Geoscience Reference
In-Depth Information
with
θ
the amplitude scale of the fluctuations in virtual potential temperature. The
r
1
/
3
dependence of
Ri
e
means that the largest eddies feel buoyancy effects the
most strongly; the small eddies are dominated by inertia and (at the smallest scales)
viscous forces. Thus, as
z/L
becomes positive from zero the largest eddies are
damped first. In a stable surface layer the relative importance of buoyant destruction
increases with increasing height; at some height the eddies become limited in size
by the stability. Loosely speaking, at this height the turbulence does not “sense”
z
,
the distance to the surface, as it does in an unstratified surface layer.
This suggests that under very stable conditions
z
drops out of the M-O governing
parameter group, leaving
g/θ
0
,Q
0
,C
0
,
and
u
∗
. The scales in this limiting case are
then
velocity:
u
∗
,
temperature:
T
∗
,
scalar:
c
∗
,
length:
L.
(10.39)
Nondimensional quantities again become universal constants. This “local
z
-less
scaling,” as it is called (
Wyngaard
,
1973
;
Chapter 12
) indicates that, for example,
∂U
∂z
∼
u
L
,
∂
∂z
∼
T
L
;
(10.40)
so that the M-O profile functions are
kz
u
∗
∂U
∂z
z
L
,
kz
T
∗
∂
∂z
z
L
.
φ
m
=
∼
h
=
∼
(10.41)
essentially linear over the entire stable range of the Kansas data.
10.4 Deviations from M-O similarity
The velocity scale of the energy-containing eddies in the convective boundary
layer with zero mean wind is expected to depend (at minimum) on
g/θ
0
,
Q
0
,and
the boundary-layer depth
z
i
. These parameters define the
free-convection velocity
scale w
∗
:
g
θ
0
Q
0
z
i
1
/
3
w
∗
=
.
(10.42)
0
.
2mKs
−
1
and
z
i
= 1000 m, typical values in fair weather over land,
Eq. (10.42)
gives
w
∗
∼
For
Q
0
=
2ms
−
1
.
From the definitions of
w
∗
and
L
we can write
w
∗
/u
∗
=
k
−
1
/
3
(
z
i
/L)
1
/
3
. Con-
−
−
vective boundary layers in the atmosphere have
z
i
/L
values up to several hundred,
so it is not unusual to find the free-convection-like state where
w
∗
u
∗
.Since
w
∗
contains
z
i
, which is not an M-O parameter, the horizontal gusts in the surface layer
due to these large eddies are not M-O similar, as pointed out by Panofsky
et al
.