Home

Context Navigation

chapter_4.1.html @ 95

Last change on this file since 95 was 83, checked in by raasch, 18 years ago

New:
---

Changed:

PALM can be generally installed on any kind of Linux-, IBM-AIX-, or NEC-SX-system by adding appropriate settings to the configuration file.

Scripts are also running under the public domain ksh.

All system relevant compile and link options as well as the host identifier (local_host) are specified in the configuration file.

Filetransfer by ftp removed (options -f removed from mrun and mbuild).

Call of (system-)FLUSH routine moved to new routine local_flush.

return_addres and return_username are read from ENVPAR-NAMELIST-file instead of using local_getenv.

Preprocessor strings for different linux clusters changed to "lc", some preprocessor directives renamed (new: intel_openmp_bug), preprocessor directives for old systems removed

advec_particles, check_open, cpu_log, cpu_statistics, data_output_dvrp, flow_statistics, header, init_dvrp, init_particles, init_1d_model, init_dvrp, init_pegrid, local_getenv, local_system, local_tremain, local_tremain_ini, modules, palm, parin, run_control

new:
local_flush

mbuild, mrun

Errors:

Property svn:keywords set to Id

File size: 137.3 KB

Rev	Line
[5]	1	<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
	2	<html><head>
[62]	3	<meta http-equiv="content-type" content="text/html; charset=ISO-8859-1"><title>PALM chapter 4.1</title></head>
[61]	4	<body><h3><a name="chapter4.1"></a>4.1
	5	Initialization parameters</h3>
	6	<br><table style="text-align: left; width: 100%;" border="1" cellpadding="2" cellspacing="2"> <tbody>
	7	<tr> <td style="vertical-align: top;"><font size="4"><b>Parameter name</b></font></td>
	8	<td style="vertical-align: top;"><font size="4"><b>Type</b></font></td>
	9	<td style="vertical-align: top;"> <p><b><font size="4">Default</font></b> <br> <b><font size="4">value</font></b></p> </td>
	10	<td style="vertical-align: top;"><font size="4"><b>Explanation</b></font></td>
	11	</tr> <tr> <td style="vertical-align: top;">
	12	<p><a name="adjust_mixing_length"></a><b>adjust_mixing_length</b></p>
	13	</td> <td style="vertical-align: top;">L</td>
	14	<td style="vertical-align: top;"><span style="font-style: italic;">.F.</span></td> <td style="vertical-align: top;"> <p style="font-style: normal;">Near-surface adjustment of the
	15	mixing length to the Prandtl-layer law.  </p> <p>Usually
	16	the mixing length in LES models l<sub>LES</sub>
[5]	17	depends (as in PALM) on the grid size and is possibly restricted
	18	further in case of stable stratification and near the lower wall (see
	19	parameter <a href="#wall_adjustment">wall_adjustment</a>).
	20	With <b>adjust_mixing_length</b> = <span style="font-style: italic;">.T.</span>
[61]	21	the Prandtl' mixing length l<sub>PR</sub> = kappa * z/phi
	22	is calculated
[5]	23	and the mixing length actually used in the model is set l = MIN (l<sub>LES</sub>,
[61]	24	l<sub>PR</sub>). This usually gives a decrease of the
	25	mixing length at
[5]	26	the bottom boundary and considers the fact that eddy sizes
[61]	27	decrease in the vicinity of the wall.  </p> <p style="font-style: normal;"><b>Warning:</b> So
	28	far, there is
[5]	29	no good experience with <b>adjust_mixing_length</b> = <span style="font-style: italic;">.T.</span> !  </p>
[61]	30	<p>With <b>adjust_mixing_length</b> = <span style="font-style: italic;">.T.</span> and the
	31	Prandtl-layer being
	32	switched on (see <a href="#prandtl_layer">prandtl_layer</a>)
	33	<span style="font-style: italic;">'(u)* 2+neumann'</span>
[5]	34	should always be set as the lower boundary condition for the TKE (see <a href="#bc_e_b">bc_e_b</a>),
	35	otherwise the near-surface value of the TKE is not in agreement with
	36	the Prandtl-layer law (Prandtl-layer law and Prandtl-Kolmogorov-Ansatz
[61]	37	should provide the same value for K<sub>m</sub>). A warning
	38	is given,
	39	if this is not the case.</p> </td> </tr> <tr>
	40	<td style="vertical-align: top;"> <p><a name="alpha_surface"></a><b>alpha_surface</b></p>
	41	</td> <td style="vertical-align: top;">R<br> </td>
	42	<td style="vertical-align: top;"><span style="font-style: italic;">0.0</span><br> </td>
	43	<td style="vertical-align: top;"> <p style="font-style: normal;">Inclination of the model domain
	44	with respect to the horizontal (in degrees).  </p> <p style="font-style: normal;">By means of <b>alpha_surface</b>
[5]	45	the model domain can be inclined in x-direction with respect to the
	46	horizontal. In this way flows over inclined surfaces (e.g. drainage
[61]	47	flows, gravity flows) can be simulated. In case of <b>alpha_surface
	48	</b>/= <span style="font-style: italic;">0</span>
	49	the buoyancy term
[5]	50	appears both in
	51	the equation of motion of the u-component and of the w-component.<br>
[61]	52	</p> <p style="font-style: normal;">An inclination
	53	is only possible in
[5]	54	case of cyclic horizontal boundary conditions along x AND y (see <a href="#bc_lr">bc_lr</a>
	55	and <a href="#bc_ns">bc_ns</a>) and <a href="#topography">topography</a> = <span style="font-style: italic;">'flat'</span>. </p>
[61]	56	<p>Runs with inclined surface still require additional
[5]	57	user-defined code as well as modifications to the default code. Please
	58	ask the <a href="http://www.muk.uni-hannover.de/%7Eraasch/PALM_group/PALM_group.html#0">PALM
[61]	59	developer  group</a>.</p> </td> </tr>
	60	<tr> <td style="vertical-align: top;"> <p><a name="bc_e_b"></a><b>bc_e_b</b></p> </td>
	61	<td style="vertical-align: top;">C * 20</td> <td style="vertical-align: top;"><span style="font-style: italic;">'neumann'</span></td>
	62	<td style="vertical-align: top;"> <p style="font-style: normal;">Bottom boundary condition of the
	63	TKE.  </p> <p><b>bc_e_b</b> may be
	64	set to <span style="font-style: italic;">'neumann'</span>
	65	or <span style="font-style: italic;">'(u) * 2+neumann'</span>.
	66	<b>bc_e_b</b>
	67	= <span style="font-style: italic;">'neumann'</span>
	68	yields to
[5]	69	e(k=0)=e(k=1) (Neumann boundary condition), where e(k=1) is calculated
[61]	70	via the prognostic TKE equation. Choice of <span style="font-style: italic;">'(u)*2+neumann'</span>
	71	also yields to
[5]	72	e(k=0)=e(k=1), but the TKE at the Prandtl-layer top (k=1) is calculated
	73	diagnostically by e(k=1)=(us/0.1)**2. However, this is only allowed if
	74	a Prandtl-layer is used (<a href="#prandtl_layer">prandtl_layer</a>).
[61]	75	If this is not the case, a warning is given and <b>bc_e_b</b>
	76	is reset
	77	to <span style="font-style: italic;">'neumann'</span>.
	78	</p> <p style="font-style: normal;">At the top
	79	boundary a Neumann
	80	boundary condition is generally used: (e(nz+1) = e(nz)).</p> </td>
	81	</tr> <tr> <td style="vertical-align: top;">
	82	<p><a name="bc_lr"></a><b>bc_lr</b></p>
	83	</td> <td style="vertical-align: top;">C * 20</td>
	84	<td style="vertical-align: top;"><span style="font-style: italic;">'cyclic'</span></td>
	85	<td style="vertical-align: top;">Boundary
	86	condition along x (for all quantities).<br> <br>
	87	By default, a cyclic boundary condition is used along x.<br> <br>
	88	<span style="font-weight: bold;">bc_lr</span> may
	89	also be
[75]	90	assigned the values <span style="font-style: italic;">'dirichlet/radiation'</span>
	91	(inflow from left, outflow to the right) or <span style="font-style: italic;">'radiation/dirichlet'</span>
[61]	92	(inflow from
[5]	93	right, outflow to the left). This requires the multi-grid method to be
	94	used for solving the Poisson equation for perturbation pressure (see <a href="http://www.muk.uni-hannover.de/%7Eraasch/PALM_group/doc/app/chapter_4.2.html#psolver">psolver</a>)
[75]	95	and it also requires cyclic boundary conditions along y (see <a href="#bc_ns">bc_ns</a>).<br> <br>
[5]	96	In case of these non-cyclic lateral boundaries, a Dirichlet condition
	97	is used at the inflow for all quantities (initial vertical profiles -
	98	see <a href="#initializing_actions">initializing_actions</a>
	99	- are fixed during the run) except u, to which a Neumann (zero
[75]	100	gradient) condition is applied. At the outflow, a radiation condition is used for all velocity components, while a Neumann (zero
	101	gradient) condition is used for the scalars. For perturbation
[5]	102	pressure Neumann (zero gradient) conditions are assumed both at the
[61]	103	inflow and at the outflow.<br> <br>
[5]	104	When using non-cyclic lateral boundaries, a filter is applied to the
	105	velocity field in the vicinity of the outflow in order to suppress any
	106	reflections of outgoing disturbances (see <a href="#km_damp_max">km_damp_max</a>
	107	and <a href="#outflow_damping_width">outflow_damping_width</a>).<br>
[61]	108	<br>
[5]	109	In order to maintain a turbulent state of the flow, it may be
	110	neccessary to continuously impose perturbations on the horizontal
	111	velocity field in the vicinity of the inflow throughout the whole run.
	112	This can be switched on using <a href="http://www.muk.uni-hannover.de/%7Eraasch/PALM_group/doc/app/chapter_4.2.html#create_disturbances">create_disturbances</a>.
	113	The horizontal range to which these perturbations are applied is
	114	controlled by the parameters <a href="#inflow_disturbance_begin">inflow_disturbance_begin</a>
	115	and <a href="#inflow_disturbance_end">inflow_disturbance_end</a>.
	116	The vertical range and the perturbation amplitude are given by <a href="http://www.muk.uni-hannover.de/%7Eraasch/PALM_group/doc/app/chapter_4.2.html#psolver">disturbance_level_b</a>,
[61]	117	<a href="http://www.muk.uni-hannover.de/%7Eraasch/PALM_group/doc/app/chapter_4.2.html#psolver">disturbance_level_t</a>,
[5]	118	and <a href="http://www.muk.uni-hannover.de/%7Eraasch/PALM_group/doc/app/chapter_4.2.html#psolver">disturbance_amplitude</a>.
	119	The time interval at which perturbations are to be imposed is set by <a href="http://www.muk.uni-hannover.de/%7Eraasch/PALM_group/doc/app/chapter_4.2.html#dt_disturb">dt_disturb</a>.<br>
[61]	120	<br>
[5]	121	In case of non-cyclic horizontal boundaries <a href="http://www.muk.uni-hannover.de/%7Eraasch/PALM_group/doc/app/chapter_4.2.html#call_psolver_at_all_substeps">call_psolver
[61]	122	at_all_substeps</a> = .T. should be used.<br> <br> <span style="font-weight: bold;">Note:</span><br>
[5]	123	Using non-cyclic lateral boundaries requires very sensitive adjustments
	124	of the inflow (vertical profiles) and the bottom boundary conditions,
	125	e.g. a surface heating should not be applied near the inflow boundary
	126	because this may significantly disturb the inflow. Please check the
[61]	127	model results very carefully.</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="bc_ns"></a><b>bc_ns</b></p>
	128	</td> <td style="vertical-align: top;">C * 20</td>
	129	<td style="vertical-align: top;"><span style="font-style: italic;">'cyclic'</span></td>
	130	<td style="vertical-align: top;">Boundary
	131	condition along y (for all quantities).<br> <br>
	132	By default, a cyclic boundary condition is used along y.<br> <br>
	133	<span style="font-weight: bold;">bc_ns</span> may
	134	also be
[75]	135	assigned the values <span style="font-style: italic;">'dirichlet/radiation'</span>
	136	(inflow from rear ("north"), outflow to the front ("south")) or <span style="font-style: italic;">'radiation/dirichlet'</span>
[5]	137	(inflow from front ("south"), outflow to the rear ("north")). This
	138	requires the multi-grid
	139	method to be used for solving the Poisson equation for perturbation
	140	pressure (see <a href="chapter_4.2.html#psolver">psolver</a>)
[61]	141	and it also requires cyclic boundary conditions along x (see<br> <a href="#bc_lr">bc_lr</a>).<br> <br>
[5]	142	In case of these non-cyclic lateral boundaries, a Dirichlet condition
	143	is used at the inflow for all quantities (initial vertical profiles -
[75]	144	see <a href="chapter_4.1.html#initializing_actions">initializing_actions</a>
	145	- are fixed during the run) except u, to which a Neumann (zero
	146	gradient) condition is applied. At the outflow, a radiation condition is used for all velocity components, while a Neumann (zero
	147	gradient) condition is used for the scalars. For perturbation
[5]	148	pressure Neumann (zero gradient) conditions are assumed both at the
[61]	149	inflow and at the outflow.<br> <br>
[5]	150	For further details regarding non-cyclic lateral boundary conditions
[61]	151	see <a href="#bc_lr">bc_lr</a>.</td> </tr>
	152	<tr> <td style="vertical-align: top;"> <p><a name="bc_p_b"></a><b>bc_p_b</b></p> </td>
	153	<td style="vertical-align: top;">C * 20</td> <td style="vertical-align: top;"><span style="font-style: italic;">'neumann'</span></td>
	154	<td style="vertical-align: top;"> <p style="font-style: normal;">Bottom boundary condition of the
	155	perturbation pressure.  </p> <p>Allowed values
	156	are <span style="font-style: italic;">'dirichlet'</span>,
	157	<span style="font-style: italic;">'neumann'</span>
	158	and <span style="font-style: italic;">'neumann+inhomo'</span>.
	159	<span style="font-style: italic;">'dirichlet'</span>
	160	sets
[5]	161	p(k=0)=0.0,  <span style="font-style: italic;">'neumann'</span>
	162	sets p(k=0)=p(k=1). <span style="font-style: italic;">'neumann+inhomo'</span>
	163	corresponds to an extended Neumann boundary condition where heat flux
	164	or temperature inhomogeneities near the
[61]	165	surface (pt(k=1))  are additionally regarded (see Shen and
	166	LeClerc
[5]	167	(1995, Q.J.R. Meteorol. Soc.,
	168	1209)). This condition is only permitted with the Prandtl-layer
	169	switched on (<a href="#prandtl_layer">prandtl_layer</a>),
[61]	170	otherwise the run is terminated.  </p> <p>Since
	171	at the bottom boundary of the model the vertical
[5]	172	velocity
[61]	173	disappears (w(k=0) = 0.0), the consistent Neumann condition (<span style="font-style: italic;">'neumann'</span> or <span style="font-style: italic;">'neumann+inhomo'</span>)
	174	dp/dz = 0 should
[5]	175	be used, which leaves the vertical component w unchanged when the
	176	pressure solver is applied. Simultaneous use of the Neumann boundary
	177	conditions both at the bottom and at the top boundary (<a href="#bc_p_t">bc_p_t</a>)
	178	usually yields no consistent solution for the perturbation pressure and
[61]	179	should be avoided.</p> </td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="bc_p_t"></a><b>bc_p_t</b></p>
	180	</td> <td style="vertical-align: top;">C * 20</td>
	181	<td style="vertical-align: top;"><span style="font-style: italic;">'dirichlet'</span></td>
	182	<td style="vertical-align: top;"> <p style="font-style: normal;">Top boundary condition of the
	183	perturbation pressure.  </p> <p style="font-style: normal;">Allowed values are <span style="font-style: italic;">'dirichlet'</span>
	184	(p(k=nz+1)= 0.0) or <span style="font-style: italic;">'neumann'</span>
	185	(p(k=nz+1)=p(k=nz)).  </p> <p>Simultaneous use
	186	of Neumann boundary conditions both at the
[5]	187	top and bottom boundary (<a href="#bc_p_b">bc_p_b</a>)
	188	usually yields no consistent solution for the perturbation pressure and
	189	should be avoided. Since at the bottom boundary the Neumann
	190	condition  is a good choice (see <a href="#bc_p_b">bc_p_b</a>),
[61]	191	a Dirichlet condition should be set at the top boundary.</p> </td>
	192	</tr> <tr> <td style="vertical-align: top;">
	193	<p><a name="bc_pt_b"></a><b>bc_pt_b</b></p>
	194	</td> <td style="vertical-align: top;">C*20</td>
	195	<td style="vertical-align: top;"><span style="font-style: italic;">'dirichlet'</span></td>
	196	<td style="vertical-align: top;"> <p style="font-style: normal;">Bottom boundary condition of the
	197	potential temperature.  </p> <p>Allowed values
	198	are <span style="font-style: italic;">'dirichlet'</span>
[5]	199	(pt(k=0) = const. = <a href="#pt_surface">pt_surface</a>
	200	+ <a href="#pt_surface_initial_change">pt_surface_initial_change</a>;
	201	the user may change this value during the run using user-defined code)
	202	and <span style="font-style: italic;">'neumann'</span>
	203	(pt(k=0)=pt(k=1)).  <br>
[61]	204	When a constant surface sensible heat flux is used (<a href="#surface_heatflux">surface_heatflux</a>), <b>bc_pt_b</b>
	205	= <span style="font-style: italic;">'neumann'</span>
[5]	206	must be used, because otherwise the resolved scale may contribute to
	207	the surface flux so that a constant value cannot be guaranteed.</p>
[61]	208	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="pc_pt_t"></a><b>bc_pt_t</b></p>
	209	</td> <td style="vertical-align: top;">C * 20</td>
	210	<td style="vertical-align: top;"><span style="font-style: italic;">'initial gradient'</span></td>
	211	<td style="vertical-align: top;"> <p style="font-style: normal;">Top boundary condition of the
	212	potential temperature.  </p> <p>Allowed are the
	213	values <span style="font-style: italic;">'dirichlet' </span>(pt(k=nz+1)
	214	does not change during the run), <span style="font-style: italic;">'neumann'</span>
	215	(pt(k=nz+1)=pt(k=nz)), and <span style="font-style: italic;">'initial_gradient'</span>.
	216	With the 'initial_gradient'-condition the value of the temperature
	217	gradient at the top is
[5]	218	calculated from the initial
[61]	219	temperature profile (see <a href="#pt_surface">pt_surface</a>,
	220	<a href="#pt_vertical_gradient">pt_vertical_gradient</a>)
[40]	221	by bc_pt_t_val = (pt_init(k=nz+1) -
	222	pt_init(k=nz)) / dzu(nz+1).<br>
[5]	223	Using this value (assumed constant during the
	224	run) the temperature boundary values are calculated as  </p>
[61]	225	<ul> <p style="font-style: normal;">pt(k=nz+1) =
	226	pt(k=nz) +
	227	bc_pt_t_val * dzu(nz+1)</p> </ul> <p style="font-style: normal;">(up to k=nz the prognostic
[40]	228	equation for the temperature is solved).<br>
[61]	229	When a constant sensible heat flux is used at the top boundary (<a href="chapter_4.1.html#top_heatflux">top_heatflux</a>),
	230	<b>bc_pt_t</b> = <span style="font-style: italic;">'neumann'</span>
[40]	231	must be used, because otherwise the resolved scale may contribute to
[61]	232	the top flux so that a constant value cannot be guaranteed.</p> </td>
	233	</tr> <tr> <td style="vertical-align: top;">
	234	<p><a name="bc_q_b"></a><b>bc_q_b</b></p>
	235	</td> <td style="vertical-align: top;">C * 20</td>
	236	<td style="vertical-align: top;"><span style="font-style: italic;">'dirichlet'</span></td>
	237	<td style="vertical-align: top;"> <p style="font-style: normal;">Bottom boundary condition of the
	238	specific humidity / total water content.  </p> <p>Allowed
	239	values are <span style="font-style: italic;">'dirichlet'</span>
[5]	240	(q(k=0) = const. = <a href="#q_surface">q_surface</a>
	241	+ <a href="#q_surface_initial_change">q_surface_initial_change</a>;
	242	the user may change this value during the run using user-defined code)
	243	and <span style="font-style: italic;">'neumann'</span>
	244	(q(k=0)=q(k=1)).  <br>
[61]	245	When a constant surface latent heat flux is used (<a href="#surface_waterflux">surface_waterflux</a>), <b>bc_q_b</b>
	246	= <span style="font-style: italic;">'neumann'</span>
[5]	247	must be used, because otherwise the resolved scale may contribute to
	248	the surface flux so that a constant value cannot be guaranteed.</p>
[61]	249	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="bc_q_t"></a><b>bc_q_t</b></p>
	250	</td> <td style="vertical-align: top;"><span style="font-style: italic;">C
	251	* 20</span></td> <td style="vertical-align: top;"><span style="font-style: italic;">'neumann'</span></td>
	252	<td style="vertical-align: top;"> <p style="font-style: normal;">Top boundary condition of the
	253	specific humidity / total water content.  </p> <p>Allowed
	254	are the values <span style="font-style: italic;">'dirichlet'</span>
[5]	255	(q(k=nz) and q(k=nz+1) do
	256	not change during the run) and <span style="font-style: italic;">'neumann'</span>.
	257	With the Neumann boundary
	258	condition the value of the humidity gradient at the top is calculated
	259	from the
[61]	260	initial humidity profile (see <a href="#q_surface">q_surface</a>,
	261	<a href="#q_vertical_gradient">q_vertical_gradient</a>)
[5]	262	by: bc_q_t_val = ( q_init(k=nz) - q_init(k=nz-1)) / dzu(nz).<br>
	263	Using this value (assumed constant during the run) the humidity
	264	boundary values
[61]	265	are calculated as  </p> <ul> <p style="font-style: normal;">q(k=nz+1) =q(k=nz) +
	266	bc_q_t_val * dzu(nz+1)</p> </ul> <p style="font-style: normal;">(up tp k=nz the prognostic
	267	equation for q is solved). </p> </td> </tr> <tr>
	268	<td style="vertical-align: top;"> <p><a name="bc_s_b"></a><b>bc_s_b</b></p> </td>
	269	<td style="vertical-align: top;">C * 20</td> <td style="vertical-align: top;"><span style="font-style: italic;">'dirichlet'</span></td>
	270	<td style="vertical-align: top;"> <p style="font-style: normal;">Bottom boundary condition of the
	271	scalar concentration.  </p> <p>Allowed values
	272	are <span style="font-style: italic;">'dirichlet'</span>
[5]	273	(s(k=0) = const. = <a href="#s_surface">s_surface</a>
	274	+ <a href="#s_surface_initial_change">s_surface_initial_change</a>;
	275	the user may change this value during the run using user-defined code)
[61]	276	and <span style="font-style: italic;">'neumann'</span>
	277	(s(k=0) =
[5]	278	s(k=1)).  <br>
[61]	279	When a constant surface concentration flux is used (<a href="#surface_scalarflux">surface_scalarflux</a>), <b>bc_s_b</b>
	280	= <span style="font-style: italic;">'neumann'</span>
[5]	281	must be used, because otherwise the resolved scale may contribute to
	282	the surface flux so that a constant value cannot be guaranteed.</p>
[61]	283	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="bc_s_t"></a><b>bc_s_t</b></p>
	284	</td> <td style="vertical-align: top;">C * 20</td>
	285	<td style="vertical-align: top;"><span style="font-style: italic;">'neumann'</span></td>
	286	<td style="vertical-align: top;"> <p style="font-style: normal;">Top boundary condition of the
	287	scalar concentration.  </p> <p>Allowed are the
	288	values <span style="font-style: italic;">'dirichlet'</span>
[5]	289	(s(k=nz) and s(k=nz+1) do
	290	not change during the run) and <span style="font-style: italic;">'neumann'</span>.
	291	With the Neumann boundary
	292	condition the value of the scalar concentration gradient at the top is
	293	calculated
[61]	294	from the initial scalar concentration profile (see <a href="#s_surface">s_surface</a>, <a href="#s_vertical_gradient">s_vertical_gradient</a>)
[5]	295	by: bc_s_t_val = (s_init(k=nz) - s_init(k=nz-1)) / dzu(nz).<br>
	296	Using this value (assumed constant during the run) the concentration
	297	boundary values
[61]	298	are calculated as </p> <ul> <p style="font-style: normal;">s(k=nz+1) = s(k=nz) +
	299	bc_s_t_val * dzu(nz+1)</p> </ul> <p style="font-style: normal;">(up to k=nz the prognostic
[5]	300	equation for the scalar concentration is
[61]	301	solved).</p> </td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="bc_uv_b"></a><b>bc_uv_b</b></p>
	302	</td> <td style="vertical-align: top;">C * 20</td>
	303	<td style="vertical-align: top;"><span style="font-style: italic;">'dirichlet'</span></td>
	304	<td style="vertical-align: top;"> <p style="font-style: normal;">Bottom boundary condition of the
	305	horizontal velocity components u and v.  </p> <p>Allowed
	306	values are <span style="font-style: italic;">'dirichlet' </span>and
	307	<span style="font-style: italic;">'neumann'</span>. <b>bc_uv_b</b>
	308	= <span style="font-style: italic;">'dirichlet'</span>
	309	yields the
[5]	310	no-slip condition with u=v=0 at the bottom. Due to the staggered grid
	311	u(k=0) and v(k=0) are located at z = - 0,5 * <a href="#dz">dz</a>
	312	(below the bottom), while u(k=1) and v(k=1) are located at z = +0,5 *
	313	dz. u=v=0 at the bottom is guaranteed using mirror boundary
[61]	314	condition:  </p> <ul> <p style="font-style: normal;">u(k=0) = - u(k=1) and v(k=0) = -
	315	v(k=1)</p> </ul> <p style="font-style: normal;">The
	316	Neumann boundary condition
[5]	317	yields the free-slip condition with u(k=0) = u(k=1) and v(k=0) =
	318	v(k=1).
	319	With Prandtl - layer switched on, the free-slip condition is not
	320	allowed (otherwise the run will be terminated)<font color="#000000">.</font></p>
[61]	321	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="bc_uv_t"></a><b>bc_uv_t</b></p>
	322	</td> <td style="vertical-align: top;">C * 20</td>
	323	<td style="vertical-align: top;"><span style="font-style: italic;">'dirichlet'</span></td>
	324	<td style="vertical-align: top;"> <p style="font-style: normal;">Top boundary condition of the
	325	horizontal velocity components u and v.  </p> <p>Allowed
	326	values are <span style="font-style: italic;">'dirichlet'</span>
	327	and <span style="font-style: italic;">'neumann'</span>.
	328	The
[5]	329	Dirichlet condition yields u(k=nz+1) = ug(nz+1) and v(k=nz+1) =
	330	vg(nz+1),
	331	Neumann condition yields the free-slip condition with u(k=nz+1) =
	332	u(k=nz) and v(k=nz+1) = v(k=nz) (up to k=nz the prognostic equations
[61]	333	for the velocities are solved).</p> </td> </tr> <tr>
	334	<td style="vertical-align: top;"><span style="font-weight: bold;"><a name="building_height"></a>building_height</span></td>
	335	<td style="vertical-align: top;">R</td> <td style="vertical-align: top;"><span style="font-style: italic;">50.0</span></td> <td>Height
	336	of a single building in m.<br> <br> <span style="font-weight: bold;">building_height</span> must
	337	be less than the height of the model domain. This parameter requires
	338	the use of <a href="#topography">topography</a>
	339	= <span style="font-style: italic;">'single_building'</span>.</td>
	340	</tr> <tr> <td style="vertical-align: top;"><span style="font-weight: bold;"><a name="building_length_x"></a>building_length_x</span></td>
	341	<td style="vertical-align: top;">R</td> <td style="vertical-align: top;"><span style="font-style: italic;">50.0</span></td> <td><span style="font-style: italic;"></span>Width of a single
	342	building in m.<br> <br>
	343	Currently, <span style="font-weight: bold;">building_length_x</span>
	344	must be at least <span style="font-style: italic;">3
	345	* </span><a style="font-style: italic;" href="#dx">dx</a> and no more than <span style="font-style: italic;">( </span><a style="font-style: italic;" href="#nx">nx</a><span style="font-style: italic;"> - 1 ) </span><span style="font-style: italic;"> * <a href="#dx">dx</a>
	346	</span><span style="font-style: italic;">- <a href="#building_wall_left">building_wall_left</a></span>.
	347	This parameter requires the use of <a href="#topography">topography</a>
	348	= <span style="font-style: italic;">'single_building'</span>.</td>
	349	</tr> <tr> <td style="vertical-align: top;"><span style="font-weight: bold;"><a name="building_length_y"></a>building_length_y</span></td>
	350	<td style="vertical-align: top;">R</td> <td style="vertical-align: top;"><span style="font-style: italic;">50.0</span></td> <td>Depth
	351	of a single building in m.<br> <br>
	352	Currently, <span style="font-weight: bold;">building_length_y</span>
	353	must be at least <span style="font-style: italic;">3
	354	* </span><a style="font-style: italic;" href="#dy">dy</a> and no more than <span style="font-style: italic;">( </span><a style="font-style: italic;" href="#ny">ny</a><span style="font-style: italic;"> - 1 ) </span><span style="font-style: italic;"> * <a href="#dy">dy</a></span><span style="font-style: italic;"> - <a href="#building_wall_south">building_wall_south</a></span>. This parameter requires
	355	the use of <a href="#topography">topography</a>
	356	= <span style="font-style: italic;">'single_building'</span>.</td>
	357	</tr> <tr> <td style="vertical-align: top;"><span style="font-weight: bold;"><a name="building_wall_left"></a>building_wall_left</span></td>
	358	<td style="vertical-align: top;">R</td> <td style="vertical-align: top;"><span style="font-style: italic;">building centered in x-direction</span></td>
	359	<td>x-coordinate of the left building wall (distance between the
	360	left building wall and the left border of the model domain) in m.<br>
	361	<br>
	362	Currently, <span style="font-weight: bold;">building_wall_left</span>
	363	must be at least <span style="font-style: italic;">1
	364	* </span><a style="font-style: italic;" href="#dx">dx</a> and less than <span style="font-style: italic;">( <a href="#nx">nx</a>
	365	- 1 ) * <a href="#dx">dx</a> -  <a href="#building_length_x">building_length_x</a></span>.
	366	This parameter requires the use of <a href="#topography">topography</a>
	367	= <span style="font-style: italic;">'single_building'</span>.<br>
	368	<br>
	369	The default value <span style="font-weight: bold;">building_wall_left</span>
	370	= <span style="font-style: italic;">( ( <a href="#nx">nx</a> +
	371	1 ) * <a href="#dx">dx</a> -  <a href="#building_length_x">building_length_x</a> ) / 2</span>
	372	centers the building in x-direction. </td> </tr> <tr>
	373	<td style="vertical-align: top;"><span style="font-weight: bold;"><a name="building_wall_south"></a>building_wall_south</span></td>
	374	<td style="vertical-align: top;">R</td> <td style="vertical-align: top;"><span style="font-style: italic;"></span><span style="font-style: italic;">building centered in y-direction</span></td>
	375	<td>y-coordinate of the South building wall (distance between the
[5]	376	South building wall and the South border of the model domain) in m.<br>
[61]	377	<br>
	378	Currently, <span style="font-weight: bold;">building_wall_south</span>
	379	must be at least <span style="font-style: italic;">1
	380	* </span><a style="font-style: italic;" href="#dy">dy</a> and less than <span style="font-style: italic;">( <a href="#ny">ny</a>
	381	- 1 ) * <a href="#dy">dy</a> -  <a href="#building_length_y">building_length_y</a></span>.
	382	This parameter requires the use of <a href="#topography">topography</a>
	383	= <span style="font-style: italic;">'single_building'</span>.<br>
	384	<br>
	385	The default value <span style="font-weight: bold;">building_wall_south</span>
	386	= <span style="font-style: italic;">( ( <a href="#ny">ny</a> +
	387	1 ) * <a href="#dy">dy</a> -  <a href="#building_length_y">building_length_y</a> ) / 2</span>
	388	centers the building in y-direction. </td> </tr> <tr>
	389	<td style="vertical-align: top;"><span style="font-weight: bold;"><a name="cloud_droplets"></a>cloud_droplets</span><br>
	390	</td> <td style="vertical-align: top;">L<br> </td>
	391	<td style="vertical-align: top;"><span style="font-style: italic;">.F.</span><br> </td>
	392	<td style="vertical-align: top;">Parameter to switch on
	393	usage of cloud droplets.<br> <br>
[5]	394	Cloud droplets require to use the particle package (<span style="font-weight: bold;">mrun</span>-option <span style="font-family: monospace;">-p particles</span>),
	395	so in this case a particle corresponds to a droplet. The droplet
	396	features (number of droplets, initial radius, etc.) can be steered with
	397	the  respective particle parameters (see e.g. <a href="#chapter_4.2.html#radius">radius</a>).
	398	The real number of initial droplets in a grid cell is equal to the
	399	initial number of droplets (defined by the particle source parameters <span lang="en-GB"><font face="Thorndale, serif"> </font></span><a href="chapter_4.2.html#pst"><span lang="en-GB"><font face="Thorndale, serif">pst</font></span></a><span lang="en-GB"><font face="Thorndale, serif">, </font></span><a href="chapter_4.2.html#psl"><span lang="en-GB"><font face="Thorndale, serif">psl</font></span></a><span lang="en-GB"><font face="Thorndale, serif">, </font></span><a href="chapter_4.2.html#psr"><span lang="en-GB"><font face="Thorndale, serif">psr</font></span></a><span lang="en-GB"><font face="Thorndale, serif">, </font></span><a href="chapter_4.2.html#pss"><span lang="en-GB"><font face="Thorndale, serif">pss</font></span></a><span lang="en-GB"><font face="Thorndale, serif">, </font></span><a href="chapter_4.2.html#psn"><span lang="en-GB"><font face="Thorndale, serif">psn</font></span></a><span lang="en-GB"><font face="Thorndale, serif">, </font></span><a href="chapter_4.2.html#psb"><span lang="en-GB"><font face="Thorndale, serif">psb</font></span></a><span lang="en-GB"><font face="Thorndale, serif">, </font></span><a href="chapter_4.2.html#pdx"><span lang="en-GB"><font face="Thorndale, serif">pdx</font></span></a><span lang="en-GB"><font face="Thorndale, serif">, </font></span><a href="chapter_4.2.html#pdy"><span lang="en-GB"><font face="Thorndale, serif">pdy</font></span></a>
[61]	400	<span lang="en-GB"><font face="Thorndale, serif">and
	401	</font></span><a href="chapter_4.2.html#pdz"><span lang="en-GB"><font face="Thorndale, serif">pdz</font></span></a><span lang="en-GB"></span><span lang="en-GB"></span>)
	402	times the <a href="#initial_weighting_factor">initial_weighting_factor</a>.<br>
	403	<br>
	404	In case of using cloud droplets, the default condensation scheme in
	405	PALM cannot be used, i.e. <a href="#cloud_physics">cloud_physics</a>
	406	must be set <span style="font-style: italic;">.F.</span>.<br>
	407	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="cloud_physics"></a><b>cloud_physics</b></p>
	408	</td> <td style="vertical-align: top;">L<br> </td>
	409	<td style="vertical-align: top;"><span style="font-style: italic;">.F.</span></td> <td style="vertical-align: top;"> <p>Parameter to switch
	410	on the condensation scheme.  </p>
	411	For <b>cloud_physics =</b> <span style="font-style: italic;">.TRUE.</span>, equations
	412	for the
[5]	413	liquid water
	414	content and the liquid water potential temperature are solved instead
	415	of those for specific humidity and potential temperature. Note
	416	that a grid volume is assumed to be either completely saturated or
	417	completely
	418	unsaturated (0%-or-100%-scheme). A simple precipitation scheme can
	419	additionally be switched on with parameter <a href="#precipitation">precipitation</a>.
[61]	420	Also cloud-top cooling by longwave radiation can be utilized (see <a href="#radiation">radiation</a>)<br> <b><br>
	421	cloud_physics =</b> <span style="font-style: italic;">.TRUE.
[75]	422	</span>requires <a href="#humidity">humidity</a>
[61]	423	=<span style="font-style: italic;"> .TRUE.</span> .<br>
[5]	424	Detailed information about the condensation scheme is given in the
	425	description of the <a href="http://www.muk.uni-hannover.de/%7Eraasch/PALM-1/Dokumentationen/Cloud_physics/wolken.pdf">cloud
[61]	426	physics module</a> (pdf-file, only in German).<br> <br>
	427	This condensation scheme is not allowed if cloud droplets are simulated
	428	explicitly (see <a href="#cloud_droplets">cloud_droplets</a>).<br>
	429	</td> </tr> <tr> <td style="vertical-align: top;"><span style="font-weight: bold;"><a name="conserve_volume_flow"></a>conserve_volume_flow</span></td>
	430	<td style="vertical-align: top;">L</td> <td style="vertical-align: top;"><span style="font-style: italic;">.F.</span></td> <td>Conservation
	431	of volume flow in x- and y-direction.<br> <br> <span style="font-weight: bold;">conserve_volume_flow</span>
	432	= <span style="font-style: italic;">.TRUE.</span>
[5]	433	guarantees that the volume flow through the xz- or yz-cross-section of
	434	the total model domain remains constant (equal to the initial value at
	435	t=0) throughout the run.<br>
[61]	436	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="cut_spline_overshoot"></a><b>cut_spline_overshoot</b></p>
	437	</td> <td style="vertical-align: top;">L</td>
	438	<td style="vertical-align: top;"><span style="font-style: italic;">.T.</span></td> <td style="vertical-align: top;"> <p>Cuts off of
	439	so-called overshoots, which can occur with the
	440	upstream-spline scheme.  </p> <p><font color="#000000">The cubic splines tend to overshoot in
[5]	441	case of discontinuous changes of variables between neighbouring grid
	442	points.</font><font color="#ff0000"> </font><font color="#000000">This
[61]	443	may lead to errors in calculating the advection tendency.</font>
	444	Choice
	445	of <b>cut_spline_overshoot</b> = <i>.TRUE.</i>
	446	(switched on by
[5]	447	default)
	448	allows variable values not to exceed an interval defined by the
	449	respective adjacent grid points. This interval can be adjusted
	450	seperately for every prognostic variable (see initialization parameters
[61]	451	<a href="#overshoot_limit_e">overshoot_limit_e</a>, <a href="#overshoot_limit_pt">overshoot_limit_pt</a>, <a href="#overshoot_limit_u">overshoot_limit_u</a>,
[5]	452	etc.). This might be necessary in case that the
	453	default interval has a non-tolerable effect on the model
[61]	454	results.  </p> <p>Overshoots may also be removed
	455	using the parameters <a href="#ups_limit_e">ups_limit_e</a>,
	456	<a href="#ups_limit_pt">ups_limit_pt</a>,
[5]	457	etc. as well as by applying a long-filter (see <a href="#long_filter_factor">long_filter_factor</a>).</p>
[61]	458	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="damp_level_1d"></a><b>damp_level_1d</b></p>
	459	</td> <td style="vertical-align: top;">R</td>
	460	<td style="vertical-align: top;"><span style="font-style: italic;">zu(nz+1)</span></td>
	461	<td style="vertical-align: top;"> <p>Height where
	462	the damping layer begins in the 1d-model
	463	(in m).  </p> <p>This parameter is used to
	464	switch on a damping layer for the
[5]	465	1d-model, which is generally needed for the damping of inertia
	466	oscillations. Damping is done by gradually increasing the value
	467	of the eddy diffusivities about 10% per vertical grid level
	468	(starting with the value at the height given by <b>damp_level_1d</b>,
[61]	469	or possibly from the next grid pint above), i.e. K<sub>m</sub>(k+1)
	470	=
[5]	471	1.1 * K<sub>m</sub>(k).
[61]	472	The values of K<sub>m</sub> are limited to 10 m**2/s at
	473	maximum.  <br>
[5]	474	This parameter only comes into effect if the 1d-model is switched on
	475	for
	476	the initialization of the 3d-model using <a href="#initializing_actions">initializing_actions</a>
[61]	477	= <span style="font-style: italic;">'set_1d-model_profiles'</span>.
	478	<br> </p> </td> </tr> <tr> <td style="vertical-align: top;"><a name="dissipation_1d"></a><span style="font-weight: bold;">dissipation_1d</span><br>
	479	</td> <td style="vertical-align: top;">C*20<br>
	480	</td> <td style="vertical-align: top;"><span style="font-style: italic;">'as_in_3d_</span><br style="font-style: italic;"> <span style="font-style: italic;">model'</span><br> </td>
	481	<td style="vertical-align: top;">Calculation method for
	482	the energy dissipation term in the TKE equation of the 1d-model.<br>
	483	<br>
	484	By default the dissipation is calculated as in the 3d-model using diss
	485	= (0.19 + 0.74 * l / l_grid) * e**1.5 / l.<br> <br>
	486	Setting <span style="font-weight: bold;">dissipation_1d</span>
	487	= <span style="font-style: italic;">'detering'</span>
	488	forces the dissipation to be calculated as diss = 0.064 * e**1.5 / l.<br>
	489	</td> </tr>
	490	<tr> <td style="vertical-align: top;"> <p><a name="dt"></a><b>dt</b></p> </td>
	491	<td style="vertical-align: top;">R</td> <td style="vertical-align: top;"><span style="font-style: italic;">variable</span></td>
	492	<td style="vertical-align: top;"> <p>Time step for
	493	the 3d-model (in s).  </p> <p>By default, (i.e.
	494	if a Runge-Kutta scheme is used, see <a href="#timestep_scheme">timestep_scheme</a>)
[5]	495	the value of the time step is calculating after each time step
	496	(following the time step criteria) and
[61]	497	used for the next step.</p> <p>If the user assigns <b>dt</b>
	498	a value, then the time step is
[5]	499	fixed to this value throughout the whole run (whether it fulfills the
	500	time step
	501	criteria or not). However, changes are allowed for restart runs,
	502	because <b>dt</b> can also be used as a <a href="chapter_4.2.html#dt_laufparameter">run
[61]	503	parameter</a>.  </p> <p>In case that the
	504	calculated time step meets the condition<br> </p> <ul>
	505	<p><b>dt</b> < 0.00001 * <a href="chapter_4.2.html#dt_max">dt_max</a> (with dt_max
	506	= 20.0)</p> </ul> <p>the simulation will be
	507	aborted. Such situations usually arise
[5]	508	in case of any numerical problem / instability which causes a
[61]	509	non-realistic increase of the wind speed.  </p> <p>A
	510	small time step due to a large mean horizontal windspeed
[5]	511	speed may be enlarged by using a coordinate transformation (see <a href="#galilei_transformation">galilei_transformation</a>),
[61]	512	in order to spare CPU time.<br> </p> <p>If the
	513	leapfrog timestep scheme is used (see <a href="#timestep_scheme">timestep_scheme</a>)
[5]	514	a temporary time step value dt_new is calculated first, with dt_new = <a href="chapter_4.2.html#fcl_factor">cfl_factor</a>
	515	* dt_crit where dt_crit is the maximum timestep allowed by the CFL and
	516	diffusion condition. Next it is examined whether dt_new exceeds or
	517	falls below the
	518	value of the previous timestep by at
	519	least +5 % / -2%. If it is smaller, <span style="font-weight: bold;">dt</span>
	520	= dt_new is immediately used for the next timestep. If it is larger,
[61]	521	then <span style="font-weight: bold;">dt </span>=
	522	1.02 * dt_prev
[5]	523	(previous timestep) is used as the new timestep, however the time
	524	step is only increased if the last change of the time step is dated
	525	back at
	526	least 30 iterations. If dt_new is located in the interval mentioned
	527	above, then dt
	528	does not change at all. By doing so, permanent time step changes as
	529	well as large
[61]	530	sudden changes (increases) in the time step are avoided.</p> </td>
	531	</tr> <tr> <td style="vertical-align: top;">
	532	<p><a name="dt_pr_1d"></a><b>dt_pr_1d</b></p>
	533	</td> <td style="vertical-align: top;">R</td>
	534	<td style="vertical-align: top;"><span style="font-style: italic;">9999999.9</span></td>
	535	<td style="vertical-align: top;"> <p>Temporal
	536	interval of vertical profile output of the 1D-model
	537	(in s).  </p> <p>Data are written in ASCII
	538	format to file <a href="chapter_3.4.html#LIST_PROFIL_1D">LIST_PROFIL_1D</a>.
[5]	539	This parameter is only in effect if the 1d-model has been switched on
	540	for the
	541	initialization of the 3d-model with <a href="#initializing_actions">initializing_actions</a>
	542	= <span style="font-style: italic;">'set_1d-model_profiles'</span>.</p>
[61]	543	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="dt_run_control_1d"></a><b>dt_run_control_1d</b></p>
	544	</td> <td style="vertical-align: top;">R</td>
	545	<td style="vertical-align: top;"><span style="font-style: italic;">60.0</span></td> <td style="vertical-align: top;"> <p>Temporal interval of
	546	runtime control output of the 1d-model
	547	(in s).  </p> <p>Data are written in ASCII
	548	format to file <a href="chapter_3.4.html#RUN_CONTROL">RUN_CONTROL</a>.
[5]	549	This parameter is only in effect if the 1d-model is switched on for the
	550	initialization of the 3d-model with <a href="#initializing_actions">initializing_actions</a>
	551	= <span style="font-style: italic;">'set_1d-model_profiles'</span>.</p>
[61]	552	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="dx"></a><b>dx</b></p>
	553	</td> <td style="vertical-align: top;">R</td>
	554	<td style="vertical-align: top;"><span style="font-style: italic;">1.0</span></td> <td style="vertical-align: top;"> <p>Horizontal grid
	555	spacing along the x-direction (in m).  </p> <p>Along
	556	x-direction only a constant grid spacing is allowed.</p> </td>
	557	</tr> <tr> <td style="vertical-align: top;">
	558	<p><a name="dy"></a><b>dy</b></p>
	559	</td> <td style="vertical-align: top;">R</td>
	560	<td style="vertical-align: top;"><span style="font-style: italic;">1.0</span></td> <td style="vertical-align: top;"> <p>Horizontal grid
[83]	561	spacing along the y-direction (in m).  </p> <p>Along y-direction only a constant grid spacing is allowed.</p> </td>
[61]	562	</tr> <tr> <td style="vertical-align: top;">
	563	<p><a name="dz"></a><b>dz</b></p>
	564	</td> <td style="vertical-align: top;">R</td>
	565	<td style="vertical-align: top;"><br> </td> <td style="vertical-align: top;"> <p>Vertical grid
	566	spacing (in m).  </p> <p>This parameter must be
	567	assigned by the user, because no
	568	default value is given.<br> </p> <p>By default, the
[5]	569	model uses constant grid spacing along z-direction, but it can be
	570	stretched using the parameters <a href="#dz_stretch_level">dz_stretch_level</a>
[61]	571	and <a href="#dz_stretch_factor">dz_stretch_factor</a>.
	572	In case of stretching, a maximum allowed grid spacing can be given by <a href="#dz_max">dz_max</a>.<br> </p> <p>Assuming
	573	a constant <span style="font-weight: bold;">dz</span>,
[5]	574	the scalar levels (zu) are calculated directly by:  </p>
[61]	575	<ul> <p>zu(0) = - dz * 0.5  <br>
	576	zu(1) = dz * 0.5</p> </ul> <p>The w-levels lie
	577	half between them:  </p> <ul> <p>zw(k) =
	578	( zu(k) + zu(k+1) ) * 0.5</p> </ul> </td> </tr>
	579	<tr><td style="vertical-align: top;"><a name="dz_max"></a><span style="font-weight: bold;">dz_max</span></td><td style="vertical-align: top;">R</td><td style="vertical-align: top;"><span style="font-style: italic;">9999999.9</span></td><td style="vertical-align: top;">Allowed maximum vertical grid
	580	spacing (in m).<br><br>If the vertical grid is stretched
	581	(see <a href="#dz_stretch_factor">dz_stretch_factor</a>
	582	and <a href="#dz_stretch_level">dz_stretch_level</a>),
	583	<span style="font-weight: bold;">dz_max</span> can
	584	be used to limit the vertical grid spacing.</td></tr><tr>
	585	<td style="vertical-align: top;"> <p><a name="dz_stretch_factor"></a><b>dz_stretch_factor</b></p>
	586	</td> <td style="vertical-align: top;">R</td>
	587	<td style="vertical-align: top;"><span style="font-style: italic;">1.08</span></td> <td style="vertical-align: top;"> <p>Stretch factor for a
	588	vertically stretched grid (see <a href="#dz_stretch_level">dz_stretch_level</a>).
	589	</p> <p>The stretch factor should not exceed a value of
	590	approx. 1.10 -
[5]	591	1.12, otherwise the discretization errors due to the stretched grid not
[61]	592	negligible any more. (refer Kalnay de Rivas)</p> </td> </tr>
	593	<tr> <td style="vertical-align: top;"> <p><a name="dz_stretch_level"></a><b>dz_stretch_level</b></p>
	594	</td> <td style="vertical-align: top;">R</td>
	595	<td style="vertical-align: top;"><span style="font-style: italic;">100000.0</span><br> </td>
	596	<td style="vertical-align: top;"> <p>Height level
	597	above which the grid is to be stretched
	598	vertically (in m).  </p> <p>The vertical grid
	599	spacings <a href="#dz">dz</a>
	600	above this level are calculated as  </p> <ul> <p><b>dz</b>(k+1)
	601	= <b>dz</b>(k) * <a href="#dz_stretch_factor">dz_stretch_factor</a></p>
	602	</ul> <p>and used as spacings for the scalar levels (zu).
	603	The
	604	w-levels are then defined as:  </p> <ul> <p>zw(k)
	605	= ( zu(k) + zu(k+1) ) * 0.5</p> </ul> </td> </tr>
	606	<tr> <td style="vertical-align: top;"><span style="font-weight: bold;"><a name="e_min"></a>e_min</span></td>
	607	<td style="vertical-align: top;">R</td> <td style="vertical-align: top;"><span style="font-style: italic;">0.0</span></td> <td>Minimum
	608	subgrid-scale TKE in m<sup>2</sup>s<sup>-2</sup>.<br>
	609	<br>This
	610	option adds artificial viscosity to the flow by ensuring that
	611	the
	612	subgrid-scale TKE does not fall below the minimum threshold <span style="font-weight: bold;">e_min</span>.</td> </tr>
	613	<tr> <td style="vertical-align: top;"> <p><a name="end_time_1d"></a><b>end_time_1d</b></p>
	614	</td> <td style="vertical-align: top;">R</td>
	615	<td style="vertical-align: top;"><span style="font-style: italic;">864000.0</span><br> </td>
	616	<td style="vertical-align: top;"> <p>Time to be
	617	simulated for the 1d-model (in s).  </p> <p>The
	618	default value corresponds to a simulated time of 10 days.
[5]	619	Usually, after such a period the inertia oscillations have completely
	620	decayed and the solution of the 1d-model can be regarded as stationary
	621	(see <a href="#damp_level_1d">damp_level_1d</a>).
	622	This parameter is only in effect if the 1d-model is switched on for the
	623	initialization of the 3d-model with <a href="#initializing_actions">initializing_actions</a>
	624	= <span style="font-style: italic;">'set_1d-model_profiles'</span>.</p>
[61]	625	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="fft_method"></a><b>fft_method</b></p>
	626	</td> <td style="vertical-align: top;">C * 20</td>
	627	<td style="vertical-align: top;"><span style="font-style: italic;">'system-</span><br style="font-style: italic;"> <span style="font-style: italic;">specific'</span></td>
	628	<td style="vertical-align: top;"> <p>FFT-method to
	629	be used.<br> </p> <p><br>
[5]	630	The fast fourier transformation (FFT) is used for solving the
	631	perturbation pressure equation with a direct method (see <a href="chapter_4.2.html#psolver">psolver</a>)
	632	and for calculating power spectra (see optional software packages,
	633	section <a href="chapter_4.2.html#spectra_package">4.2</a>).</p>
[61]	634	<p><br>
[5]	635	By default, system-specific, optimized routines from external
	636	vendor libraries are used. However, these are available only on certain
	637	computers and there are more or less severe restrictions concerning the
[61]	638	number of gridpoints to be used with them.<br> </p> <p>There
	639	are two other PALM internal methods available on every
[5]	640	machine (their respective source code is part of the PALM source code):</p>
[61]	641	<p>1.: The <span style="font-weight: bold;">Temperton</span>-method
[5]	642	from Clive Temperton (ECWMF) which is computationally very fast and
	643	switched on with <b>fft_method</b> = <span style="font-style: italic;">'temperton-algorithm'</span>.
	644	The number of horizontal gridpoints (nx+1, ny+1) to be used with this
[61]	645	method must be composed of prime factors 2, 3 and 5.<br> </p>
[5]	646	2.: The <span style="font-weight: bold;">Singleton</span>-method
	647	which is very slow but has no restrictions concerning the number of
[61]	648	gridpoints to be used with, switched on with <b>fft_method</b>
	649	= <span style="font-style: italic;">'singleton-algorithm'</span>.
	650	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="galilei_transformation"></a><b>galilei_transformation</b></p>
	651	</td> <td style="vertical-align: top;">L</td>
	652	<td style="vertical-align: top;"><i>.F.</i></td>
	653	<td style="vertical-align: top;">Application of a
	654	Galilei-transformation to the
[5]	655	coordinate
[61]	656	system of the model.<br><p>With <b>galilei_transformation</b>
	657	= <i>.T.,</i> a so-called
[5]	658	Galilei-transformation is switched on which ensures that the coordinate
	659	system of the model is moved along with the geostrophical wind.
	660	Alternatively, the model domain can be moved along with the averaged
	661	horizontal wind (see <a href="#use_ug_for_galilei_tr">use_ug_for_galilei_tr</a>,
	662	this can and will naturally change in time). With this method,
	663	numerical inaccuracies of the Piascek - Williams - scheme (concerns in
	664	particular the momentum advection) are minimized. Beyond that, in the
	665	majority of cases the lower relative velocities in the moved system
	666	permit a larger time step (<a href="#dt">dt</a>).
	667	Switching the transformation on is only worthwhile if the geostrophical
	668	wind (ug, vg)
	669	and the averaged horizontal wind clearly deviate from the value 0. In
	670	each case, the distance the coordinate system has been moved is written
	671	to the file <a href="chapter_3.4.html#RUN_CONTROL">RUN_CONTROL</a>.
[61]	672	</p> <p>Non-cyclic lateral boundary conditions (see <a href="#bc_lr">bc_lr</a>
	673	and <a href="#bc_ns">bc_ns</a>), the specification
	674	of a gestrophic
[5]	675	wind that is not constant with height
	676	as well as e.g. stationary inhomogeneities at the bottom boundary do
[61]	677	not allow the use of this transformation.</p> </td> </tr>
	678	<tr> <td style="vertical-align: top;"> <p><a name="grid_matching"></a><b>grid_matching</b></p>
	679	</td> <td style="vertical-align: top;">C * 6</td>
	680	<td style="vertical-align: top;"><span style="font-style: italic;">'match'</span></td> <td style="vertical-align: top;">Variable to adjust the
	681	subdomain
	682	sizes in parallel runs.<br> <br>
[5]	683	For <b>grid_matching</b> = <span style="font-style: italic;">'strict'</span>,
	684	the subdomains are forced to have an identical
	685	size on all processors. In this case the processor numbers in the
	686	respective directions of the virtual processor net must fulfill certain
	687	divisor conditions concerning the grid point numbers in the three
[61]	688	directions (see <a href="#nx">nx</a>, <a href="#ny">ny</a>
[5]	689	and <a href="#nz">nz</a>).
	690	Advantage of this method is that all PEs bear the same computational
[61]	691	load.<br> <br>
[5]	692	There is no such restriction by default, because then smaller
	693	subdomains are allowed on those processors which
	694	form the right and/or north boundary of the virtual processor grid. On
	695	all other processors the subdomains are of same size. Whether smaller
	696	subdomains are actually used, depends on the number of processors and
	697	the grid point numbers used. Information about the respective settings
	698	are given in file <a href="file:///home/raasch/public_html/PALM_group/home/raasch/public_html/PALM_group/doc/app/chapter_3.4.html#RUN_CONTROL">RUN_CONTROL</a>.<br>
[61]	699	<br>
[5]	700	When using a multi-grid method for solving the Poisson equation (see <a href="http://www.muk.uni-hannover.de/%7Eraasch/PALM_group/doc/app/chapter_4.2.html#psolver">psolver</a>)
	701	only <b>grid_matching</b> = <span style="font-style: italic;">'strict'</span>
[61]	702	is allowed.<br> <br> <b>Note:</b><br>
[5]	703	In some cases for small processor numbers there may be a very bad load
	704	balancing among the
[61]	705	processors which may reduce the performance of the code.</td> </tr>
	706	<tr> <td style="vertical-align: top;"><a name="inflow_disturbance_begin"></a><b>inflow_disturbance_<br>
	707	begin</b></td> <td style="vertical-align: top;">I</td>
	708	<td style="vertical-align: top;"><span style="font-style: italic;">MIN(10,</span><br style="font-style: italic;"> <span style="font-style: italic;">nx/2 or ny/2)</span></td>
	709	<td style="vertical-align: top;">Lower
[5]	710	limit of the horizontal range for which random perturbations are to be
[61]	711	imposed on the horizontal velocity field (gridpoints).<br> <br>
[5]	712	If non-cyclic lateral boundary conditions are used (see <a href="#bc_lr">bc_lr</a>
	713	or <a href="#bc_ns">bc_ns</a>),
	714	this parameter gives the gridpoint number (counted horizontally from
	715	the inflow)  from which on perturbations are imposed on the
	716	horizontal velocity field. Perturbations must be switched on with
	717	parameter <a href="chapter_4.2.html#create_disturbances">create_disturbances</a>.</td>
[61]	718	</tr> <tr> <td style="vertical-align: top;"><a name="inflow_disturbance_end"></a><b>inflow_disturbance_<br>
	719	end</b></td> <td style="vertical-align: top;">I</td>
	720	<td style="vertical-align: top;"><span style="font-style: italic;">MIN(100,</span><br style="font-style: italic;"> <span style="font-style: italic;">3/4nx or</span><br style="font-style: italic;"> <span style="font-style: italic;">3/4ny)</span></td> <td style="vertical-align: top;">Upper
[5]	721	limit of the horizontal range for which random perturbations are
[61]	722	to be imposed on the horizontal velocity field (gridpoints).<br> <br>
[5]	723	If non-cyclic lateral boundary conditions are used (see <a href="#bc_lr">bc_lr</a>
	724	or <a href="#bc_ns">bc_ns</a>),
	725	this parameter gives the gridpoint number (counted horizontally from
	726	the inflow)  unto which perturbations are imposed on the
	727	horizontal
	728	velocity field. Perturbations must be switched on with parameter <a href="chapter_4.2.html#create_disturbances">create_disturbances</a>.</td>
[61]	729	</tr> <tr> <td style="vertical-align: top;">
	730	<p><a name="initializing_actions"></a><b>initializing_actions</b></p>
	731	</td> <td style="vertical-align: top;">C * 100</td>
	732	<td style="vertical-align: top;"><br> </td> <td style="vertical-align: top;"> <p style="font-style: normal;">Initialization actions
	733	to be carried out.  </p> <p style="font-style: normal;">This parameter does not have a
[5]	734	default value and therefore must be assigned with each model run. For
	735	restart runs <b>initializing_actions</b> = <span style="font-style: italic;">'read_restart_data'</span>
	736	must be set. For the initial run of a job chain the following values
[61]	737	are allowed:  </p> <p style="font-style: normal;"><span style="font-style: italic;">'set_constant_profiles'</span>
	738	</p> <ul> <p>A horizontal wind profile consisting
	739	of linear sections (see <a href="#ug_surface">ug_surface</a>,
	740	<a href="#ug_vertical_gradient">ug_vertical_gradient</a>,
	741	<a href="#ug_vertical_gradient_level">ug_vertical_gradient_level</a>
	742	and <a href="#vg_surface">vg_surface</a>, <a href="#vg_vertical_gradient">vg_vertical_gradient</a>,
	743	<a href="#vg_vertical_gradient_level">vg_vertical_gradient_level</a>,
[5]	744	respectively) as well as a vertical temperature (humidity) profile
	745	consisting of
[61]	746	linear sections (see <a href="#pt_surface">pt_surface</a>,
	747	<a href="#pt_vertical_gradient">pt_vertical_gradient</a>,
	748	<a href="#q_surface">q_surface</a>
[5]	749	and <a href="#q_vertical_gradient">q_vertical_gradient</a>)
	750	are assumed as initial profiles. The subgrid-scale TKE is set to 0 but K<sub>m</sub>
[61]	751	and K<sub>h</sub> are set to very small values because
	752	otherwise no TKE
	753	would be generated.</p> </ul> <p style="font-style: italic;">'set_1d-model_profiles' </p>
	754	<ul> <p>The arrays of the 3d-model are initialized with
	755	the
[5]	756	(stationary) solution of the 1d-model. These are the variables e, kh,
	757	km, u, v and with Prandtl layer switched on rif, us, usws, vsws. The
	758	temperature (humidity) profile consisting of linear sections is set as
	759	for 'set_constant_profiles' and assumed as constant in time within the
	760	1d-model. For steering of the 1d-model a set of parameters with suffix
[61]	761	"_1d" (e.g. <a href="#end_time_1d">end_time_1d</a>,
	762	<a href="#damp_level_1d">damp_level_1d</a>)
	763	is available.</p> </ul> <p><span style="font-style: italic;">'by_user'</span></p><p style="margin-left: 40px;">The initialization of the arrays
	764	of the 3d-model is under complete control of the user and has to be
	765	done in routine <a href="chapter_3.5.1.html#user_init_3d_model">user_init_3d_model</a>
	766	of the user-interface.<span style="font-style: italic;"></span></p><p><span style="font-style: italic;">'initialize_vortex'</span>
	767	</p> <div style="margin-left: 40px;">The initial
	768	velocity field of the
[5]	769	3d-model corresponds to a
	770	Rankine-vortex with vertical axis. This setting may be used to test
	771	advection schemes. Free-slip boundary conditions for u and v (see <a href="#bc_uv_b">bc_uv_b</a>, <a href="#bc_uv_t">bc_uv_t</a>)
	772	are necessary. In order not to distort the vortex, an initial
	773	horizontal wind profile constant
	774	with height is necessary (to be set by <b>initializing_actions</b>
	775	= <span style="font-style: italic;">'set_constant_profiles'</span>)
	776	and some other conditions have to be met (neutral stratification,
	777	diffusion must be
	778	switched off, see <a href="#km_constant">km_constant</a>).
	779	The center of the vortex is located at jc = (nx+1)/2. It
	780	extends from k = 0 to k = nz+1. Its radius is 8 * <a href="#dx">dx</a>
	781	and the exponentially decaying part ranges to 32 * <a href="#dx">dx</a>
[61]	782	(see init_rankine.f90). </div> <p><span style="font-style: italic;">'initialize_ptanom'</span>
	783	</p> <ul> <p>A 2d-Gauss-like shape disturbance
	784	(x,y) is added to the
[5]	785	initial temperature field with radius 10.0 * <a href="#dx">dx</a>
	786	and center at jc = (nx+1)/2. This may be used for tests of scalar
	787	advection schemes
	788	(see <a href="#scalar_advec">scalar_advec</a>).
	789	Such tests require a horizontal wind profile constant with hight and
	790	diffusion
	791	switched off (see <span style="font-style: italic;">'initialize_vortex'</span>).
	792	Additionally, the buoyancy term
	793	must be switched of in the equation of motion  for w (this
[61]	794	requires the user to comment out the call of <span style="font-family: monospace;">buoyancy</span> in the
	795	source code of <span style="font-family: monospace;">prognostic_equations.f90</span>).</p>
	796	</ul> <p style="font-style: normal;">Values may be
[5]	797	combined, e.g. <b>initializing_actions</b> = <span style="font-style: italic;">'set_constant_profiles
[61]	798	initialize_vortex'</span>, but the values of <span style="font-style: italic;">'set_constant_profiles'</span>,
	799	<span style="font-style: italic;">'set_1d-model_profiles'</span>
	800	, and <span style="font-style: italic;">'by_user'</span>
	801	must not be given at the same time.</p> <p style="font-style: italic;"> </p> </td> </tr>
	802	<tr> <td style="vertical-align: top;"> <p><a name="km_constant"></a><b>km_constant</b></p>
	803	</td> <td style="vertical-align: top;">R</td>
	804	<td style="vertical-align: top;"><i>variable<br>
	805	(computed from TKE)</i></td> <td style="vertical-align: top;"> <p>Constant eddy
	806	diffusivities are used (laminar
	807	simulations).  </p> <p>If this parameter is
	808	specified, both in the 1d and in the
[5]	809	3d-model constant values for the eddy diffusivities are used in
	810	space and time with K<sub>m</sub> = <b>km_constant</b>
	811	and K<sub>h</sub> = K<sub>m</sub> / <a href="chapter_4.2.html#prandtl_number">prandtl_number</a>.
	812	The prognostic equation for the subgrid-scale TKE is switched off.
	813	Constant eddy diffusivities are only allowed with the Prandtl layer (<a href="#prandtl_layer">prandtl_layer</a>)
[61]	814	switched off.</p> </td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="km_damp_max"></a><b>km_damp_max</b></p>
	815	</td> <td style="vertical-align: top;">R</td>
	816	<td style="vertical-align: top;"><span style="font-style: italic;">0.5*(dx
	817	or dy)</span></td> <td style="vertical-align: top;">Maximum
[5]	818	diffusivity used for filtering the velocity field in the vicinity of
[61]	819	the outflow (in m<sup>2</sup>/s).<br> <br>
[5]	820	When using non-cyclic lateral boundaries (see <a href="#bc_lr">bc_lr</a>
	821	or <a href="#bc_ns">bc_ns</a>),
	822	a smoothing has to be applied to the
	823	velocity field in the vicinity of the outflow in order to suppress any
	824	reflections of outgoing disturbances. Smoothing is done by increasing
	825	the eddy diffusivity along the horizontal direction which is
	826	perpendicular to the outflow boundary. Only velocity components
	827	parallel to the outflow boundary are filtered (e.g. v and w, if the
	828	outflow is along x). Damping is applied from the bottom to the top of
[61]	829	the domain.<br> <br>
[5]	830	The horizontal range of the smoothing is controlled by <a href="#outflow_damping_width">outflow_damping_width</a>
	831	which defines the number of gridpoints (counted from the outflow
	832	boundary) from where on the smoothing is applied. Starting from that
	833	point, the eddy diffusivity is linearly increased (from zero to its
	834	maximum value given by <span style="font-weight: bold;">km_damp_max</span>)
	835	until half of the damping range width, from where it remains constant
	836	up to the outflow boundary. If at a certain grid point the eddy
	837	diffusivity calculated from the flow field is larger than as described
[61]	838	above, it is used instead.<br> <br>
[5]	839	The default value of <span style="font-weight: bold;">km_damp_max</span>
[61]	840	has been empirically proven to be sufficient.</td> </tr> <tr>
	841	<td style="vertical-align: top;"> <p><a name="long_filter_factor"></a><b>long_filter_factor</b></p>
	842	</td> <td style="vertical-align: top;">R</td>
	843	<td style="vertical-align: top;"><i>0.0</i></td>
	844	<td style="vertical-align: top;"> <p>Filter factor
	845	for the so-called Long-filter.<br> </p> <p><br>
[5]	846	This filter very efficiently
	847	eliminates 2-delta-waves sometimes cauesed by the upstream-spline
	848	scheme (see Mahrer and
	849	Pielke, 1978: Mon. Wea. Rev., 106, 818-830). It works in all three
[61]	850	directions in space. A value of <b>long_filter_factor</b>
	851	= <i>0.01</i>
[5]	852	sufficiently removes the small-scale waves without affecting the
[61]	853	longer waves.<br> </p> <p>By default, the filter is
	854	switched off (= <i>0.0</i>).
[5]	855	It is exclusively applied to the tendencies calculated by the
	856	upstream-spline scheme (see <a href="#momentum_advec">momentum_advec</a>
	857	and <a href="#scalar_advec">scalar_advec</a>),
	858	not to the prognostic variables themselves. At the bottom and top
	859	boundary of the model domain the filter effect for vertical
	860	2-delta-waves is reduced. There, the amplitude of these waves is only
	861	reduced by approx. 50%, otherwise by nearly 100%.  <br>
[61]	862	Filter factors with values > <i>0.01</i> also
	863	reduce the amplitudes
[5]	864	of waves with wavelengths longer than 2-delta (see the paper by Mahrer
	865	and
[62]	866	Pielke, quoted above). </p> </td> </tr> <tr><td style="vertical-align: top;"><a name="loop_optimization"></a><span style="font-weight: bold;">loop_optimization</span></td><td style="vertical-align: top;">C*16</td><td style="vertical-align: top;"><span style="font-style: italic;">see right</span></td><td>Method used to optimize loops for solving the prognostic equations .<br><br>By
	867	default, the optimization method depends on the host on which PALM is
	868	running. On machines with vector-type CPUs, single 3d-loops are used to
	869	calculate each tendency term of each prognostic equation, while on all
	870	other machines, all prognostic equations are solved within one big loop
	871	over the two horizontal indices<span style="font-family: Courier New,Courier,monospace;"> i </span>and<span style="font-family: Courier New,Courier,monospace;"> j </span>(giving a good cache uitilization).<br><br>The default behaviour can be changed by setting either <span style="font-weight: bold;">loop_optimization</span> = <span style="font-style: italic;">'vector'</span> or <span style="font-weight: bold;">loop_optimization</span> = <span style="font-style: italic;">'cache'</span>.</td></tr><tr>
[61]	872	<td style="vertical-align: top;"><a name="mixing_length_1d"></a><span style="font-weight: bold;">mixing_length_1d</span><br>
	873	</td> <td style="vertical-align: top;">C*20<br>
	874	</td> <td style="vertical-align: top;"><span style="font-style: italic;">'as_in_3d_</span><br style="font-style: italic;"> <span style="font-style: italic;">model'</span><br> </td>
	875	<td style="vertical-align: top;">Mixing length used in the
	876	1d-model.<br> <br>
	877	By default the mixing length is calculated as in the 3d-model (i.e. it
	878	depends on the grid spacing).<br> <br>
	879	By setting <span style="font-weight: bold;">mixing_length_1d</span>
	880	= <span style="font-style: italic;">'blackadar'</span>,
[5]	881	the so-called Blackadar mixing length is used (l = kappa * z / ( 1 +
	882	kappa * z / lambda ) with the limiting value lambda = 2.7E-4 * u_g / f).<br>
[75]	883	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="humidity"></a><b>humidity</b></p>
[61]	884	</td> <td style="vertical-align: top;">L</td>
	885	<td style="vertical-align: top;"><i>.F.</i></td>
	886	<td style="vertical-align: top;"> <p>Parameter to
	887	switch on the prognostic equation for specific
	888	humidity q.<br> </p> <p>The initial vertical
	889	profile of q can be set via parameters <a href="chapter_4.1.html#q_surface">q_surface</a>, <a href="chapter_4.1.html#q_vertical_gradient">q_vertical_gradient</a>
[5]	890	and <a href="chapter_4.1.html#q_vertical_gradient_level">q_vertical_gradient_level</a>.
	891	Boundary conditions can be set via <a href="chapter_4.1.html#q_surface_initial_change">q_surface_initial_change</a>
	892	and <a href="chapter_4.1.html#surface_waterflux">surface_waterflux</a>.<br>
[61]	893	</p>
[5]	894	If the condensation scheme is switched on (<a href="chapter_4.1.html#cloud_physics">cloud_physics</a>
	895	= .TRUE.), q becomes the total liquid water content (sum of specific
[61]	896	humidity and liquid water content).</td> </tr>
	897	<tr> <td style="vertical-align: top;"> <p><a name="momentum_advec"></a><b>momentum_advec</b></p>
	898	</td> <td style="vertical-align: top;">C * 10</td>
	899	<td style="vertical-align: top;"><i>'pw-scheme'</i></td>
	900	<td style="vertical-align: top;"> <p>Advection
	901	scheme to be used for the momentum equations.<br> <br>
[5]	902	The user can choose between the following schemes:<br>
[61]	903	<br> <br> <span style="font-style: italic;">'pw-scheme'</span><br>
	904	</p> <div style="margin-left: 40px;">The scheme of
	905	Piascek and
[5]	906	Williams (1970, J. Comp. Phys., 6,
	907	392-405) with central differences in the form C3 is used.<br>
	908	If intermediate Euler-timesteps are carried out in case of <a href="#timestep_scheme">timestep_scheme</a>
[61]	909	= <span style="font-style: italic;">'leapfrog+euler'</span>
	910	the
[5]	911	advection scheme is - for the Euler-timestep - automatically switched
[61]	912	to an upstream-scheme.<br> </div> <p> </p> <p><span style="font-style: italic;">'ups-scheme'</span><br>
	913	</p> <div style="margin-left: 40px;">The
	914	upstream-spline scheme is
[5]	915	used
	916	(see Mahrer and Pielke,
	917	1978: Mon. Wea. Rev., 106, 818-830). In opposite to the
	918	Piascek-Williams scheme, this is characterized by much better numerical
	919	features (less numerical diffusion, better preservation of flow
	920	structures, e.g. vortices), but computationally it is much more
	921	expensive. In
	922	addition, the use of the Euler-timestep scheme is mandatory (<a href="#timestep_scheme">timestep_scheme</a>
[61]	923	= <span style="font-style: italic;">'</span><i>euler'</i>),
	924	i.e. the
[5]	925	timestep accuracy is only of first order.
	926	For this reason the advection of scalar variables (see <a href="#scalar_advec">scalar_advec</a>)
	927	should then also be carried out with the upstream-spline scheme,
	928	because otherwise the scalar variables would
	929	be subject to large numerical diffusion due to the upstream
[61]	930	scheme.  </div> <p style="margin-left: 40px;">Since
	931	the cubic splines used tend
[5]	932	to overshoot under
	933	certain circumstances, this effect must be adjusted by suitable
	934	filtering and smoothing (see <a href="#cut_spline_overshoot">cut_spline_overshoot</a>,
[61]	935	<a href="#long_filter_factor">long_filter_factor</a>,
	936	<a href="#ups_limit_pt">ups_limit_pt</a>, <a href="#ups_limit_u">ups_limit_u</a>, <a href="#ups_limit_v">ups_limit_v</a>, <a href="#ups_limit_w">ups_limit_w</a>).
[5]	937	This is always neccessary for runs with stable stratification,
	938	even if this stratification appears only in parts of the model domain.<br>
[61]	939	</p> <div style="margin-left: 40px;">With stable
	940	stratification the
[5]	941	upstream-spline scheme also
	942	produces gravity waves with large amplitude, which must be
	943	suitably damped (see <a href="chapter_4.2.html#rayleigh_damping_factor">rayleigh_damping_factor</a>).<br>
[61]	944	<br> <span style="font-weight: bold;">Important: </span>The
[5]	945	upstream-spline scheme is not implemented for humidity and passive
[75]	946	scalars (see <a href="#humidity">humidity</a>
[5]	947	and <a href="#passive_scalar">passive_scalar</a>)
	948	and requires the use of a 2d-domain-decomposition. The last conditions
	949	severely restricts code optimization on several machines leading to
	950	very long execution times! The scheme is also not allowed for
	951	non-cyclic lateral boundary conditions (see <a href="#bc_lr">bc_lr</a>
[61]	952	and <a href="#bc_ns">bc_ns</a>).</div> </td>
	953	</tr> <tr> <td style="vertical-align: top;"><a name="netcdf_precision"></a><span style="font-weight: bold;">netcdf_precision</span><br>
	954	</td> <td style="vertical-align: top;">C*20<br>
	955	(10)<br> </td> <td style="vertical-align: top;"><span style="font-style: italic;">single preci-</span><br style="font-style: italic;"> <span style="font-style: italic;">sion for all</span><br style="font-style: italic;"> <span style="font-style: italic;">output quan-</span><br style="font-style: italic;"> <span style="font-style: italic;">tities</span><br> </td>
	956	<td style="vertical-align: top;">Defines the accuracy of
	957	the NetCDF output.<br> <br>
	958	By default, all NetCDF output data (see <a href="chapter_4.2.html#data_output_format">data_output_format</a>)
	959	have single precision  (4 byte) accuracy. Double precision (8
	960	byte) can be choosen alternatively.<br>
	961	Accuracy for the different output data (cross sections, 3d-volume data,
	962	spectra, etc.) can be set independently.<br> <span style="font-style: italic;">'<out>_NF90_REAL4'</span>
	963	(single precision) or <span style="font-style: italic;">'<out>_NF90_REAL8'</span>
	964	(double precision) are the two principally allowed values for <span style="font-weight: bold;">netcdf_precision</span>,
	965	where the string <span style="font-style: italic;">'<out>'
	966	</span>can be chosen out of the following list:<br> <br>
	967	<table style="text-align: left; width: 284px; height: 234px;" border="1" cellpadding="2" cellspacing="2"> <tbody>
	968	<tr> <td style="vertical-align: top;"><span style="font-style: italic;">'xy'</span><br> </td>
	969	<td style="vertical-align: top;">horizontal cross section<br>
	970	</td> </tr> <tr> <td style="vertical-align: top;"><span style="font-style: italic;">'xz'</span><br> </td>
	971	<td style="vertical-align: top;">vertical (xz) cross
	972	section<br> </td> </tr> <tr> <td style="vertical-align: top;"><span style="font-style: italic;">'yz'</span><br> </td>
	973	<td style="vertical-align: top;">vertical (yz) cross
	974	section<br> </td> </tr> <tr> <td style="vertical-align: top;"><span style="font-style: italic;">'2d'</span><br> </td>
	975	<td style="vertical-align: top;">all cross sections<br>
	976	</td> </tr> <tr> <td style="vertical-align: top;"><span style="font-style: italic;">'3d'</span><br> </td>
	977	<td style="vertical-align: top;">volume data<br> </td>
	978	</tr> <tr> <td style="vertical-align: top;"><span style="font-style: italic;">'pr'</span><br> </td>
	979	<td style="vertical-align: top;">vertical profiles<br>
	980	</td> </tr> <tr> <td style="vertical-align: top;"><span style="font-style: italic;">'ts'</span><br> </td>
	981	<td style="vertical-align: top;">time series, particle
	982	time series<br> </td> </tr> <tr> <td style="vertical-align: top;"><span style="font-style: italic;">'sp'</span><br> </td>
	983	<td style="vertical-align: top;">spectra<br> </td>
	984	</tr> <tr> <td style="vertical-align: top;"><span style="font-style: italic;">'prt'</span><br> </td>
	985	<td style="vertical-align: top;">particles<br> </td>
	986	</tr> <tr> <td style="vertical-align: top;"><span style="font-style: italic;">'all'</span><br> </td>
	987	<td style="vertical-align: top;">all output quantities<br>
	988	</td> </tr> </tbody> </table> <br> <span style="font-weight: bold;">Example:</span><br>
[5]	989	If all cross section data and the particle data shall be output in
[61]	990	double precision and all other quantities in single precision, then <span style="font-weight: bold;">netcdf_precision</span> = <span style="font-style: italic;">'2d_NF90_REAL8'</span>, <span style="font-style: italic;">'prt_NF90_REAL8'</span>
	991	has to be assigned.<br> </td> </tr>
	992	<tr> <td style="vertical-align: top;"> <p><a name="npex"></a><b>npex</b></p> </td>
	993	<td style="vertical-align: top;">I</td> <td style="vertical-align: top;"><br> </td> <td style="vertical-align: top;"> <p>Number of processors
	994	along x-direction of the virtual
[5]	995	processor
[61]	996	net.  </p> <p>For parallel runs, the total
	997	number of processors to be used
[5]	998	is given by
[61]	999	the <span style="font-weight: bold;">mrun</span>
	1000	option <a href="http://www.muk.uni-hannover.de/software/mrun_beschreibung.html#Opt-X">-X</a>.
[5]	1001	By default, depending on the type of the parallel computer, PALM
	1002	generates a 1d processor
	1003	net (domain decomposition along x, <span style="font-weight: bold;">npey</span>
[61]	1004	= <span style="font-style: italic;">1</span>) or a
	1005	2d-net (this is
[5]	1006	favored on machines with fast communication network). In case of a
	1007	2d-net, it is tried to make it more or less square-shaped. If, for
	1008	example, 16 processors are assigned (-X 16), a 4 * 4 processor net is
[61]	1009	generated (<span style="font-weight: bold;">npex</span>
	1010	= <span style="font-style: italic;">4</span>, <span style="font-weight: bold;">npey</span>
[5]	1011	= <span style="font-style: italic;">4</span>).
	1012	This choice is optimal for square total domains (<a href="#nx">nx</a>
	1013	= <a href="#ny">ny</a>),
	1014	since then the number of ghost points at the lateral boundarys of
	1015	the subdomains is minimal. If <span style="font-weight: bold;">nx</span>
[61]	1016	nd <span style="font-weight: bold;">ny</span>
	1017	differ extremely, the
[5]	1018	processor net should be manually adjusted using adequate values for <span style="font-weight: bold;">npex</span> and <span style="font-weight: bold;">npey</span>.  </p>
[61]	1019	<p><b>Important:</b> The value of <span style="font-weight: bold;">npex</span> * <span style="font-weight: bold;">npey</span> must exactly
	1020	correspond to the
[5]	1021	value assigned by the <span style="font-weight: bold;">mrun</span>-option
[61]	1022	<tt>-X</tt>.
[5]	1023	Otherwise the model run will abort with a corresponding error
	1024	message.  <br>
	1025	Additionally, the specification of <span style="font-weight: bold;">npex</span>
[61]	1026	and <span style="font-weight: bold;">npey</span>
	1027	may of course
[5]	1028	override the default setting for the domain decomposition (1d or 2d)
	1029	which may have a significant (negative) effect on the code performance.
[61]	1030	</p> </td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="npey"></a><b>npey</b></p>
	1031	</td> <td style="vertical-align: top;">I</td>
	1032	<td style="vertical-align: top;"><br> </td> <td style="vertical-align: top;"> <p>Number of processors
	1033	along y-direction of the virtual
[5]	1034	processor
[61]	1035	net.  </p> <p>For further information see <a href="#npex">npex</a>.</p> </td> </tr>
	1036	<tr> <td style="vertical-align: top;"> <p><a name="nsor_ini"></a><b>nsor_ini</b></p>
	1037	</td> <td style="vertical-align: top;">I</td>
	1038	<td style="vertical-align: top;"><i>100</i></td>
	1039	<td style="vertical-align: top;"> <p>Initial number
	1040	of iterations with the SOR algorithm.  </p> <p>This
	1041	parameter is only effective if the SOR algorithm was
[5]	1042	selected as the pressure solver scheme (<a href="chapter_4.2.html#psolver">psolver</a>
[61]	1043	= <span style="font-style: italic;">'sor'</span>)
	1044	and specifies the
[5]	1045	number of initial iterations of the SOR
	1046	scheme (at t = 0). The number of subsequent iterations at the following
	1047	timesteps is determined
	1048	with the parameter <a href="#nsor">nsor</a>.
[61]	1049	Usually <b>nsor</b> < <b>nsor_ini</b>,
	1050	since in each case
[5]	1051	subsequent calls to <a href="chapter_4.2.html#psolver">psolver</a>
	1052	use the solution of the previous call as initial value. Suitable
	1053	test runs should determine whether sufficient convergence of the
	1054	solution is obtained with the default value and if necessary the value
[61]	1055	of <b>nsor_ini</b> should be changed.</p> </td>
	1056	</tr> <tr> <td style="vertical-align: top;">
	1057	<p><a name="nx"></a><b>nx</b></p>
	1058	</td> <td style="vertical-align: top;">I</td>
	1059	<td style="vertical-align: top;"><br> </td> <td style="vertical-align: top;"> <p>Number of grid
	1060	points in x-direction.  </p> <p>A value for this
	1061	parameter must be assigned. Since the lower
[5]	1062	array bound in PALM
	1063	starts with i = 0, the actual number of grid points is equal to <b>nx+1</b>.
	1064	In case of cyclic boundary conditions along x, the domain size is (<b>nx+1</b>)*
[61]	1065	<a href="#dx">dx</a>.</p> <p>For
	1066	parallel runs, in case of <a href="#grid_matching">grid_matching</a>
	1067	= <span style="font-style: italic;">'strict'</span>,
	1068	<b>nx+1</b> must
[5]	1069	be an integral multiple
	1070	of the processor numbers (see <a href="#npex">npex</a>
	1071	and <a href="#npey">npey</a>)
	1072	along x- as well as along y-direction (due to data
[61]	1073	transposition restrictions).</p> </td> </tr> <tr>
	1074	<td style="vertical-align: top;"> <p><a name="ny"></a><b>ny</b></p>
	1075	</td> <td style="vertical-align: top;">I</td>
	1076	<td style="vertical-align: top;"><br> </td> <td style="vertical-align: top;"> <p>Number of grid
	1077	points in y-direction.  </p> <p>A value for this
	1078	parameter must be assigned. Since the lower
[5]	1079	array bound in PALM starts with i = 0, the actual number of grid points
[61]	1080	is equal to <b>ny+1</b>. In case of cyclic boundary
	1081	conditions along
[5]	1082	y, the domain size is (<b>ny+1</b>) * <a href="#dy">dy</a>.</p>
[61]	1083	<p>For parallel runs, in case of <a href="#grid_matching">grid_matching</a>
	1084	= <span style="font-style: italic;">'strict'</span>,
	1085	<b>ny+1</b> must
[5]	1086	be an integral multiple
	1087	of the processor numbers (see <a href="#npex">npex</a>
	1088	and <a href="#npey">npey</a>)
	1089	along y- as well as along x-direction (due to data
[61]	1090	transposition restrictions).</p> </td> </tr> <tr>
	1091	<td style="vertical-align: top;"> <p><a name="nz"></a><b>nz</b></p>
	1092	</td> <td style="vertical-align: top;">I</td>
	1093	<td style="vertical-align: top;"><br> </td> <td style="vertical-align: top;"> <p>Number of grid
	1094	points in z-direction.  </p> <p>A value for this
	1095	parameter must be assigned. Since the lower
[5]	1096	array bound in PALM
	1097	starts with k = 0 and since one additional grid point is added at the
[61]	1098	top boundary (k = <b>nz+1</b>), the actual number of grid
	1099	points is <b>nz+2</b>.
	1100	However, the prognostic equations are only solved up to <b>nz</b>
	1101	(u,
[5]	1102	v)
[61]	1103	or up to <b>nz-1</b> (w, scalar quantities). The top
	1104	boundary for u
	1105	and v is at k = <b>nz+1</b> (u, v) while at k = <b>nz</b>
	1106	for all
	1107	other quantities.  </p> <p>For parallel
	1108	runs,  in case of <a href="#grid_matching">grid_matching</a>
	1109	= <span style="font-style: italic;">'strict'</span>,
	1110	<b>nz</b> must
[5]	1111	be an integral multiple of
	1112	the number of processors in x-direction (due to data transposition
[61]	1113	restrictions).</p> </td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="omega"></a><b>omega</b></p>
	1114	</td> <td style="vertical-align: top;">R</td>
	1115	<td style="vertical-align: top;"><i>7.29212E-5</i></td>
	1116	<td style="vertical-align: top;"> <p>Angular
	1117	velocity of the rotating system (in rad s<sup>-1</sup>).
	1118	</p> <p>The angular velocity of the earth is set by
	1119	default. The
[5]	1120	values
[61]	1121	of the Coriolis parameters are calculated as:  </p> <ul>
	1122	<p>f = 2.0 * <b>omega</b> * sin(<a href="#phi">phi</a>)
	1123	<br>f* = 2.0 * <b>omega</b> * cos(<a href="#phi">phi</a>)</p>
	1124	</ul> </td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="outflow_damping_width"></a><b>outflow_damping_width</b></p>
	1125	</td> <td style="vertical-align: top;">I</td>
	1126	<td style="vertical-align: top;"><span style="font-style: italic;">MIN(20,
[5]	1127	nx/2</span> or <span style="font-style: italic;">ny/2)</span></td>
[61]	1128	<td style="vertical-align: top;">Width of
[5]	1129	the damping range in the vicinity of the outflow (gridpoints).<br>
[61]	1130	<br>
[5]	1131	When using non-cyclic lateral boundaries (see <a href="chapter_4.1.html#bc_lr">bc_lr</a>
	1132	or <a href="chapter_4.1.html#bc_ns">bc_ns</a>),
	1133	a smoothing has to be applied to the
	1134	velocity field in the vicinity of the outflow in order to suppress any
	1135	reflections of outgoing disturbances. This parameter controlls the
	1136	horizontal range to which the smoothing is applied. The range is given
	1137	in gridpoints counted from the respective outflow boundary. For further
	1138	details about the smoothing see parameter <a href="chapter_4.1.html#km_damp_max">km_damp_max</a>,
[61]	1139	which defines the magnitude of the damping.</td> </tr>
	1140	<tr> <td style="vertical-align: top;"> <p><a name="overshoot_limit_e"></a><b>overshoot_limit_e</b></p>
	1141	</td> <td style="vertical-align: top;">R</td>
	1142	<td style="vertical-align: top;"><i>0.0</i></td>
	1143	<td style="vertical-align: top;"> <p>Allowed limit
	1144	for the overshooting of subgrid-scale TKE in
[5]	1145	case that the upstream-spline scheme is switched on (in m<sup>2</sup>/s<sup>2</sup>).
[61]	1146	</p> <p>By deafult, if cut-off of overshoots is switched
	1147	on for the
[5]	1148	upstream-spline scheme (see <a href="#cut_spline_overshoot">cut_spline_overshoot</a>),
	1149	no overshoots are permitted at all. If <b>overshoot_limit_e</b>
	1150	is given a non-zero value, overshoots with the respective
	1151	amplitude (both upward and downward) are allowed.  </p>
[61]	1152	<p>Only positive values are allowed for <b>overshoot_limit_e</b>.</p>
	1153	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="overshoot_limit_pt"></a><b>overshoot_limit_pt</b></p>
	1154	</td> <td style="vertical-align: top;">R</td>
	1155	<td style="vertical-align: top;"><i>0.0</i></td>
	1156	<td style="vertical-align: top;"> <p>Allowed limit
	1157	for the overshooting of potential temperature in
[5]	1158	case that the upstream-spline scheme is switched on (in K).  </p>
[61]	1159	<p>For further information see <a href="#overshoot_limit_e">overshoot_limit_e</a>.
	1160	</p> <p>Only positive values are allowed for <b>overshoot_limit_pt</b>.</p>
	1161	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="overshoot_limit_u"></a><b>overshoot_limit_u</b></p>
	1162	</td> <td style="vertical-align: top;">R</td>
	1163	<td style="vertical-align: top;"><i>0.0</i></td>
	1164	<td style="vertical-align: top;">Allowed limit for the
[5]	1165	overshooting of
	1166	the u-component of velocity in case that the upstream-spline scheme is
[61]	1167	switched on (in m/s). <p>For further information see <a href="#overshoot_limit_e">overshoot_limit_e</a>.
	1168	</p> <p>Only positive values are allowed for <b>overshoot_limit_u</b>.</p>
	1169	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="overshoot_limit_v"></a><b>overshoot_limit_v</b></p>
	1170	</td> <td style="vertical-align: top;">R</td>
	1171	<td style="vertical-align: top;"><i>0.0</i></td>
	1172	<td style="vertical-align: top;"> <p>Allowed limit
	1173	for the overshooting of the v-component of
[5]	1174	velocity in case that the upstream-spline scheme is switched on
[61]	1175	(in m/s).  </p> <p>For further information see <a href="#overshoot_limit_e">overshoot_limit_e</a>.
	1176	</p> <p>Only positive values are allowed for <b>overshoot_limit_v</b>.</p>
	1177	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="overshoot_limit_w"></a><b>overshoot_limit_w</b></p>
	1178	</td> <td style="vertical-align: top;">R</td>
	1179	<td style="vertical-align: top;"><i>0.0</i></td>
	1180	<td style="vertical-align: top;"> <p>Allowed limit
	1181	for the overshooting of the w-component of
[5]	1182	velocity in case that the upstream-spline scheme is switched on
[61]	1183	(in m/s).  </p> <p>For further information see <a href="#overshoot_limit_e">overshoot_limit_e</a>.
	1184	</p> <p>Only positive values are permitted for <b>overshoot_limit_w</b>.</p>
	1185	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="passive_scalar"></a><b>passive_scalar</b></p>
	1186	</td> <td style="vertical-align: top;">L</td>
	1187	<td style="vertical-align: top;"><i>.F.</i></td>
	1188	<td style="vertical-align: top;"> <p>Parameter to
	1189	switch on the prognostic equation for a passive
	1190	scalar. <br> </p> <p>The initial vertical profile
	1191	of s can be set via parameters <a href="#s_surface">s_surface</a>,
	1192	<a href="#s_vertical_gradient">s_vertical_gradient</a>
[5]	1193	and  <a href="#s_vertical_gradient_level">s_vertical_gradient_level</a>.
	1194	Boundary conditions can be set via <a href="#s_surface_initial_change">s_surface_initial_change</a>
[61]	1195	and <a href="#surface_scalarflux">surface_scalarflux</a>.
	1196	</p> <p><b>Note:</b> <br>
	1197	With <span style="font-weight: bold;">passive_scalar</span>
	1198	switched
[75]	1199	on, the simultaneous use of humidity (see <a href="#humidity">humidity</a>)
[61]	1200	is impossible.</p> </td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="phi"></a><b>phi</b></p>
	1201	</td> <td style="vertical-align: top;">R</td>
	1202	<td style="vertical-align: top;"><i>55.0</i></td>
	1203	<td style="vertical-align: top;"> <p>Geographical
	1204	latitude (in degrees).  </p> <p>The value of
	1205	this parameter determines the value of the
[5]	1206	Coriolis parameters f and f*, provided that the angular velocity (see <a href="#omega">omega</a>)
[61]	1207	is non-zero.</p> </td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="prandtl_layer"></a><b>prandtl_layer</b></p>
	1208	</td> <td style="vertical-align: top;">L</td>
	1209	<td style="vertical-align: top;"><i>.T.</i></td>
	1210	<td style="vertical-align: top;"> <p>Parameter to
	1211	switch on a Prandtl layer.  </p> <p>By default,
	1212	a Prandtl layer is switched on at the bottom
[5]	1213	boundary between z = 0 and z = 0.5 * <a href="#dz">dz</a>
	1214	(the first computational grid point above ground for u, v and the
	1215	scalar quantities).
	1216	In this case, at the bottom boundary, free-slip conditions for u and v
	1217	(see <a href="#bc_uv_b">bc_uv_b</a>)
	1218	are not allowed. Likewise, laminar
	1219	simulations with constant eddy diffusivities (<a href="#km_constant">km_constant</a>)
[61]	1220	are forbidden.  </p> <p>With Prandtl-layer
	1221	switched off, the TKE boundary condition <a href="#bc_e_b">bc_e_b</a>
	1222	= '<i>(u)*2+neumann'</i> must not be used and is
	1223	automatically
	1224	changed to <i>'neumann'</i> if necessary.  Also,
	1225	the pressure
[5]	1226	boundary condition <a href="#bc_p_b">bc_p_b</a>
	1227	= <i>'neumann+inhomo'</i>  is not allowed. </p>
[61]	1228	<p>The roughness length is declared via the parameter <a href="#roughness_length">roughness_length</a>.</p>
	1229	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="precipitation"></a><b>precipitation</b></p>
	1230	</td> <td style="vertical-align: top;">L</td>
	1231	<td style="vertical-align: top;"><span style="font-style: italic;">.F.</span></td> <td style="vertical-align: top;"> <p>Parameter to switch
	1232	on the precipitation scheme.<br> </p> <p>For
	1233	precipitation processes PALM uses a simplified Kessler
[5]	1234	scheme. This scheme only considers the
	1235	so-called autoconversion, that means the generation of rain water by
	1236	coagulation of cloud drops among themselves. Precipitation begins and
	1237	is immediately removed from the flow as soon as the liquid water
[72]	1238	content exceeds the critical value of 0.5 g/kg.</p><p>The precipitation rate and amount can be output by assigning the runtime parameter <a href="chapter_4.2.html#data_output">data_output</a> = <span style="font-style: italic;">'prr'</span> or <span style="font-style: italic;">'pra'</span>, respectively. The time interval on which the precipitation amount is defined can be controlled via runtime parameter <a href="chapter_4.2.html#precipitation_amount_interval">precipitation_amount_interval</a>.</p> </td> </tr>
[61]	1239	<tr><td style="vertical-align: top;"><a name="pt_reference"></a><span style="font-weight: bold;">pt_reference</span></td><td style="vertical-align: top;">R</td><td style="vertical-align: top;"><span style="font-style: italic;">use horizontal average as
	1240	refrence</span></td><td style="vertical-align: top;">Reference
	1241	temperature to be used in all buoyancy terms (in K).<br><br>By
	1242	default, the instantaneous horizontal average over the total model
	1243	domain is used.</td></tr><tr> <td style="vertical-align: top;"> <p><a name="pt_surface"></a><b>pt_surface</b></p>
	1244	</td> <td style="vertical-align: top;">R</td>
	1245	<td style="vertical-align: top;"><i>300.0</i></td>
	1246	<td style="vertical-align: top;"> <p>Surface
	1247	potential temperature (in K).  </p> <p>This
	1248	parameter assigns the value of the potential temperature
	1249	pt at the surface (k=0)<b>.</b> Starting from this value,
	1250	the
[5]	1251	initial vertical temperature profile is constructed with <a href="#pt_vertical_gradient">pt_vertical_gradient</a>
[61]	1252	and <a href="#pt_vertical_gradient_level">pt_vertical_gradient_level
	1253	</a>.
[5]	1254	This profile is also used for the 1d-model as a stationary profile.</p>
[61]	1255	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="pt_surface_initial_change"></a><b>pt_surface_initial</b>
	1256	<br> <b>_change</b></p> </td> <td style="vertical-align: top;">R</td> <td style="vertical-align: top;"><span style="font-style: italic;">0.0</span><br> </td>
	1257	<td style="vertical-align: top;"> <p>Change in
	1258	surface temperature to be made at the beginning of
[5]	1259	the 3d run
[61]	1260	(in K).  </p> <p>If <b>pt_surface_initial_change</b>
	1261	is set to a non-zero
[5]	1262	value, the near surface sensible heat flux is not allowed to be given
	1263	simultaneously (see <a href="#surface_heatflux">surface_heatflux</a>).</p>
[61]	1264	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="pt_vertical_gradient"></a><b>pt_vertical_gradient</b></p>
	1265	</td> <td style="vertical-align: top;">R (10)</td>
	1266	<td style="vertical-align: top;"><i>10 * 0.0</i></td>
	1267	<td style="vertical-align: top;"> <p>Temperature
	1268	gradient(s) of the initial temperature profile (in
[5]	1269	K
[61]	1270	/ 100 m).  </p> <p>This temperature gradient
	1271	holds starting from the height
[5]	1272	level defined by <a href="#pt_vertical_gradient_level">pt_vertical_gradient_level</a>
	1273	(precisely: for all uv levels k where zu(k) >
	1274	pt_vertical_gradient_level,
	1275	pt_init(k) is set: pt_init(k) = pt_init(k-1) + dzu(k) * <b>pt_vertical_gradient</b>)
	1276	up to the top boundary or up to the next height level defined
	1277	by <a href="#pt_vertical_gradient_level">pt_vertical_gradient_level</a>.
	1278	A total of 10 different gradients for 11 height intervals (10 intervals
	1279	if <a href="#pt_vertical_gradient_level">pt_vertical_gradient_level</a>(1)
[61]	1280	= <i>0.0</i>) can be assigned. The surface temperature is
	1281	assigned via <a href="#pt_surface">pt_surface</a>.
	1282	</p> <p>Example:  </p> <ul> <p><b>pt_vertical_gradient</b>
	1283	= <i>1.0</i>, <i>0.5</i>,  <br>
	1284	<b>pt_vertical_gradient_level</b> = <i>500.0</i>,
	1285	<i>1000.0</i>,</p> </ul> <p>That
	1286	defines the temperature profile to be neutrally
[5]	1287	stratified
	1288	up to z = 500.0 m with a temperature given by <a href="#pt_surface">pt_surface</a>.
[61]	1289	For 500.0 m < z <= 1000.0 m the temperature gradient is
	1290	1.0 K /
[5]	1291	100 m and for z > 1000.0 m up to the top boundary it is
	1292	0.5 K / 100 m (it is assumed that the assigned height levels correspond
[61]	1293	with uv levels). </p> </td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="pt_vertical_gradient_level"></a><b>pt_vertical_gradient</b>
	1294	<br> <b>_level</b></p> </td> <td style="vertical-align: top;">R (10)</td> <td style="vertical-align: top;"> <p><i>10 *</i>
	1295	<span style="font-style: italic;">0.0</span><br>
	1296	</p> </td> <td style="vertical-align: top;">
	1297	<p>Height level from which on the temperature gradient defined by
	1298	<a href="#pt_vertical_gradient">pt_vertical_gradient</a>
	1299	is effective (in m).  </p> <p>The height levels
	1300	are to be assigned in ascending order. The
[5]	1301	default values result in a neutral stratification regardless of the
	1302	values of <a href="#pt_vertical_gradient">pt_vertical_gradient</a>
	1303	(unless the top boundary of the model is higher than 100000.0 m).
	1304	For the piecewise construction of temperature profiles see <a href="#pt_vertical_gradient">pt_vertical_gradient</a>.</p>
[61]	1305	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="q_surface"></a><b>q_surface</b></p>
	1306	</td> <td style="vertical-align: top;">R</td>
	1307	<td style="vertical-align: top;"><i>0.0</i></td>
	1308	<td style="vertical-align: top;"> <p>Surface
	1309	specific humidity / total water content (kg/kg).  </p> <p>This
	1310	parameter assigns the value of the specific humidity q at
	1311	the surface (k=0).  Starting from this value, the initial
	1312	humidity
[5]	1313	profile is constructed with  <a href="#q_vertical_gradient">q_vertical_gradient</a>
	1314	and <a href="#q_vertical_gradient_level">q_vertical_gradient_level</a>.
	1315	This profile is also used for the 1d-model as a stationary profile.</p>
[61]	1316	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="q_surface_initial_change"></a><b>q_surface_initial</b>
	1317	<br> <b>_change</b></p> </td> <td style="vertical-align: top;">R<br> </td> <td style="vertical-align: top;"><i>0.0</i></td>
	1318	<td style="vertical-align: top;"> <p>Change in
	1319	surface specific humidity / total water content to
[5]	1320	be made at the beginning
[61]	1321	of the 3d run (kg/kg).  </p> <p>If <b>q_surface_initial_change</b><i>
	1322	</i>is set to a
[5]	1323	non-zero value the
	1324	near surface latent heat flux (water flux) is not allowed to be given
	1325	simultaneously (see <a href="#surface_waterflux">surface_waterflux</a>).</p>
[61]	1326	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="q_vertical_gradient"></a><b>q_vertical_gradient</b></p>
	1327	</td> <td style="vertical-align: top;">R (10)</td>
	1328	<td style="vertical-align: top;"><i>10 * 0.0</i></td>
	1329	<td style="vertical-align: top;"> <p>Humidity
	1330	gradient(s) of the initial humidity profile
	1331	(in 1/100 m).  </p> <p>This humidity gradient
	1332	holds starting from the height
[5]	1333	level  defined by <a href="#q_vertical_gradient_level">q_vertical_gradient_level</a>
	1334	(precisely: for all uv levels k, where zu(k) >
	1335	q_vertical_gradient_level,
	1336	q_init(k) is set: q_init(k) = q_init(k-1) + dzu(k) * <b>q_vertical_gradient</b>)
	1337	up to the top boundary or up to the next height level defined
	1338	by <a href="#q_vertical_gradient_level">q_vertical_gradient_level</a>.
	1339	A total of 10 different gradients for 11 height intervals (10 intervals
	1340	if <a href="#q_vertical_gradient_level">q_vertical_gradient_level</a>(1)
[61]	1341	= <i>0.0</i>) can be asigned. The surface humidity is
	1342	assigned
[5]	1343	via <a href="#q_surface">q_surface</a>. </p>
[61]	1344	<p>Example:  </p> <ul> <p><b>q_vertical_gradient</b>
	1345	= <i>0.001</i>, <i>0.0005</i>,  <br>
	1346	<b>q_vertical_gradient_level</b> = <i>500.0</i>,
	1347	<i>1000.0</i>,</p> </ul>
[5]	1348	That defines the humidity to be constant with height up to z =
	1349	500.0
	1350	m with a
	1351	value given by <a href="#q_surface">q_surface</a>.
[61]	1352	For 500.0 m < z <= 1000.0 m the humidity gradient is
	1353	0.001 / 100
[5]	1354	m and for z > 1000.0 m up to the top boundary it is
	1355	0.0005 / 100 m (it is assumed that the assigned height levels
	1356	correspond with uv
[61]	1357	levels). </td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="q_vertical_gradient_level"></a><b>q_vertical_gradient</b>
	1358	<br> <b>_level</b></p> </td> <td style="vertical-align: top;">R (10)</td> <td style="vertical-align: top;"> <p><i>10 *</i>
	1359	<i>0.0</i></p> </td> <td style="vertical-align: top;"> <p>Height level from
[75]	1360	which on the humidity gradient defined by <a href="#q_vertical_gradient">q_vertical_gradient</a>
[61]	1361	is effective (in m).  </p> <p>The height levels
	1362	are to be assigned in ascending order. The
[5]	1363	default values result in a humidity constant with height regardless of
	1364	the values of <a href="#q_vertical_gradient">q_vertical_gradient</a>
	1365	(unless the top boundary of the model is higher than 100000.0 m). For
	1366	the piecewise construction of humidity profiles see <a href="#q_vertical_gradient">q_vertical_gradient</a>.</p>
[61]	1367	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="radiation"></a><b>radiation</b></p>
	1368	</td> <td style="vertical-align: top;">L</td>
	1369	<td style="vertical-align: top;"><i>.F.</i></td>
	1370	<td style="vertical-align: top;"> <p>Parameter to
	1371	switch on longwave radiation cooling at
	1372	cloud-tops.  </p> <p>Long-wave radiation
	1373	processes are parameterized by the
[5]	1374	effective emissivity, which considers only the absorption and emission
	1375	of long-wave radiation at cloud droplets. The radiation scheme can be
	1376	used only with <a href="#cloud_physics">cloud_physics</a>
[61]	1377	= .TRUE. .</p> </td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="random_generator"></a><b>random_generator</b></p>
	1378	</td> <td style="vertical-align: top;">C * 20</td>
	1379	<td style="vertical-align: top;"> <p><i>'numerical</i><br>
	1380	<i>recipes'</i></p> </td> <td style="vertical-align: top;"> <p>Random number
	1381	generator to be used for creating uniformly
	1382	distributed random numbers. <br> </p> <p>It is
	1383	used if random perturbations are to be imposed on the
[5]	1384	velocity field or on the surface heat flux field (see <a href="chapter_4.2.html#create_disturbances">create_disturbances</a>
	1385	and <a href="chapter_4.2.html#random_heatflux">random_heatflux</a>).
	1386	By default, the "Numerical Recipes" random number generator is used.
	1387	This one provides exactly the same order of random numbers on all
	1388	different machines and should be used in particular for comparison runs.<br>
[61]	1389	<br>
[5]	1390	Besides, a system-specific generator is available ( <b>random_generator</b>
[61]	1391	= <i>'system-specific')</i> which should particularly be
	1392	used for runs
[5]	1393	on vector parallel computers (NEC), because the default generator
	1394	cannot be vectorized and therefore significantly drops down the code
[61]	1395	performance on these machines.<br> </p> <span style="font-weight: bold;">Note:</span><br>
[5]	1396	Results from two otherwise identical model runs will not be comparable
[61]	1397	one-to-one if they used different random number generators.</td> </tr>
	1398	<tr> <td style="vertical-align: top;"> <p><a name="random_heatflux"></a><b>random_heatflux</b></p>
	1399	</td> <td style="vertical-align: top;">L</td>
	1400	<td style="vertical-align: top;"><i>.F.</i></td>
	1401	<td style="vertical-align: top;"> <p>Parameter to
	1402	impose random perturbations on the internal two-dimensional near
	1403	surface heat flux field <span style="font-style: italic;">shf</span>.
	1404	<br> </p>If a near surface heat flux is used as bottom
	1405	boundary
[5]	1406	condition (see <a href="#surface_heatflux">surface_heatflux</a>),
	1407	it is by default assumed to be horizontally homogeneous. Random
[61]	1408	perturbations can be imposed on the internal
	1409	two-dimensional heat flux field <span style="font-style: italic;">shf</span> by assigning <b>random_heatflux</b>
	1410	= <i>.T.</i>. The disturbed heat flux field is calculated
	1411	by
[5]	1412	multiplying the
	1413	values at each mesh point with a normally distributed random number
	1414	with a mean value and standard deviation of 1. This is repeated after
[61]	1415	every timestep.<br> <br>
	1416	In case of a non-flat <a href="#topography">topography</a>, assigning
	1417	<b>random_heatflux</b>
	1418	= <i>.T.</i> imposes random perturbations on the
	1419	combined heat
	1420	flux field <span style="font-style: italic;">shf</span>
	1421	composed of <a href="#surface_heatflux">surface_heatflux</a>
	1422	at the bottom surface and <a href="#wall_heatflux">wall_heatflux(0)</a>
	1423	at the topography top face.</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="rif_max"></a><b>rif_max</b></p>
	1424	</td> <td style="vertical-align: top;">R</td>
	1425	<td style="vertical-align: top;"><i>1.0</i></td>
	1426	<td style="vertical-align: top;"> <p>Upper limit of
	1427	the flux-Richardson number.  </p> <p>With the
	1428	Prandtl layer switched on (see <a href="#prandtl_layer">prandtl_layer</a>),
	1429	flux-Richardson numbers (rif) are calculated for z=z<sub>p</sub>
	1430	(k=1)
[5]	1431	in the 3d-model (in the 1d model for all heights). Their values in
	1432	particular determine the
	1433	values of the friction velocity (1d- and 3d-model) and the values of
	1434	the eddy diffusivity (1d-model). With small wind velocities at the
	1435	Prandtl layer top or small vertical wind shears in the 1d-model, rif
[61]	1436	can take up unrealistic large values. They are limited by an upper (<span style="font-weight: bold;">rif_max</span>) and lower
	1437	limit (see <a href="#rif_min">rif_min</a>)
	1438	for the flux-Richardson number. The condition <b>rif_max</b>
	1439	> <b>rif_min</b>
	1440	must be met.</p> </td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="rif_min"></a><b>rif_min</b></p>
	1441	</td> <td style="vertical-align: top;">R</td>
	1442	<td style="vertical-align: top;"><i>- 5.0</i></td>
	1443	<td style="vertical-align: top;"> <p>Lower limit of
	1444	the flux-Richardson number.  </p> <p>For further
	1445	explanations see <a href="#rif_max">rif_max</a>.
	1446	The condition <b>rif_max</b> > <b>rif_min </b>must
	1447	be met.</p> </td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="roughness_length"></a><b>roughness_length</b></p>
	1448	</td> <td style="vertical-align: top;">R</td>
	1449	<td style="vertical-align: top;"><i>0.1</i></td>
	1450	<td style="vertical-align: top;"> <p>Roughness
	1451	length (in m).  </p> <p>This parameter is
	1452	effective only in case that a Prandtl layer
[5]	1453	is switched
	1454	on (see <a href="#prandtl_layer">prandtl_layer</a>).</p>
[61]	1455	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="scalar_advec"></a><b>scalar_advec</b></p>
	1456	</td> <td style="vertical-align: top;">C * 10</td>
	1457	<td style="vertical-align: top;"><i>'pw-scheme'</i></td>
	1458	<td style="vertical-align: top;"> <p>Advection
	1459	scheme to be used for the scalar quantities.  </p> <p>The
	1460	user can choose between the following schemes:<br> </p> <p><span style="font-style: italic;">'pw-scheme'</span><br>
	1461	</p> <div style="margin-left: 40px;">The scheme of
	1462	Piascek and
[5]	1463	Williams (1970, J. Comp. Phys., 6,
	1464	392-405) with central differences in the form C3 is used.<br>
	1465	If intermediate Euler-timesteps are carried out in case of <a href="#timestep_scheme">timestep_scheme</a>
[61]	1466	= <span style="font-style: italic;">'leapfrog+euler'</span>
	1467	the
[5]	1468	advection scheme is - for the Euler-timestep - automatically switched
[61]	1469	to an upstream-scheme. <br> </div> <br> <p><span style="font-style: italic;">'bc-scheme'</span><br>
	1470	</p> <div style="margin-left: 40px;">The Bott
	1471	scheme modified by
[5]	1472	Chlond (1994, Mon.
	1473	Wea. Rev., 122, 111-125). This is a conservative monotonous scheme with
	1474	very small numerical diffusion and therefore very good conservation of
	1475	scalar flow features. The scheme however, is computationally very
	1476	expensive both because it is expensive itself and because it does (so
	1477	far) not allow specific code optimizations (e.g. cache optimization).
	1478	Choice of this
	1479	scheme forces the Euler timestep scheme to be used for the scalar
	1480	quantities. For output of horizontally averaged
	1481	profiles of the resolved / total heat flux, <a href="chapter_4.2.html#data_output_pr">data_output_pr</a>
[61]	1482	= <i>'wptBC'</i> / <i>'wptBC' </i>should
	1483	be used, instead of the
[5]	1484	standard profiles (<span style="font-style: italic;">'wpt'</span>
[61]	1485	and <span style="font-style: italic;">'wpt'</span>)
	1486	because these are
[5]	1487	too inaccurate with this scheme. However, for subdomain analysis (see <a href="#statistic_regions">statistic_regions</a>)
	1488	exactly the reverse holds: here <i>'wptBC'</i> and <i>'wptBC'</i>
[61]	1489	show very large errors and should not be used.<br> <br>
[5]	1490	This scheme is not allowed for non-cyclic lateral boundary conditions
	1491	(see <a href="#bc_lr">bc_lr</a>
[61]	1492	and <a href="#bc_ns">bc_ns</a>).<br> <br>
	1493	</div> <span style="font-style: italic;">'ups-scheme'</span><br>
	1494	<p style="margin-left: 40px;">The upstream-spline-scheme
	1495	is used
[5]	1496	(see Mahrer and Pielke,
	1497	1978: Mon. Wea. Rev., 106, 818-830). In opposite to the Piascek
	1498	Williams scheme, this is characterized by much better numerical
	1499	features (less numerical diffusion, better preservation of flux
	1500	structures, e.g. vortices), but computationally it is much more
	1501	expensive. In
	1502	addition, the use of the Euler-timestep scheme is mandatory (<a href="#timestep_scheme">timestep_scheme</a>
[61]	1503	= <span style="font-style: italic;">'</span><i>euler'</i>),
	1504	i.e. the
[5]	1505	timestep accuracy is only first order. For this reason the advection of
	1506	momentum (see <a href="#momentum_advec">momentum_advec</a>)
	1507	should then also be carried out with the upstream-spline scheme,
	1508	because otherwise the momentum would
	1509	be subject to large numerical diffusion due to the upstream
[61]	1510	scheme.  </p> <p style="margin-left: 40px;">Since
	1511	the cubic splines used tend
[5]	1512	to overshoot under
	1513	certain circumstances, this effect must be adjusted by suitable
	1514	filtering and smoothing (see <a href="#cut_spline_overshoot">cut_spline_overshoot</a>,
[61]	1515	<a href="#long_filter_factor">long_filter_factor</a>,
	1516	<a href="#ups_limit_pt">ups_limit_pt</a>, <a href="#ups_limit_u">ups_limit_u</a>, <a href="#ups_limit_v">ups_limit_v</a>, <a href="#ups_limit_w">ups_limit_w</a>).
[5]	1517	This is always neccesssary for runs with stable stratification,
	1518	even if this stratification appears only in parts of the model
[61]	1519	domain.  </p> <p style="margin-left: 40px;">With
	1520	stable stratification the
[5]	1521	upstream-upline scheme also produces gravity waves with large
	1522	amplitude, which must be
	1523	suitably damped (see <a href="chapter_4.2.html#rayleigh_damping_factor">rayleigh_damping_factor</a>).<br>
[61]	1524	</p> <p style="margin-left: 40px;"><span style="font-weight: bold;">Important: </span>The
[5]	1525	upstream-spline scheme is not implemented for humidity and passive
[75]	1526	scalars (see <a href="#humidity">humidity</a>
[5]	1527	and <a href="#passive_scalar">passive_scalar</a>)
	1528	and requires the use of a 2d-domain-decomposition. The last conditions
	1529	severely restricts code optimization on several machines leading to
	1530	very long execution times! This scheme is also not allowed for
	1531	non-cyclic lateral boundary conditions (see <a href="#bc_lr">bc_lr</a>
[61]	1532	and <a href="#bc_ns">bc_ns</a>).</p><br>A
	1533	differing advection scheme can be choosed for the subgrid-scale TKE
	1534	using parameter <a href="chapter_4.1.html#use_upstream_for_tke">use_upstream_for_tke</a>.</td>
	1535	</tr> <tr> <td style="vertical-align: top;">
	1536	<p><a name="statistic_regions"></a><b>statistic_regions</b></p>
	1537	</td> <td style="vertical-align: top;">I</td>
	1538	<td style="vertical-align: top;"><i>0</i></td>
	1539	<td style="vertical-align: top;"> <p>Number of
	1540	additional user-defined subdomains for which
[5]	1541	statistical analysis
[61]	1542	and corresponding output (profiles, time series) shall be
	1543	made.  </p> <p>By default, vertical profiles and
	1544	other statistical quantities
[5]	1545	are calculated as horizontal and/or volume average of the total model
	1546	domain. Beyond that, these calculations can also be carried out for
[61]	1547	subdomains which can be defined using the field <a href="chapter_3.5.3.html">rmask </a>within the
	1548	user-defined software
[5]	1549	(see <a href="chapter_3.5.3.html">chapter
	1550	3.5.3</a>). The number of these subdomains is determined with the
[61]	1551	parameter <b>statistic_regions</b>. Maximum 9 additional
	1552	subdomains
[5]	1553	are allowed. The parameter <a href="chapter_4.3.html#region">region</a>
	1554	can be used to assigned names (identifier) to these subdomains which
	1555	are then used in the headers
[61]	1556	of the output files and plots.</p><p>If the default NetCDF
	1557	output format is selected (see parameter <a href="chapter_4.2.html#data_output_format">data_output_format</a>),
	1558	data for the total domain and all defined subdomains are output to the
	1559	same file(s) (<a href="chapter_3.4.html#DATA_1D_PR_NETCDF">DATA_1D_PR_NETCDF</a>,
	1560	<a href="chapter_3.4.html#DATA_1D_TS_NETCDF">DATA_1D_TS_NETCDF</a>).
	1561	In case of <span style="font-weight: bold;">statistic_regions</span>
	1562	> <span style="font-style: italic;">0</span>,
[5]	1563	data on the file for the different domains can be distinguished by a
	1564	suffix which is appended to the quantity names. Suffix 0 means data for
[61]	1565	the total domain, suffix 1 means data for subdomain 1, etc.</p><p>In
	1566	case of <span style="font-weight: bold;">data_output_format</span>
	1567	= <span style="font-style: italic;">'profil'</span>,
	1568	individual local files for profiles (<a href="chapter_3.4.html#PLOT1D_DATA">PLOT1D_DATA</a>) are
	1569	created for each subdomain. The individual subdomain files differ by
	1570	their name (the
[5]	1571	number of the respective subdomain is attached, e.g.
[48]	1572	PLOT1D_DATA_1). In this case the name of the file with the data of
	1573	the total domain is PLOT1D_DATA_0. If no subdomains
[61]	1574	are declared (<b>statistic_regions</b> = <i>0</i>),
	1575	the name
[48]	1576	PLOT1D_DATA is used (this must be considered in the
[61]	1577	respective file connection statements of the <span style="font-weight: bold;">mrun</span> configuration
	1578	file).</p> </td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="surface_heatflux"></a><b>surface_heatflux</b></p>
	1579	</td> <td style="vertical-align: top;">R</td>
	1580	<td style="vertical-align: top;"><span style="font-style: italic;">no prescribed<br>
	1581	heatflux<br> </span></td> <td style="vertical-align: top;"> <p>Kinematic sensible
	1582	heat flux at the bottom surface (in K m/s).  </p> <p>If
	1583	a value is assigned to this parameter, the internal two-dimensional
	1584	surface heat flux field <span style="font-style: italic;">shf</span>
	1585	is initialized with the value of <span style="font-weight: bold;">surface_heatflux</span> as
	1586	bottom (horizontally homogeneous) boundary condition for the
[5]	1587	temperature equation. This additionally requires that a Neumann
	1588	condition must be used for the potential temperature (see <a href="#bc_pt_b">bc_pt_b</a>),
	1589	because otherwise the resolved scale may contribute to
	1590	the surface flux so that a constant value cannot be guaranteed. Also,
	1591	changes of the
	1592	surface temperature (see <a href="#pt_surface_initial_change">pt_surface_initial_change</a>)
	1593	are not allowed. The parameter <a href="#random_heatflux">random_heatflux</a>
[61]	1594	can be used to impose random perturbations on the (homogeneous) surface
	1595	heat
[5]	1596	flux field <span style="font-style: italic;">shf</span>. </p>
[61]	1597	<p>
	1598	In case of a non-flat <a href="#topography">topography</a>, the
	1599	internal two-dimensional surface heat
	1600	flux field <span style="font-style: italic;">shf</span>
	1601	is initialized with the value of <span style="font-weight: bold;">surface_heatflux</span>
	1602	at the bottom surface and <a href="#wall_heatflux">wall_heatflux(0)</a>
	1603	at the topography top face. The parameter<a href="#random_heatflux"> random_heatflux</a>
	1604	can be used to impose random perturbations on this combined surface
	1605	heat
	1606	flux field <span style="font-style: italic;">shf</span>.
	1607	</p> <p>If no surface heat flux is assigned, <span style="font-style: italic;">shf</span> is calculated
	1608	at each timestep by u<sub></sub> theta<sub>*</sub>
	1609	(of course only with <a href="#prandtl_layer">prandtl_layer</a>
	1610	switched on). Here, u<sub>*</sub>
	1611	and theta<sub>*</sub> are calculated from the Prandtl law
	1612	assuming
[5]	1613	logarithmic wind and temperature
	1614	profiles between k=0 and k=1. In this case a Dirichlet condition (see <a href="#bc_pt_b">bc_pt_b</a>)
[61]	1615	must be used as bottom boundary condition for the potential temperature.</p><p>See
	1616	also <a href="#top_heatflux">top_heatflux</a>.</p>
	1617	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="surface_pressure"></a><b>surface_pressure</b></p>
	1618	</td> <td style="vertical-align: top;">R</td>
	1619	<td style="vertical-align: top;"><i>1013.25</i></td>
	1620	<td style="vertical-align: top;"> <p>Atmospheric
	1621	pressure at the surface (in hPa).  </p>
[5]	1622	Starting from this surface value, the vertical pressure
	1623	profile is calculated once at the beginning of the run assuming a
	1624	neutrally stratified
	1625	atmosphere. This is needed for
	1626	converting between the liquid water potential temperature and the
	1627	potential temperature (see <a href="#cloud_physics">cloud_physics</a><span style="text-decoration: underline;"></span>).</td>
[61]	1628	</tr> <tr> <td style="vertical-align: top;">
	1629	<p><a name="surface_scalarflux"></a><b>surface_scalarflux</b></p>
	1630	</td> <td style="vertical-align: top;">R</td>
	1631	<td style="vertical-align: top;"><i>0.0</i></td>
	1632	<td style="vertical-align: top;"> <p>Scalar flux at
	1633	the surface (in kg/(m<sup>2</sup> s)).  </p>
	1634	<p>If a non-zero value is assigned to this parameter, the
[5]	1635	respective scalar flux value is used
	1636	as bottom (horizontally homogeneous) boundary condition for the scalar
[61]	1637	concentration equation. This additionally requires that a
	1638	Neumann
[5]	1639	condition must be used for the scalar concentration (see <a href="#bc_s_b">bc_s_b</a>),
	1640	because otherwise the resolved scale may contribute to
	1641	the surface flux so that a constant value cannot be guaranteed. Also,
	1642	changes of the
	1643	surface scalar concentration (see <a href="#s_surface_initial_change">s_surface_initial_change</a>)
[61]	1644	are not allowed. <br> </p> <p>If no surface scalar
	1645	flux is assigned (<b>surface_scalarflux</b>
[5]	1646	= <i>0.0</i>),
	1647	it is calculated at each timestep by u<sub></sub> s<sub>*</sub>
	1648	(of course only with Prandtl layer switched on). Here, s<sub>*</sub>
	1649	is calculated from the Prandtl law assuming a logarithmic scalar
	1650	concentration
	1651	profile between k=0 and k=1. In this case a Dirichlet condition (see <a href="#bc_s_b">bc_s_b</a>)
	1652	must be used as bottom boundary condition for the scalar concentration.</p>
[61]	1653	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="surface_waterflux"></a><b>surface_waterflux</b></p>
	1654	</td> <td style="vertical-align: top;">R</td>
	1655	<td style="vertical-align: top;"><i>0.0</i></td>
	1656	<td style="vertical-align: top;"> <p>Kinematic
	1657	water flux near the surface (in m/s).  </p> <p>If
	1658	a non-zero value is assigned to this parameter, the
[5]	1659	respective water flux value is used
	1660	as bottom (horizontally homogeneous) boundary condition for the
	1661	humidity equation. This additionally requires that a Neumann
	1662	condition must be used for the specific humidity / total water content
	1663	(see <a href="#bc_q_b">bc_q_b</a>),
	1664	because otherwise the resolved scale may contribute to
	1665	the surface flux so that a constant value cannot be guaranteed. Also,
	1666	changes of the
	1667	surface humidity (see <a href="#q_surface_initial_change">q_surface_initial_change</a>)
[61]	1668	are not allowed.<br> </p> <p>If no surface water
	1669	flux is assigned (<b>surface_waterflux</b>
[5]	1670	= <i>0.0</i>),
	1671	it is calculated at each timestep by u<sub></sub> q<sub>*</sub>
	1672	(of course only with Prandtl layer switched on). Here, q<sub>*</sub>
	1673	is calculated from the Prandtl law assuming a logarithmic temperature
	1674	profile between k=0 and k=1. In this case a Dirichlet condition (see <a href="#bc_q_b">bc_q_b</a>)
	1675	must be used as the bottom boundary condition for the humidity.</p>
[61]	1676	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="s_surface"></a><b>s_surface</b></p>
	1677	</td> <td style="vertical-align: top;">R</td>
	1678	<td style="vertical-align: top;"><i>0.0</i></td>
	1679	<td style="vertical-align: top;"> <p>Surface value
	1680	of the passive scalar (in kg/m<sup>3</sup>). <br>
	1681	</p>
[5]	1682	This parameter assigns the value of the passive scalar s at
	1683	the surface (k=0)<b>.</b> Starting from this value, the
	1684	initial vertical scalar concentration profile is constructed with<a href="#s_vertical_gradient">
	1685	s_vertical_gradient</a> and <a href="#s_vertical_gradient_level">s_vertical_gradient_level</a>.</td>
[61]	1686	</tr> <tr> <td style="vertical-align: top;">
	1687	<p><a name="s_surface_initial_change"></a><b>s_surface_initial</b>
	1688	<br> <b>_change</b></p> </td> <td style="vertical-align: top;">R</td> <td style="vertical-align: top;"><i>0.0</i></td>
	1689	<td style="vertical-align: top;"> <p>Change in
	1690	surface scalar concentration to be made at the
[5]	1691	beginning of the 3d run (in kg/m<sup>3</sup>).  </p>
[61]	1692	<p>If <b>s_surface_initial_change</b><i> </i>is
	1693	set to a
[5]	1694	non-zero
	1695	value, the near surface scalar flux is not allowed to be given
	1696	simultaneously (see <a href="#surface_scalarflux">surface_scalarflux</a>).</p>
[61]	1697	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="s_vertical_gradient"></a><b>s_vertical_gradient</b></p>
	1698	</td> <td style="vertical-align: top;">R (10)</td>
	1699	<td style="vertical-align: top;"><i>10 * 0</i><i>.0</i></td>
	1700	<td style="vertical-align: top;"> <p>Scalar
	1701	concentration gradient(s) of the initial scalar
[5]	1702	concentration profile (in kg/m<sup>3 </sup>/
[61]	1703	100 m).  </p> <p>The scalar gradient holds
	1704	starting from the height level
[5]	1705	defined by <a href="#s_vertical_gradient_level">s_vertical_gradient_level
[61]	1706	</a>(precisely: for all uv levels k, where zu(k) >
[5]	1707	s_vertical_gradient_level, s_init(k) is set: s_init(k) = s_init(k-1) +
[61]	1708	dzu(k) * <b>s_vertical_gradient</b>) up to the top
	1709	boundary or up to
[5]	1710	the next height level defined by <a href="#s_vertical_gradient_level">s_vertical_gradient_level</a>.
	1711	A total of 10 different gradients for 11 height intervals (10 intervals
	1712	if <a href="#s_vertical_gradient_level">s_vertical_gradient_level</a>(1)
[61]	1713	= <i>0.0</i>) can be assigned. The surface scalar value is
	1714	assigned
	1715	via <a href="#s_surface">s_surface</a>.<br> </p>
	1716	<p>Example:  </p> <ul> <p><b>s_vertical_gradient</b>
	1717	= <i>0.1</i>, <i>0.05</i>,  <br>
	1718	<b>s_vertical_gradient_level</b> = <i>500.0</i>,
	1719	<i>1000.0</i>,</p> </ul> <p>That
	1720	defines the scalar concentration to be constant with
[5]	1721	height up to z = 500.0 m with a value given by <a href="#s_surface">s_surface</a>.
	1722	For 500.0 m < z <= 1000.0 m the scalar gradient is 0.1
[61]	1723	kg/m<sup>3 </sup>/ 100 m and for z > 1000.0 m up to
	1724	the top
	1725	boundary it is 0.05 kg/m<sup>3 </sup>/ 100 m (it is
	1726	assumed that the
[5]	1727	assigned height levels
	1728	correspond with uv
[61]	1729	levels).</p> </td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="s_vertical_gradient_level"></a><b>s_vertical_gradient_</b>
	1730	<br> <b>level</b></p> </td> <td style="vertical-align: top;">R (10)</td> <td style="vertical-align: top;"> <p><i>10 *</i>
	1731	<i>0.0</i></p> </td> <td style="vertical-align: top;"> <p>Height level from
	1732	which on the scalar gradient defined by <a href="#s_vertical_gradient">s_vertical_gradient</a>
	1733	is effective (in m).  </p> <p>The height levels
	1734	are to be assigned in ascending order. The
[5]	1735	default values result in a scalar concentration constant with height
	1736	regardless of the values of <a href="#s_vertical_gradient">s_vertical_gradient</a>
	1737	(unless the top boundary of the model is higher than 100000.0 m). For
	1738	the
	1739	piecewise construction of scalar concentration profiles see <a href="#s_vertical_gradient">s_vertical_gradient</a>.</p>
[61]	1740	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="timestep_scheme"></a><b>timestep_scheme</b></p>
	1741	</td> <td style="vertical-align: top;">C * 20</td>
	1742	<td style="vertical-align: top;"> <p><i>'runge</i><br>
	1743	<i>kutta-3'</i></p> </td> <td style="vertical-align: top;"> <p>Time step scheme to
	1744	be used for the integration of the prognostic
	1745	variables.  </p> <p>The user can choose between
	1746	the following schemes:<br> </p> <p><span style="font-style: italic;">'runge-kutta-3'</span><br>
	1747	</p> <div style="margin-left: 40px;">Third order
	1748	Runge-Kutta scheme.<br>
[5]	1749	This scheme requires the use of <a href="#momentum_advec">momentum_advec</a>
	1750	= <a href="#scalar_advec">scalar_advec</a>
[61]	1751	= '<i>pw-scheme'</i>. Please refer to the <a href="../tec/numerik.heiko/zeitschrittverfahren.pdf">documentation
	1752	on PALM's time integration schemes (28p., in German)</a>
	1753	fur further details.<br> </div> <p><span style="font-style: italic;">'runge-kutta-2'</span><br>
	1754	</p> <div style="margin-left: 40px;">Second order
	1755	Runge-Kutta scheme.<br>
[5]	1756	For special features see <b>timestep_scheme</b> = '<i>runge-kutta-3'</i>.<br>
[61]	1757	</div> <br> <span style="font-style: italic;"><span style="font-style: italic;">'leapfrog'</span><br>
	1758	<br> </span> <div style="margin-left: 40px;">Second
	1759	order leapfrog scheme.<br>
[5]	1760	Although this scheme requires a constant timestep (because it is
	1761	centered in time),  is even applied in case of changes in
	1762	timestep. Therefore, only small
	1763	changes of the timestep are allowed (see <a href="#dt">dt</a>).
	1764	However, an Euler timestep is always used as the first timestep of an
	1765	initiali run. When using the Bott-Chlond scheme for scalar advection
	1766	(see <a href="#scalar_advec">scalar_advec</a>),
	1767	the prognostic equation for potential temperature will be calculated
	1768	with the Euler scheme, although the leapfrog scheme is switched
	1769	on.  <br>
	1770	The leapfrog scheme must not be used together with the upstream-spline
	1771	scheme for calculating the advection (see <a href="#scalar_advec">scalar_advec</a>
	1772	= '<i>ups-scheme'</i> and <a href="#momentum_advec">momentum_advec</a>
[61]	1773	= '<i>ups-scheme'</i>).<br> </div> <br>
	1774	<span style="font-style: italic;">'</span><span style="font-style: italic;"><span style="font-style: italic;">leapfrog+euler'</span><br>
	1775	<br> </span> <div style="margin-left: 40px;">The
	1776	leapfrog scheme is used, but
[5]	1777	after each change of a timestep an Euler timestep is carried out.
	1778	Although this method is theoretically correct (because the pure
	1779	leapfrog method does not allow timestep changes), the divergence of the
	1780	velocity field (after applying the pressure solver) may be
	1781	significantly larger than with <span style="font-style: italic;">'leapfrog'</span>.<br>
[61]	1782	</div> <br> <span style="font-style: italic;">'euler'</span><br>
	1783	<br> <div style="margin-left: 40px;">First order
	1784	Euler scheme.  <br>
[5]	1785	The Euler scheme must be used when treating the advection terms with
	1786	the upstream-spline scheme (see <a href="#scalar_advec">scalar_advec</a>
[61]	1787	= <span style="font-style: italic;">'ups-scheme'</span>
	1788	and <a href="#momentum_advec">momentum_advec</a>
[5]	1789	= <span style="font-style: italic;">'ups-scheme'</span>).</div>
[61]	1790	<br><br>A differing timestep scheme can be choosed for the
	1791	subgrid-scale TKE using parameter <a href="#use_upstream_for_tke">use_upstream_for_tke</a>.<br>
	1792	</td> </tr> <tr> <td style="text-align: left; vertical-align: top;"><span style="font-weight: bold;"><a name="topography"></a></span><span style="font-weight: bold;">topography</span></td>
	1793	<td style="vertical-align: top;">C * 40</td> <td style="vertical-align: top;"><span style="font-style: italic;">'flat'</span></td> <td>
	1794	<p>Topography mode.  </p> <p>The user can
	1795	choose between the following modes:<br> </p> <p><span style="font-style: italic;">'flat'</span><br> </p>
	1796	<div style="margin-left: 40px;">Flat surface.</div> <p><span style="font-style: italic;">'single_building'</span><br>
	1797	</p> <div style="margin-left: 40px;">Flow
	1798	around a single rectangular building mounted on a flat surface.<br>
[5]	1799	The building size and location can be specified with the parameters <a href="#building_height">building_height</a>, <a href="#building_length_x">building_length_x</a>, <a href="#building_length_y">building_length_y</a>, <a href="#building_wall_left">building_wall_left</a> and <a href="#building_wall_south">building_wall_south</a>.</div>
[61]	1800	<span style="font-style: italic;"></span> <p><span style="font-style: italic;">'read_from_file'</span><br>
	1801	</p> <div style="margin-left: 40px;">Flow around
	1802	arbitrary topography.<br>
	1803	This mode requires the input file <a href="chapter_3.4.html#TOPOGRAPHY_DATA">TOPOGRAPHY_DATA</a><font color="#000000">. This file contains </font><font color="#000000"><font color="#000000">the </font></font><font color="#000000">arbitrary topography </font><font color="#000000"><font color="#000000">height
	1804	information</font></font><font color="#000000">
	1805	in m. These data <span style="font-style: italic;"></span>must
	1806	exactly match the horizontal grid.</font> </div> <span style="font-style: italic;"><br> </span><font color="#000000">
	1807	Alternatively, the user may add code to the user interface subroutine <a href="chapter_3.5.1.html#user_init_grid">user_init_grid</a>
	1808	to allow further topography modes.<br> <br>
	1809	All non-flat <span style="font-weight: bold;">topography</span>
	1810	modes </font>require the use of <a href="#momentum_advec">momentum_advec</a>
[5]	1811	= <a href="#scalar_advec">scalar_advec</a>
[61]	1812	= '<i>pw-scheme'</i>, <a href="chapter_4.2.html#psolver">psolver</a>
	1813	= <i>'poisfft'</i> or '<i>poisfft_hybrid'</i>,
	1814	<i> </i><a href="#alpha_surface">alpha_surface</a>
	1815	= 0.0, <a href="#bc_lr">bc_lr</a> = <a href="#bc_ns">bc_ns</a> = <span style="font-style: italic;">'cyclic'</span>, <a style="" href="#galilei_transformation">galilei_transformation</a>
[75]	1816	= <span style="font-style: italic;">.F.</span>, <a href="#cloud_physics">cloud_physics </a> = <span style="font-style: italic;">.F.</span>,  <a href="#cloud_droplets">cloud_droplets</a> = <span style="font-style: italic;">.F.</span>,  <a href="#humidity">humidity</a> = <span style="font-style: italic;">.F.</span>, and <a href="#prandtl_layer">prandtl_layer</a> = .T..<br>
[61]	1817	<font color="#000000"><br>
	1818	Note that an inclined model domain requires the use of <span style="font-weight: bold;">topography</span> = <span style="font-style: italic;">'flat'</span> and a
	1819	nonzero </font><a href="#alpha_surface">alpha_surface</a>.</td>
	1820	</tr> <tr><td style="vertical-align: top;"><a name="top_heatflux"></a><span style="font-weight: bold;">top_heatflux</span></td><td style="vertical-align: top;">R</td><td style="vertical-align: top;"><span style="font-style: italic;">no prescribed<br>
	1821	heatflux</span></td><td style="vertical-align: top;"><p>Kinematic
	1822	sensible heat flux at the top boundary (in K m/s).  </p>
	1823	<p>If a value is assigned to this parameter, the internal
	1824	two-dimensional surface heat flux field <span style="font-family: monospace;">tswst</span> is
	1825	initialized with the value of <span style="font-weight: bold;">top_heatflux</span> as
	1826	top (horizontally homogeneous) boundary condition for the
[40]	1827	temperature equation. This additionally requires that a Neumann
	1828	condition must be used for the potential temperature (see <a href="chapter_4.1.html#bc_pt_t">bc_pt_t</a>),
	1829	because otherwise the resolved scale may contribute to
	1830	the top flux so that a constant value cannot be guaranteed.<span style="font-style: italic;"></span> </p>
[61]	1831	<p><span style="font-weight: bold;">Note:</span><br>The
	1832	application of a top heat flux additionally requires the setting of
	1833	initial parameter <a href="#use_top_fluxes">use_top_fluxes</a>
	1834	= .T..<span style="font-style: italic;"></span><span style="font-weight: bold;"></span> </p><p>No
	1835	Prandtl-layer is available at the top boundary so far.</p><p>See
	1836	also <a href="#surface_heatflux">surface_heatflux</a>.</p>
	1837	</td></tr><tr> <td style="vertical-align: top;">
	1838	<p><a name="ug_surface"></a><span style="font-weight: bold;">ug_surface</span></p>
	1839	</td> <td style="vertical-align: top;">R<br> </td>
	1840	<td style="vertical-align: top;"><span style="font-style: italic;">0.0</span><br> </td>
	1841	<td style="vertical-align: top;">u-component of the
	1842	geostrophic
	1843	wind at the surface (in m/s).<br> <br>
[5]	1844	This parameter assigns the value of the u-component of the geostrophic
	1845	wind (ug) at the surface (k=0). Starting from this value, the initial
	1846	vertical profile of the <br>
[61]	1847	u-component of the geostrophic wind is constructed with <a href="#ug_vertical_gradient">ug_vertical_gradient</a>
	1848	and <a href="#ug_vertical_gradient_level">ug_vertical_gradient_level</a>.
	1849	The
[5]	1850	profile constructed in that way is used for creating the initial
	1851	vertical velocity profile of the 3d-model. Either it is applied, as it
	1852	has been specified by the user (<a href="#initializing_actions">initializing_actions</a>
	1853	= 'set_constant_profiles') or it is used for calculating a stationary
	1854	boundary layer wind profile (<a href="#initializing_actions">initializing_actions</a>
[61]	1855	= 'set_1d-model_profiles'). If ug is constant with height (i.e. ug(k)=<span style="font-weight: bold;">ug_surface</span>)
	1856	and  has a large
[5]	1857	value, it is recommended to use a Galilei-transformation of the
	1858	coordinate system, if possible (see <a href="#galilei_transformation">galilei_transformation</a>),
[61]	1859	in order to obtain larger time steps.<br> </td> </tr>
	1860	<tr> <td style="vertical-align: top;"> <p><a name="ug_vertical_gradient"></a><span style="font-weight: bold;">ug_vertical_gradient</span></p>
	1861	</td> <td style="vertical-align: top;">R(10)<br>
	1862	</td> <td style="vertical-align: top;"><span style="font-style: italic;">10
	1863	* 0.0</span><br> </td> <td style="vertical-align: top;">Gradient(s) of the initial
	1864	profile of the  u-component of the geostrophic wind (in
	1865	1/100s).<br> <br>
[5]	1866	The gradient holds starting from the height level defined by <a href="#ug_vertical_gradient_level">ug_vertical_gradient_level</a>
	1867	(precisely: for all uv levels k where zu(k) > <a href="#ug_vertical_gradient_level">ug_vertical_gradient_level</a>,
[61]	1868	ug(k) is set: ug(k) = ug(k-1) + dzu(k) * <span style="font-weight: bold;">ug_vertical_gradient</span>)
	1869	up to the top
	1870	boundary or up to the next height level defined by <a href="#ug_vertical_gradient_level">ug_vertical_gradient_level</a>.
	1871	A
[5]	1872	total of 10 different gradients for 11 height intervals (10
	1873	intervals  if <a href="#ug_vertical_gradient_level">ug_vertical_gradient_level</a>(1)
[61]	1874	= 0.0) can be assigned. The surface geostrophic wind is assigned by <a href="#ug_surface">ug_surface</a>. <br> </td>
	1875	</tr> <tr> <td style="vertical-align: top;">
	1876	<p><a name="ug_vertical_gradient_level"></a><span style="font-weight: bold;">ug_vertical_gradient_level</span></p>
	1877	</td> <td style="vertical-align: top;">R(10)<br>
	1878	</td> <td style="vertical-align: top;"><span style="font-style: italic;">10
	1879	* 0.0</span><br> </td> <td style="vertical-align: top;">Height level from which on the
[5]	1880	gradient defined by <a href="#ug_vertical_gradient">ug_vertical_gradient</a>
[61]	1881	is effective (in m).<br> <br>
[5]	1882	The height levels are to be assigned in ascending order. For the
	1883	piecewise construction of a profile of the u-component of the
	1884	geostrophic wind component (ug) see <a href="#ug_vertical_gradient">ug_vertical_gradient</a>.<br>
[61]	1885	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="ups_limit_e"></a><b>ups_limit_e</b></p>
	1886	</td> <td style="vertical-align: top;">R</td>
	1887	<td style="vertical-align: top;"><i>0.0</i></td>
	1888	<td style="vertical-align: top;"> <p>Subgrid-scale
	1889	turbulent kinetic energy difference used as
[5]	1890	criterion for applying the upstream scheme when upstream-spline
	1891	advection is switched on (in m<sup>2</sup>/s<sup>2</sup>).
[61]	1892	</p> <p>This variable steers the appropriate
	1893	treatment of the
[5]	1894	advection of the subgrid-scale turbulent kinetic energy in case that
	1895	the uptream-spline scheme is used . For further information see <a href="#ups_limit_pt">ups_limit_pt</a>.  </p>
[61]	1896	<p>Only positive values are allowed for <b>ups_limit_e</b>.
	1897	</p> </td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="ups_limit_pt"></a><b>ups_limit_pt</b></p>
	1898	</td> <td style="vertical-align: top;">R</td>
	1899	<td style="vertical-align: top;"><i>0.0</i></td>
	1900	<td style="vertical-align: top;"> <p>Temperature
	1901	difference used as criterion for applying
	1902	the upstream scheme when upstream-spline advection  is
	1903	switched on
	1904	(in K).  </p> <p>This criterion is used if the
	1905	upstream-spline scheme is
[5]	1906	switched on (see <a href="#scalar_advec">scalar_advec</a>).<br>
	1907	If, for a given gridpoint, the absolute temperature difference with
	1908	respect to the upstream
	1909	grid point is smaller than the value given for <b>ups_limit_pt</b>,
	1910	the upstream scheme is used for this gridpoint (by default, the
	1911	upstream-spline scheme is always used). Reason: in case of a very small
	1912	upstream gradient, the advection should cause only a very small
	1913	tendency. However, in such situations the upstream-spline scheme may
	1914	give wrong tendencies at a
	1915	grid point due to spline overshooting, if simultaneously the downstream
	1916	gradient is very large. In such cases it may be more reasonable to use
	1917	the upstream scheme. The numerical diffusion caused by the upstream
	1918	schme remains small as long as the upstream gradients are small.<br>
[61]	1919	</p> <p>The percentage of grid points for which the
	1920	upstream
[5]	1921	scheme is actually used, can be output as a time series with respect to
	1922	the
[61]	1923	three directions in space with run parameter (see <a href="chapter_4.2.html#dt_dots">dt_dots</a>, the
	1924	timeseries names in the NetCDF file are <i>'splptx'</i>, <i>'splpty'</i>,
	1925	<i>'splptz'</i>). The percentage
	1926	of gridpoints  should stay below a certain limit, however, it
	1927	is
[5]	1928	not possible to give
	1929	a general limit, since it depends on the respective flow.  </p>
[61]	1930	<p>Only positive values are permitted for <b>ups_limit_pt</b>.<br>
	1931	</p>
[5]	1932	A more effective control of
[61]	1933	the “overshoots” can be achieved with parameter <a href="#cut_spline_overshoot">cut_spline_overshoot</a>.
	1934	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="ups_limit_u"></a><b>ups_limit_u</b></p>
	1935	</td> <td style="vertical-align: top;">R</td>
	1936	<td style="vertical-align: top;"><i>0.0</i></td>
	1937	<td style="vertical-align: top;"> <p>Velocity
	1938	difference (u-component) used as criterion for
[5]	1939	applying the upstream scheme
	1940	when upstream-spline advection is switched on (in m/s).  </p>
[61]	1941	<p>This variable steers the appropriate treatment of the
[5]	1942	advection of the u-velocity-component in case that the upstream-spline
	1943	scheme is used. For further
[61]	1944	information see <a href="#ups_limit_pt">ups_limit_pt</a>.
	1945	</p> <p>Only positive values are permitted for <b>ups_limit_u</b>.</p>
	1946	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="ups_limit_v"></a><b>ups_limit_v</b></p>
	1947	</td> <td style="vertical-align: top;">R</td>
	1948	<td style="vertical-align: top;"><i>0.0</i></td>
	1949	<td style="vertical-align: top;"> <p>Velocity
	1950	difference (v-component) used as criterion for
[5]	1951	applying the upstream scheme
	1952	when upstream-spline advection is switched on (in m/s).  </p>
[61]	1953	<p>This variable steers the appropriate treatment of the
[5]	1954	advection of the v-velocity-component in case that the upstream-spline
	1955	scheme is used. For further
[61]	1956	information see <a href="#ups_limit_pt">ups_limit_pt</a>.
	1957	</p> <p>Only positive values are permitted for <b>ups_limit_v</b>.</p>
	1958	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="ups_limit_w"></a><b>ups_limit_w</b></p>
	1959	</td> <td style="vertical-align: top;">R</td>
	1960	<td style="vertical-align: top;"><i>0.0</i></td>
	1961	<td style="vertical-align: top;"> <p>Velocity
	1962	difference (w-component) used as criterion for
[5]	1963	applying the upstream scheme
	1964	when upstream-spline advection is switched on (in m/s).  </p>
[61]	1965	<p>This variable steers the appropriate treatment of the
[5]	1966	advection of the w-velocity-component in case that the upstream-spline
	1967	scheme is used. For further
[61]	1968	information see <a href="#ups_limit_pt">ups_limit_pt</a>.
	1969	</p> <p>Only positive values are permitted for <b>ups_limit_w</b>.</p>
	1970	</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="use_surface_fluxes"></a><b>use_surface_fluxes</b></p>
	1971	</td> <td style="vertical-align: top;">L</td>
	1972	<td style="vertical-align: top;"><i>.F.</i></td>
	1973	<td style="vertical-align: top;"> <p>Parameter to
	1974	steer the treatment of the subgrid-scale vertical
	1975	fluxes within the diffusion terms at k=1 (bottom boundary).<br> </p>
	1976	<p>By default, the near-surface subgrid-scale fluxes are
[5]	1977	parameterized (like in the remaining model domain) using the gradient
	1978	approach. If <b>use_surface_fluxes</b>
[61]	1979	= <i>.TRUE.</i>, the user-assigned surface fluxes are used
	1980	instead
	1981	(see <a href="#surface_heatflux">surface_heatflux</a>,
	1982	<a href="#surface_waterflux">surface_waterflux</a>
	1983	and <a href="#surface_scalarflux">surface_scalarflux</a>)
	1984	<span style="font-weight: bold;">or</span> the
	1985	surface fluxes are
[5]	1986	calculated via the Prandtl layer relation (depends on the bottom
[61]	1987	boundary conditions, see <a href="#bc_pt_b">bc_pt_b</a>,
	1988	<a href="#bc_q_b">bc_q_b</a>
	1989	and <a href="#bc_s_b">bc_s_b</a>).<br> </p>
	1990	<p><b>use_surface_fluxes</b>
	1991	is automatically set <i>.TRUE.</i>, if a Prandtl layer is
	1992	used (see <a href="#prandtl_layer">prandtl_layer</a>).
	1993	</p> <p>The user may prescribe the surface fluxes at the
	1994	bottom
	1995	boundary without using a Prandtl layer by setting <span style="font-weight: bold;">use_surface_fluxes</span> =
	1996	<span style="font-style: italic;">.T.</span> and <span style="font-weight: bold;">prandtl_layer</span> = <span style="font-style: italic;">.F.</span>. If , in this
	1997	case, the
	1998	momentum flux (u<sub>*</sub><sup>2</sup>)
	1999	should also be prescribed,
[5]	2000	the user must assign an appropriate value within the user-defined code.</p>
[61]	2001	</td> </tr> <tr><td style="vertical-align: top;"><a name="use_top_fluxes"></a><span style="font-weight: bold;">use_top_fluxes</span></td><td style="vertical-align: top;">L</td><td style="vertical-align: top;"><span style="font-style: italic;">.F.</span></td><td style="vertical-align: top;"> <p>Parameter to steer
	2002	the treatment of the subgrid-scale vertical
	2003	fluxes within the diffusion terms at k=nz (top boundary).</p><p>By
	2004	default, the fluxes at nz are calculated using the gradient approach.
	2005	If <b>use_top_fluxes</b>
	2006	= <i>.TRUE.</i>, the user-assigned top fluxes are used
	2007	instead
	2008	(see <a href="chapter_4.1.html#top_heatflux">top_heatflux</a>).</p><p>Currently,
	2009	only a value for the sensible heatflux can be assigned. In case of <span style="font-weight: bold;">use_top_fluxes</span> = <span style="font-style: italic;">.TRUE.</span>, the latent
	2010	heat flux at the top will be automatically set to zero.</p></td></tr><tr>
	2011	<td style="vertical-align: top;"> <p><a name="use_ug_for_galilei_tr"></a><b>use_ug_for_galilei_tr</b></p>
	2012	</td> <td style="vertical-align: top;">L</td>
	2013	<td style="vertical-align: top;"><i>.T.</i></td>
	2014	<td style="vertical-align: top;"> <p>Switch to
	2015	determine the translation velocity in case that a
	2016	Galilean transformation is used.<br> </p> <p>In
	2017	case of a Galilean transformation (see <a href="#galilei_transformation">galilei_transformation</a>),
	2018	<b>use_ug_for_galilei_tr</b>
[5]	2019	= <i>.T.</i>  ensures
	2020	that the coordinate system is translated with the geostrophic windspeed.<br>
[61]	2021	</p> <p>Alternatively, with <b>use_ug_for_galilei_tr</b>
	2022	= <i>.F</i>.,
[5]	2023	the
	2024	geostrophic wind can be replaced as translation speed by the (volume)
	2025	averaged velocity. However, in this case the user must be aware of fast
	2026	growing gravity waves, so this
[61]	2027	choice is usually not recommended!</p> </td> </tr> <tr><td align="left" valign="top"><a name="use_upstream_for_tke"></a><span style="font-weight: bold;">use_upstream_for_tke</span></td><td align="left" valign="top">L</td><td align="left" valign="top"><span style="font-style: italic;">.F.</span></td><td align="left" valign="top">Parameter to choose the
	2028	advection/timestep scheme to be used for the subgrid-scale TKE.<br><br>By
[5]	2029	default, the advection scheme and the timestep scheme to be used for
[61]	2030	the subgrid-scale TKE are set by the initialization parameters <a href="#scalar_advec">scalar_advec</a> and <a href="#timestep_scheme">timestep_scheme</a>,
	2031	respectively. <span style="font-weight: bold;">use_upstream_for_tke</span>
	2032	= <span style="font-style: italic;">.T.</span>
[5]	2033	forces the Euler-scheme and the upstream-scheme to be used as timestep
	2034	scheme and advection scheme, respectively. By these methods, the strong
	2035	(artificial) near-surface vertical gradients of the subgrid-scale TKE
	2036	are significantly reduced. This is required when subgrid-scale
	2037	velocities are used for advection of particles (see particle package
	2038	parameter <a href="chapter_4.2.html#use_sgs_for_particles">use_sgs_for_particles</a>).</td></tr><tr>
[61]	2039	<td style="vertical-align: top;"> <p><a name="vg_surface"></a><span style="font-weight: bold;">vg_surface</span></p>
	2040	</td> <td style="vertical-align: top;">R<br> </td>
	2041	<td style="vertical-align: top;"><span style="font-style: italic;">0.0</span><br> </td>
	2042	<td style="vertical-align: top;">v-component of the
	2043	geostrophic
	2044	wind at the surface (in m/s).<br> <br>
[5]	2045	This parameter assigns the value of the v-component of the geostrophic
	2046	wind (vg) at the surface (k=0). Starting from this value, the initial
	2047	vertical profile of the <br>
[61]	2048	v-component of the geostrophic wind is constructed with <a href="#vg_vertical_gradient">vg_vertical_gradient</a>
	2049	and <a href="#vg_vertical_gradient_level">vg_vertical_gradient_level</a>.
	2050	The
[5]	2051	profile
	2052	constructed in that way is used for creating the initial vertical
	2053	velocity profile of the 3d-model. Either it is applied, as it has been
	2054	specified by the user (<a href="#initializing_actions">initializing_actions</a>
	2055	= 'set_constant_profiles')
	2056	or it is used for calculating a stationary boundary layer wind profile
[61]	2057	(<a href="#initializing_actions">initializing_actions</a>
	2058	=
[5]	2059	'set_1d-model_profiles'). If vg is constant
	2060	with height (i.e. vg(k)=<span style="font-weight: bold;">vg_surface</span>)
	2061	and  has a large value, it is
	2062	recommended to use a Galilei-transformation of the coordinate system,
	2063	if possible (see <a href="#galilei_transformation">galilei_transformation</a>),
	2064	in order to obtain larger
[61]	2065	time steps.</td> </tr> <tr> <td style="vertical-align: top;"> <p><a name="vg_vertical_gradient"></a><span style="font-weight: bold;">vg_vertical_gradient</span></p>
	2066	</td> <td style="vertical-align: top;">R(10)<br>
	2067	</td> <td style="vertical-align: top;"><span style="font-style: italic;">10
	2068	* 0.0</span><br> </td> <td style="vertical-align: top;">Gradient(s) of the initial
	2069	profile of the  v-component of the geostrophic wind (in
	2070	1/100s).<br> <br>
[5]	2071	The gradient holds starting from the height level defined by <a href="#vg_vertical_gradient_level">vg_vertical_gradient_level</a>
	2072	(precisely: for all uv levels k where zu(k)
	2073	> <a href="#vg_vertical_gradient_level">vg_vertical_gradient_level</a>,
	2074	vg(k) is set: vg(k) = vg(k-1) + dzu(k)
[61]	2075	* <span style="font-weight: bold;">vg_vertical_gradient</span>)
	2076	up to
[5]	2077	the top boundary or up to the next height
	2078	level defined by <a href="#vg_vertical_gradient_level">vg_vertical_gradient_level</a>.
	2079	A total of 10 different
[61]	2080	gradients for 11 height intervals (10 intervals  if <a href="#vg_vertical_gradient_level">vg_vertical_gradient_level</a>(1)
	2081	=
[5]	2082	0.0) can be assigned. The surface
	2083	geostrophic wind is assigned by <a href="#vg_surface">vg_surface</a>.</td>
[61]	2084	</tr> <tr> <td style="vertical-align: top;">
	2085	<p><a name="vg_vertical_gradient_level"></a><span style="font-weight: bold;">vg_vertical_gradient_level</span></p>
	2086	</td> <td style="vertical-align: top;">R(10)<br>
	2087	</td> <td style="vertical-align: top;"><span style="font-style: italic;">10
	2088	* 0.0</span><br> </td> <td style="vertical-align: top;">Height level from which on the
[5]	2089	gradient defined by <a href="#vg_vertical_gradient">vg_vertical_gradient</a>
[61]	2090	is effective (in m).<br> <br>
[5]	2091	The height levels are to be assigned in ascending order. For the
	2092	piecewise construction of a profile of the v-component of the
	2093	geostrophic wind component (vg) see <a href="#vg_vertical_gradient">vg_vertical_gradient</a>.</td>
[61]	2094	</tr> <tr> <td style="vertical-align: top;">
	2095	<p><a name="wall_adjustment"></a><b>wall_adjustment</b></p>
	2096	</td> <td style="vertical-align: top;">L</td>
	2097	<td style="vertical-align: top;"><i>.T.</i></td>
	2098	<td style="vertical-align: top;"> <p>Parameter to
	2099	restrict the mixing length in the vicinity of the
[5]	2100	bottom
[61]	2101	boundary.  </p> <p>With <b>wall_adjustment</b>
	2102	= <i>.TRUE., </i>the mixing
[5]	2103	length is limited to a maximum of  1.8 * z. This condition
	2104	typically affects only the
[61]	2105	first grid points above the bottom boundary.</p> </td> </tr>
	2106	<tr> <td style="vertical-align: top;"><span style="font-weight: bold;"><a name="wall_heatflux"></a>wall_heatflux</span></td>
	2107	<td style="vertical-align: top;">R(5)</td> <td style="vertical-align: top;"><span style="font-style: italic;">5 * 0.0</span></td> <td>Prescribed
	2108	kinematic sensible heat flux in W m<sup>-2</sup>
	2109	at the five topography faces:<br> <br> <div style="margin-left: 40px;"><span style="font-weight: bold;">wall_heatflux(0)
	2110	</span>top face<br> <span style="font-weight: bold;">wall_heatflux(1)
	2111	</span>left face<br> <span style="font-weight: bold;">wall_heatflux(2)
	2112	</span>right face<br> <span style="font-weight: bold;">wall_heatflux(3)
	2113	</span>south face<br> <span style="font-weight: bold;">wall_heatflux(4)
	2114	</span>north face</div> <br>
	2115	This parameter applies only in case of a non-flat <a href="#topography">topography</a>. The
	2116	parameter <a href="#random_heatflux">random_heatflux</a>
	2117	can be used to impose random perturbations on the internal
	2118	two-dimensional surface heat
	2119	flux field <span style="font-style: italic;">shf</span>
	2120	that is composed of <a href="#surface_heatflux">surface_heatflux</a>
	2121	at the bottom surface and <span style="font-weight: bold;">wall_heatflux(0)</span>
	2122	at the topography top face. </td> </tr> </tbody>
	2123	</table><br>
	2124	<p style="line-height: 100%;"><br><font color="#000080"><font color="#000080"><a href="chapter_4.0.html"><font color="#000080"><img name="Grafik1" src="left.gif" align="bottom" border="2" height="32" width="32"></font></a><a href="index.html"><font color="#000080"><img name="Grafik2" src="up.gif" align="bottom" border="2" height="32" width="32"></font></a><a href="chapter_4.2.html"><font color="#000080"><img name="Grafik3" src="right.gif" align="bottom" border="2" height="32" width="32"></font></a></font></font></p>
	2125	<p style="line-height: 100%;"><i>Last
	2126	change: </i> $Id: chapter_4.1.html 83 2007-04-19 16:27:07Z raasch $ </p>
	2127	<br><br>
[5]	2128	</body></html>

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

Context Navigation

source: palm/trunk/DOC/app/chapter_4.1.html @ 95

Download in other formats: