Tropical deep convection and tropical tropopause - Academia Sinica

Tropical deep convection
and tropical tropopause
Pao K. Wang
Department of Atmospheric and Oceanic Sciences
University of Wisconsin-Madison
8 July 2009 at ECRC, Academia Sinica, Taipei, Taiwan

Tropical deep convective clouds
• Deep convective clouds play important
roles in many atmospheric processes
• Tropical deep convection plays pivotal role
in the transport of momentum and many
major trace chemicals.
• This talk concerns the tropical deep
convections that penetrate through the
tropopause and enter the stratosphere.
• But we will use midlatitude cases to
illustrate the penetrative convection first.

Deep convection in the atmosphere
• Horizontal velocity in the
atmosphere
• Large scale – ~ 10 m/s
• Mesoscale – > 50 m/s
• Vertical velocity in the
atmosphere
• Large scale – w ~ a few
cm/s.
• Shallow convections –
w max ~ a few m/s
• Deep convections – w max
~ 30 - 60 m/s, even more
in some extreme cases
• Deep convections usually
organize into a few narrow
belts (such as ITCZ and
frontal bands)

Vertical Temperature Profile
• The scale height H
(= kT/mg) of the
earth’s atmosphere
is ~ 8 km
• A convection is
considered deep if
its vertical scale is
comparable to the
scale height.

Deep convective storms

Traditional view of deep convective storms
tropopause

A conceptual global water vapor
transport model (Holton et al.,1995)

Satellite observation of middle
latitude deep convective storms
plume
Anvil
Storms over Balearic Islands

GOES visible images—nearly every
active cell is associated with plumes

Penetrative convection
• “Penetration” here does
not merely mean that
the cloud top height
exceeds the local
tropopause level.
Irreversible mass
transfer is required
• The mere presence of
an overshooting top
does not necessarily
qualify a storm as a
penetrative one
because it may just
distort the tropopause
temporarily but causes
no irreversible
transport of materials.

Instability and Wave Breaking
• Convection-induced instability and gravity wave breaking at
the storm top send H 2 O through the tropopause to enter the
stratosphere.
Overshooting top plumes
Anvil wave breaking
Wang (2007)

Fujita (1982, 1989) observed jumping cirrus above
severe storms – they are also due to wave breaking
Similar shape, size, orientation and occur at similar relative location
jumping cirrus
jumping cirrus
Overshooting top
Overshooting top
Photo courtesy of Martin Setvak
Modeled CCOPE storm cloud top
From: Wang (2004, GRL)

Jumping cirrus taken by webcam in Zurich
(courtesy of Willi Schmid, ETH, Zurich)
wind

Some recent observations of the tropopause
• So according to recent observations and
numerical modeling, midlatitude deep convective
storms can penetrate the tropopause and inject
water substance into the stratosphere.
• On the other hand, Highwood and Hoskins (1998)
stated that few of tropical “hot towers” are
observed to be taller than 14 km. there are even
observations indicating sinking motion in the
layer above the top of these hot towers (loosely
called the tropical tropopause layer, or TTL).
• This implies that the deep convective clouds may
not be the water vapor source in the tropics.

• Some suggest that the tropical crosstropopause
transport of H2O is performed
by large scale ascent (e.g., Küpper et al.,
2004). But there is no observations to
substantiate this claim.
• If even the layer above the strong
convection contains sinking motion, how
can we expect rising motion in large scale?

These questions cast doubts on the
global cycle chart
? ? ?
? ? ?

Modified Holton chart
do you see the differences?
NCAR-ACD

Wind shear environment in midlatitudes and
tropics
• In the tropics, wind shears are
usually weaker
• Can deep convection penetrate the
tropopause in a weak shear
environment?
• So we performed a model study on
the cloud top behavior of a deep
convective storm in the absence of
wind shear.

Simulated CCOPE supercell without wind shear
and winds. The results are pretty much the same.

CCOPE supercell without wind shear-central cross-section

CCOPE supercell without wind shear –cloud top view

A pancake cloud above a Cb over
Taiwan
(courtesy of Po-Hsiung Lin)

Satellite observation
(courtesy of Martin Setvak)

Oscillatory TTL

Weak wind shear – cold ring as seen in satellite IR
images– so the model is able to reproduce
observed cloud top features.

Tropical tropopause height and
deep convective clouds
• However, in real tropics, the tropopause is
higher. It is usually taken to be ~ 17 km.
• Tropical tropopause is thought to be
approximately maintained by convectiveradiative
equilibrium
• Yet few of tropical “hot towers” are
observed to be taller than 14 km
(Highwood and Hoskins, 1998)
• So how can the tropical tropopause be
maintained at 17 km level?
• Also, would H2O go thru the tropopause?

A tropical case (Tiwi Islands, Australia)
30000
30000
25000
25000
20000
20000
Height (m)
15000
Height (m)
15000
10000
10000
5000
5000
0
-5 0 5 10 15 20
0
250 450 650 850 1050
qv (g/kg)
Theta (K)
Slightly modified so that the cold point
tropopause is at 17 km. The level of
neutral buoyancy (LNB) is ~ 10-15 km
depending on where the parcel starts.
Lane and Reed, 2001

Formation of vertically segmented convective
clouds

Such segmented clouds have been observed
(a deep convective cloud in New Mexico)
Courtesy of Ken Bowman

Effect of cloud top gravity waves
(orientation of anvils tends to line up with waves)

3-D renderings showing the multiple anvil structure

A maritime Cb near Taiwan

Oscillating TTL

General behavior of the cloud
• The updraft causes the cloud top layer to
become unstable
• A patch of the cloud material breaks off
and penetrates into the stratosphere due
to cloud top instability. It gradually
spreads out and probably eventually
dissipates.
• The remaining cloud body “shrinks” back
to the troposphere and oscillates about
the LNB (~ 14-15 km).
• The 14-17 km TTL layer also oscillates
vertically. This oscillatory behavior has
been observed before (e.g., Danielsen,
1993).

One possible explanation
• Tropical deep convections may have
adequate energy to reach ~17 km, but
the LNB is often at ~ 14 km due to
humidity structure.
• The LNB cloud tops are more stable and
are the ones most often observed. On the
other hand, the clouds in the TTL (14-17
km) tend to be more disorganized and
probably thinner, hence are not often
observed.
• Thus we observe tropical deep convective
clouds often reaching ~ 14 km but the
actual tropopause height is at ~ 17 km.

Another revision?
? ? ?

Conclusions
• Tropical deep convective storms can penetrate through the
tropopause and transport water into the stratosphere.
• This transport is due to the gravity wave breaking and
instability at the cloud top. The forms of such transport
may look different in midlatitudes and tropics due to
different wind shear structure.
• The stable cloud top layer is ~ 14-15 km. The vertical
motions in the TTL are oscillatory and generate gravity
waves.
• Both the H 2 O and momentum transported vertically into the
stratosphere may have implications on the global climate
and chemical processes. They may be related to QBO as
well.
• But many questions remain to be answered (e.g., the H2O-
QBO-wind shear connection).

The End
Thank you