Seasonal and QBO variations of ascent rate
in the tropical lower stratosphere as inferred
from UARS HALOE trace gas data
Masanori Niwano
Department of Geophysics, Kyoto University, Kyoto, Japan
Koji Yamazaki
Graduate School of Environmental Earth Science, Hokkaido University, Sapporo, Japan
Masato Shiotani
Radio Science Center for Space and Atmosphere, Kyoto University, Kyoto, Japan
Received 17 June 2003; revised 22 September 2003; accepted 9 October 2003; published 27 December 2003.
[1] Seasonal and interannual variations in ascent rates are investigated as a function of
latitude and height, using water vapor (H2O) and methane (CH4) data from the
stratospheric measurements of the Halogen Occultation Experiment (HALOE). The ascent
rate is inferred from the ascending signal of variations in the entry value of [H2O] +
2[CH4] (Hˆ). Within ±15! of the equator the derived ascent rate exhibits two kinds of
dominant variations with a clear latitudinal structure, seasonal variation, and the quasi-
biennial oscillation (QBO). The seasonal cycle exhibits a vertically in-phase variation,
with a northern winter maximum of 0.2–0.4 mm s!1 and a summer minimum of
"0.2 mm s!1 in the 20–60 hPa layer. The latitudinal structure is characterized by an early
appearance of a subtropical summer maximum of the ascent rate and by double peaks at
10–15!N and S during the northern winter season. The QBO component of the ascent
rate shows tropically confined anomalies with a rapid downward propagation, but mass
attenuation anomalies estimated from the ascent rate show a much slower downward
propagation. The descent anomalies exhibit a well-structured and equatorially
symmetric variation, while the ascent anomalies have a tendency to propagate
latitudinally. This might be connected with the phase dependency of the QBO
acceleration. An examination of the phase and amplitude of the ascent rate and
temperature for both the seasonal and QBO components emphasizes that the radiative
damping timescale is considerably long (40–100 days) below 40 hPa. INDEX TERMS:
0341 Atmospheric Composition and Structure: Middle atmosphere—constituent transport and chemistry
(3334); 3334 Meteorology and Atmospheric Dynamics: Middle atmosphere dynamics (0341, 0342); 3319
Meteorology and Atmospheric Dynamics: General circulation; KEYWORDS: mean meridional circulation,
transport, water vapor
Citation: Niwano, M., K. Yamazaki, and M. Shiotani, Seasonal and QBO variations of ascent rate in the tropical lower stratosphere
as inferred from UARS HALOE trace gas data, J. Geophys. Res., 108(D24), 4794, doi:10.1029/2003JD003871, 2003.
1. Introduction
[2] Mean meridional circulation in the tropical lower
stratosphere has an important role in determining the
stratospheric transport of air of tropospheric origin. Air
mass is distributed to the whole middle atmosphere by
tropical upwelling and extratropical poleward flow in the
lower stratosphere, and by meridional flow from the
summer to the winter hemisphere in the upper stratosphere
and the mesosphere [e.g., Dunkerton, 1978]. The mean
meridional circulation in the middle atmosphere is produced
by planetary waves, synoptic waves, and gravity waves, as
illustrated in Plumb [2002]. Additional tropical thermal
forcing is also necessary to drive the observed upwelling
maximum on the summer side of the lower stratosphere
[Plumb and Eluszkiewicz, 1999; Semeniuk and Shepherd,
2001]. In the lower stratosphere, longwave heating by
ozone absorption of upward radiative flux from the tropo-
sphere could deform tropical upwelling [Norton, 2001].
[3] The stratospheric meridional circulation in the tropics
is mainly modulated by two dominant kinds of variations:
seasonal and quasi-biennial. The seasonal variation of
upwelling in the equatorial lower stratosphere shows a
maximum during the northern fall-winter season, which is
explained by a superposition of upward wave momentum
JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 108, NO. D24, 4794, doi:10.1029/2003JD003871, 2003
Copyright 2003 by the American Geophysical Union.
0148-0227/03/2003JD003871$09.00
ACL 12 - 1

flux in both hemispheres. The maximum momentum flux in
the Northern Hemisphere is observed during November–
March and in the Southern Hemisphere during September–
November, so the total momentum flux of both hemispheres
shows continuous large values from September to March
[e.g., Randel et al., 2002, hereinafter referred to as R02].
This wave drag variation produces the corresponding sea-
sonal variation of temperature in the lower stratosphere
[Reed and Vlcek, 1969; Yulaeva et al., 1994]. The quasi-
biennial oscillation (QBO) in the meridional circulation
shows a well-defined two-cell structure symmetric with
respect to the equator, related to the zonal wind direction
and temperature [Plumb and Bell, 1982; Dunkerton, 1985].
However, an asymmetric structure is pronounced during the
solstitial seasons, such that the winter cell extends to higher
latitudes while the summer one almost disappears. The
asymmetry can be explained by meridional transport of
the QBO anomalies of angular momentum due to the
meridional wind from the summer to the winter hemisphere
[e.g., Baldwin et al., 2001].
[4] These kinds of variations in tropical upwelling are
indirectly estimated from the examination of radiative heat-
ing rates or wave forcing for seasonal cycle [e.g., Rosenlof
and Holton, 1993; Rosenlof, 1995; Eluszkiewicz et al.,
1996; R02] and for the QBO [Randel et al., 1999]. How-
ever, both the calculations include large uncertainties in the
equatorial lower stratosphere. The radiative heating rate is
quite small because the equatorial lower stratosphere
approaches the equilibrium temperature. So the temperature
difference near the QBO shear region or the tropical
tropopause region leads to larger percentage difference of
heating rate and also the residual circulation in the lower
stratosphere. There are other major uncertainties in aerosol
heating effect and tropospheric cloudiness in the equatorial
lower stratosphere [Olaguer et al., 1992; Eluszkiewicz et al.,
1997]. With respect to the calculation of wave forcing, on
the other hand, the absolute vorticity vanishes in the deep
tropics, which makes it difficult to calculate the latitudinal
distribution of the tropical upwelling.
[5] Stratospheric measurements of water vapor (H2O) and
methane (CH4) by the Halogen Occultation Experiment
(HALOE) on board the Upper Atmosphere Research Satel-
lite (UARS) have revealed seasonal variation of total
hydrogen [H2O] + 2[CH4] (Hˆ) since 1991 (Figure 1) and
provide a valuable opportunity to estimate the dynamical
properties, such as ascent rate, horizontal and vertical
mixing, and empirical age spectrum [e.g., Mote et al.,
1996]. Mote et al. [1998] deduced a time mean vertical
distribution of upwelling, vertical diffusion and horizontal
attenuation simultaneously from Hˆ and methane data by
applying a simple one-dimensional model, and showed
vertical velocity in a good agreement with the results
derived from the calculation of radiative heating. Some
extensive studies succeeded in deriving seasonal cycle and
the QBO component in the tropical upwelling using water
vapor profiles of the HALOE and in situ measurements,
respectively [Mote et al., 1996; Andrews et al., 1999;
Niwano and Shiotani, 2001]. However, these studies did
not make a full spectral analysis of ascent rate of Hˆ as a
function of latitude and altitude.
[6] In this study, we infer ascent rate from the annually
varying signature of Hˆ in the tropical lower stratosphere,
and describe the variations in terms of latitude, altitude, and
frequency. The analysis of the frequency is focused on
seasonal and interannual variations using HALOE data.
The tropical isolation from midlatitudes, reported by various
studies [Plumb, 1996; Andrews et al., 2001; Haynes and
Shuckburgh, 2000], backs up our estimation of the ascent
rate from Hˆ data in the tropical lower stratosphere within the
subtropical barriers. First, the satellite, rawinsonde and
reanalysis data used in this study and the method of
estimating the ascent rate are described in section 2. In
section 3 the derived ascent rate then is shown for both the
seasonal and QBO components. The validity of estimating
ascent rate, and the relationship between variations in ascent
rate and temperature are discussed in section 4. Finally, a
summary is given in section 5.
2. Data and Methods
2.1. HALOE Trace Gas Data
[7] In this study, we use the H2O and CH4 profiles by the
HALOE version 19 data from November 1991 to December
1999. Long term observations from HALOE provide con-
tinuous and high quality data, as summarized by Russell et
al. [1993]. The V19 data of H2O and CH4 have an accuracy
of ±5–10% [Dessler and Kim, 1999], which is similar to
that of the V17 data reported by Harries et al. [1996] and
Park et al. [1996]. The previous H2O retrieval method
(V18) had large differences between sunrise and sunset
observations, while the sunrise/sunset differences for the
V19 H2O data have been significantly reduced (E. E.
Remsberg, personal communication, 2002). A further com-
parison of HALOE V19 H2O with data from other satellite,
balloonborne, and aircraft measurements, is included in the
Stratospheric Processes and Their Role in Climate (SPARC)
water vapor assessment report [Kley et al., 2000].
[8] The HALOE water vapor data have a dry bias of 5–
20% between 60 and 100 hPa compared with aircraft or
balloon measurements [Kley et al., 2000]. Randel et al.
[2001] noted that this dry bias may be partly caused by
errors in the temperature-dependent O2 continuum absorp-
tion, which is used in the V19 HALOE retrieval. The V19
retrieval of HALOE data shows that Hˆ at lower latitudes
decreases above 40 hPa in the stratosphere with height [Kley
et al., 2000], and correspondingly that the effective yield of
H2O from CH4 destruction is less than 2 in the lower
stratosphere [Dessler and Kim, 1999], as previously sug-
gested [Le Texier et al., 1988]. However, it should be
emphasized that time mean Hˆ may be vertically constant
throughout the stratosphere within the accuracy of HALOE
data.
[9] The HALOE data (level 2) are available on pressure
levels of 1000 # 10!(i/30) hPa (i = 0, 1, . . .), corresponding
to a vertical spacing of about 0.5 km. HALOE instrument
retrieves data in the tropics approximately 10 times per a
year at sunrise and sunset events separately. For details of
constructing zonal mean values with a latitudinal grid
interval of 2.5 degrees, refer to Niwano and Shiotani
[2001]. The mean seasonal cycle is calculated by sampling
raw data on each calendar month, based on the data only
spanning January 1993 and December 1999, because data
are sparse under a condition of enhanced Pinatubo aerosol
concentration in the lower stratosphere. If there are fewer
ACL 12 - 2 NIWANO ET AL.: VARIATIONS IN ASCENT RATE