Changes between Version 21 and Version 22 of doc/tec/bc

Timestamp:

May 4, 2016 9:56:19 AM (9 years ago)

Author:

Giersch

Comment:

--

doc/tec/bc

@@ -275 +275 @@
 
 == Topography ==
+The Cartesian topography in PALM is generally based on the mask method ([#briscolini1989 Briscolini and Santangelo, 1989]) and allows for explicitly resolving solid obstacles such as buildings and orography. The implementation makes use of the following simplifications:
+
+ 1. the obstacle shape is approximated by (an appropriate number of) full grid cells to fit the grid, i.e., a grid cell is either 100% fluid or 100% obstacle,
+
+ 2. so far, only bottom surface-mounted obstacles are permitted (no holes or overhanging structures), and
+
+ 3. the obstacles are fixed (not moving).
+
+These simplifications transform the 3-D obstacle dimension to a 2.5-D topography. This reduced dimension format is conform to the Digital Elevation Model (DEM) format. DEMs of city morphologies have become increasingly available worldwide due to advances in remote sensing technologies. Consequently, it is sufficient to provide 2-D topography height data to mask obstacles and their faces in PALM. The model domain is then separated into three subdomains (see Fig. 3):
+
+ A. grid points in free fluid without adjacent walls, where the standard PALM code is executed,
+
+ B. grid points next to walls that require extra code (e.g., wall functions), and
+
+ C. grid points within obstacles that are excluded from calculations
+
+[[Image(04.png,600px,border=1)]]
+
+Figure 3: Sketch of the 2.5-D implementation of topography using the mask method (here for ''w''). The yellow and red lines represent the limits of the arrays ''nzb_w_inner'' and ''nzb_w_outer'' as described in Sect. [wiki:doc/tec/topography topography implementation], respectively.
+
+Additional topography code is only executed in grid volumes of subdomain B. The faces of the obstacles are always located where the
+respective wall-normal velocity components ''u'', ''v'', and ''w'' are defined (cf. Fig. 1 in Sect. [wiki:doc/tec/discret discretization]) so that the impermeability boundary condition can be implemented by setting the respective wall-normal velocity component to zero.
+
+An exception is made for the 5th-order advection scheme, where the numerical stencil at grid points adjacent to obstacles would require data within the obstacle. In order to avoid this behavior, the order of the advection scheme is successively degraded at respective grid volumes adjacent to obstacles, i.e., from the 5th-order to 3rd-order at the second grid point above/beside an obstacle and from the 3rd-order to a 2nd-order at grid points directly adjacent to an obstacle.
+
+Wall surfaces in PALM can be aligned horizontally (bottom surface or rooftop, i.e., always facing upwards) or vertically (facing north,
+east, south or west direction). At horizontal surfaces, PALM allows to either specify the surface values (''θ, q,,v,,, s'') or
+to prescribe their respective surface fluxes. The latter is the only option for vertically oriented surfaces. Simulations with topography
+require the application of MOST between each wall surface and the first computational grid point. For vertical walls, neutral
+stratification is assumed for MOST. The topography implementation has been validated by [#letzel2008 Letzel et al. (2008)] and [#kanda2013 Kanda et al. (2013)}. [#park2013 Park and Baik (2013)] have recently extended the vertical wall boundary conditions for non-neutral
+stratifications and validated their results against wind tunnel data. Up to now, however, these modifications are not included in PALM 4.0. Figure 4 shows exemplarily the development of turbulence structures induced by a densely built-up artificial island off the coast of Macau, China (see also animation in [#knoop2014 Knoop et al., 2014]). The approaching flow above the sea exhibits relatively weak turbulence due to the smooth water surface. Within the building areas, strong turbulence is generated by additional wind shear (due to the walls of isolated buildings) and due to a general increase in surface roughness.
+
+[[Image(05.png,600px,border=1)]]
+
+Figure 4: Snapshot of the absolute value of the 3-D rotation vector of the velocity field (red to white colors) for a simulation of the city of Macau, including a newly built-up artificial island (left). Buildings are displayed in blue. A neutrally-stratified flow was simulated with the mean flow direction from the upper-left to the bottom-right, i.e. coming from the open sea and flowing from the artificial island to the city of Macau. The figure shows only a subregion of the simulation domain that spanned a horizontal model domain of about ''6.1x2.0x1 km^3^'', and with an equidistant grid spacing of ''8''m}$. The copyright for the underlying satellite image is held by Cnes / Spot Image, Digitalglobe. For more details, see associated animation [#knoop2014 Knoop et al. (2014)].
+
+The technical realization of the topography will be outlined in Sect. [wiki:doc/tec/topography topography implementation]. 
+
 
 = References =
@@ -294 +332 @@
 * [=#orlanski1976] '''Orlanski I.''' 1976. A simple boundary condition for unbounded hyperbolic flows. J. Comput. Phys. 21: 251–269.
 
-* [=#miller1981] ''' Miller MJ, Thorpe AJ.''' 1981. Radiation conditions for the lateral boundaries of limited-area numerical models. Q. J. Roy. Meteor. Soc. 107: 615–628.
-
+* [=#miller1981] '''Miller MJ, Thorpe AJ.''' 1981. Radiation conditions for the lateral boundaries of limited-area numerical models. Q. J. Roy. Meteor. Soc. 107: 615–628.
+
+* [=#briscolini1989] '''Briscolini M, and Santangelo P.''' 1989. Development of the mask method for incompressible unsteady flows. J. Comput. Phys. 84: 57–75.
+
+* [=#letzel2008] '''Letzel MO, Krane M, Raasch S.''' 2008. High resolution urban large-eddy simulation studies from street canyon to neighbourhood scale. Atmos. Environ. 42: 8770–8784.
+
+* [=#kanda2013] '''Kanda M, Inagaki A, Miyamoto T, Gryschka M, Raasch S.''' 2013. A new aerodynamic parameterization for real urban surfaces.
+  Bound.-Lay. Meteorol. 148: 357–377.
+
+* [=#park2013] '''Park SB, and Baik J.''' 2013. A large-eddy simulation study of thermal effects on turbulence coherent structures in and above a building array. J. Appl. Meteorol. 52: 1348–1365.
+
+* [=#knoop2014] '''Knoop H, Keck M, Raasch S.''' 2014. Urban large-eddy simulation - influence of a densely build-up artificial island on the turbulent flow in the city of Macau, Computer animation. doi:10.5446/14368.
+