Content uploaded by Hanh Nguyen
Author content
All content in this area was uploaded by Hanh Nguyen on Feb 16, 2018
Content may be subject to copyright.
Advanced Review
The expanding tropics: a critical
assessment of the observational
and modeling studies
Christopher Lucas,∗Bertrand Timbal and Hanh Nguyen
This review provides comprehensive coverage of the tropical expansion literature
to date. The primary focus is on the annual- and zonal-mean behavior of the
phenomenon. An idealized model that identifies the mean meridional circulation
as a hemisphere-wide structure with significant tropical–extratropical interaction
is introduced as background for the understanding of the expansion and the
methodologies used for detection. A variety of metrics from different data sources
have been used to identify an expansion of the global tropics since 1979 by 1◦–3◦
latitude in each hemisphere, an average trend of approximately 0.5◦–1.0◦decade−1.
The symmetry of this expansion—whether Northern and Southern hemispheres
are expanding at the same rate—is unclear. Limitations of observational datasets,
including reanalyses, prevent a more precise determination at this time. General
circulation models are able to qualitatively reproduce this expansion, but generally
underestimate its magnitude. Multiple factors have been identified as potential
drivers of the expansion, including increasing greenhouses gases, stratospheric
ozone depletion, and anthropogenic aerosols. No single factor by itself appears to
explain the full expansion, perhaps a shortcoming of the models or experiment
design. It may be that some combination of these forcings is producing the change,
but the relative contribution of each forcing to the expansion is currently unknown.
The key issues remaining to be resolved are briefly summarized at the end. ©2013
John Wiley & Sons, Ltd.
How to cite this article:
WIREs Clim Change 2013. doi: 10.1002/wcc.251
SUBTROPICAL DRYING AND
TROPICAL EXPANSION
In recent decades, observations suggest that the
hydroclimate in many regions has been chang-
ing. Reduced precipitation and more frequent
droughts have been observed in southern portions
of Australia,1,2 much of the Mediterranean region,3,4
portions of North America, particularly the south-
western United States,5the Altiplano in South
America,6and northern China.7Similar tendencies
have also been noted in Africa and southeast Asia.8,9
∗Correspondence to: c.lucas@bom.gov.au
Centre for Australian Weather and Climate Research/Bureau of
Meteorology, Melbourne, Vic, Australia
Conflict of interest: The authors have declared no conflicts of
interest for this article.
While some places may only show seasonal rain-
fall declines, these deficits can drive extreme heat in
later seasons.10 Droughts and precipitation have many
natural influences across multiple timescales; telecon-
nections from changes in sea surface temperature
(SST) driven by phenomena like the El Ni ˜
no-Southern
Oscillation (ENSO) or other decadal and centennial
variability8,11,12 have a particularly noteworthy effect.
However, some portion of the apparent trend may be
linked to changes in the broader circulation patterns of
the globe caused by anthropogenic global warming,8
trends that are projected to continue through the 21st
century.9
A focus of these drying trends is found in the
subtropics, among the driest regions on Earth. The
subtropics are broadly associated with the downward
branch of the Hadley circulation (HC), a planetary-
scale tropical overturning circulation spanning half the
©2013 John Wiley & Sons, Ltd.
Advanced Review wires.wiley.com/climatechange
globe. Conceptually, this part of the circulation can
be considered the ‘edge of the tropics’. One projected
consequence of anthropogenic climate change is that
this edge is expected to expand toward the poles,
shifting precipitation patterns, and broadening the
area covered by frequent drought.13,14 During the last
decade, numerous studies using a variety of metrics to
identify the tropical edge have emerged, suggesting
just such a shift referred to as an expansion of
the tropics. Circulation patterns outside the tropics
have also been related to the expected hydroclimatic
changes.1–4These extratropical changes may, in fact,
be closely related to tropical expansion.
Earlier reviews13,14 summarized the basic
understanding of tropical expansion. However, much
remains unclear, including the exact amount of
tropical expansion and the physical forcings driving it.
A significant amount of new research has subsequently
improved the understanding. Herein, we provide a
comprehensive review of the state of knowledge in
relation to tropical expansion.
Understanding tropical expansion necessitates
knowledge of the components of the mean meridional
circulation (MMC) and the interactions between
them. A key factor lies in defining the ‘edge of the trop-
ics’, a nebulous concept that is somewhat resistant to
precise definition. Part of this nebulosity arises because
the edge is not fixed, varying with both the change of
season and geographic location. We will focus primar-
ily on the zonally and annually averaged circulation
in this review, ignoring some of the finer detail that
ultimately may be important to understanding the
issue at hand. To provide a background, an idealized
model of the MMC is described. This qualitative
model presents the MMC as a pair of hemisphere-wide
structures in which the extratropics play a crucial role.
The observational studies of tropical expansion are
then summarized and organized around the features
of the MMC that are used to identify the tropical
edge. The various definitions of the tropical edge that
have been used are given. A discussion of the sources
of observational uncertainty in the different estimates
follows. Many results are derived from reanalysis-
based data, which may contain inhomogeneities and
other artifacts that could affect the understanding of
the long-term behavior of the climate. Potential uncer-
tainty introduced by the individual methodologies of
the studies is also described. The mechanisms and
climate forcings behind tropical expansion are then
explored, largely described from modeling studies.
The forcings are generally anthropogenic in origin,
although natural factors do influence the tropical
edge. A synthesis of the observations and modeling
results is then presented, where model-predicted
impacts are compared to the observed results to pro-
vide evidence for the relative roles of the prospective
drivers. The final section addresses some remaining
questions and directions for future research into
this topic.
AN IDEALIZED MODEL OF THE MMC
Figure 1 presents a schematic diagram of the idealized
MMC, depicting the zonally and annually averaged
POLAR FRONT
FIGURE 1|Schematic diagram detailing the components and physical processes of the idealized mean meridional circulation (MMC). Diagram
represents the hemispheric annual and zonal-mean flow; a mirror image about the equator would represent the global circulation. Labels in all
capitals represent the main features of the MMC. Bold labels refer to subcomponents of the circulation. Text in italics identify important physical
processes. See text for further discussion.
©2013 John Wiley & Sons, Ltd.
WIREs Climate Change The expanding tropics
circulation. Only one hemisphere is shown; the full
global MMC would be represented by a ‘mirror image’
in the other hemisphere. This view of the flow is devel-
oped from ‘classical’ meteorological concepts and
studies using the (dry) isentropic streamfunction15,16
to diagnose the flow. Paluis et al.17,18 have developed
a conceptually similar model based on a moist
isentropic view of the circulation. These models differ
from the more conventional ‘three-cell’ model based
on the isobaric streamfunction, while still sharing
many characteristics. The most notable differences
are the enhanced extratropical circulation and the
idea of a complex hemisphere-wide overturning
circulation. This diagram and its discussion provide
a qualitative framework for the methodologies and
results discussed in this article.
In the tropics, many similarities to the three-cell
model are observed. The flow is characterized by a
deep meridional overturning circulation known as
the Hadley circulation (HC). Mean ascent, driven by
tropical mesoscale convective systems,19 is observed
near the equator in the Intertropical Convergence
Zone (ITCZ) with a broad compensating subsidence
near 30◦latitude. Radiative cooling drives much
of the descent. In the lower-troposphere, the trade
winds move air toward the equator; in the upper-
troposphere, the flow moves poleward and terminates
in the subtropical jet (STJ). Webster20 provides a
thorough elementary review of the characteristics
and physics of the HC. The circulation varies with
season and with the rising branch generally located
off the equator by 5◦latitude or more in the summer
hemisphere. The descending branch also migrates
with the season. The circulation in the winter
hemisphere is significantly stronger than that in the
summer hemisphere.21
Idealized axisymmetric models of the HC22,23
suggest that the HC width depends on the size and
rotation rate of the Earth, the tropical tropopause
height, and the pole-to-equator temperature gradient.
In idealized aquaplanet simulations,24 the width of the
HC increases with increasing mean temperature and
temperature gradient. Other experiments25 indicate
more complex relationships, with feedbacks between
many different processes. No single parameter solely
determines the scale of the HC. One important factor,
often neglected in simpler models, is the effect of
extratropical baroclinic eddies.26 A significant role for
these eddies has been suggested since the middle of
the 20th century; Palm´
en and Newton27 summarize
this earlier work. Allowing for baroclinic instability
suggests a dependence of the HC width on the gross
static stability,28 a relationship broadly supported in
both idealized and realistic general circulation model
(GCM) simulations.24,29,30 This relationship has been
found to apply in all seasons.31
In Figure 1, eddies dominate the circulation in
the ‘polar front’, a symbiosis of the eddy-driven polar
front jet (PFJ) and baroclinic waves. The polar front
marks the position of the ‘storm track’, whose location
and variability is related to the ‘annular modes’.32
When the storm track is active, an overturning
circulation similar to the Ferrel circulation is observed
in association with the baroclinic waves and transient
weather systems.33 On average, the flow in the
extratropics is quasi-isentropic. Moist air ascends in
the polar front, resulting in enhanced precipitation,
and returns equatorward, its moisture largely
removed.34 This circulation also varies seasonally,
generally moving poleward and strengthening in the
winter hemisphere, a response to greater baroclinicity.
Significant longitudinal variability in wave activity
arises from surface topographic features like mountain
ranges and land-ocean boundaries.27
The exact nature of the relationship between the
extratropical eddies and the HC remains unclear and
may also depend on season. Bordoni and Schneider35
suggest that the impact of eddies on the HC is variable
during the year. In their study of the Asian monsoon,
they find that direct thermal driving becomes more
important than eddies with the onset of the monsoon.
Kang and Polvani36 found a strong interannual
relationship of the HC edge and the position of
the eddy-driven jet in the Southern Hemisphere (SH)
during summer, with the HC edge moving 1◦latitude
for a 2◦shift in the position of the westerly jet. Ding
et al.37 show a linkage between tropical SST and the
SH annular mode (SAM) through tropically forced
Rossby wave trains.
In the idealized MMC model (Figure 1),
tropical–extratropical interactions have the greatest
impact in the subtropics, loosely defined as the region
between the HC and the polar front. In general, the
region is dominated by subsidence and the subtropical
ridge (STR). In a relative sense, the air is dry and
precipitation is low. Much of the variability here,
on both seasonal and interannual time scales, is due
to the continual evolution of tropical–extratropical
interactions and their response to different forcings.
For example, deep tropical convection—modulated
by the Madden Julian Oscillation—has been shown
to produce Rossby wave trains which extend well into
the higher latitudes.38
The stratosphere appears to play a significant
role in the MMC. Much of its effect is related to
ozone, and its observed depletion in recent decades.39
The Antarctic ozone hole which forms every spring
in particular has a significant impact on the MMC.
©2013 John Wiley & Sons, Ltd.
Advanced Review wires.wiley.com/climatechange
Increasing wind shear in the lower stratosphere has
an effect on baroclinic waves, increasing their phase
speed.40 Stratospheric water vapor concentrations are
known to affect the decadal rate of warming of the
surface climate41 and these changing concentrations
could drive changes in the tropospheric MMC.42 This
is consistent with evidence that suggests that the
stratospheric circulation changes precede anomalous
tropospheric weather regimes. These stratospheric
changes are translated to the troposphere through
the annular modes.43
The model represents the ‘typical’ structure of
the MMC. At any given time or place, the represented
features may be more or less prominent and/or
in a different position. As noted earlier, seasonal
variation is particularly important. Also, the seasonal
evolution of the MMC is quite different in the two
hemispheres; the variability of intensity and extent
of the circulation is much larger in the Northern
Hemisphere (NH) compared with the SH.21 The
distribution of land, more prevalent in the NH, is an
important factor in this different evolution.44 Quasi-
stationary waves, associated with mountain ranges,
also play a role.45 Rind and Perlwitz46 suggest that
topography is a factor in HC width based on a
set of paleoclimate simulations. These factors also
contribute to significant longitudinal variability in the
behavior of the MMC, particularly in the extratropics.
OBSERVATIONS OF TROPICAL
EXPANSION
A variety of datasets and/or metrics have been
used in tropical expansion studies. From a practical
standpoint, the definition of the ‘edge of the tropics’ is
critical. With most metrics the point of transition from
tropics to extratropics is neither intuitively obvious
nor unambiguous, but rather is a transition occurring
over a range of latitudes (i.e., the subtropics);
defining ‘the edge’ is then a matter of choosing some
point within this transition zone that is (hopefully)
representative of the whole. The methods for making
this choice can be subjective (but rationally decided)
or objective (e.g., chosen through an algorithm), but
ultimately the distinction is arbitrary. The subtleties of
a complex region of the atmosphere are reduced to a
single number that may or may not be representative of
the behavior of the whole MMC. There is no universal
definition of the tropical edge that is applicable across
all datasets; each sees the transition zone in a different
way and as such, the responses of a given metric
show different sensitivities. Adding further complexity
to these metrics are the effects of seasonal and/or
longitudinal variability of the MMC; the zonal and
annual averaging used in many studies may introduce
uncertainty or mask important processes in its own
right. These aspects, along with the different time
periods covered by different studies, confound the
interpretation and comparison of the trends.
This section is organized around the different
features of the MMC and/or datasets that have been
used to identify the tropical edge. For each, a brief
overview of the methodology is given and the results
briefly discussed. A table is produced that summarizes
the information from each study, including the data
sources, the exact metric used, and the annual
observed rate of expansion, given by hemisphere when
available. As many studies here use the various global
reanalysis products, references to those products and
their acronyms are shown in Table 1.
Tropopause Methods
The tropopause height frequency (THF) method-
ology55 is frequently used in tropical expansion
studies. Outside of the subtropics, annual histograms
of THF have a single peak. This peak is found at
15–17 km in the tropics and at 11–13 km in the
extratropics. Within the subtropics, the THF is
bimodal, an indicator used to define the tropical edge.
The number of tropical tropopause days (TTD) is
defined as the number of days that the tropopause
exceeds a specified height threshold representative of
the tropics. An array of TTD in time-latitude space is
created and given values contoured. The slope of the
contours provides an estimate of the long-term trend.
Lu et al.56 also applied this basic methodology.
The definition of the various thresholds is
crucial. Generally, tropopause height thresholds from
TABLE 1 Acronyms and References for the Reanalysis Products
Discussed in the Article
Acronym Expansion and References
NCEP National Center for Environmental Prediction
(NCEP)/National Center for Atmospheric
Research (NCAR) Reanalysis47
ERA40 European Centre for Medium-Range Weather
Forecasting (ECMWF) 40-year Reanalysis48
NCEP2 NCEP/Department of Energy (DOE) Reanalysis49
JRA25 Japanese 25-year Reanalysis50
ERA-I ECMWF Interim Reanalysis51
CFSR NCEP Climate Forecast System Reanalysis52
20CR NOAA-CIRES 20th Century Reanalysis53
MERRA National Aeronautics and Space Administration
(NASA) Modern Era Reanalysis for Research
and Applications54
©2013 John Wiley & Sons, Ltd.
WIREs Climate Change The expanding tropics
14.5 to ∼16 km have been chosen, although there is
some sensitivity to the choice of tropical tropopause
threshold and the definition of the tropical edge.57
This sensitivity arises, at least in part, from the
observed global rise in tropopause heights58 of
∼40–80 m decade−1,59–61 and the use of a fixed
height threshold could act to confound the tropical
expansion signal. Some studies57,58 define relative
tropical tropopause thresholds to try and remove the
effect of the ‘background’ tropopause trends. These
methodologies reduce the magnitude of the tropical
expansion trends. However, the amount of trend
reduction has little apparent relationship with the
time tendencies in the tropopause thresholds and the
results are not robust.
A comparison of rates of THF-based tropical
expansion in the SH computed directly from historical
radiosonde data and from four reanalyses62 suggests
that much of the sensitivity is related to issues with the
underlying data quality; reanalysis-derived tropopause
heights63 are subject to strong biases in the subtropics.
Global tropopause trends were estimated to have a
small effect on the estimated expansion trends (less
than 0.1◦decade−1).62
A blended tropopause height dataset61 derived
from the two ECMWF reanalyses combines thermal
and dynamic definitions of the tropopause is also used
to examine expansion rates. This dataset reduces the
bias in the subtropics. The THF methodology is not
used in this study, but expansion rates are consistent
with those studies. A non-THF tropopause-based
methodology—the potential temperature difference
between the tropopause and the surface (θ ), a dry
bulk stability parameter—has also been used to iden-
tify the edges of the tropics and its long-term trend.64
Table 2 summarizes the results from studies
using a tropopause-based metric. For the most part,
these indicate an expansion of less than 1◦decade−1
in each hemisphere. Many THF-based studies indicate
that the SH has expanded more rapidly than the NH.
Results with relative thresholds57,58 generally show
smaller trends and even contraction in some cases.
The results are dependent on the reanalysis chosen.
In several studies, significant interannual variability
is identified,56,58,62 with effects due to ENSO and
large volcanic eruptions noted. Seasonal differences
in tropopause-based trends may also be present,
generally suggesting a weak tendency toward greater
TABLE 2 Summary of Studies Using the Tropopause-Based Methodologies
Investigators Data Sources Thresholds Time Period
Expansion
(◦Latitude Decade−1)
Siedel and Randel55 NCEP, ERA40
radiosondes
15.5 km/TTD =300, 200, or
100
1979–2005 Both: 1.7–3.1
Equal between
hemispheres
Lu et al.56 NCEP and ERA40 120 hPa/200 days 1958–1999 SH: 0.5–0.7
NH: 0.05–0.2
Birner57 NCEP, NCEP2, ERA40,
and JRA25
Variable/equivalent latitude 1979–2008 SH: −0.4 to 0.7
NH: −0.2 to 0.0
NCEP, NCEP2, ERA40,
and JRA25
Variable 2/equivalent
latitude
1979–2008 SH: 0.0–0.6
NH: −0.2 to 0.2
Wilcox et al.61 S ERA-I, ERA40 blended
tropopause
Average of 15.0, 15.5, and
16.0 km/NA
ERA40: 1958–2001
ERA-I: 1989–2007
ERA40 Both: 0.7
ERA-I Both: 0.9
(60–80% in SH)
Davis and Rosenlof58 S NCEP, CFSR, JRA25,
and MERRA
15.0 km/equivalent latitude 1979–2009 SH: 0.2–0.8
NH: −0.1 to 0.6
S NCEP, CFSR, JRA25,
and MERRA
Variable, 1.5 km below
tropical mean/equivalent
latitude
1979–2009 SH: 0.1–0.4
NH: −0.3 to 0.3
Lucas et al.62 Radiosonde 14.5 km/200 days 1979–2010 SH: 0.4, regionally varies
0.2–0.6
Davis and Birner64 MERRA, ERA-I, and
NCEP
Subtropical stability maxima 1979–2010 Both: −0.5 to 0.9
Positive values equal expansion regardless of hemisphere. Thresholds refer to tropical tropopause height threshold and the number of TTD used to define the
edge. An ‘s’ after the investigators indicates the study examined the seasonality of the expansion.
NH, Northern Hemisphere; SH, Southern Hemisphere.
©2013 John Wiley & Sons, Ltd.
Advanced Review wires.wiley.com/climatechange
expansion in the summer and autumn months of each
hemisphere58,61 although these trends are not always
statistically significant.58 Longitudinal variability in
the amount of tropical expansion was also noted with
this methodology.62
Satellite Methods
Varied metrics derived from satellite-based platforms
have been widely used to investigate tropical
expansion. One common method is analogous to
the THF methodology; zonal averages of the metric
are arranged in a time-latitude array, a definition
of the tropical edge chosen, and trends estimated
from the slope of chosen contours. As with the THF
methodology, background trends in the quantity of
interest may mask the true rate of tropical expansion.
A common metric in satellite studies is outgoing
longwave radiation (OLR).58,65,66 Since 1979, broad
trends have been noted in many OLR datasets, on
the order of 1–3 W m−2decade−1, depending on the
dataset and the area examined.58,67–69 Several recent
studies70,71 have examined cloudiness measurements
from the International Satellite Cloud Climatology
Project (ISCCP)72 and the Pathfinder Atmosphere
(PATMOS)73 datasets, defining the position of the
edge from either a fixed threshold or the minimum
in subtropical cloudiness. One study74 examined the
long-term changes in different NH ozone regimes
derived from satellite data, whose boundaries are
closely linked with upper-tropospheric fronts.75 A
later study76 used this approach to extend these results
into the SH. Trends in upper-troposphere and lower-
stratosphere temperature data retrieved from the
(Advanced) Microwave Sounding Unit (MSU/AMSU)
have also been used to infer tropical expansion,
either through analysis of global temperature trends77
or changes to the idealized temperature structure
associated with the STJ.78
Table 3 summarizes the satellite-based studies.
Large differences exist in many estimates using the
same or similar data sources. Four OLR datasets each
shows a general widening trend in both hemispheres,
greater in the NH.65,66 However, differences between
the datasets are large, in some cases statistically
significant.58 Using a relative threshold reduces the
TABLE 3 Summary of Studies of Tropical Expansion Using Satellite-Based Methodologies
Investigators Data Sources Metric Time Period
Expansion
(◦Latitude Decade−1)
Hudson et al.75 Satellite-derived ozone Total ozone 1979–2003
1979–1991
NH: about 1◦
NH: 1.1
Fu et al.77 MSU Global temperature trends 1979–2005 Estimated total shift of
1◦per hemisphere
Hu and Fu65 S HIRS, ISSCP, and GEWEX Monthly OLR, 250 W m−21984–2004
1980–2003 (HIRS)
SH: 0.3–0.9
NH: 0.6–1.1
Johanson and Fu79 HIRS, ISSCP, and GEWEX Multidataset OLR,
250 W m−2
1984–2004
1979–2003 (HIRS)
Both: 3.0
Hu et al.66 S NOAA OLR, 250 W m−21979–2009 SH: 0.3
NH: 0.9
Fu and Lin78 3 MSU datasets Lower stratospheric
temperature
1979–2009 SH: 0.2
NH: 0.3
Zhou et al.70 S ISSCP cloudiness Subtropical 60% coverage 1984–2006 SH: 0.7–1.3
NH: 1.3–1.6
S ICCSP cloudiness Subtropical cloud cover
minimum
1984–2006 SH: −0.1 to −0.6
NH: 0.3–1.6
Davis and Rosenlof58 S NOAA, HIRS, GEWEX,
and ISSCP
OLR/variable: 20 W m−2
below subtropical
maximum
1979–2005,
1979–2009 for
NOAA
SH: −0.3 to 0.4
NH: 0.1–0.5
Allen et al.71 ISSCP, AVHRR cloud
climatology
Subtropical cloud cover
minimum
1983–2008 SH: 0.36
NH: 0.39
Hudson76 Satellite-derived ozone Total ozone 1979–2010 SH: 2.0
NH: 1.2
Positive values equal expansion regardless of hemisphere. An ‘S’ after the investigators indicates the study examined the seasonality of the expansion.
NH, Northern Hemisphere; SH, Southern Hemisphere.
©2013 John Wiley & Sons, Ltd.
WIREs Climate Change The expanding tropics
magnitude of the trends, such that two OLR datasets
suggest contracting trends in the SH.58 The studies
using minimum cloudiness metrics show considerably
different results,70,71 with differing signs in the SH.
Global temperature trends suggested a total shift in
the jet stream of about 1◦in each hemisphere,77 later
refined to identify total expansions of 1◦in the NH and
0.6◦in the SH.78 Studies using ozone data74,76 indicate
a broadening of about 1◦decade−1in the NH, partic-
ularly between 1979 and 1991. In the SH, expansion
using this methodology has averaged 2.0◦decade−1
between 1979 and 2010. The OLR-based studies show
only weak signals of seasonal variability.58,65,66 Inter-
annual variability was not discussed.
Streamfunction Methods
One common metric is the edge of the HC as derived
from calculations of the (isobaric) mass streamfunc-
tion ().21,58,64,65,71,79,80The mass streamfunction is a
vertical integral of the meridional wind and produces
the familiar three-cell model of the MMC. The edge
of the tropics is generally taken as the latitude of
the subtropical zero isopleth in the mid-troposphere.
The monthly position of the HC is determined, which
can be averaged to the desired resolution and tracked
in time to estimate the expansion. This metric,
annually averaged, for eight reanalysis products is
shown in Figure 2. Visual examination of the time
FIGURE 2|Time series of the position of the subtropical edge of
the Hadley cell as defined from the streamfunction in the eight modern
reanalyses for the Northern (top) and Southern (bottom) Hemisphere.
(Reprinted with permission from Ref 21. Copyright 2013 The American
Meteorological Society). Time series from ERA40 (post 2002) and 20CR
(post 2008) are partially extrapolated from statistical relations with the
other reanalyses.
series shows an expansionary trend is apparent in
each hemisphere—particularly the NH—although
considerable scatter exists in the estimates.
Table 4 summarizes the -based results. Over-
all, expansion rates are below 0.8◦decade−1, with
little difference between the hemispheres.21,58,65 Stud-
ies that examine both hemispheres combined64,79,80
suggest expansion rates of 0.4 and 3.2◦decade−1.
No studies using this methodology indicate con-
traction. Seasonally, expansion is more rapid in
the respective summer and autumn seasons of each
hemisphere,21,58,65 when the HC is extended furthest
poleward and its intensity is lowest. Despite the
large trends, the NH expansion in many reanalyses
is not statistically significant due to high variability
within the time series.21 The SH expansion is more
robust across the different products, with a shift in
the location indicated across all reanalyses in the late
1990s (Figure 2).
Jet Stream Methods
Another metric used to examine tropical expansion is
the change in position of the various jet streams,
which can be defined in different ways. Several
studies58,81 identify jet streams in several reanalyses
from the mass-weighted wind between 400 and
100 hPa. Jet streams have also been defined from
the zonal-mean zonal wind at 850 hPa and from the
location of the speed maxima of the jet streams as
defined above.58,64 The NH 850 to 300 hPa maximum
wind derived from radiosonde data has also been
investigated.71 While not explicitly considered in this
section, the satellite temperature study noted earlier
interpret their data based on an idealized model
of the STJ, although no wind data are actually
used there.78
Table 5 shows the trends based on these metrics.
For the jet streams defined from the mean wind, trends
of less than 0.2◦decade−1are generally observed.
No consistent difference is noted between upper and
lower jet streams in these metrics.58,81 When using
the latitude of the maximum mean wind, the time
series are much more variable and hence have greater
uncertainty in the trend values.58 This is particularly
notable for the 850 hPa zonal winds in the SH, where
some individual reanalyses identify contractive trends
on an annual basis.58 The seasonality of the trends
is inconsistent across the metrics. For zonal-mean
wind metrics, the seasonality tends to be weaker. For
those based on maxima, the seasonality is larger with
changes of sign and strength throughout the year.
For the 850 hPa maximum wind defined edge in the
SH, expansion is noted in summer and autumn while
strong contraction is noted in winter and spring.58
©2013 John Wiley & Sons, Ltd.
Advanced Review wires.wiley.com/climatechange
TABLE 4 Summary of Studies Using Streamfunction () Methods
Investigators Data Sources Metric Time Period
Expansion
(◦Latitude Decade−1)
Hu and Fu65 S NCEP, ERA40, and NCEP2 Isobaric mass streamfunction 1979–2005 SH: 0.4–1.2
NH: 1.0–2.3
Johanson and Fu79 NCEP, ERA40, and NCEP2 Multidataset average isobaric
mass streamfunction
1979–2004 Both: 3.2
Stachnik and Schumacher80 S NCEP, ERA40, NCEP2,
ERA-I, MERRA, CSFR,
JRA25, and 20CR
Isobaric mass streamfunction 1979–2008 Both: 1.10 (ensemble)
Range: 0.29–1.48
Davis and Rosenlof58 S NCEP, JRA25, CFSR, and
MERRA
Isobaric mass streamfunction 1979–2009 SH: 0.1–0.8
NH: 0.1–0.8
Allen et al.71 NCEP, NCEP2, ERA40,
MERRA, and CFSR
Multidataset average isobaric
mass streamfunction
1979–1999 SH: 0.36 ±0.3
NH: 0.42 ±0.2
Nguyen et al.21 S NCEP, ERA40, NCEP2,
ERA-I, MERRA, CFSR,
JRA25, and 20CR
Isobaric mass streamfunction 1979–2009 SH: 0.1–0.8
NH: 0.2–1.0
Davis and Birner64 MERRA, ERA-I, and NCEP Isobaric mass streamfunction 1979–2010 Both: 0.4–2.0
Positive values equal expansion regardless of hemisphere. An ‘S’ after the investigators indicates the study examined the seasonality of the expansion.
NH, Northern Hemisphere; SH, Southern Hemisphere.
TABLE 5 Summary of Studies Using Jet Stream Methods
Investigators Data Sources Metric Time Period
Expansion
(◦Latitude Decade−1)
Archer and Caldeira81 NCEP and ERA40 Mean latitude of mean 400
to 100 hPa wind
1979–2001 SH: 0.06–0.12
NH: 0.17–0.19
Davis and Rosenlof58 S NCEP, JRA25, CFSR, and
MERRA
Mean latitude of mean 400
to 100 hPa wind
1979–2009 SH: 0.0–0.1
NH: 0.0–0.1
S Mean latitude of mean
850 hPa wind
1979–2009 SH: 0.0–0.2
NH: 0.1–0.3
S Latitude of maximum 400
to 100 hPa wind
1979–2009 SH: 0.1–0.3
NH: 0.0–0.2
S Latitude of maximum
850 hPa wind
1979–2009 SH: −0.4 to 0.2
NH: 0.0–0.2
Allen et al.71 Radiosonde Latitude of 850 to 300 hPa
wind
1979–2009 SH: 0.20
NH: 0.08
Davis and Birner64 MERRA, ERA-I, and NCEP Latitude of maximum wind 1979–2010 Both: 0.3–0.6
Positive values equal expansion regardless of hemisphere. An ‘S’ after the investigators indicates the study examined the seasonality of the expansion.
NH, Northern Hemisphere; SH, Southern Hemisphere.
Surface-Based Methods
Several studies rely upon surface-based variables to
investigate tropical expansion. Several studies66,70,71
utilize the Global Precipitation Climatology Project
(GPCP) monthly dataset82,83 to examine shifts in the
locations/boundaries of the subtropical dry zones, as
defined by the location of the precipitation minimum
or the 2.4 mm day−1contour (Ref 70 only). Several
studies have examined metrics based on P-E (precip-
itation minus evaporation) from several sources; an
observationally based composite P-E dataset derived
from GPCP and Woods Hole Oceanographic Institute
oceanic evaporation data71 and the output from
several reanalysis products.58 The global position of
the STR, derived from sea level pressure data from
three reanalyses and the HadSLP2 dataset,84 has also
been used to estimate tropical expansion.66
Table 6 summarizes these results of the surface-
based studies. The multiple studies using GPCP data
to examine subtropical precipitation minimum have
conflicting results. One study finds that the tropical
edge is moving poleward around 0.5–0.7◦decade−1in
©2013 John Wiley & Sons, Ltd.
WIREs Climate Change The expanding tropics
TABLE 6 Summary of Surface-Based Methodologies
Investigators Data Sources Metric Time Period
Expansion
(◦Latitude Decade−1)
Hu et al.66 S GPCP Subtropical precipitation
minimum
1979–2009 SH: 0.52
NH: 0.71
S NCEP, ERA40, NCEP2,
and HadSLP2
Subtropical ridge location 1979–2009 SH: 0.18
NH: 0.22
Davis and Rosenlof58 S NCEP, JRA25, CFSR, and
MERRA
Subtropical P-E =0 1979–2009 SH:−0.7 to 2.0
NH: 0.0–0.5
Zhou et al.70 S GPCP Subtropical 2.4 mm day−1
isopleth
1979–2007 SH: 0.4
NH: 0.5
S GPCP Subtropical minimum
precipitation
1979–2007 SH: 0.1
NH: 0.0
Allen et al.71 GPCP and WHOI
evaporation
Subtropical P-E =0 1979–2009 SH: 0.31
NH: 0.52
GPCP Subtropical precipitation
minimum
1979–1999 SH: 0.42
NH: 0.31
Positive values equal expansion regardless of hemisphere. An ‘S’ after the investigators indicates the study examined the seasonality of the expansion.
NH, Northern Hemisphere; SH, Southern Hemisphere.
the SH, slightly more in the NH.66 Another indicates
a slower rate of expansion, with slightly more seen
in the SH.71 The third reports little expansion in
either hemisphere using the annual metric, but show
considerable expansion in the NH summer; this study
reports similar results from estimates using the edge
the subtropical dry zone.70 The STR metrics suggest
a poleward trend of around 0.2◦decade−1in each
hemisphere, although there is considerable difference
between datasets.66 Expansion estimates using P-E as
a metric show considerable scatter, particularly in the
SH.58,71 Little consensus exists among these metrics
regarding the strength or timing of the seasonality of
the expansion.
Intensity Changes
Changes to the intensity of large-scale tropical
circulations—both the HC and the zonal Walker
circulation (WC)—have also been reported by several
studies. These changes, while not the main focus of
the article, are useful in interpreting the expansion
results. Table 7 briefly summarizes these findings.
For the HC, the intensity is often estimated using
the peak value of within the cell. Figure 3 shows
the time series of this metric for eight reanalyses.
In the NH, a significant strengthening of the HC
is indicated in five of eight reanalyses. In the SH,
the results are mixed, with both strengthening and
weakening being identified in different products; only
four of eight define a significant trend in either
direction. Seasonally the most robust intensification
in these data is observed in NH winter and spring
(i.e., DJF and MAM), with the SH and other seasons
producing weaker trends, often of a different sign.21
Other studies85,89 using different approaches generally
support these findings, particularly the changes to the
strength of the NH winter (DJF) cell. An even stronger
trend in the intensity of the DJF winter cell was noted
in the ERA40 reanalysis.86 A general intensification in
the annual average HC intensities in both hemispheres
is suggested by a reanalysis ensemble average; two
exceptions were noted in the SH.80 The interannual
variability of the HC using global radiosonde data
shows no obvious trend in the data.86,87 Satellite
data, in particular OLR and reflected shortwave,
were used to infer an intensification of the HC.66
The interpretation of these results was questioned,90
and a subsequent investigation91 significantly lowered
the estimates of radiative imbalance, a result of
adjustments for satellite altitude errors and instrument
drift. Satellite-derived estimates of water vapor
transport indicate an intensification of the HC during
DJF, also suggested by reanalysis data.88
Studies of the trends in WC are contradictory;
decadal variability appears to play a significant
role. Weakening trends have been noted in the
NCEP reanalysis85 and an independent historical
sea level pressure dataset.92 However, satellite water
vapor transport observations suggest that the WC is
strengthening since 1979.88 On an interannual basis,
radiosonde estimates87 suggest that the strength of the
HC and WC varied in opposition to one another, an
observation also supported by satellite data.88
©2013 John Wiley & Sons, Ltd.
Advanced Review wires.wiley.com/climatechange
TABLE 7 Summary of the Observed Changes to the Intensity of the Tropical Circulations
Investigator Subject Data Source Comments
Chen et al.67 Broad tropics Satellite shortwave and
longwave fluxes
Intensity shift not quantified; subtle changes to satellite
radiances suggestive of intensification of the HC
and/or WC
Tanaka et al.85 HC NCEP velocity potential DJF winter cell is intensifying. No significant trend in JJA
WC WC weakening since (at least) 1979. A strong link the
SOI
Mitas and Clement86 DJF winter cell NCEP/ERA40/NCEP2,
isobaric streamfunction
Trends in peak : 7.5/21.2/1.2 ×109kg s−1decade−1
for NCEP/ERA40/NCEP2
Radiosonde [extended Oort
and Yienger (1996)
series] Ref 87
Trend in peak :0.3×109kg s−1decade−1
GCM multimodel ensemble Modeled trends over the 20th century mostly range −10
to +4×109kg s−1decade−1, suggest weakening
Stachnik and
Schumacher80
HC Eight reanalysis ensemble NH: 4.0 ×109kg s−1decade−1
Ranges from 0.3 to 1.43 ×109kg s−1decade−1
SH: 0.7 ×109kg s−1decade−1
Ranges from −13.9 to +5.4 ×109kg s−1decade−1
Sohn and Park88 DJF HC NCEP, ERA40, and SSM/I Intensification of NH winter cell of ∼15–25% since 1979
DJF WC Intensification of Pacific WC of ∼25% since
1979/significant decadal variability, with decrease
from mid-20th century to 1979
Nguyen et al.21
(see also Figure 3)
HC Eight reanalyses NH: −0.5 to 12.7 ×109kg s−1decade−1
SH: −3.2to9.0×109kg s−1decade−1
Positive values equal expansion regardless of hemisphere. An ‘S’ after the investigators indicates the study examined the seasonality of the expansion.
HC, Hadley circulation; NH, Northern Hemisphere; SH, Southern Hemisphere; WC, Walker circulation.
FIGURE 3|Time series of the peak intensity (vertically-averaged
maximum ) of the Hadley cell in the eight reanalyses for the Northern
(top) and Southern (bottom) Hemispheres. (Reprinted with permission
from Ref 21. Copyright 2013 The American Meteorological Society).
Time series from ERA40 (post 2002) and 20CR (post 2008) are partially
extrapolated from statistical relations with the other reanalyses .
SOURCES OF OBSERVATIONAL
UNCERTAINTY
As seen in the previous section, considerable
uncertainty exists in the observations. While the
broad consensus of studies suggests that the tropics
are expanding, results from different datasets and
methodologies do not necessarily agree. In this section,
possible sources of this uncertainty are discussed. In
particular, the potential shortcomings of the reanalysis
products are addressed, followed by a description of
some of the limitations associated with the individual
metrics and methodologies.
Reanalysis Considerations
The majority of observational studies of tropical
expansion rely on one or more of the eight available
reanalysis data products, and this potentially has a
wide-ranging effect on the results. These products are a
hybrid of numerical weather prediction model output
and assimilated data from various observational
platforms. They are very useful for climate analysis as
they incorporate historical observations in a physically
©2013 John Wiley & Sons, Ltd.
WIREs Climate Change The expanding tropics
consistent manner.93 Generally, reanalyses adequately
reproduce many mean ‘dynamic’ quantities like wind
and temperature, albeit with some systematic error.94
Perhaps most importantly, their globally complete
spatial and temporal coverage provides information
in otherwise data-sparse regions.
However, potential shortcomings remain. A
lack of observations may not fully constrain the
reanalysis, and allow possibly unrealistic model
variability to dominate.95 The observational network,
particularly the satellite-based components, has
evolved considerably over the previous decades,
which may introduce inhomogeneities into the
reanalyses.90,96,97 Comparisons of radiosonde and
reanalysis-based estimates of tropical expansion using
the THF methodology62 indicate that inhomogeneities
in reanalysis products may be behind the differences
in estimates of tropical expansion.
Further, differences in the physics of the
underlying models used within the various reanalyses
can result in different representations of the
circulation. The strong HC intensification in the
ERA40 was demonstrated to be a result of
a (unrealistic) tropical mid-tropospheric cooling
trend.98 The NH winter cells in the NCEP and
ERA40 reanalyses show significant differences, driven
by the differing cloud amounts and radiation budgets
predicted by the two products.99 Newer reanalyses
are not free from these shortcomings; considerable
variability is present in the strength and width of the
HC within the different products.21,80
These issues conceivably account for the
differences observed in tropical expansion trend
estimates across reanalysis products. The differences
in performance are readily apparent in Figures 2 and 3,
where considerable scatter is noted in the position and
strength of the NH and SH Hadley cells. Simpler
metrics, like changes to the mean jet stream position,
are better-captured by the reanalyses and have less
scatter in the estimates. More complex metrics, like
tropopause height, or P-E, are more subject to the
vagaries of the reanalysis models and cover a broader
range, even within a given metric. The scatter in the
estimates is often larger in the SH, where fewer in situ
observations are available and the reanalyses are more
reliant on their internal dynamics and assimilated
satellite data.100 The available evidence, while not
conclusive at this time, indicates that these effects are
potentially quite significant and cannot be ignored.
Methodological Concerns
In addition to underlying problems with the
reanalyses, differences in the expansion rates between
the different methodologies and datasets are, at
least to some degree, a manifestation of the
differing physics represented by the different metrics.58
Further, individual datasets and/or methodologies
may introduce their own unique shortcomings. These
are described below.
Tropopause-based methods are the most widely
used and the most widely critiqued methodology
employed to date.57,58 Results using this technique
are broadly consistent across the range of studies,
the initial effort of Seidel and Randel55 being the
exception. The interpretation of the results from this
methodology is relatively straightforward, without
relying on too many assumptions, although seasonal
variability could confound the results. Identifying the
tropopause in the subtropics can be problematic in
the reanalyses, but the use of blended tropopause data
is promising.61 The reanalyses do broadly capture
the interannual variability of TTD, although the
possibility of inhomogeneities and/or other reanalysis
issues cannot be discounted.62 The use of appropriate
relative thresholds57,58 is likely important, but needs
further work to enhance their robustness. The use of a
fixed threshold may account for up to 20% of the total
trend due to global background tropopause trends.62
Satellite-based data are subject to many
uncertainties. In a study of the shifts of the midlatitude
storm track using ISCCP cloud cover data, correction
of data problems significantly reduced or even
eliminated the trends in the data.101 Similar issues
may affect estimates of tropical expansion that use
satellite-based metrics and potentially account for the
differences in cloudiness-based estimates of tropical
expansion, where both corrected71 and uncorrected70
data are used.
Satellite estimates using broadband OLR are
a particular concern. The observational data have a
strong trend, while modeling studies102,103 suggest
that net changes to broadband OLR should be
relatively small, on the order of 0.5 K or less. One
potential factor responsible for the trend—overlooked
in tropical expansion studies to date—is a change
in the equatorial crossing time (ECT) of the polar-
orbiting satellites.104 This alters the time of day
when the satellite views a scene and has been
shown to produce significant biases, particularly over
land.105–107 Figure 4 explores the possible impact on
tropical expansion calculations. The left panel shows
the uncorrected NOAA OLR data108; expansionary
trends of the 250 W m−2contours of 0.82 and −0.32◦
decade−1are seen in the NH and SH, respectively.
These values are similar to those previously reported
using the same dataset and analysis protocol.66 Using a
dataset that has a statistically based ECT correction107
©2013 John Wiley & Sons, Ltd.
Advanced Review wires.wiley.com/climatechange
FIGURE 4|Time-latitude profiles of annually averaged NOAA AVHRR OLR for uncorrected (left) and the equatorial crossing time (ECT)-corrected
data (Ref 107) for the period 1979–2008. Vertical dashed lines indicate the changes in the satellite. Contours are every 10 Wm−2, with the
250 W m−2contour colored in red. The trend along this contour is also indicated with a dashed line.
applied effectively removes any expansionary trends
in the data. While it cannot be absolutely stated
that the statistical correction algorithm has not
removed a valid long-term climate signal, these
results raise questions about the true magnitude of
OLR-based estimates of tropical expansion. Clearly,
more investigation into this ECT bias and its effects
on tropical expansion calculations is warranted.
Further, a physically based model for the
interpretation of the OLR in terms of the ‘edge
of the tropics’ is required, as opposed to a simple
climatological definition. The expected behavior of
OLR in response to a changing climate is complex.
While changes to overall broadband OLR were small,
changes in specific spectral regions could be quite
large.103 Changes to the surface, tropospheric and
stratospheric temperatures all play a role, as does the
variability in water vapor, clouds, and aerosol. These
effects vary in different portions of the spectrum,
with sources often making opposite contributions.
Consideration of these effects and the examination of
specific spectral regions could lead to more reliable
estimates of tropical expansion, and explain any
differences in trends between hemispheres.
At first glance, the use of to identify the HC
and the edge of the tropics would appear ideal. It
provides an unambiguous, quantitatively defined defi-
nition of the tropical edge. In practice, these definitions
are sensitive to the differences in the reanalyses, which
show significant variations in the strength, width, and
position of the HC (Figures 2 and 3). Further, the
HC shows a larger degree of seasonality in the trends
than other metrics, particularly in the NH. The HC
during summer and early autumn is generally weak.21
The meridional gradient about the zero-contour is
‘flat’ and as a consequence the identified position of
the tropical edge is sensitive to small perturbations.
These seasons are when the tropical expansion trends
with this metric are the largest.21,58,65 Conceivably,
even small errors/uncertainties in the reanalyses could
overwhelm the comparatively weak signals at this
time of year and result in an artificial trend. Temporal
inhomogeneities may also be important here, as
significant changes were made to the global observing
system in the early 21st century,51,54 particularly the
addition of new satellite data sources. Conceivably,
these could be behind the SH shift noted in Figure 2.
Further, the seasonal cycle of width defined from is
different than those defined from the jet streams and
θ stability parameter, with different amplitude and
timing.64 More investigation of these effects and their
impact on the results is required.
The jet stream methods from reanalysis data58,81
and radiosonde data71 appear to produce reasonably
robust results. These results indicate very small (and
often not statistically significant) amounts of expan-
sion. These studies tend to be based on the annual or
seasonal mean winds; care must be taken to identify
the SH STJ during austral summer.64 Studies based on
maximum winds produce noisier results.58 Jet streams
objectively identified on daily time scales suggest
a more complex behavior of these phenomena,109
with more than one jet often apparent and a distinct
annual cycle in the position and strength of the jet in
both hemispheres. How well the mean quantities used
in tropical expansion studies represent this complex
behavior and what its impact may be are unknown.
The suitability of the other methodologies that
have been used to assess the rate of tropical expansion
is unclear. In many cases, there are simply not enough
studies to support (or refute) their robustness. The
source of the disparity between results using the
©2013 John Wiley & Sons, Ltd.
WIREs Climate Change The expanding tropics
GPCP subtropical minimum precipitation is unclear,
and may reflect the use of different versions of
the GPCP dataset (version 2.1 vs 2.2). If this is
the case, it suggests a strong sensitivity to the
input data. Further, the use of absolute thresholds70
with this data needs to be carefully considered;
as with the OLR and THF methods, the chosen
threshold should be physically linked with the tropical
edge, beyond just a climatological coincidence. Other
metrics derived from reanalyses (like P-E) are subject
to reanalysis uncertainties, particularly in the data-
poor SH and should be treated with caution until
better understood. The effects of seasonal variability
also need consideration with these methods.
MECHANISMS AND FORCINGS
BEHIND TROPICAL EXPANSION
In this section, the influences on the position of
the tropical edge are examined. Tropical expansion
and changes to the HC have been examined in
numerous GCMs using paleoclimate, 20th century,
and climate change scenarios. From these simulations,
physical mechanisms have been proposed. Simulations
from simplified GCMs with idealized forcing are
instructive. Natural climate forcing factors, as seen
through observational studies, can have a significant
effect. Modeling also suggests a significant role for
anthropogenic influences. Experiments examining the
role of individual forcings are described; these climate
forcings include greenhouse gases, stratospheric ozone
depletion, and aerosols.
Tropical Expansion in CMIP3 and Other
Simulations
The GCMs used in the World Climate Research
Programme (WCRP) Coupled Model Intercompar-
ison Project phase 3 (CMIP3) multimodel dataset
project110 show significant changes to the tropical cir-
culation during the 21st century as a result of increas-
ing greenhouse gas concentrations (generally either the
A1B or A2 scenarios), a poleward expansion of the HC
of 1o–2oover the 21st century along with a decrease
in the intensity of the HC and WC.30,67,111,112 Closely
associated to this are a shift in the subtropical dry zone
and changes to the hydrologic cycle, with wet regions
becoming wetter and dry regions drier.113 Much of
this change is hypothesized to be driven by the ther-
modynamic effects of increasing specific humidity,
although dynamic changes also allow the poleward
flanks of the HC to dry.114 A recent study115 suggest
that the poleward shifts of the MMC are more respon-
sible for subtropical drying than the thermodynamic
effect. On an annual basis, these projected changes are
largely symmetric about the equator.56
While CMIP3 simulations generally show an
expansion of the tropics, uncertainties remain. The
strength of the modeled HC across the suite of
models varies by a factor of 2.112 Other GCMs
outside of the CMIP3 framework are more equivocal
in their response. Several studies46,116 investigated
the HC in simulations of paleo, present day, and
future climate simulations, covering a wide ‘parameter
space’ in overall temperature, surface characteristics,
topography, and solar forcing. Despite the range,
changes in the characteristics of the HC were noted
between these different eras were generally small.
These results highlight the model dependency in
the results, owing (at least in part) to the various
formulations, parameterizations, and assumptions
made within models. One of the studies116 reports
that the formulation of the ocean model plays a strong
role in determining the final characteristics of the HC.
The climate sensitivity of a given model, related to the
strength of the feedbacks within the model, can also
have an impact on future trends in the circulation.117
As noted in the idealized model of the MMC, a
strong interaction exists between the tropics and extra-
tropics. Hence, it seems plausible that variability in the
extratropical circulation has an effect on the tropical
edge (or vice versa). This is consistent with results36
that identify an interannual relationship between
the HC edge and the latitude of the eddy-driven
jet in the SH during summer. Similar relationships
between the HC extent and the annular modes in both
hemispheres have been identified in most reanalysis
products, although again not in JJA.21 The CMIP3
simulations suggest that tropical expansion is to some
degree a consequence of circulation changes in the
extratropical portion of the MMC, rather than a
driver of those changes. Up to half of the projected
change to the subtropical dry zones can be associated
with changes to the annular mode.118 Midlatitude
dynamics are likely important to the expansion.119
Hadley cell scaling arguments indicate that the extra-
tropical tropopause heights are a better predictor of
modeled tropical expansion than tropical tropopause
heights.56 Overall, the role of tropical–extratropical
interactions in tropical expansion remains uncertain,
but it appears to be significant.
Proposed Mechanisms of Tropical
Expansion
Two potential dynamical mechanisms for tropical
expansion have been proposed.29 Both mechanisms
require tropical–extratropical interactions and invoke
©2013 John Wiley & Sons, Ltd.
Advanced Review wires.wiley.com/climatechange
a poleward shift in the position of extratropical jet
stream. The first mechanism involves a reduction
in baroclinicity in the subtropics, which dampens
eddy activity there and extends the HC further
poleward. This reduction is a result of an increase
in the (dry) static stability, related to a quasi-moist
adiabatic adjustment to increasing specific humidity.
A recent study120 suggests that changing wind shear
characteristics may be more important than static
stability changes.
The second mechanism relates to an increase
in the phase speed of upper-tropospheric baroclinic
waves. The faster moving waves are unable to
penetrate as far equatorward, producing a poleward
shift in eddy momentum flux convergence and an
associated shift in the position of the eddy-driven
jet.121,122 This may be related to a rising tropopause
in the subtropics.123 An examination124 of this
mechanism in greater detail suggests that a transition
in the wave type (equatorward- vs poleward-breaking
waves40) may also play a role. Changes to the eddy
length scale may be a factor,125 perhaps in response
to an increase in upper-tropospheric baroclinicity in
the midlatitudes.126
Idealized Simulations
Experiments with simplified GCMs and idealized
heating profiles have been used to explore tropical
expansion and the shift in the eddy-driven jet stream.
Using idealized heating profiles reminiscent of those
expected from global warming scenarios, Butler
et al.127 found that both tropical upper-tropospheric
heating and stratospheric cooling resulted in a
poleward shift in midlatitude jet and an expansion of
the HC. Although qualitatively similar in terms of the
jet response, the resulting circulations due to forcings
from the two regions were different. Allen et al.128
performed a suite of simulations using a range of
idealized heating profiles, both stratospheric cooling
and tropospheric heating. In their experiments, heat-
ing in the deep tropics resulted in little change in jet
stream and HC, while midlatitude heating produced a
poleward jet shift and expansion of the HC. Tandon
et al.120 examined the circulation response to a series
of tropospheric heating profiles of varying widths.
Heating near the equator produced a contraction
of the tropics and poleward shift in the jet stream.
Subtropical heating produced the opposite responses.
While the model simulations provide a degree
of insight into the dynamical mechanisms of tropical
expansion, some uncertainty remains. In the idealized
runs, the results are sensitive to the choice of
initial conditions and the details of the forcing used.
However, the meridional extent of the forcing appears
to be particularly important factor in determining the
response of the HC in these idealized experiments.
Natural Variability
The position of the tropical edge varies on interannual
(and longer) time scales. A primary source of climate
variability is ENSO. Both the width and intensity of
the HC varies with the phase of ENSO.86 During
the warm El Ni ˜
no phase, the HC is narrower
and more intense; during La Ni ˜
na, the opposite is
observed. This general tendency is also suggested in
GCM simulations,29 which also show an equatorial
shift of the STJ during El Ni ˜
no. The various
reanalysis products indicate an intensification80 and
contraction21 of the HC during El Ni ˜
no compared
with Neutral and La Ni ˜
na phases in reanalyses.
Using THF metrics, the expansion and variability
were more pronounced in the extratropical portions
of the subtropics, less so toward the deep tropics, with
changes in width from La Ni ˜
na to El Ni ˜
no in the order
of 1◦–2◦latitude.62 The Pacific Decadal Oscillation
(PDO) has also been hypothesized to impact the width
of the tropics. During cool phases of the PDO, the
HC is ∼1◦latitude further poleward compared with
its warm phase position. This is primarily observed
during the equinox seasons.129
Globally significant volcanic eruptions can also
have an impact on the tropics. Eruptions can increase
stratospheric aerosol, warming the lower stratosphere
and lowering tropopause height.130 Tropical precip-
itation can be lowered by ∼5%131 and particularly
large eruptions may act as a trigger for El Ni ˜
no.132
A signal of tropical contraction following large
eruptions is apparent, particularly in studies using
tropopause-based methodologies. The magnitude of
this effect is estimated to be ∼1◦latitude.62
Greenhouse Gases and SST
The relative roles of radiative forcing and SST
changes on tropical expansion from 1958 to 2000
have been examined using the THF methodology.56
Expansion trends were accurately replicated by the
model, with more pronounced trends in the SH. The
expansion was attributed to changes in atmospheric
radiative forcing; indirect SST changes had little effect.
Stratospheric cooling was found to be particularly
important in producing the observed trends, although
no distinction between the effects of greenhouse
gases and stratospheric ozone depletion could be
made due to the design of the experiment. Seasonal
changes to SST do not appear to account for tropical
expansion in CMIP3 simulations, suggesting that
©2013 John Wiley & Sons, Ltd.
WIREs Climate Change The expanding tropics
extratropical eddies have induced the expansion.119
Deser and Phillips133 found that the direct effects of
atmospheric radiative forcing led to a strengthening
and intensification of midlatitude westerly winds in
the SH over the latter half of the 20th century. In
the NH, SST changes, particularly over the tropics,
were responsible for the intensification of the Aleutian
low and a weakened WC. While the HC was not
specifically examined, an expansionary trend in the
SH is implied by the poleward shift of the westerlies.
Other studies indicate that as a whole the indirect SST
change is the main driver of circulation change from
both the preindustrial (1870) to the present period
and extending into the future, although factors like
ozone depletion may have a significant, but temporary
impact on the hemispheric circulation.134 Hu et al.66
also emphasized that historical SST changes played
a significant, but incomplete role in reproducing the
observed expansion.
Stratospheric Ozone Depletion
Numerous studies have suggested that stratospheric
ozone depletion plays a role in tropical expansion
of the SH. For the 20th century, the GCMs in
CMIP3 with ozone depletion consistently showed
greater tropical expansion through a variety of metrics
than those without.135 In future scenarios, GCMs
with ozone recovery showed on average near zero
trends in the tropopause, jet location, HC, and SAM
index, in contrast to runs without ozone recovery.
These results were only qualitative, as details of
the ozone forcing were unavailable. Using seven
reanalyses and a multimodel GCM dataset, optimal
fingerprinting techniques identified a separable signal
of stratospheric ozone depletion in SH tropical
expansion during DJF, but not in other seasons.136
A multimodel ensemble of 17 climate-chemistry
models (CCM) has been used to understand the
response of the SH circulation to the effects of
explicitly simulated stratospheric ozone depletion.137
This study found that ozone depletion in the late-
spring led to a poleward shift and intensification of
the SH tropospheric jet and an expansion of the
HC. These changes to the HC were primarily found
during the SH summer. Similar results were obtained
using a coupled CCM,124 suggesting that the observed
changes in the 20th century were better explained by
stratospheric ozone changes compared to changing
GHG concentrations.
Model simulations using separate ozone and
GHG forcings have also suggested that stratospheric
ozone depletion was the main driver of 20th century
SH climate change. In particular, ozone depletion
resulted in a widening of the HC and an expansion
of the SH subtropical dry zones.138 In the first half of
the 21st century, ozone recovery may approximately
cancel the effects of GHG on the MMC, resulting
in a reduced rate of expansion.139 Other models
have suggested that the distinct effects of polar
stratospheric ozone depletion could extend well into
the SH subtropics, resulting in a precipitation increase
at those latitudes.140
Aerosol
Aerosol forcing of the climate is complex, but quite
significant. They can directly affect the radiative
flux through scattering (primarily sulfate aerosol)
or absorption (e.g., black carbon), contributing to
surface cooling and/or atmospheric solar heating.141
Semi-direct and indirect aerosol effects may also
impact the climate.142 The semi-direct effect arises
when the heating associated with absorbing aerosols
changes relative humidity, impacting the lifetime
of clouds. The indirect effects of aerosol involve
their interactions with cloud properties, including
increasing their albedo and lifetime and alteration of
their precipitation characteristics. Modeling studies143
suggest that these indirect effects can be approximately
of equal magnitude (but oppositely signed) to the
forcing by GHG. Further complicating the role of
aerosol is their observed regional distribution and
relatively short lifetime.144
Earlier studies have examined aerosol effects
on aspects of the tropical circulation, particularly on
regional scales. The direct effects of increasing black
carbon may intensify the Asian summer monsoon,
while increases of only scattering aerosol weakens the
circulation.145 The indirect aerosol effect may drive
a southward shift in the ITCZ, potentially associated
with past Sahelian drought.146 While aerosol loadings
are regional in scope, they can have circulation impacts
at remote regions around the globe.147,148
Aerosol impacts on the broader MMC have
also been reported. The direct effects of black carbon
included a northward displacement of the ITCZ
and a strengthened (weakened) HC in the Northern
(Southern) Hemisphere.148 Anthropogenic aerosols
(largely absorbing) weaken the HC and shift the peak
streamfunction northward. Natural aerosols (mostly
reflecting) had the opposite effects. Opposing shifts
in the STJ were also noted, with anthropogenic
aerosols shifting the jet stream poleward in both
hemispheres.149 The observed shift in the NH tropics
has been attributed to midlatitude heating by black
carbon aerosol and tropospheric ozone. The modeled
response was notably less than that observed in
reality.71 Model simulations that included smoke
©2013 John Wiley & Sons, Ltd.
Advanced Review wires.wiley.com/climatechange
aerosol from landscape fires showed a widening and
weakening of the HC.150
The response of the tropical circulation to
both direct and indirect aerosol effects has been
modeled.151 Radiative cooling from aerosol is most
prominent in the NH, creating an interhemispheric
radiative imbalance. The atmosphere attempts to
moderate this asymmetry by altering the zonal-mean
circulation in the tropics, resulting in a weaker
(stronger) HC in the Northern (Southern) Hemi-
sphere. The asymmetrical portions of the circulation
(i.e., the WC) strengthen in both hemispheres. In
the zonal-mean, the subtropics (defined here as
20◦–30◦latitude) of both hemisphere become wetter.
However, the authors conclude that the atmospheric
responses may be overestimated because of a lack of
an interactive ocean model in these simulations.
SYNTHESIS
A variety of methodologies have been used to examine
tropical expansion. Most studies begin at the start of
the satellite era (1979) and extend into the first decade
of the 21st century. Figure 5 summarizes the rates
of expansion estimated by these studies, stratified
into the methodological categories discussed earlier.
The numerous studies indicate a robust poleward
expansion of the tropical edges of both hemispheres
over the last ∼30 years. Almost all studies suggest that
the rate of expansion in each hemisphere has been less
than 1◦decade−1, with many suggesting rates around
half of that value. Many of these widening trends
may not be statistically significant at the 95% level.58
Tropopause-based studies generally suggest a greater
widening in the SH; satellite-based studies indicate
the opposite. Studies using suggest a greater
seasonality and an (highly uncertain) intensification
of the HC. As a whole, jet stream metrics suggest a
generally smaller rate of expansion. Taken together, a
total expansion of 1◦–3◦latitude in each hemisphere
is implied since 1979.
The seasonality of the changes, if any, is unclear.
Some metrics, notably the -based findings, show
strong seasonality with the largest changes in each
hemisphere’s respective summer and autumn.21,65
Jet stream metrics based on the position of the
peak zonal-mean 850 hPa zonal wind show notable
seasonal differences in the SH.58 Seasonal variability
is suggested in other metrics,58,70 but many of these
differences are not statistically significant.58 As noted
earlier, some of the metrics, particularly during the
warm season, may not produce robust results.
Modeling studies suggest that multiple climate
forcings can result in an expansion of the
FIGURE 5|Summary of result of observational tropical expansion
studies, broken down by categories. Pluses (+) represent SH values,
crosses (×) NH values of tropical expansion trend. Symbols in the
corresponding hemisphere indicates expansion, symbols in the
‘opposite’ hemisphere indicate contraction. Within a particular
methodology, the vertical position of each symbol is randomized to
improve clarity of individual symbols; no other meaning is implied.
Colors of symbols refer to source and/or methodology, with legends for
each category in boxes to side of plot, where numbers represent the
cited reference.
tropics, including increasing GHG concentrations and
associated changes to the SST, stratospheric ozone
depletion (particularly in the SH) and the effects
of aerosols, both direct and indirect. Each of these
forcings results in subtly different impacts on the
tropical edge, as summarized in Table 8. A comparison
of these projections with the relevant climatological
variables can provide a qualitative assessment of the
relative importance of these factors.
Tropical expansion is noted in both the NH
and the SH, suggesting some degree of symmetry in
the response. This suggests that the observed increase
in specific humidity (water vapor),152–154 associated
with the increase in GHG and global temperature, is
playing a role in tropical expansion. However, the
observations of HC intensity, while highly uncertain,
suggest no trend or a slight intensification, which
is inconsistent with this response. Some studies111,151
suggest that this thermodynamic effect may impact the
zonal WC more than the HC. Most studies, but not all,
report that the WC has weakened in recent decades,
although internal climate variability may account for
a significant fraction of the observed weakening.155
During the 20th century, zonal-mean precip-
itation has increased in the SH tropics and sub-
tropics, while decreasing in the NH.156 Coupled
climate models indicate that these changes are a result
of anthropogenic radiative forcing, with GHG and
sulfate aerosol forcing specifically noted. Kang et al.140
also report a similar tendency in the subtropical
©2013 John Wiley & Sons, Ltd.
WIREs Climate Change The expanding tropics
TABLE 8 Summary of Tropical Expansion Projections Following Different Forcings
Forcing Citations Projected Effects
A2, A1B (many contain ozone, but
generally O3 recovery when
evaluated)
Lu et al.,56 Seager et al.114 Expansion rate: 0.1–0.2◦decade−1
Symmetric about equator
Weakening and expansion of HC
Reduction of precipitation in subtropics
Enhancement of rain in ITCZ/monsoon
Poleward shift in jet
Stratospheric ozone depletion Son et al.,135 Son et al.,137 Polvani
et al.,139 Kang et al.,140 Staten et al.,134
Min and Son136
Greater impact in SH spring and summer
SH subtropical precipitation increase
HC widening and poleward shift in SH
Direct aerosol (sulfate) Wang,148 Allen et al.71 Strengthened HC
Equatorward shift in jet stream
Direct aerosol (black carbon) Wang,148 Allen et al.71 Weakening of HC
Northward shift in ITCZ
Poleward shift of jet stream
Indirect aerosol Rostayn and Lohmann,146 Rotstayn
et al.,147 Ming and Ramaswamy151
Asymmetric response—intensification of SH
HC, weakening in NH
Precipitation increase in subtropics
Southward shift in ITCZ
Sea surface temperature Hu et al., 2011,65 Deser and Phillips,133
Staten et al.134
Poleward shift in jet stream
precipitation in the SH, but attribute the increase
to stratospheric ozone depletion. As noted in earlier, a
considerable amount of modeling evidence exists for a
significant role of stratospheric ozone depletion in pro-
ducing SH tropical expansion, although these changes
are largely confined to the SH spring and summer
seasons. Some degree of SH asymmetry in expansion
rates is observed in tropopause-based metrics, but not
in others (Figure 5).
Aerosols, particularly from volcanoes, clearly
play a role in tropical expansion on interannual
time scales, but longer term effects are unlikely.
The asymmetric expansion (in some results) and
precipitation response is reminiscent of the indirect
aerosol effect,143 a forcing that has not been included
in many climate simulations to date, including the
CMIP3 archive. However, the evidence for this factor
is unclear; little change has been observed in either a
proxy for the global ITCZ21 or in the satellite-detected
eastern Pacific ITCZ157 and the apparent HC
intensification has occurred in both hemispheres, more
in the NH. The role of direct aerosol effects is also
incomplete; while Allen et al.71 explicitly attribute NH
expansion to black carbon aerosol and tropospheric
ozone, the projected response is considerably less that
the observed increase.
The dynamical mechanisms behind tropical
expansion are unclear, but many hypotheses proposed
to date29,122,124 indicate that the extratropics and
baroclinic eddies play a significant role. One
manifestation of this is a poleward shift in the
storm track, closely associated with the position of
the eddy-driven jet stream. Yin158 notes this as a
response to climate change in simulations of the 21st
century. Bender et al.101 identified such a shift in
recent decades, albeit with some reservations due
to data issues. Overall, a shift in the storm track
results in less precipitation over subtropical regions,
with this mechanism possibly more important than
the thermodynamic effect.115 Other modeling studies
have indicated a close linkage of the jet position
and the HC.135,137 Relationships between the HC
edge (identified using the methodology) and the jet
stream position have been identified for some seasons,
particularly in the SH.21,36 A poleward shift in the
jet stream is also closely related to poleward shift in
the annular mode. Up to half of the projected change
to the subtropical dry zones may be associated with
changes to the annular mode.118
To summarize, no clear consensus exists
regarding the primary forcing mechanism behind
tropical expansion. Any singular forcing generally
results in an under-prediction of the observed
amount of tropical expansion.79 This may reflect
a shortcoming of the modeling approaches used
to date; for example, the models simply are not
correct or sensitive enough to reproduce the observed
changes. Certainly, given the varied response of the
©2013 John Wiley & Sons, Ltd.
Advanced Review wires.wiley.com/climatechange
different models in play, this suggestion cannot be
dismissed out-of-hand. However, it may be that
tropical expansion simply does not have a single cause.
Rather, it may arise from a variety of causes that have
acted in concert to produce the observed response,
with different factors acting in the NH and SH. Other
factors not thoroughly considered to date by modeling
may also play a role. The atmosphere is complex, and
an attempt to attribute the whole change to a single
forcing may not be appropriate.
OUTSTANDING ISSUES
As summarized in this review, considerable efforts
have been made within the climate community
to understand the causes and effects of tropical
expansion. From this research a coherent picture is
emerging, but many gaps remain in our knowledge.
These are briefly summarized below.
A key issue is the determination of the true rate
of tropical expansion. A broad consensus from current
studies suggests the value is below 1◦decade−1, but
which hemisphere is expanding at a faster rate (if
either) remains unclear. Also unclear is the seasonality
of the expansion, with significant differences identified
between the various methodologies. Developing a
clearer picture requires improving our understanding
and interpretation of current methodologies used to
estimate tropical expansion, as well as the formulation
of new ones. Clarifying the linkages between different
metrics is crucial. For example, do THF methods tell
us the same thing as methods? Addressing data
quality issues through homogenized datasets and/or
reanalysis products is paramount.
One particular shortcoming in our understand-
ing is the role of tropical–extratropical interaction
in the MMC. To what degree does tropical expan-
sion arise from changes in the extratropical circu-
lation? Several studies discussed herein suggest that
the changes to the extratropics may be the source
of the tropical expansion. Traditionally in the atmo-
spheric sciences, the tropics, and extratropics have
been treated as separate entities. To overcome this,
a new conceptual framework of understanding for
the hemispheric (and global) MMC is required. The
idealized model shown earlier provides a qualitative
description of such a framework, a useful starting
point.
Another significant issue is the understanding of
the climate factors responsible for tropical expansion.
This includes not only identifying the most significant
climate forcings, but also the dynamical mechanisms
by which they act. This requires the ongoing efforts
of model development and evaluation, as well as
observational studies to propose new hypotheses and
verify the results. A sharper focus on the seasonality
and/or longitudinal variability of the expansion
could improve the understanding of the hypothesized
forcings. The current understanding of many potential
forcings (e.g., the role of aerosol in the climate)
remains uncertain. It is helpful to understand the trop-
ical expansion observed to date, which can then lead
to a better interpretation of projected climate futures.
Finally, a better understanding of the impacts
of tropical expansion on the climate and society is
required. Much of the research to date has focused
on the global or zonal-mean responses to expansion,
e.g., the subtropical dry zone will expand. Some
studies have identified this as a factor in recently
observed SH rainfall declines.159 However, questions
remain about what this means at a given location
in the subtropics. For instance, GCM simulations
often indicate a significant regional variability in the
response. The reality of the MMC is more complicated
than that presented in a simple zonal-mean picture.
For purposes of planning and climate adaptation, a
better understanding of regional responses to tropical
expansion is required.
ACKNOWLEDGMENTS
This research was funded as part of the South Eastern Australia Climate initiative Phase 2 (SEACI 2).
Interpolated OLR data were provided by the NOAA/OAR/ESRL PSD, Boulder, CO, USA, from their Web site
http://www.esrl.noaa.gov/psd/. The ECT-corrected version of the same dataset was provided by Matt Sapiano.
Tim Cowan, Pandora Hope, and two anonymous reviewers provided helpful comments on the manuscript.
REFERENCES
1. CSIRO (Commonwealth Scientific and Indus-
trial Research Organisation). Climate and Water
Availability in South-eastern Australia: A Synthesis
of Findings from Phase 2 of the South Eastern
Australian Climate Initiative (SEACI). Australia:
CSIRO; 2012, 41 pp.
©2013 John Wiley & Sons, Ltd.
WIREs Climate Change The expanding tropics
2. Indian Ocean Climate Initiative. Western Australia’s
Weather and Climate: A Synthesis of Indian Ocean
Climate Initiative Stage 3 Research. Australia: CSIRO
and BoM; 2012.
3. Hoerling M, Eisheid J, Perlwitz J, Quan X,
Zhang T, Pegion P. On the increased frequency of
Mediterranean drought. J Clim 2012, 25:2146–2161.
doi: 10.1175/JCLI-D-11-00296.1.
4. Sousa PM, Trigo RM, Aizpurua P, Nieto R,
Gimeno L, Garcia-Herrera R. Trends and extremes
of drought indices throughout the 20th century in
the Mediterranean. Nat Hazards Earth Syst Sci 2011,
11:33–51.
5. Cayan DR, Das T, Pierce DW, Barnett TP, Tyree M,
Gershunov A. Future dryness in the southwest US and
the hydrology of the early 21st century drought. Proc
Natl Acad Sci U S A 2010, 107:21271–21276. doi:
10.1073/pnas.0912391107.
6. Morales MS, Christie DA, Villalba R, Argollo J,
Pacajes J, Silva JS, Alvarez CA, Llancbure JC, Soliz
Gamboa CC. Precipitation changes in the South
American Altiplano since 1300 AD reconstructed by
tree-rings. Clim Past 2012, 8:653–666. doi: 10.5194/
cp-8-653-2012.
7. Ye J-S. Trend and variability of China’s summer
precipitation during 1955–2008. Int J Climatol 2013.
doi: 10.1002/joc.3705.
8. Dai A. Drought under global warming: a review.
WIREs Clim Change 2011, 2:45–65. doi: 10.1002/
wcc.81.
9. Dai A. Increasing drought under global warming in
observations and models. Nat Clim Change 2012, 3:
52–58. doi: 10.1038/nclimate1633.
10. Mueller B, Seneviratne S. Hot days induced by pre-
cipitation deficits at the global scale. Proc Natl Acad
Sci U S A 2012, 109:12398–12403. doi: 10.1073/pnas.
1204330109.
11. Mishra AK, Singh VP. A review of drought concepts.
JHydrol2010, 391:202–216. doi: 10.1016/j.hydrol.
2010.07.012.
12. Kenyon J, Hegerl GC. Influence of modes of climate
variability on global precipitation extremes. J Clim
2010, 23:6248–6262. doi: 10.1175/2010JCLI3617.1.
13. Seidel DJ, Fu Q, Randel WJ, Reichler TJ. Widening
of the tropical belt in a changing climate. Nat Geosci
2008, 1:21–24. doi: 10.1038/ngeo.2007.38.
14. Reichler T. Changes in the atmospheric circulation as
an indicator of climate change. In: Letcher TM, ed.
Climate Change: Observed Impacts on Planet Earth.
Amsterdam: Elseveier; 2009, 145–164.
15. Townsend RD, Johnson DR. A diagnostic study of
isentropic zonally averaged mass circulation during
the First GARP Global Experiment. J Atmos Sci 1985,
42:1565–1579. doi: 10.1175/1520-0469(1985)042
<1565:ADSOTI>2.0.CO;2.
16. Schneider T. The general circulation of the atmo-
sphere. Annu Rev Earth Planet Sci 2006, 34:655–688.
doi: 10.1146/annurev.earth.34.031405.125144.
17. Pauluis OA, Czaja A, Korty R. The global atmospheric
circulation on moist isentropes. Science 2008,
321:1075–1078. doi: 10.1126/science.1159649.
18. Pauluis OA, Czaja A, Korty R. The global atmo-
spheric circulation in moist isentropic coordinates. J
Clim 2010, 23:3077–3093. doi: 10.1175/2009JCLI
12789.1.
19. Fierro AO, Simpson J, LeMone MA, Straka JM, Smull
BF. On how hot towers fuel the Hadley cell: an obser-
vational and modelling study of line-organized convec-
tion in the equatorial trough from TOGA-COARE.
J Atmos Sci 2009, 66:2730–2746. doi: 10.1175/
2009JAS3017.1.
20. Webster PJ. The elementary Hadley circulation. In:
Diaz HF, Bradley RS, eds. The Hadley Circulation:
Past, Present and Future. Dordrecht: Kluwer Academic
Publishers; 2005, 9–60.
21. Nguyen H, Evans A, Lucas C, Smith I, Timbal B. The
Hadley circulation in reanalyses: climatology, variabil-
ity and expansion. J Clim 2013, 26:3357–3376. doi:
10.1175/JCLI-D-12-00224.
22. Schneider EK, Lindzen RS. Axially symmetric steady-
state models of the basic state for instability and
climate studies. Part I: linearized calculations. J
Atmos Sci 1977, 34:263–279. doi: 10.1175/1520-
0469(1977)034<0263:ASSSMO>2.0.CO;2.
23. Held IM, Hou AY. Nonlinear axially symmetric
circulations in a nearly inviscid atmosphere. J
Atmos Sci 1980, 37:515–533. doi: 10.1175/1520-
0469(1980)037<0515:NASCIA>2.0.CO;2.
24. Frierson DMW, Lu J, Chen G. Width of Hadley
cell in simple and comprehensive general circulation
models. Geophys Res Lett 2007, 34:L18804. doi:
10.1029/2007GL031115.
25. Rind D, Rossow WB. The effects of physical processes
on the Hadley circulation. J Atmos Sci 1984,
41:479–507. doi: 10.1175/1520-0469(1984)041
<0479:TEOPPO>2.0.CO;2.
26. Walker C, Schneider T. Eddy influences on
Hadley circulations: Simulations with an idealized
GCM. J Atmos Sci 2006, 63:3333–3350. doi:
10.1175/JAS3821.1.
27. Palm´
en E, Newton CW. Atmospheric Circulation
Systems. New York: Academic Press; 1969, 603.
28. Held IM. The General Circulation of the Atmo-
sphere. Presented at the 2000 Woods Hole Oceano-
graphic Institute Geophysical Fluid Dynamics Pro-
gram. Woods Hole, MA: Woods Hole Oceanogr. Inst.;
2000. Available at: http://www.whoi.edu/fileserver.
do?id=21464&pt=10&p=17332. (Accessed Septem-
ber 9 2013)
29. Lu J, Chen G, Frierson DMW. Response of the
zonal mean atmospheric circulation to El Ni ˜
no versus
©2013 John Wiley & Sons, Ltd.
Advanced Review wires.wiley.com/climatechange
global warming. J Clim 2008, 21:5835–5851. doi:
10.1175/2008JCLI2200.1.
30. Lu J, Vecchi GA, Reichler T. Expansion of the Hadley
cell under global warming. Geophys Res Lett 2007,
34:L06805. doi: 10.1029/2006GL028443.
31. Kang SM, Lu J. Expansion of the Hadley Cell under
global warming: winter versus summer. J Clim 2012,
25:8387–8393. doi: 10.1175/JCLI-D-12-00323.1.
32. Limpasuvan V, Hartmann DL. Wave-maintained
annular modes of climate variability. J Clim
2000, 13:4414–4429. doi: 10.1175/1520-0442(2000)
013<4414:WMAMOC>2.0.CO;2.
33. Karoly DJ, McIntosh PC, Berrisford P, McDougall
TJ, Hirst AC. Similarities of the Deacon cell in
the Southern Ocean and the Ferrel cells in the
atmosphere. Q J R Meteorol Soc 1997, 123:519–526.
doi: 10.1002/qj.49712353813.
34. Galewsky J, Sobel A, Held I. Diagnosis of
subtropical humidity dynamics using tracers of last
saturation. J Atmos Sci 2005, 62:3353–3367. doi:
10.1175/JAS3533.1.
35. Bordoni S, Schneider T. Monsoons as eddy-mediated
regime circulations of the tropical overturning circu-
lation. Nat Geosci 2008, 1:515–519. doi: 10.1038/
ngeo248.
36. Kang SM, Polvani LM. The interannual relationship
between the latitude on the eddy-driven jet and the
edge of the Hadley Cell. J Clim 2011, 24:563–568.
doi: 10.1175/2010JCLI4077.1.
37. Ding Q, Steig EJ, Battisti DS, Wallace JM. Influence
of the tropics on the Southern Annular Mode. J
Clim 2012, 25:6330–6348. doi: 10.1175/JCLI-D-11-
00523.1.
38. Moore RW, Martius O, Spengler T. The modulation
of the subtropical and extratropical atmosphere in
the Pacific basin in response to the Madden-Julian
Oscillation. Mon Weather Rev 2010, 138:2761–2779.
doi: 10.1175/2010MWR3194.1.
39. Solomon S. Stratospheric ozone depletion: a review of
concepts and history. Rev Geophys 1999, 37:275–316.
doi: 10.1029/1999RG900008.
40. Wittman MAH, Charlton AJ, Polvani LM. The
effect of lower stratospheric shear on baroclinic
instability. J Atmos Sci 2007, 64:479–496. doi:
10.1175/JAS3828.1.
41. Solomon S, Rosenlof KH, Portman RW, Daniel JS,
Davis SM, Sanford TJ, Plattner G-K. Contributions of
stratospheric water vapour to decadal changes in the
rate of global warming. Science 2010, 327:1219–1222.
doi: 10.1126/science.1182488.
42. Maycock AC, Joshi MM, Shine KP, Scaife AA.
The circulation response to idealized changes in
stratospheric water vapour. J Clim 2012, 26:545–561.
doi: 10.1175/JCLI-D-12-00155.1.
43. Baldwin MP, Dunkerton TJ. Stratospheric harbingers
of anomalous weather regimes. Science 2001,
294:581–584. doi: 10.1126/science.1063315.
44. Cook KH. Role of continents in driving the Hadley
cells. J Atmos Sci 2003, 60:957–976. doi: 10.1175/
1520-0469(2003)060<0957:ROCIDT>2.0.CO;2.
45. Becker E, Schmitz G. Interaction between extratropical
stationary waves and the zonal mean circulation.
J Atmos Sci 2001, 58:462–480. doi: 10.1175/1520-
0469(2001)058<0462:IBESWA>2.0.CO;2.
46. Rind D, Perlwitz J. The response of the Hadley
circulation to climate changes, past and future. In:
Diaz HF, Bradley RS, eds. The Hadley Circulation:
Past, Present and Future. Dordrecht: Kluwer Academic
Publishers; 2004, 399–436.
47. Kalnay E, Kanamitsu M, Kistler R, Collins W,
Deaven D, Gandin L, Iredell M, Saha S, White G,
Woollen J, et al. The NMC/NCAR 40-year reanalysis
project. Bull Am Meteorol Soc 1996, 77:437–471.
doi: 10.1175/1520-0477(1996)077<0437:TNYRP>
2.0.CO;2.
48. Uppala SM, Kallberg PW, Simmons AJ, Andrae U,
Da Costa BV, Fiorino M, Gibson JK, Haseler J,
Hernandez A, Kelly GA, et al. The ERA-40 reanalysis.
Q J R Meteorol Soc 2005, 131:2961–3012. doi:
10.1256/qj.04.176.
49. Kanamitsu M, Ebisuzakai W, Woollen J, Yang S-K,
Hnilo JJ, Fiorino M, Potter GL. NCEP-DOE AMIP-
II reanalysis (R-2). Bull Am Meteorol Soc 2002,
83:1631–1643. doi: 10.1175/BAMS-83-11-1631.
50. Onogi K, Tsutsui J, Koide H, Sakamoto M, Kobayashi
S, Hatsushika H, Matsumoto T, Yamazaki N,
Kamahori H, Takahashi K, et al. The JRA-25
reanalysis. J Meteorol Soc Jpn 2007, 85:369–432.
51. Dee DP, Uppala SM, Simmons AJ, Berrisford P, Poli P,
Kobayashi S, Andrae U, Balmaseda MA, Balsamo G,
Bauer P, et al. The ERA-Interim reanalysis: configura-
tion and performance of the data assimilation system.
Q J R Meteorol Soc 2011, 137:553–597. doi: 10.1002/
qj.828.
52. Saha S, Moorthi S, Pan H-L, Wu X, Wang J, Nadiga
S, Tripp P, Kistler R, Woollen J, Behringer D,
et al. The NCEP climate forecast system reanalysis.
Bull Am Meteorol Soc 2010, 91:1015–1057. doi:
10.1175/2010BAMS3001.1.
53. Compo GP, Whitaker JS, Sardeshmukh PD, Matsui N,
Allan RJ, Yin X, Gleason BE, Vose RS, Rutledge G,
Bessemoulin P, et al. The twentieth century reanalysis
project. Q J R Meteorol Soc 2011, 137:1–28. doi:
10.1002/qj.776.
54. Rienecker MM, Suarez MJ, Gelaro R, Todling
R, Bacmeister J, Liu E, Bosilovish MG, Schubert
SD, Takacs L, Kim G-K, et al. MERRA: NASA’s
modern-era retrospective analysis for research and
applications. J Clim 2011, 24:3624–3648. doi:
10.1175/JCLI-D-11-00015.1.
©2013 John Wiley & Sons, Ltd.
WIREs Climate Change The expanding tropics
55. Seidel DJ, Randel WJ. Variability and trends in
the global tropopause estimated from radiosonde
data. J Geophys Res 2007, 111:D21101. doi:
10.1029/2006JD007363.
56. Lu J, Deser C, Reichler T. Cause of the widening of
the tropical belt since 1958. Geophys Res Lett 2009,
36:L03803. doi: 10.1029/2008GL036076.
57. Birner T. Recent widening of the tropical belt from
global tropopause statistics: sensitivities. J Geophys
Res 2010, 115:D23109. doi: 10.1029/2010JD014664.
58. Davis SM, Rosenlof KH. A multi-diagnostic inter-
comparison of tropical width time series using
reanalyses and satellite observations. J Clim 2012,
25:1061–1078. doi: 10.1175/JCLI-D-11-00127.1.
59. Seidel DJ, Randel WJ. Variability and trends in
the global tropopause estimated from radiosonde
data. J Geophys Res 2006, 111:D21101. doi:
10.1029/2006JD007363.
60. Schmidt T, Wickert J, Beyerle G, Heise S. Global
tropopause heights estimated from GPS radio
occultation. Geophys Res Lett 2008, 35:L23802. doi:
10.1029/2008GL035820.
61. Wilcox LJ, Hoskins BJ, Shine KP. A global blended
tropopause based on ERA data. Part II: trends and
tropical broadening. Q J R Meteorol Soc 2012, 138:
576–584. doi: 10.1002/qj.910.
62. Lucas C, Nguyen H, Timbal B. An observational
analysis of Southern Hemisphere tropical expansion.
J Geophys Res 2012, 117:D17112. doi: 10.1029/
2011JD017033.
63. Reichler T, Dameris M, Sausen R. Determining
the tropopause height from gridded data. Geophys
Res Lett 2003, 30:2042–2046. doi: 10.1029/
2003GL018240.
64. Davis NA, Birner T. Seasonal to multi-decadal
variability of the width of the tropical belt. J Geophys
Res 2013, 118:7773–7787. doi: 10.1002/jgrd.50610.
65. Hu Y, Fu Q. Observed poleward expansion of the
Hadley circulation since 1979. Atmos Chem Phys
2007, 7:5229–5236. doi: 10.5194/acp-7-5229-2007.
66. Hu Y, Zhou C, Liu J. Observational evidence for the
poleward expansion of the Hadley circulation. Adv
Atmos Sci 2011, 28:33–44. doi: 10.1007/s00376-010-
0032-1.
67. Chen J, Carlson BE, Del Genio AD. Evidence
for strengthening of the tropical general circulation
in the 1990s. Science 2002, 295:838–841. doi:
10.1126/science.1065835.
68. Wielecki BA, Wong T, Allen RP, Slingo A, Kiehl JT,
Soden BJ, Gordon CT, Miller AJ, Yang S-K, Randall
DA, et al. Evidence for large decadal variability in the
tropical mean radiative energy budget. Science 2002,
295:841–844. doi: 10.1126/science.1065837.
69. Hatzidimitriou D, Vardavas I, Pavlakis KG,
Hatzianastassiou N, Matsoukas C, Drakakis E. On
the decadal increase in the tropical mean outgoing
longwave radiation for the period 1984–2000. Atmos
Chem Phys 2004, 2004:1419–1425. doi: 10.5194/acp-
4-1419-2004.
70. Zhou YP, Xu K-M, Sud YC, Betts AK. Recent trends
of the tropical hydrological cycle inferred from Global
Precipitation Climatology Project and International
Satellite Cloud Climatology Project data. J Geophys
Res 2011, 116:D09101. doi: 10.1029/2010JD015197.
71. Allen RJ, Sherwood SC, Norris JR, Zender CS. Recent
Northern Hemisphere tropical expansion primarily
driven by black carbon and tropospheric ozone.
Nature 2012, 485:350–354. doi: 10.1038/nature
11097.
72. Rossow WB, Schiffer RA. ISCCP cloud data
products. Bull Am Meteorol Soc 1991, 72:2–20.
doi: 10.1175/1520-0477(1991)072<0002:ICDP>2.0.
CO;2.
73. Jacobowitz H, Stowe LL, Ohring G, Heidinger
A, Knapp K, Nalli NR. The Advanced Very
High Resolution Radiometer Pathfinder Atmosphere
(PATMOS) climate data set: a resource for climate
research. Bull Am Meteorol Soc 2003, 84:785–793.
doi: 10.1175/BAMS-84-6-785.
74. Hudson RD, Andrade MF, Follette MB, Frolov AD.
The total ozone field separated into meteorological
regimes—Part II: Northern hemisphere mid-latitude
total ozone trends. Atmos Chem Phys 2006,
6:5183–5191. doi: 10.5194/acp-6-5183-2006.
75. Hudson RD, Frolov AD, Andrade MF, Follette
MB. 2003: The total ozone field separated into
meteorological regimes. Part I: defining the regions.
J Atmos Sci 2003, 60:1669–1677. doi: 10.1175/1520-
0469(2003)060<1669:TTOFSI>2.0.CO;2.
76. Hudson RD. Measurements of the movement of the jet
streams at mid-latitudes, in the Northern and Southern
Hemispheres, 1979 to 2010. Atmos Chem Phys 2012,
12:7797–7808. doi: 10.5194/acp-12-7797-2012.
77. Fu Q, Johanson CM, Wallace JM, Reichler T.
Enhanced mid-latitude tropospheric warming in
satellite measurements. Science 2006, 312:1179. doi:
10.1126/science.1125566.
78. Fu Q, Lin P. Poleward shift of subtropical jets
inferred from satellite-observed lower-stratospheric
temperatures. J Clim 2011, 24:5597–5603. doi:
10.1175/JCLI-D-11-00027.1.
79. Johanson CM, Fu Q. Hadley cell widening: model
simulations versus observations. J Clim 2009,
22:2713–2725. doi: 10.1175/2008JCLI2620.1.
80. Stachnik JP, Schumacher C. A comparison of the
Hadley circulation in modern reanalyses. J Geophys
Res 2011, 116:D22102. doi: 10.1029/2011JD016677.
81. Archer CL, Caldeira K. Historical trends in the jet
streams. Geophys Res Lett 2008, 35:L08803. doi:
10.1029/2008GL033614.
©2013 John Wiley & Sons, Ltd.
Advanced Review wires.wiley.com/climatechange
82. Adler RF, Huffman GJ, Chang A, Ferraro R, Xie
P-P, Janowiak J, Rudolf B, Schneider U, Curtis S,
Bolvin D, et al. The version 2 Global Precipita-
tion Climatology Project (GPCP) monthly precipita-
tion analysis (1979-Present). J Hydrometeor 2003,
4:1147–1167. doi: 10.1175/1525-7541(2003)004
<1147:TVGPCP>2.0.CO;2.
83. Huffman GJ, Adler RF, Bolvin DF, Gu G. Improving
the global precipitation record: GPCP version
2.1. Geophys Res Lett 2009, 36:L17808. doi:
10.1029/2009GL040000.
84. Allan RJ, Ansell TJ. A new globally complete
monthly historical mean sea level pressure data set
(HadSLP2):1850-2004. J Clim 2006, 19:5816–5842.
doi: 10.1175/JCLI3937.1.
85. Tanaka HL, Ishizaki N, Kitoh A. Trend and
interannual variability of Walker, monsoon and
Hadley circulations defined by velocity potential in
the upper troposphere. Tellus 2004, 56A:250–269.
doi: 10.1111/j.1600-0870.2004.00049.x.
86. Mitas CM, Clement A. Has the Hadley cell been
strengthening in recent decades? Geophys Res Lett
2005, 32:L03809. doi: 10.1029/2004GL021765.
87. Oort AH, Yienger JJ. Observed interannual variability
in the Hadley circulation and its connection to ENSO.
J Clim 1996, 9:2751–2767. doi: 10.1175/1520-
0442(1996)009<2751:OIVITH>2.0.CO;2.
88. Sohn BJ, Park S-C. Strengthened tropical circulations
in past three decades inferred from water vapour
transport. J Geophys Res 2010, 115:D15112. doi:
10.1029/2009JD013713.
89. Quan X-W, Diaz HF, Hoerling MP. Change in
the tropical Hadley cell since 1950. In: Diaz HF,
Bradley RS, eds. The Hadley Circulation: Past, Present
and Future. Dordrecht: Kluwer Academic Publishers;
2004, 85–120.
90. Trenberth KE. Changes in tropical clouds and radi-
ation. Science 2095, 2002:296. doi: 10.1126/sci-
ence.296.5576.2095a.
91. Wong T, Wielicki BA, Lee RB, Smith GL, Bush KA,
Willis JK. Reexamination of the observed decadal
variability of the Earth radiation budget suing altitude-
corrected ERBE/ERBS nonscanner WFOV data. JClim
2006, 19:4028–4040. doi: 10.1175/JCLI3838.1.
92. Vecchi GA, Soden BJ, Wittenberg AT, Held IM,
Leetmaa A, Harrison MJ. Weakening of tropical
Pacific atmospheric circulation due to anthropogenic
forcing. Nature 2006, 441:73–76. doi: 10.1038/nature
04744.
93. Thorne PW, Vose RS. Reanalyses suitable for
characterizing long-term trends. Bull Am Meteorol Soc
2010, 91:353–361. doi: 10.1175/2009BAMS2858.1.
94. Reichler T, Kim J. Uncertainties in the climate mean
state of global observations, reanalyses and the GFDL
climate model. J Geophys Res 2008, 113:D05106. doi:
10.1029/2007JD009278.
95. Sterl A. On the (in)homogeneity of reanalysis products.
J Clim 2004, 17:3866–3873. doi: 10.1175/1520-
0442(2004)017<3866:OTIORP>2.0.CO;2.
96. Trenberth KE, Stepaniak DP, Hurrell JW, Fiorno M.
Quality of reanalysis in the tropics. J Clim 2001,
14:1499–1510. doi: 10.1175/1520-0442(2001)014<
1499:QORITT>2.0.CO;2.
97. Bengtsson L, Hagemann S, Hodges KI. Can climate
trends be calculated from reanalysis data? J Geophys
Res 2004, 109:D11111. doi: 10.1029/2004JD004536.
98. Mitas CM, Clement A. Recent behaviour of Hadley
Cell and tropical thermodynamics in climate models
and re-analyses. Geophys Res Lett 2006, 33:L01810.
doi: 10.1029/2005GL024406.
99. Song H, Zhang M. Changes of the boreal winter
Hadley Circulation in the NCEP-NCAR and ECMWF
Reanalyses: a comparative study. J Clim 2007,
20:5191–5200. doi: 10.1175/JCLI4260.1.
100. Bengtsson L, Hodges KI, Hagemann S. Sensitivity of
the ERA40 reanalysis to the observing systems: deter-
mination of the global atmospheric circulation from
reduced observations. Tellus 2004, 56A:456–471. doi:
10.1111/j.1600-0870.2004.00079.x.
101. Bender FA-M, Ramanathan V, Tselioudis G. Changes
in extratropical storm track cloudiness 1983-2008:
observational support for a poleward shift. Clim
Dyn 2012, 38:2037–2053. doi: 10.1007/s00382-011-
1065-6.
102. Huang Y, Ramaswamy V, Soden B. An investigation
of the sensitivity of the clear-sky outgoing longwave
radiation to atmospheric temperature and water
vapour. J Geophys Res 2007, 112:D05104. doi:
10.1029/2005JD006906.
103. Huang Y, Ramaswamy V. Evolution and trend of the
outgoing longwave radiation spectrum. J Clim 2009,
22:4637–4651. doi: 10.1175/2009JCLI2874.1.
104. Ignatov A, Laszlo I, Harrod ED, Kidwell KB,
Goodrum GP. Equator crossing times for NOAA,
ERS and EOS sun-synchronous satellites. Int J
Remote Sensing 2004, 25:5255–5266. doi: 10.1080/
01431160410001712981.
105. Waliser DE, Zhou W. Removing satellite equatorial
crossing time biases from the OLR and HRC datasets.
J Clim 1997, 10:2125–2146. doi: 10.1175/1520-
0442(1997)010<2125:RSECTB>2.0.CO;2.
106. Lee H-T, Gruber A, Ellingson RG, Laszlo I.
Development of the HIRS outgoing longwave
radiation climate dataset. J Atmos Oceanic Tech 2007,
24:2029–2047. doi: 10.1175/2007JTECHA989.1.
107. Sapiano MRP, Janowiak JE, Smith TM, Arkin PA, Xie
P, Lee H. Correction for temporal discontinuities in
the OPI. J Atmos Oceanic Tech 2010, 27:457–469.
doi: 10.1175/2009JTECHA1366.1.
108. Liebmann B, Smith CA. Description of a complete
(Interpolated) outgoing longwave radiation dataset.
Bull Am Meteorol Soc 1996, 77:1275–1277.
©2013 John Wiley & Sons, Ltd.
WIREs Climate Change The expanding tropics
109. Koch P, Wernli H, Davies HUW. An event-based jet-
stream climatology and typology. Int J Climatol 2006,
26:283–301. doi: 10.1002/joc.1255.
110. Meehl GA, Covey C, Delworth T, Latif M, McAvaney
B, Mitchell JFB, Stouffer RJ, Taylor KE. The
WCRP CMIP3 multi-model dataset: a new era in
climate change research. Bull Am Meteorol Soc 2007,
88:1383–1394. doi: 10.1175/BAMS-88-9-1383.
111. Vecchi GA, Soden BJ. Global warming and the
weakening of the tropical circulation. JClim2007,
20:4316–4340. doi: 10.1175/JCLI4258.1.
112. Gastineau G, Le Truet H, Li L. Hadley circulation
changes under global warming conditions indicated by
coupled climate models. Tellus 2008, 60A:863–884.
doi: 10.1111/j.1600-0870.2008.00344.x.
113. Held I, Soden BJ. Robust responses of the hydrological
cycle to global warming. J Clim 2006, 19:5686–5699.
doi: 10.1175/JCLI3990.1.
114. Seager R, Niak N, Vecchi GA. Thermodynamic
and dynamic mechanisms for large-scale changes
in the hydrological cycle in response to global
warming. J Clim 2010, 23:4651–4668. doi: 10.1175/
2010JCLI13655.1.
115. Scheff J, Frierson D. Twenty-first century multi-
model subtropical precipitation declines are mostly
midlatitude shifts. J Clim 2012, 25:4330–4347. doi:
10.1175/JCLI-D-11-00393.1.
116. Otto-Bliesner BL, Clement A. The sensitivity of
the Hadley circulation to past and future forcings
in two climate models. In: Diaz HF, Bradley RS,
eds. The Hadley Circulation: Past, Present and
Future. Dordreacht: Kluwer Academic Publishers;
2005, 437–464.
117. Arblaster JM, Meehl GA, Karoly DJ. Future climate
change in the Southern Hemisphere: competing effects
of ozone and greenhouse gases. Geophys Res Lett
2011, 38:L02701. doi: /10.1029/2010GL045384.
118. Previdi M, Liepert BG. Annular modes and Hadley cell
expansion under global warming. Geophys Res Lett
2007, 34:L22701. doi: 10.1029/2007GL031243.
119. Sobel AH, Camargo SJ. Projected future seasonal
changes in tropical summer climate. J Clim 2011,
24:473–487. doi: 10.1175/2010JCLI3748.1.
120. Tandon NF, Gerber EP, Sobel AH, Polvani LM.
Understanding Hadley cell expansion vs. contraction:
insights from simplified models and implications for
recent observations. J Clim 2013, 26:4304–4321. doi:
10.1175/JCLI-D-12-00598.1.
121. Chen G, Held IM. Phase speed spectra and the
recent poleward shift of Southern Hemisphere surface
westerlies. Geophys Res Lett 2007, 34:L21805. doi:
10.1029/2007GL031200.
122. Chen G, Lu J, Frierson DMW. Phase speed spectra
and the latitude of surface westerlies: Interannual
variability and the global warming trend. J Clim 2008,
21:5942–5959. doi: 10.1175/2008JCLI2306.1.
123. Lorenz DJ, DeWeaver ET. Tropopause height and
zonal wind response to global warming in the
IPCC scenario integrations. J Geophys Res 2007,
112:D10119. doi: 10.1029/2006JD008087.
124. McLandress C, Shepherd T, Scinocca J, Plummer
D, Sigmond M, Jonsson A, Reader C. Separat-
ing the dynamical effects of climate change and
ozone depletion: Part 2. Southern hemisphere tro-
posphere. J Clim 2011, 24:1850–1868. doi: 10.1175/
2010JCL3958.1.
125. Kidston J, Dean SM, Renwick JA, Vallis GK. A robust
increase in the eddy length scale in the simulation of
future climates. Geophys Res Lett 2010, 37:L03806.
doi: 10.1029/2009GL041615.
126. Rivi`
ere G. A dynamical interpretation of the
poleward shift of the jet stream in global warming
scenarios. J Atmos Sci 2011, 2011:1253–1272. doi:
10.1175/2011JAS3641.1.
127. Butler AH, Thompson DWJ, Heikes R. The steady-
state atmospheric circulation response, to climate
change-like thermal forcings in a simple general
circulation model. J Clim 2010, 23:3474–3496. doi:
10.1175/2010JCLI3228.1.
128. Allen RJ, Sherwood SC, Norris JR, Zender CS. The
equilibrium response to idealized thermal forcings in a
comprehensive GCM: implications for recent tropical
expansion. Atmos Chem Phys 2012, 12:4795–4816.
doi: 10.5194/acp-12-4795-2012.
129. Grassi B, Redaelli G, Canziani PO, Visconti G. Effects
of the PDO Phase on the tropical belt width. J
Clim 2012, 25:3282–3290. doi: 10.1175/JCLI-D-11-
00244.1.
130. Santer BD, Wehner MF, Wigley TML, Sausen R,
Meehl GA, Taylor KE, Ammann C, Arblaster J, Wash-
ington WM, Boyle JS, et al. Contributions of anthro-
pogenic and natural forcing to recent tropopause
height changes. Science 2003, 301:479–483. doi:
10.1126/science.1084123.
131. Gu G, Adler RF, Huffman GJ, Curtis S. Tropical
rainfall variability on interannual-to-interdecadal and
longer time scales derived from the GPCP monthly
product. J Clim 2007, 20:4033–4046. doi: 10.1175/
JCLI14227.1.
132. Emile-Geay J, Seager R, Cane MA, Cook ER, Haug
GH. Volcanoes and ENSO over the past millen-
nium. J Clim 2008, 21:3134–3148. doi: 10.1175/
2007JCLI1884.1.
133. Deser C, Phillips AS. Atmospheric circulation
trends, 1950–2000: the relative role of sea surface
temperature forcing and direct atmospheric radiative
forcing. J Clim 2009, 22:396–413. doi: 10.1175/
2008JCLI2453.1.
134. Staten PW, Rutz JJ, Reichler T, Lu J. Breaking down
the tropospheric circulation response by forcing. Clim
Dyn 2012, 39:2361–2375. doi: 10.1007/s00382-011-
1267-y.
©2013 John Wiley & Sons, Ltd.
Advanced Review wires.wiley.com/climatechange
135. Son S-W, Tandon NF, Polvani LM, Waugh DW.
Ozone hole and southern hemisphere climate change.
Geophys Res Lett 2009, 36:L15705. doi: 101.1029/
2009GL038671.
136. Min S-K, Son S-W. Multimodel attribution of the
Southern Hemisphere Hadley cell widening: Major
role of ozone depletion. J Geophys Res 2013,
118:3007–3015. doi: 10.1002/jgrd.50232.
137. Son S-W, Gerber EP, Polvani LM, Gillett NP,
Seo K-H, Eyring V, Shepherd TG, Waugh D,
Akiyoshi A, Austin J, et al. Impact of stratospheric
ozone on Southern Hemisphere circulation change:
a multimodel assessment. J Geophys Res 2010,
115:D00M07. doi: 10.1029/2010JD014271.
138. Polvani LM, Waugh DW, Correa GJP, Son S-W.
Stratospheric ozone depletion: the main driver of
twentieth-century atmospheric circulation changes in
the Southern Hemisphere. J Clim 2011, 24:795–812.
doi: 10.1175/2010JCLI3772.1.
139. Polvani LM, Previdi M, Deser C. Large cancellation,
due to ozone recovery, of future Southern Hemisphere
atmospheric circulation trends. Geophys Res Lett
2011, 38:L04707. doi: 10.1029/2011GL046712.
140. Kang SM, Polvani LM, Fyfe JC, Sigmond M. Impact
of polar ozone depletion on subtropical precipi-
tation. Science 2011, 332:951–954. doi: 10.1126/
science.1202131.
141. Ramanathan V, Crutzen PJ, Kiehl JT, Rosenfeld D.
Aerosols, climate and the hydrological cycle. Science
2001, 294:2119–2124. doi: 10.1126/science.1064034.
142. Lohmann U, Feichter J. Global indirect aerosol effects:
areview.Atmos Chem Phys 2005, 5:715–737. doi:
10.5194/acp-5-715-2005.
143. Ming Y, Ramaswamy V. Nonlinear climate and
hydrological responses to aerosol effects. J Clim 2009,
22:1329–1339. doi: 10.1175/2008JCLI2362.1.
144. Chung CE, Ramanathan V, Kim D, Podgorny IA.
Global anthropogenic aerosol direct forcing derived
from satellite and ground-based observations. J
Geophys Res 2005, 110:D24207. doi: 10.1029/
2005JD006536.
145. Randles CA, Ramaswamy V. Absorbing aerosols over
Asia: a geophysical fluid dynamics laboratory general
circulation models sensitivity study of model response
to aerosol optical depth and aerosol absorption.
J Geophys Res 2008, 113:D21203. doi: 10.1029/
2008jd010140.
146. Rotstayn LD, Lohmann U. Tropical rainfall trends
and the indirect aerosol effect. J Clim 2002,
15:2103–2116. doi: 10.1175/1520-0442(2002)015<
2103:TRTATI>2.0.CO;2.
147. Rotstayn LD, Cai W, Dix MR, Farquhar GD, Feng Y,
Ginoux P, Herzog M, Ito A, Penner JE, Roderick ML,
et al. Have Australian rainfall and cloudiness increased
due to the remote effects of Asian anthropogenic
aerosols? J Geophys Res 2007, 112:D09202. doi:
10.1029/2006JD007712.
148. Wang C. Impact of direct radiative forcing
of black carbon aerosols on tropical convective
precipitation. Geophys Res Lett 2007, 34:L05709.
doi: 10.1029/2006GL028416.
149. Allen RJ, Sherwood RC. The impact of natural versus
anthropogenic aerosols on atmospheric circulation in
the Community Atmospheric Model. Clim Dyn 2011,
36:1959–1978. doi: 10.1007/s00382-010-0898-8.
150. Tosca MG, Randerson JT, Zender CS. Global impact
of smoke aerosols from landscape fires on climate
and the Hadley circulation. Atmos Chem Phys 2013,
13:5227–5241. doi: 10.5194/acp-13-5227-2013.
151. Ming Y, Ramaswamy V. A model investigation of
aerosol-induced changes in tropical circulation. J Clim
2011, 24:5125–5133. doi: 10.1175/2011JCLI4108.1.
152. Willett KM, Gillett NP, Jones PD, Thorne PW.
Attribution of observed surface humidity changes to
human influence. Nature 2007, 449:710–713. doi:
10.1038/nature06207.
153. Dai A. Recent climatology, variability and trends in
global surface humidity. J Clim 2006, 19:3589–3606.
doi: 10.1175/JCLI3816.1.
154. Lucas, C. A high-quality humidity database for
Australia. CAWCR Technical Report No. 024;
2010, 162 pp. Available at: http://www.cawcr.gov.au/
publications/technicalreports.php. (Accessed Septem-
ber 02, 2013)
155. Power SB, Kociuba G. What caused the observed
twentieth-century weakening of the Walker Circu-
lation? J Clim 2011, 24:6501–6514. doi: 10.1175/
2011JCLI4101.1.
156. Zhang X, Zwiers FW, Hegerl GC, Lambert FH,
Gillett NP, Solomon S, Stott PA, Nozawa T.
Detection of human influence on twentieth century
precipitation trends. Nature 2007, 448:461–465. doi:
10.1038/nature06025.
157. Bain CL, De Paz J, Kramer J, Magnusdottir G,
Smyth P, Stern H, Wang C-C. Detecting the ITCZ
in instantaneous satellite data using spatiotemporal
statistical modelling: ITCZ climatology in the
east Pacific. J Clim 2011, 24:216–230. doi:
10.1175/2010JCLI13716.1.
158. Yin JH. A consistent poleward shift of the storm
tracks in simulations of 21st century climate
change. Geophys Res Lett 2005, 32:L18701. doi:
10.1029/2005GL023684.
159. Cai W, Cowan T, Thatcher M. Rainfall reductions
over Southern Hemisphere semi-arid regions: the role
of subtropical dry zone expansion. Sci Rep 2012,
2:702. doi: 10.1038/srep00702.
©2013 John Wiley & Sons, Ltd.