Density currents and slope winds

1
Density currents and slope winds H. van Dop
Atmospheric boundary layers, Les Houches, 17-27
June 2008
2
Katabatic Flows (adapted from Polous, 2005)
control air-surface exchange of
heat, momentum and trace gases are common
(and significant) in gently sloping terrain
(almost everywhere, except in the
Netherlands) are crucial to an understanding
of (local and global) climate
3
However
difficult to represent well in numerical
models a significant complication to the
study of the NBL
4
Characteristics
clear skies and weak synoptic
conditions ground-based inversion a very
low level jet (z/h ltlt 1) significant for slope
angles of a few degrees
5
(No Transcript)
6
(No Transcript)
7
Christian Reuten and his tank
8
Similarity between field and tank experiment
9
Tank experiment
10
Example (Zhong and Whiteman, 2004)
11
Earlier Studies
Prandtl, 1942 (analytic) Ball, 1956
(bulk) Manins and Sawford, 1979 (bulk) van
den Broeke, 1994 (bulk) Denby, 2001 (RANS 2nd
order) Oerlemans and Grisogono, 2002
(analytic, with height dependent eddy
diffusivity) Skyllingstad, 2002 (LES)
12
Modeling of Katabatic Flows
13
Equations of motion
Stably stratified environment
14
Transition to slope coordinates, s,y,n
z
n
x
?
s
and introduction of temperature deficit variable,
?
15
In the s, y, n-frame
16
(No Transcript)
17
The set of equations is consistent with
Schumann (1990) (for anabatic flow) With
appropriate approximations it reduces to
Prandtls model SFS closure is required
18
(No Transcript)
19
(No Transcript)
20
(No Transcript)
21
(No Transcript)
22
(No Transcript)
23
(No Transcript)
24
(No Transcript)
25
(No Transcript)
26
LES model
Domain 250 x 250 x 50 m Resolution 96 x 96 x
128 Periodic boundary conditions at the
sides Spongy top (25) Monin Obukhov surface
conditions TKE-equation closure constants
based on spectral filtering (filterscale ?)
and Kolmogorov theory, assuming Ck 1.5
Runs to steady state
27
Results
28
Velocity temperature

• Wind speed max overestimated
• Overestimated mixing

29
(No Transcript)
30
Conclusions and outlook
Main turbulence production is near the
surface, where the resolution is (always) not
sufficient dependency on filter scale and
grid scale Vertical mixing is too strong in
the model Numerical diffusion is
significant (25)
31
Outlook
Use higher order advection schemes (in
progress) Further investigations into the
effects of stratification and slope angle on
height and strength of the jet, etc
Inventarization of the turbulent structure
of katabatic flow A scaling theory for
katabatic flow?