Houze Robert A., Jr. Cloud Dynamics
Подождите немного. Документ загружается.
5.4 Altostratus and Altocumulus 189
Results obtained for a basic-state vertical motion
WB = 2 em
S-I
are shown in
Figs. 5.38-5.40. Most evident in Fig. 5.38 is the strong effect of the fall speed of
the crystals that comprise the cloud. With time, the height of the cloud base and
the height and amount of the maximum ice concentration decrease substantially,
especially during the first 20 min
of
the lifetime of the ice-cloud layer. Figure 5.39
shows the vertical profiles of heating in the cloud layer associated with latent heat
release as well as infrared, solar, and net radiation at three different stages of the
lifetime
of
the cirriform cloud layer. The latent heating by vapor deposition where
the cloud has formed and cooling by sublimation of ice below the cloud layer are
generally
of
the same order of magnitude as the radiative heating and cooling. A
somewhat surprising result is that the net radiative heating curve indicates little if
any destabilization of the cloud layer. These properties
of
the model of cirriform
cloud formation suggest that the assumptions we made in Sec. 5.3.3, which in-
volved viewing a cirriform cloud layer as a radiatively driven mixed layer, may
not always be met. In that treatment, the fall speed of the ice particles was tacitly
neglected and the profile of heating used was one appropriate for an extremely
high layer
of
cirriform cloud. 153 From Fig. 5.40, it is evident that despite the lack
of
strong radiative destabilization, embedded cellular structure remains evident
throughout the lifetime
of
the cirriform cloud layer; however, it decreases in
intensity in the later stages.
5.4 Altostratus and Altocumulus
5.4.1 Altostratus and Altocumulus Produced as Remnants
of Other Clouds
Like high-level cirriform cloud, middle-level altostratus and altocumulus may be
produced as remnants
of
nimbostratus or cumulonimbus. Figure 5.41 is the result
of a numerical simulation (using a model of the type to be discussed in Sec. 7.5.3)
of a growing tropical cumulonimbus, which can be seen to issue protruding layers
of cloud in middle (as well as lower) levels.!" This behavior is related to the
variation of the horizontal wind field in and around the cloud.
5.4.2 Altocumulus as High-Based Convective Clouds
Altostratus and altocumulus may also form in layers aloft, completely removed
from any deep nimbostratus or cumulonimbus source. One important type is
simply cumulus or cumulonimbus with middle-level bases (i.e., altocumulus cas-
15l The fact that these assumptions may not always be met was recognized and pointed out by Lilly
(1988).
154 That these simulations represent realistic cloud structures over the equatorial Atlantic Ocean
was verified by means of stereoscopic cloud photography by Warner
et al. (1980).
....:...
-
oc-----_
4
,
,
,
,
,
'.osw
~X>
C'
,~<-QC
r
,
,
,
a
-4
I
I
55·60
min;
,
,
,
.
.
4
-4 a
°C day-1
"
\~
25-30min; !
OIR
/
/
/
(
"
-820
OR
____
osw
::':.~
...
8
OIR
~
...
)
e
7
,....-4'
~
<,
I
N
OC
''\(
6
10-15min:
-20
a
Figure 5.39
8.5
(a) t = 15
min
8.0
~.::==:=::.:
7 .5
10---../,,"'-"
7.0
t==:===~==::tJJ
6.5
6.0
E
-"
t = 30
min
8.5 (c) t = 60
min
.00
6.0
0.5
1.0
1.5
2.0 2.5
3.0
3.5
4.0 4.5
5.0
5.5
6.0
x
(km)
Figure 5.40
5.4 Altostratus and Altocumulus
191
8.0
6.4
~4.8
E
~3.2
N
~
0
1.6
0.0 0
3.6 7.8 10.8
y(km)
14.4
Figure 5.41 Model simulation of a cumulus congestus cloud
over
the tropical eastern Atlantic
Ocean. Cross section runs
north-south
through a three-dimensional domain. Heavy contours show
liquid
water
content in g rn". Thin isopleths are for vertical velocity at 1 m
S-I
intervals; dashed
contours are for downdrafts. (From Simpson and van Helvoirt, 1980.)
tellanus, Fig. 1.14d). The dynamics of these cloud elements are well described by
the cumulus and cumulonimbus dynamics discussed in Chapters 7 and 8.
5.4.3 Altostratus and Altocumulus as Shallow-Layer
Clouds Aloft
Many altostratus and altocumulus clouds are neither the remnants of other clouds
nor cumulus or cumulonimbus clouds aloft. These clouds include altostratus lay-
ers such as the one seen in Fig. 1.13 and altocumulus stratiformis (Fig. 1.14a and
b). These are layer clouds that closely resemble stratus and stratocumulus. How-
ever, they do not occur in the planetary boundary layer, and their dynamics are
therefore not exactly analogous to those of stratus and stratocumulus (Sees. 5.1
and 5.2). They are more similar to the cirriform cloud layers described in Sec. 5.3
in that they occur in shallow layers of air aloft, separate from the boundary layer.
The dynamics of this type of altostratus and altocumulus can be investigated by
means of model calculations like those used in Sec. 5.3.4 to describe a thin layer of
cirriform cloud. That model has been modified to middle-level cloud conditions by
changing the fall velocity relationship and radiative absorption coefficients to
make them appropriate for a cloud of liquid water drops and calculating condensa-
tion and evaporation based on 100% relative humidity with respect to liquid wa-
Figure 5.39 Results of model of cirriform cloud formation in a layer of air in which the basic-state
vertical motion
W
o
= 2 em
S-I.
Vertical profiles of horizontally averaged heating by phase changes of
water (QC), infrared radiation (QIR), absorption of solar radiation (QSW), and net radiative processes
(QR). (From
Starr
and Cox, 1985a. Reprinted with permission from the American Meteorological
Society.)
Figure 5.40 Results of a model of cirriform cloud formation in a layer of air in which the
basic-state vertical motion
W
o
= 2 em
S-I.
Ice-water
mixing ratio field in ""g
s'
at various times.
Contours for 20 and 40 ""g g" are shown by shading. (From Starr and Cox, 1985a. Reprinted with
permission from the American Meteorological Society.)
192 5 Shallow-Layer Clouds
ter.
155
In calculations illustrated in Fig. 5.42, the fall velocity was assumed to be
0.9 em
s-I_two
orders of magnitude less than that assumed in the case of cirri-
form cloud
(-1
m S-I, as in Fig. 5.37). Because of the smaller particle fall speeds,
the simulated middle-level cloud retains its hydrometeors throughout the cloud
lifetime, has larger mixing ratios of hydrometeor mass, and is contained in a
shallower layer than the cirriform cloud (cf. Fig. 5.40 and Fig. 5.42). As in the
cirriform cloud layer, the latent heat release and radiative heating are of the same
order of magnitude (Fig. 5.43). However, unlike the model cirriform cloud layer,
the radiative heating produces destabilization in the middle-level cloud, strongly
cooling the upper part of the cloud layer and warming it below. Consequently,
turbulent kinetic energy is maintained throughout the cloud lifetime (Fig. 5.44).
155 This calculation was carried out by Starr and Cox (1985a,bl as part of the same study in which
they modeled the thin layer of cirrus. Although this calculation was originally described as represent-
ing altostratus, it probably better represents altocumulus stratiformis, since altostratus is often deeper
than the cloud assumed in the simulation and is more often than not glaciated while altocumulus
stratiformis is shallow, contains overturning convective elements, and tends to consist of supercooled
liquid water (at least during its earlier stages).
8.5
:-
Figure 5.42
8.0
~7.5
...
7.0
Figure 5.43
QR
-80 -60 -40 -20 0 20 40 60
°C
day-1
-
...
~100
~
-
~
50
Figure 5.44
10 20 30 40 50 60
t (min)
5.4 Altostratus and Altocumulus
193
Thus, a middle-level cloud layer (altocumulus stratiformis) could be described
appropriately by the theory of a cloud-filled radiatively driven mixed layer aloft,
which was used in Sec. 5.3.3 to describe the layer of ice cloud generated by
cumulonimbus.
The interpretation of altocumulus stratiformis as a cloud-filled radiatively
driven mixed layer helps to explain its structure since the elements of altocumulus
stratiformis often appear to be produced by Rayleigh-Benard convection (as de-
scribed in Sec. 2.9.3) in the form of cells (Fig. 1.14a) or rolls (Fig. 1.14b). The
differential heating mechanism driving the Rayleigh-Benard convection is evi-
dently the infrared radiation, and the presence or absence of shear likely deter-
mines whether or not the elements take the form of rolls (Fig. 5.20). The average
distance between altocumulus (and cirrocumulus) rolls has been reported to be
<0.25 km in 39% of cases, <0.5 km in 78% of cases, and <0.75 km in 93% of
cases.
156 The observed depths of the cloud layers moreover indicated a horizontal-
to-vertical aspect ratio
-I:
I,
which is consistent with the rolls being of the
Rayleigh-Benard type.
In some cases, altocumulus rolls are produced by shear instability, probably of
the Kelvin-Helmholtz type (Sec. 2.9.2), or they may be the result of mixed
thermal and shear instability. These distinctions can be difficult to make in visual
observations of clouds.
5.4.4 Ice Particle Generation by Altocumulus Elements
The individual elements (rolls or cells) of altocumulus stratiformis tend to glaciate
and produce fallstreaks of precipitating ice particles in the later stages of their
lifetimes (Fig. 5.45). The process appears to be similar to that which occurs in
stratocumulus (Fig. 5.12). In their glaciating stage, the elements of altocumulus
156 Data reported by Suring (1941) and discussed further by Borovikov et al. (1963). The book of
Borovikov
et al. (1963) is of interest as one of the first attempts at a comprehensive treatment of cloud
dynamics.
Figure 5.42 Results of a model of middle-level cloud formation. Nighttime conditions were
assumed and the basic-state vertical motion was
WB = 2 em S-I. Liquid-water mixing ratio field in
contours of 0.001,
I, 50, 100, and 150
fJ-g
g-I. (From Starr and Cox, 1985b. Reprinted with permission
from the American Meteorological Society.)
Figure 5.43 Results of a model of middle-level cloud formation. Nighttime conditions were
assumed and the basic-state vertical motion was
WB = 2 em S-I. Vertical profiles of horizontally
averaged heating by phase changes of water (QCl and infrared radiation (QR). (From Starr and Cox,
1985b. Reprinted with permission from the American Meteorological Society.)
Figure 5.44 Results of a model of middle-level cloud formation. Nighttime conditions were
assumed and the basic-state vertical motion was
WB = 2 em S-I. Domain-averaged turbulent kinetic
energy as a function of time. (From Starr and Cox, 1985b. Reprinted with permission from the
American Meteorological Society.)
194 5 Shallow-Layer Clouds
I---II---r-rl
,711
IJP'I~)J/~
Ice
Strands
Fallstreaks
(a) (b) (c) (d) (e)
Figure 5.45 Schematic showing progressive stages in the glaciation of altocumulus clouds. The
time interval between the young supercooled water clouds depicted in (a) and the ice particle
fallstreaks in (e) can be several hours. (From Hobbs and Rangno, 1985. Reprinted with permission
from the American Meteorological Society.)
(0)
...L
Depositional
Growth
T
No Growth
(Possible' Evaporation)
!
Small Ice
Particles from
Fallstreok
(-5
L-' and
<0.001 mm h-')
Needles
(>4
L-')
...L
Depositional
Growth
T
Fallstreok No Growth
(Possible Evaporation)
!
(b)
Figure 5.46 Schematic of interaction of altocumulus and stratocumulus clouds. (a) Dendritic
crystals in a fallstreak from the altocumulus fall into the stratocumulus, where they grow by riming and
reach the ground as light precipitation. (b) Small ice particles in a fallstreak from the altocumulus grow
into needles in the stratocumulus and contribute to the precipitation at the ground. (From Locatelli
et
al.,
1983. Reproduced with permission from the American Meteorological Society.)
may have the appearance of altocumulus floccus (Fig. 1.14c), which are similar to
cirrus floccus (Fig. 1.15b). Or they may develop fallstreaks and appear somewhat
similar to cirrus spissatus with virga (Fig. 1.16) or, in very late stages after the
original cloud element has disappeared, to cirrus uncinus (Fig. LISe). The more
cumuliform altocumulus castellanus (Fig. 1.14d) is also observed to produce fall-
streaks of precipitating ice particles in its later stages.
5.4.5 Interaction of Altocumulus and Lower Cloud Layers
When a layer of stratus or stratocumulus cloud is present below altocumulus
elements, ice particles may fall into the lower cloud from the altocumulus layer.
This process is illustrated in Fig. 5.46. In this case, a layer of stratocumulus was
stimulated to precipitate as a result of the interaction of the two cloud layers.
5.4 Altostratus and Altocumulus
195
Fallstreaks produced during the latter stages of the altocumulus elements (as in
Fig. 5.45) penetrated through the lower stratocumulus cloud layer. The ice parti-
cles in the fallstreaks grew by riming and deposition while they passed through the
supercooled stratus cloud. Thus, water condensed within the stratus layer was
transferred to the ice particles produced aloft and ended up on the ground as part
of the precipitation. The water in the stratus cloud would probably otherwise have
never reached the ground as precipitation since the drops would have remained
too small to fall out of the lower-level cloud.
Chapter
6
Nimbostratus
"Now
downwards by the world's attraction driven,
That tends to earth, which had upris'n to
heaven;"!"
The shallow-layer clouds, which we considered in Chapter 5, do not produce
much precipitation because they do not have sufficient vertical extent.
The
few
particles that grow large enough to precipitate usually fall out of the cloud layer
before they exceed the size of small drizzle. In this chapter, we consider deep
stratiform clouds from which significant amounts of rain or snow fall. These
clouds, referred to as
nimbostratus (see Sec. 1.2.2.2),
occur
extensively in both
the midlatitudes and the tropics. They are associated primarily with the wide-
spread continuous clouds of mesoscale convective systems, hurricanes, and ex-
tratropical cyclones (e.g., Figs. 1.27-1.29).158 The structure and dynamics of these
cloud systems will be considered in Chapters
9-11.
These storms are large and
complex, involving air motions on a range of scales and various mixtures of
convective and stratiform cloud structures. As background for these later chap-
ters, the present
chapter
isolates and discusses certain generic properties
of
the
nimbostratus clouds
that
occur
within these cloud systems. Chapters 7 and 8 will
describe the properties of purely convective clouds. This preliminary examination
of both the nimbostratus
and
convective clouds by themselves is prerequisite to
our
later discussions
of
complex cloud systems in Chapters
9-11.
Since the widespread vertical air motions essential to the production of nimbo-
stratus cloud are largely those of the parent storm (i.e., the mesoscale convective
system, hurricane, or extratropical cyclone), the dynamics of nimbostratus cannot
be fully treated in this chapter. That treatment must await the later chapters. The
emphasis here will be on the microphysical and convective processes that produce
nimbostratus, regardless of the nature of the parent storm. What is known about
these aspects of nimbostratus has been inferred largely from observations of the
cloud microphysics and radar-echo structure of this cloud type and its precipita-
tion.
From
radar
and microphysical data, certain characteristics of the air motions
that must be present within nimbostratus to produce the radar echoes and micro-
physical structure become evident. In particular, it can be deduced that nimbo-
stratus does not
occur
alone
but
rather
is closely associated, in each of its various
contexts, with convective cloud processes.
157 Goethe on precipitation faIling from a cloud.
158 It is a common misconception that nimbostratus is a midlatitude phenomenon associated primar-
ily with extratropical cyclones.
196
6.1 Stratiform Precipitation
197
We begin the chapter by examining the observed radar-echo structure and
microphysics of the precipitation that falls from nimbostratus (Sec. 6.1). In Sees.
6.2 and 6.3, we explore the relationship of the nimbostratus to convection, which
appears to be essential to the precipitation processes in nimbostratus. Finally, in
Sec. 6.4, we examine briefly the radiative processes and turbulent mixing that
occur within nimbostratus.
6.1 Stratiform Precipitation
6.1.1 Definition and Distinction from Convective
Precipitation
Precipitation is generally considered to be of two clearly distinguishable
types-
stratiform and convective. Stratiform precipitation falls from nimbostratus clouds,
while convective precipitation falls from cumulus and cumulonimbus clouds. A
simple comparison of the salient features of these two basic types of precipitation
is provided by Fig. 6.1. We will use this diagram to aid in making more precise
definitions of stratiform and convective precipitation. Although our present pur-
pose is to examine stratiform precipitation, with convective clouds and precipita-
tion being the subject of Chapters
7-9,
it is useful here to define both types of
precipitation, as the definitions are better understood in contrast with each other.
Stratiform and convective precipitation can be defined in terms of their vertical
velocity scales. Stratiform precipitation is defined as a precipitation process in
which the vertical air motion is small compared to the fall velocity of ice crystals
and snow; more specifically, the vertical velocity of the air
w satisfies the con-
dition
Iwl
< Vice (6.1)
where
Vice represents the scale of the terminal fall velocity of ice crystals and
snow
(-1-3
m
S-I).
Ice particles in the upper levels of the cloud play an important
role in the precipitation process, and in (6.1) the magnitude of the vertical air
motion is small compared to the magnitude of the fall velocity of the ice particles.
It
is possible for warm stratiform clouds (tops below the
O°C
level) to produce
precipitation. We have already seen in Chapter 5 that stratus and stratocumulus
may produce drizzle. When low-level warm stratiform clouds are a little deeper or
more vigorous, they may produce some light rain. However, by far the bulk of
stratiform precipitation falls from nimbostratus that reaches well above
O°C
level.
The stratiform precipitation process in deep nimbostratus containing ice parti-
cles aloft is depicted schematically in Fig. 6.1a. Precipitation particles that eventu-
ally fall to the ground as raindrops have their early history as ice particles in the
upper parts of the cloud. We will defer until later a discussion of the important
question of how these particles first appear in the upper reaches of the cloud. For
now, we note only that they may form
in situ; they may be introduced from a
198 6 Nimbostratus
RADAR-ECHOBOUNDARY
1
----:::=:;::::;::::;;=-~;;;;;~=:::;:;::::=.=:::;:==;::=.:::::;:::;:::::;:0-CLOU
D
f\
TOP
\
E1km
\
E1/8-1
km
"BRIGHT
BAND"
EARTH'S
SURFACE
(a)
'-
_...J
---EARTH'S
SURFACE
1
4
TIME--
- 5 km
(b)
Figure 6.1 (a) Characteristics of stratiform precipitation. (b) Characteristics of convective
precipitation. Shading shows higher intensities of radar echo, with hatching indicating the strongest
echo. In (b) cloud is shown at a succession of times 1
0
,
•••
,
In'
Growing precipitation particle is carried
upward by strong updrafts until
12 and then falls relative to the ground, reaching the surface
just
after
1
5'
After 15, the cloud may die or continue for a considerable time in a steady state before dissipation
sets in at
In-I
and
In'
The dashed boundary indicates an evaporating cloud. (From Houze,
1981.
© American Geophysical Union.)
source located to the side of or above the nimbostratus; or they may originate
from a source embedded within the cloud itself. Once introduced into the nimbo-
stratus cloud, the ice particles begin to grow. Upward air velocity maintains
supersaturation by condensing vapor, which is deposited onto the ice particles
(Sec.
3.2.3). Although w must be large enough to maintain supersaturation, it must
be small enough not to violate (6.1). The ice particles in the upper levels of
nimbostratus must fall; they cannot be suspended or carried aloft by the air
motions as they grow. Thus, the general in-cloud vertical air motion in pure
nimbostratus normally does not exceed a few tens of centimeters per second.
These vertical velocities support growth of particles by vapor deposition.
The higher the level at which the ice particles are formed or introduced from an
outside source, the longer they will be able to grow by deposition of the vapor
made available as a result of mean upward air motion in the nimbostratus. The
time available for growth of the ice particles falling from cloud top is
-1-3
h (the
time it takes a particle falling at
-1-3
m s
-I
to descend 10 km). This time is
sufficient for the particles to grow by deposition. When the ice particles falling and
growing by deposition descend to within about 2.5 km of the
O°C
level, aggrega-