sup>2</sup></span>.
344	If more than one user-defined
345	quantity shall be output, the following steps have to be carried out in
346	the
347	same way for each of the quantities. First, steps 1 - 6
348	of <a href="chapter_3.5.4.html#2D_and_3D"><span style="font-weight: bold;">part 3.</span> </a> <a href="chapter_3.5.4.html#2D_and_3D">(2d cross
349	section or </a><a href="chapter_3.5.4.html#2D_and_3D">3d
350	volume data)</a> are required. Second, the gridpoint
351	data
352	of the quantity has to
353	be resorted to array <span style="font-family: monospace;">local_pf</span>
354	in subroutine <a href="chapter_3.5.1.html#user_data_output_dvrp"><span style="font-family: monospace;">user_data_output_dvrp</span></a>
355	as follows:<br><br><code>
356	CASE ( '<span style="color: rgb(255, 0, 0);">u2</span>',
357	'<span style="color: rgb(255, 0, 0);">u2</span>_xy',
358	'<span style="color: rgb(255, 0, 0);">u2</span>_xz',
359	'<span style="color: rgb(255, 0, 0);">u2</span>_yz'
360	)<br>
361	DO  i = nxl, nxr+1<br>
362	DO  j = nys, nyn+1<br>
363	DO  k = nzb, nz_do3d<br>
364	local_pf(i,j,k) = <span style="color: rgb(255, 0, 0);">u2</span>(k,j,i)<br>
365	ENDDO<br>
366	ENDDO<br>
367	ENDDO</code><br><br>Only those parts in <span style="color: rgb(255, 0, 0);">red</span> have to be
368	adjusted by the user.<br><br>After performing these steps,
369	the user-defined quantity <code>'u2'</code> can be
370	selected like standard model quantities by the <a href="chapter_4.2.html#dvrp_graphics">dvrp_graphics</a>
371	package parameter <a href="chapter_4.2.html#mode_dvrp">mode_dvrp</a>.<br><br><h4><a name="spectra"></a>5. Output of user-defined spectra</h4>This
372	example shows the output of the
373	quantity "turbulent resolved-scale horizontal momentum flux" (uv). If
374	more than one user-defined
375	quantity shall be output, the following steps have to be carried out in
376	the
377	same way for each of the quantities.<br><br><ol><li>The
378	calculation of user-defined spectra is closely linked with the
379	calculation of <a href="chapter_3.5.4.html#vertical_profiles"><span style="font-weight: bold;">part 1.</span>
380	(user-defined
381	</a><a href="chapter_3.5.4.html#vertical_profiles">vertical
382	profiles</a><a href="chapter_3.5.4.html#vertical_profiles">)</a><span style="font-weight: bold;"></span> and <a href="chapter_3.5.4.html#2D_and_3D"><span style="font-weight: bold;">part 3.</span> (user-defined</a><a href="chapter_3.5.4.html#2D_and_3D"> </a><a href="chapter_3.5.4.html#2D_and_3D">3d volume data</a><a href="chapter_3.5.4.html#2D_and_3D">)</a><span style="font-weight: bold;"></span>. Therefore, the
383	following prerequisites apply
384	for each user-defined
385	spectra quantity:</li></ol><ol style="list-style-type: upper-alpha;"><ol style="list-style-type: upper-alpha;"><li>From <a href="chapter_3.5.4.html#vertical_profiles"><span style="font-weight: bold;">part 1.</span>
386	(user-defined
387	</a><a href="chapter_3.5.4.html#vertical_profiles">vertical
388	profiles</a><a href="chapter_3.5.4.html#vertical_profiles">)</a>
389	steps 2 and 3. <br>See the sample code (as comment lines) for <span style="font-style: italic;">'uv'</span> and <code>ustvst</code>,
390	respectively. (Actual output of vertical profiles - step 1 - is not
391	required.)</li><li>From <a href="chapter_3.5.4.html#2D_and_3D"><span style="font-weight: bold;">part 3.</span> (user-defined</a><a href="chapter_3.5.4.html#2D_and_3D"> </a><a href="chapter_3.5.4.html#2D_and_3D">3d volume data</a><a href="chapter_3.5.4.html#2D_and_3D">)</a> steps 2, 3,
392	4, 5, and 6. <br>See the sample code (as comment lines) for <span style="font-style: italic;">'uv'</span> and <code>ustvst</code>,
393	respectively. (Actual output of 3d volume data - step 1 - is not
394	required.)</li><li>The quantity has to be given a unique
395	string identifier, e.g. <span style="font-style: italic;">'uv'</span>.
396	This identifier must be different from the identifiers used for the
397	PALM standard output (see list in description of package parameter <a href="chapter_4.2.html#data_output_sp">data_output_sp</a>).
398	To switch on output of this quantity, the user has to assign the string
399	identifier to the package parameter <a href="chapter_4.2.html#data_output_sp">data_output_sp</a>,
400	eg.:<br><br><span style="font-family: monospace;">
401	data_output_sp</span> =  <span style="font-style: italic;">'uv'</span><br><br></li></ol>A.
402	and B. as prerequisites for C. require a naming convention of identical
403	identifiers, e.g. data_output_pr_user = 'uv', data_output_user =
404	'uv' and data_output_sp = 'uv'. This naming convention applies only
405	in case of user-defined spectra.</ol><ol start="2"><li>Edit
406	the subroutine <a href="chapter_3.5.1.html#user_spectra">user_spectra</a>
407	(only those parts in <span style="color: rgb(255, 0, 0);">red</span>)
408	as follows:<br><br><code>
409	IF ( mode == 'preprocess' )  THEN<br><br>
410	SELECT CASE ( TRIM( data_output_sp(m) ) )<br>
411	<br>
412	CASE ( 'u', 'v', 'w', 'pt', 'q' )<br>!--
413	Not allowed here since these are the standard quantities used in <br>!--
414	preprocess_spectra.<br>
415	<br><span style="color: rgb(255, 0, 0); background-color: rgb(255, 255, 255);">
416	CASE ( 'uv' )</span><br style="color: rgb(255, 0, 0); background-color: rgb(255, 255, 255);"><span style="color: rgb(255, 0, 0); background-color: rgb(255, 255, 255);">
417	pr = pr_palm+1</span><br style="color: rgb(255, 0, 0); background-color: rgb(255, 255, 255);"><span style="color: rgb(255, 0, 0); background-color: rgb(255, 255, 255);">
418	d(nzb+1:nzt,nys:nyn,nxl:nxr) = ustvst(nzb+1:nzt,nys:nyn,nxl:nxr)</span><br style="color: rgb(255, 0, 0); background-color: rgb(255, 255, 255);">
419	<br>
420	CASE DEFAULT<br>
421	PRINT*, '+++ user_spectra/preprocess: Spectra of ', &<br>
422	TRIM( data_output_sp(m) ), ' can not be calculated'<br>
423	<br>
424	END SELECT<br><br>    ELSEIF (
425	mode == 'data_output' )  THEN<br><br>
426	SELECT CASE ( TRIM( data_output_sp(m) ) )<br><br>
427	CASE ( 'u', 'v', 'w', 'pt', 'q' )<br>!--
428	Not allowed here since these are the standard quantities used in <br>!--
429	data_output_spectra.<br><br><span style="color: rgb(255, 0, 0);">
430	CASE ( 'uv' )</span><br style="color: rgb(255, 0, 0);"><span style="color: rgb(255, 0, 0);">
431	pr = 6</span><br style="color: rgb(255, 0, 0);"><br>
432	CASE DEFAULT<br>
433	PRINT*, '+++ user_spectra/data_output: Spectra of ', &<br>
434	TRIM( data_output_sp(m) ), ' are not defined'<br><br>
435	END SELECT<br><br>    ENDIF</code><br></li></ol>Note
436	that spectra output is an <a href="chapter_3.7.html">optional
437	software package</a> (see <a href="chapter_3.7.html">chapter
438	3.7</a>). Therefore user-defined spectra also require the package
439	activation via <code>mrun -p spectra</code>.<hr>
440	<p style="line-height: 100%;"><br><font color="#000080"><font color="#000080"><a href="chapter_3.5.3.html"><font color="#000080"><img src="left.gif" name="Grafik1" align="bottom" border="2" height="32" width="32"></font></a><a href="index.html"><font color="#000080"><img src="up.gif" name="Grafik2" align="bottom" border="2" height="32" width="32"></font></a><a href="chapter_3.5.5.html"><font color="#000080"><img style="border: 2px solid ; width: 32px; height: 32px;" alt="" src="right.gif" name="Grafik3"></font></a></font></font></p><p style="line-height: 100%;"><i>Last
441	change: </i> $Id: chapter_3.5.4.html 136 2007-11-26
442	02:47:32Z letzel $</p>
443	</body></html>

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

Download in other formats:

| Impressum | ©Leibniz Universität Hannover |