Cargese International School, COST ACTION 723, UTLS, 3-15 Oct., 2005

1
Cargese International School, COST ACTION 723,
UTLS, 3-15 Oct., 2005
1
Water Vapour Variation in the Upper Troposphere
and Lower Stratosphere over the Asian Summer
Monsoon Nawo Eguchi eguchi.nawo_at_nies.go.jp Nationa
l Institute for Environmental Studies (NIES),
Japan
Abstract The water vapour variation over the
Asian summer monsoon, especially with a focus on
the vertical structure, was investigated using
the Upper Atmosphere Research Satellite (UARS) /
Microwave Limb Sounder (MLS) water vapour data
(version 7.02) during the summer 1992. The wetter
air intrudes into the tropical lowermost
stratosphere (LS) over the convective region
(70-180ºE, 25-35ºN) and extends to the
extra-tropical LS along the isentropic surfaces
below 380 K around the Asian subtropical jet
(ASJ). In order to estimate both water vapour
fluxes passing through and along the isentropic
surfaces, the water vapour budget is calculated
diagnostically using the conservation equation of
water vapour for the isentropic coordinate. The
time (seasonal) mean water vapour flux
divergence, amounting to the maximum value in the
ASJ, is the main term and the structure
resembles to a stationary Rossby wave which has
the wave train along the ASJ. The term of
vertical flux divergence on the 380 K isentropic
surface over the convective region is large and
it is about the same value as the seasonal mean
flux divergence in the ASJ on the 350 K
isentropic surface. I came to a conclusion that
the diabatic effects and the stationary Rossby
wave associated with the Asian summer monsoon
play an important role in both the water vapour
variation and the exchange process between the
upper troposphere (UT) and LS in the tropical and
extra-tropical region.
2
2
1. Introduction

• There are two main entrances of mass and chemical
  species from the troposphere to the stratosphere
  (?, ? in Figure A).
• Stratosphere-Troposphere Exchange (STE) between
  the tropical upper troposphere (also known as the
  Tropical Tropopause Layer TTL) and the tropical
  lower stratosphere passing through the isentropic
  surface (the tropical tropopause).
• STE between the tropical upper troposphere (or
  TTL) and the extra-tropical lower stratosphere
  along isentropic surfaces.

?
(gt380 K)
?
(Tropopause380 K)
Figure A Schematic latitude-height
cross-section showing the basic ideas of STE
processes.
Holton et al.,1995

• The previous studies showed that the Asian
  summer monsoon impacts the STE process around the
  tropopause region, and the distributions of water
  vapour and other trace gases.
• STE is associated with intraseasonal variations
  of the Asian subtropical jet (ASJ), especially at
  the northeast side of the Asian monsoon high
  Terao,1998 Dethof et al.,1999 Dunkerton,1995.
• However, there is less understanding about the
  STE process and water vapour variation around the
  tropopause region during the Asian summer monsoon.

3
3
2. Purpose

• What is the water vapour variation like,
  especially the vertical direction, in the
  tropopause region over the Asian summer monsoon ?
• What is the dominant STE process which affects
  the water vapour variation in the tropopause
  region, and what kind of time scales is dominated?

3. Data

• Water Vapour data from Upper Atmosphere Research
  Satellite (UARS) / Microwave Limb Sounder (MLS)
  (Version 7.02) cf. Read et al., 2004
• MLS measures the atmosphere in the limb
  direction with a vertical resolution of nearly
  3km.
• Water vapour mixing ratio ppmv or g/kg,
  relative humidity with respect to ice between
  316 - 46 hPa (11 levels)
• May September 1992
• Cirrus Clouds Frequency data from UARS /
  Cryogenic Limb Array Etalon Spectrometry (CLAES)
• At 100 and 146 hPa, cirrus clouds are inferred
  from CLAES aerosol (revision 9) using a threshold
  extinction value 3.3e-3 km-1 cf. Mergenthaler et
  al., 1999.
• Meteorological data from ECMWF 40 Years
  Re-Analysis (ERA-40)
• Temperature, horizontal wind, vertical wind,
  potential vorticity

4
4
4. Water vapour distribution around the
tropopause over the Asian summer monsoon
(a)
Figure 1. Horizontal map of (a) water vapour
mixing ratio ppmv (shade), relative humidity
with respect to ice (black contour with 10
intervals) and Outgoing Longwave Radiation
(OLR ) W/m2 (blue contour with 10 W/m2
intervals from 190 W/m2), (b) temperature K
(shade and black contour larger than 204 K with
10 K intervals), cirrus clouds frequency
(yellow contour, interval is with 10 from 10
) at 100 hPa in August 1992. Vectors show the
horizontal wind m/s.
The Asian subtropical jet (ASJ)
(b)
The Asian mosoon high

• At 100 hPa (the lowermost stratosphere), the
  moist area exist at the north side of the
  convective regions, such as the Indian
  subcontinent and the western Pacific around 25ºN.
• The dry air extends from the middle-latitude to
  the equatorial region in the central Pacific. The
  dry region also exists at the south of the
  equator in the western Pacific and Indian Ocean.
• The anti-cyclonic circulation is located at the
  northwest side of the convective region (the
• center is 80ºE, 30ºN), the Asian subtropical jet
  (ASJ) forms the northern part of the circulation.
• The cirrus clouds occur frequently over the cold
  region (60ºE120ºE, 20ºN).

5
5
Figure 2. Latitude-pressure section of monthly
mean water vapour mixing ratio ppmv (shade),
temperature K (blue dotted contour with 10 K
interval from 190 K), potential temperature K
(red line) and zonal wind m/s (yellow contour,
the dotted line indicates negative value
(easterly wind).) averaged between 70ºE and 120ºE
in August 1992. Black dot-dash-line indicates the
lapse rate tropopause obtained from NCEP
reanalysis.
ppmv

• The wet air (gt 4.5 ppmv) extends up to 83 hPa
  passing through the tropopause (and 380 K
  potential temperature) around 25ºN without
  crossing the coldest region (lt 190 K) where the
  cirrus clouds occur frequently.
• At the north side of westerly wind (the Asian
  subtropical jet) core (gt40ºN), wet region spreads
  along the isentropic surfaces (below 380 K).

6
6
Figure 3. Same as Figure 1, but for
longitude-pressure section averaged between 15ºN
and 30ºN. Orange marks at 100 and 147 hPa
indicate the frequency of cirrus clouds larger
than 10 , red marks at bottom of this figure
show the convective region (lower than 230 W/m2
OLR from NOAA). Vectors show the zonal and
vertical wind components m/s, hPa/s.

• The isentropic surfaces above 365 K tilts
  westward increasing height between 60ºE and
  200ºE and the easterly wind related with the
  anti-cyclonic circulation is dominated over
  there.
• The wet air intrudes into the lowermost
  stratosphere passing through the 380 K isentropic
  surface especially over the Asian monsoon region.

7
7
5. Calculation of water vapour flux divergence

• In order to estimate both fluxes across and
  along the isentropic surfaces in the tropics and
  middle latitude shown in Figure 2, the water
  vapour budget is calculated diagnostically using
  the conservation equation of water vapour for the
  isentropic coordinate.

(1) (2) (3)
Seasonal mean Intraseasonal time scales
(20-80 days) Synoptic time scales (less than 20
days)
q water vapor mixing ration kg/kg v
horizontal wind m/s

• C in equation (1) indicates condensation which
  is supposed to be ignored in this study.
• Horizontal flux divergence is
  divided into three time scales, seasonal mean,
  intraseasonal (20-80 days) and synoptic (8-20
  days) time scales (Equation (2)).
• R is defined as the vertical flux divergence
  passing through the isentropic surfaces which is
  associated with the sensitive heating of the
  convective clouds and radiative heating release
  (Equation (3)).

8
8
convergence
divergence
Figure 4. Map for water vapour flux (vectors),
(a) the stream function (shade) and (b) potential
function at 100hPa in August 1992.

• The stream function (rotational component) of
  water vapour flux is about four times as large as
  the potential function (divergence component)
  over the Asian monsoon region.
• The divergence component is enhanced at the
  south side of the Asian monsoon high (lt 30ºN)
  especially over the western Pacific.

9
9
(a)
(a)
?
(b)
(b)
? -2.32e-5 10-9g/kg/s
?
? -4.76e-6 10-9g/kg/s
Figure 5. Map for (a) R and horizontal wind
(vectors) on the 380 K isentropic surface and (b)
seasonal mean water vapour flux (vectors) and the
divergence (shade) on the 355 K
isentropic surface. The contour indicates
potential vorticity from 2 to 4 PVU
106Km2/kg/s.

• At both isentropic surfaces, the seasonal mean
  water vapour flux divergence becomes the main
  component, and as a result, the distribution of
  R term (diabatic effects) is quite similar to the
  seasonal mean water vapour flux divergence field.
• The seasonal mean water vapour flux divergence
  has a quasi-Rossby wave structure which has the
  wave train along the Asian subtropical jet ( ?
  shown in Fig. 5 (b)).
• The upward water vapour flux divergence (R) is
  large over the convective region ( ? in Fig. 5
  (a)), where the isentropic surfaces tilts
  westward increasing height as shown in Fig. 3.

10
10
6. Summary and Discussion

• The water vapour variation over the Asian summer
  monsoon was investigated with a focus on the
  vertical direction using the UARS MLS water
  vapour data (version 7.02) during the summer
  1992.
• The latitude (longitude)-height sections of
  seasonal mean water vapour show that the wetter
  air intrudes into the tropical lowermost
  stratosphere over the convective region
  (70-180ºE,25-35ºN) passing through the isentropic
  surfaces (e.g., 380 K), and extends to the
  mid-latitude lower stratosphere passing through
  the subtropical tropopause along the isentropic
  surfaces (e.g.,350 K). This suggests that both
  vertical and horizontal stratosphere -
  troposphere exchanges existed potentially around
  the Asian summer monsoon region.
• In order to estimate both water vapour fluxes
  passing through the isentropic surfaces (over the
  convective area) and along the isentropic
  surfaces at the north side of the monsoon high,
  the water vapour budget is calculated
  diagnostically by using the conservation equation
  of water vapour for the isentropic coordinate.
• The time (seasonal) mean water vapour flux
  divergence is larger than the other components,
  such as intraseasoanl and synoptic time scales
  flux divergences. The horizontal structure of the
  seasonal mean water vapour flux divergence
  resembles a quasi-stationary Rossby wave
  structure which has the wave train along the
  Asian subtropical jet (ASJ), especially between
  the north side of the Asian monsoon high and the
  easternmost Pacific.
• The isentropic surfaces larger than 365 K tilt
  westward increasing height over the convective
  region (the southeast side of the anticyclone)
  where the easterly wind is enhanced. The water
  vapour vertical flux divergence associated with
  the diabatic effects is large over the convective
  region.
• I came to a conclusion that the diabatic effects
  and the stationary Rossby wave associated with
  the Asian summer monsoon play an important role
  in both the water vapour variation and the
  exchange between the upper troposphere and lower
  stratosphere in the tropical and extra-tropical
  region.

11
11
References
Bannister et al., Q. J. R. Meteorol. Soc., 130,
1531-1554, 2004. Dethof,A, et al., Q. J. R.
Meteorol. Soc., 125, 1079-1106,
1999. Dunkerton,T.J., J. Geophy. Res. Atmosphere,
100, D8, 16,675-16,688, 1995. Mergenthaler, J..L.
et al., J. Geophy. Res. Atmosphere, 104, D18,
22,183-22,194, 1999. Park M. et al., J. Geophy.
Res. Atmosphere, 109 (D3)Art. No. D03302 2004.
Randel,W. and M. Park 2005. Read, W.G. et
al., J. Geophy. Res. Atmosphere, 109, D6,
doi10.1029/2003_JD004056, 2004. Terao, T., J.
Meteor. Soc. Japan, 77, 1271-1286, 1999.
Acknowledgements
The author would like to thank Dr. Toru Terao at
Osaka Kyoiku University, Dr. Hatsuki Fujinami at
Nagoya University, Dr. Yoshiyuki Kajikawa at
University of Hawaii, Prof. Masato Shiotani, Drs.
Hisahiro Takashima and Noriyuki Nishi at Kyoto
University for their beneficial comments and
valuable suggestions. I also thank UARS MLS
science staffs , especially William G. Read for
their giving us key data. I acknowledge the UARS
CLAES science team, the European Centre for
Medium-Range Weather Forecasts (ECMWF) for the
datasets. Most figures of this study were drawn
by using the GFD-DENNOU Library.