Prominent Midlatitude Circulation Signature in High
Asia’s Surface Climate During Monsoon
Thomas Mölg
1
, Fabien Maussion
2
, Emily Collier
1
, John C. H. Chiang
3
,
and Dieter Scherer
4
1
Climate System Research Group, Institute of Geography, Friedrich-Alexander-University Erlangen-Nürnberg (FAU),
Erlangen, Germany,
2
Institute for Atmospheric and Cryospheric Sciences, University of Innsbruck, Innsbruck, Austria,
3
Department of Geography, Center for Atmospheric Sciences, University of California, Berkeley, CA, USA,
4
Chair of
Climatology, Technische Universität Berlin, Berlin, Germany
Abstract High Asia has experienced strong environmental changes in recent decades, as evident in
records of glaciers, lakes, tree rings, and vegetation. The multiscale understanding of the climatic drivers,
however, is still incomplete. In particular, few systematic assessments have evaluated to what degree, if at all,
the midlatitude westerly circulation modifies local surface climates in the reach of the Indian Summer
Monsoon. This paper shows that a southward shift of the upper-tropospheric westerlies contributes
significantly to climate variability in the core monsoon season (July–September) by two prominent dipole
patterns at the surface: cooling in the west of High Asia contrasts with warming in the east, while moist
anomalies in the east and northwest occur with drying along the southwestern margins. Circulation
anomalies help to understand the dipoles and coincide with shifts in both the westerly wave train and the
South Asian High, which imprint on air mass advection and local energy budgets. The relation of the
variabilities to a well-established index of midlatitude climate dynamics allows future research on climate
proxies to include a fresh hypothesis for the interpretation of environmental changes.
1. Introduction
Climatic studies of High Asia have flourished in recent years. The drivers of this interest are the important
roles of the High Asian landmass for the hemisphere-scale atmospheric circulation (Chiang et al., 2015)
and the Asian water cycle (Zhang et al., 2013), together with widespread environmental changes in recent
decades that illustrate strong effects of climate change in this part of the world. The observed changes
concern glaciers (e.g., Bolch et al., 2012), lakes (e.g., Kropáček et al., 2012), streamflow (e.g., Zhang et al.,
2013), and vegetation characteristics (e.g., Zhang et al., 2013)—phenomena that qualify as so-called climate
proxies. To date, however, the climatic drivers of the observed variability and changes in proxy data are insuf-
ficiently understood, as emphasized by Bolch et al. (2012). In particular, the link between large-scale dynamics
and the local or regional surface climates deserves more attention, which would extend the room for physical
interpretations of the environmental changes.
Among the existing and largely statistical efforts, relating proxy data to local meteorological time series is the
dominant approach (e.g., Grießinger et al., 2016; Liang et al., 2016). A second tendency is the linkage with
remote large-scale mechanisms of climate variability in the tropics (e.g., Pacific and Indian Ocean conditions)
and extratropics (e.g., North Atlantic variability) (Hochreuther et al., 2016; Liang et al., 2016), which makes
sense in view of High Asia’s location at the interface of tropical and extratropical climates. The location aspect
also matters in a third group of studies; these relate climate proxy data to the “in situ”large-scale dynamics,
the extratropical westerlies and the Indian Summer Monsoon (ISM) (e.g., Benn & Owen, 1998; Li et al., 2016;
Mölg et al., 2012; Yao et al., 2012). However, the said research typically regards the driver role of the two sys-
tems in separation and to be unequal. Since westerlies blow over High Asia constantly only in winter
(Maussion et al., 2014), but the strongest signals in proxies all over High Asia are associated with the summer
months (Grießinger et al., 2016; Liang et al., 2016; Zhu et al., 2011) during the northward progression of the
monsoon, the ISM is favored in the literature for explaining climatic fluctuations at all time scales; only for
individual mountain chains in northern High Asia are the westerlies deemed important (Yao et al., 2012).
Yet the separation concept is not consistent with the findings from atmospheric research that the
westerlies do not fully “withdraw”in summer (see below). Only a few statistical studies searched for imprints
of both the westerlies and the ISM in their proxy records (Joswiak et al., 2013), and even fewer through
MÖLG ET AL. CIRCULATION INFLUENCES ON HIGH ASIA 12,702
PUBLICATION
S
Journal of Geophysical Research: Atmospheres
RESEARCH ARTICLE
10.1002/2017JD027414
Key Points:
•An established large-scale index of
westerly waves helps to reveal
significant anomalies in High Asia’s
local and regional surface climate
•These anomalies appear as
widespread, distinct dipole patterns in
surface-air temperature and moisture
•Midlatitude climate dynamics can be
incorporated for explaining
surface-climate variability in
monsoonal High Asia
Supporting Information:
•Supporting Information S1
Correspondence to:
T. Mölg,
thomas.moelg@fau.de
Citation:
Mölg, T., Maussion, F., Collier, E., Chiang,
J. C. H., & Scherer, D. (2017). Prominent
midlatitude circulation signature in
High Asia’s surface climate during
monsoon. Journal of Geophysical
Research: Atmospheres,122,
12,702–12,712. https://doi.org/10.1002/
2017JD027414
Received 14 JUL 2017
Accepted 12 NOV 2017
Accepted article online 18 NOV 2017
Published online 11 DEC 2017
©2017. American Geophysical Union.
All Rights Reserved.
process-based models (Mölg et al., 2014). Deepening our understanding of the combined westerlies-ISM
contribution to the summertime surface-climate and environmental variability in High Asia remains an untra-
veled path in the recent research.
This is surprising on the one side, since atmospheric dynamics literature has long since emphasized that the
interplay of both circulation systems shapes the summer climate of monsoonal Asia, starting in the 1960s
(Ramaswamy, 1962) and continuing to the present (e.g., Bothe et al., 2011; Curio et al., 2015; Krishnan
et al., 2009). In particular, the midlatitude westerlies over Asia are part of a circumglobal wave train in the
upper troposphere (Ding & Wang, 2005), which creates linkages to the upstream climate and the North
Atlantic as a consequence (Bothe et al., 2011; Liu et al., 2015; Mölg et al., 2014). Further themes of the many
studies on large-scale dynamics that are relevant to the westerlies-monsoon interplay have been teleconnec-
tion routes (e.g., Bothe et al., 2011), the westerly jet stream (e.g., Schiemann et al., 2009), air-pressure patterns
at the surface (e.g., Saeed et al., 2011) and in the upper troposphere (e.g., Wei et al., 2017), and the role of
synoptic-scale troughs and ridges (e.g., Krishnan et al., 2009). The local patterns of climate variability over
High Asia and their specific relation to the westerly influence, however, are less well studied.
Thus, and on the other side, a quantitative framework that allows climate proxy research to consider westerly
impacts in a similar fashion as ISM influences does not exist. For instance, the routine usage of an index
with understandable relations to both the large-scale dynamics and regional-to-local climate variability in
High Asia is not well developed for the westerlies. Hence, we postulate that valuable information about
the past and recent interaction between westerlies and ISM is largely unrevealed from climate proxy
archives. The goal of the present research is to attempt a step toward establishing a more systematic fra-
mework. We examine whether an index of the large-scale westerly circulation has the potential to capture
significant surface-climate variability in High Asia during June–September. This goal follows the hypothesis
that the activity of the midlatitude westerlies in summer appears prominently beyond the free atmo-
sphere in the surface climate of High Asia as well. We focus on the temperature and moisture fields, which
contain the variables that correlate most strongly with proxy signals (e.g., Grießinger et al., 2016;
Hochreuther et al., 2016; Liang et al., 2016; Zhu et al., 2011). The findings are presented in a concise
way, which intends to stimulate fresh lines of thinking in future research on proxy data from a climate
dynamics angle.
2. Data and Methods
2.1. Reanalysis and Index Definitions
Reanalysis data represent a replicate of the observed atmospheric state on a global scale. The National
Centers for Environmental Prediction (NCEP)/National Center for Atmospheric Research (NCAR) reanalysis
(Kalnay et al., 1996) provides records since 1948, but at a rather coarse horizontal resolution of 2.5° that limits
its usage for surface-climate analyses. Thus, we only consider the tropospheric levels at and above 500 hPa
from this product for studying the upper half of the troposphere, which is well above the surface of High
Asia (apart from localized high peaks). Humidity is only available up to 300 hPa, and thus, our wind analyses
for the upper troposphere are done at this level (the results for 200 hPa winds were very similar for all
analyses). The European reanalysis Era-Interim (Dee et al., 2011) is available since 1979 at a 0.75° resolution,
while NASA’s reanalysis MERRA, version 2 (Bosilovich et al., 2015), has a resolution of 0.5° to 0.625° (latitude
and longitude, respectively) and covers 1980 onward. We include these two products in the modeling
(section 2.2) and the surface-climate analyses, respectively. See Text S1 in the supporting information for
data download links.
For the state of the westerlies, we follow the Global Wave Train Index (GTI) of Ding and Wang (2005). They
demonstrated in detail that the GTI captures a major component of climate variability in the upper-
tropospheric circulation, which is associated with the westerly jet stream. Therefore, we do not attempt to
invent a further index. The many correlations of the GTI with leading modes of climate variability over
High Asia, as we will see in the results (section 3) and Table S1 in the supporting information, corroborate this
choice in the given context. The GTI is simply the anomaly of the 200 hPa geopotential height, area-averaged
over 35°–40°N and 60°–70°E. We multiply the GTI by 1 for easier interpretation in accordance with the
theme of this study (clarified in Figure 1 caption) and will refer to as the “westerly index”hereafter. Two
further indices are considered in the discussion of our results. First, the subtropical westerly jet position
Journal of Geophysical Research: Atmospheres 10.1002/2017JD027414
MÖLG ET AL. CIRCULATION INFLUENCES ON HIGH ASIA 12,703
index defined by Zhao et al. (2014) is another one in the upper tropo-
sphere: area-averaged zonal wind at 200 hPa in a southern region
(30°–45°N, 50°–80°E) minus the one in a northern region (45°–60°N,
50°–80°E). Positive deviations in the normalized time series, therefore,
imply a more southerly jet position. Second is the All-India rainfall
index, since it is the most widely used ISM index; its time series is readily
available online (see Text S1).
Figure 1 illustrates that there were about 15 distinctly positive and
negative phases of the westerly index since 1948, both in June (early
monsoon season) and in the July–September (JAS) season. Positive
ones coincide with an anomalously low 200 hPa surface and thus
appreciable influence of the westerlies at subtropical latitudes (hence-
forth, W+). Negative phases indicate reduced southward influence of
the west winds (W). The difference between contrasting composites
(e.g., Ding & Wang, 2005; Saeed et al., 2011), here W+ minus W,
therefore reveals the westerly “signature”in the climatic variability.
Among all the employed indices, only the GTI June time series and
the ISM JAS index showed a statistically significant trend over
1948–2014 (p<0.05) and were linearly detrended to focus on the
interannual variability.
2.2. Atmospheric Modeling
For all years with W+ and Wcases (Figure 1) since 1980, we simulated
the June–September season with version 3.8.1 of the Weather Research
and Forecasting (WRF) model (Skamarock and Klemp, 2008); we used a
20 km horizontal grid spacing over a large domain that contains High
Asia (Figure S1 in the supporting information) and Era-Interim as
forcing data. The model settings are from a recent study and were
carefully evaluated with in situ measurements from an intensive field experiment in the Himalaya (Collier
& Immerzeel, 2015). Table S2 and Text S2 reveal the relevant model details.
2.3. Station Observations
We accessed all Global Historical Climatology Network station data available through the “climate explorer”
platform (Table S3). The distribution of stations is sharply biased toward eastern High Asia; this feature also
exists if further (nonpublically available station data) are incorporated (Liu et al., 2015). Therefore, we pre-
ferred the direct use of station data over gridded products that are based on ground observations, since
the gridded fields cannot contain additional information in station-free regions. W+ and Wcomposites
from stations typically include years before 1980 or lack some years in the 1980–2014 (MERRA and WRF) per-
iod, which demonstrates variable data availability (Table S3). However, all stations provide W+ and Wyears
that stretch over several decades, which should ascertain a reasonable degree of robustness in light of multi-
decadal patterns that are of interest here.
2.4. Statistical Applications
We employ the widely used empirical orthogonal function (EOF) analysis to extract the leading modes of
atmospheric variability over High Asia, which yields the spatial patterns and the associated time series
(expansion coefficient). Spatial patterns are illustrated by the so-called loadings, which measure (in dimen-
sionless units) the strength of the relation between expansion coefficient and the local variability. Since
reanalysis grid cells have different areas due to their degree resolution, we apply the cosine correction prior
to the EOF to ensure that an equal area has the same contribution to the total variance of the data (Ding &
Wang, 2005). We utilize North’s rule of thumb to assess the uniqueness of the modes in the final EOF step
(North et al., 1982).
The significance for temporal correlations is based on a two-sided ttest. Those for the spatial differences in
data rely on daily mean fields (to increase the sample size and make more robust assessments) and on the
Kolmogorov-Smirnov test (which can accommodate strongly skewed distributions like daily precipitation).
a
June
1950 1960 1970 1980 1990 2000 2010
Units of standard deviation
-3
-2
-1
0
1
2
3
bJAS
1950 1960 1970 1980 1990 2000 2010
Units of standard deviation
-3
-2
-1
0
1
2
3
Figure 1. States of the westerly wave train. Standardized mean wave train index
for (a) June and (b) July–September from 1948 to 2014. Symbols highlight
distinctly positive phases (red dots; ≥0.75) and negative phases (black stars,
≤0.75). For convenience of interpretation, the original wave train index (GTI) is
multiplied by 1 in this paper, such that the positive phases coincide with W+.
Journal of Geophysical Research: Atmospheres 10.1002/2017JD027414
MÖLG ET AL. CIRCULATION INFLUENCES ON HIGH ASIA 12,704
We evaluate the Pvalues from the latter test with the false discovery rate (FDR) field-significance approach of
Wilks (2016). The control level is set to 0.01, so only grid cells with p≪0.01 will be significant in a spatial
context (typically, an order of magnitude smaller in our case).
3. Results and Discussion
To set the context, we begin by looking at the atmospheric anomalies due to westerly influence in the upper
troposphere, where the ISM-westerlies interaction originates (Bothe et al., 2011; Ding & Wang, 2005; Krishnan
et al., 2009; Saeed et al., 2011). For these levels we utilize the NCEP/NCAR reanalysis product, which provides
relatively long time series among available reanalyses (since 1948). Below we highlight results for the main
monsoon season, JAS, but also performed all analyses for June. June conditions reflect ISM onset character-
istics and may strongly impact the annual state of particular proxies (Li et al., 2016; Mölg et al., 2014), an
aspect that should be considered in the discussion.
Figure 2 summarizes the patterns of the atmospheric anomalies, without addressing their significance at this
stage. Strong westerlies cool the troposphere at all latitudes above West High Asia (Figure 2a) and coincide
with widespread drying over classical ISM areas (Arabian Sea, India) and the southern rim regions of High Asia
(Figure 2b). The associated flow anomalies reside on two action centers: an anticyclonic anomaly at higher
latitudes (denoted “A”in Figure 2c) and a cyclonic anomaly at ~37°N (“B”in Figure 2c). Consequently, a strong
northerly wind anomaly extends from the high latitudes to the subtropics, while the southern flank of the
anomalous circulation entails enhanced westerlies at 30–35°N (Figure 2c). These meridional and zonal com-
ponents favor the advection of cold and dry air typical of the westerly influence (Krishnan et al., 2009;
Krishnan & Sugi, 2001; Zhao et al., 2014). A similar flow configuration was found to be important for climate
variability of the Tarim Basin north of High Asia (Zhao et al., 2014). Tarim anomalies “involve matching circu-
lation anomalies in the tropics and mid-latitudes”state Zhao et al. (2014), which underlines the result in
Figure 2c that coupled anomalies across major climate zones matter. The spatially confined warm and moist
anomalies in High Asia (Figures 2a and 2b) are discussed further below, as well as the aforementioned role of
large-scale advection.
However, we need to pose an essential question: are the patterns in Figure 2 significant or merely a part of
random variability? The EOF analysis reveals that at least one of the two major patterns of each tropospheric
temperature, humidity, and wind variability over High Asia correlates significantly (at the 1% level) with the
westerly index (Table S1). Figure 3 illustrates the most relevant spatial variability patterns. W+ influence goes
Figure 2. Effects of southward-penetrating westerlies in the upper half of the troposphere during boreal summer.
Differences between W+ and Wcomposites in JAS, 1948–2014, for (a) geopotential thickness of the 500–200 hPa layer
(reduced thickness indicates lower layer temperature), (b) vertically integrated specific humidity of the 500–300 hPa layer,
and (c) wind vectors at 300 hPa, where positive zonal (westerly) and negative meridional (northerly) anomalies are indicated
by the grey and blue shadings of the thick arrow, respectively, to illustrate the main pattern; the two anomaly centers
connecting the system are labeled (A and B). Note that the wind arrows represent anomalies, which act upon the mean flow
(Figure S2). Also note that specific humidity in NCEP/NCAR is only available up to 300 hPa, which determines the layer top for
Figure 2b. The dashed line in every plot is the 2,000 m contour of terrain elevation, which roughly delineates High Asia.
Journal of Geophysical Research: Atmospheres 10.1002/2017JD027414
MÖLG ET AL. CIRCULATION INFLUENCES ON HIGH ASIA 12,705
along with a distinct dipole pattern with cooling in the west but warming in the northeast of High Asia
(Figure 3a), which closely resembles the pattern in Figure 2a. Also, the circulation anomalies of Figure 2c reap-
pear in the leading EOF patterns of zonal and meridional winds. The zonal pattern reveals increasing wester-
lies during W+, except in the northwest of High Asia (Figure 3c). The meridional pattern features prominent
southerly wind anomalies (orange in Figure 3d), which agree with the eastern flank of the anomaly center “B”
in Figure 2c. According to the North test, these three patterns are unique since their error bars in Figure 4
(black entries) do not overlap with the adjacent modes. The situation is different for EOF2 of humidity, which
correlates appreciably with the westerly index (r= 0.60) but is degenerate (Figure 4). We therefore rotated the
first three EOFs of humidity with the varimax method, which is standard to extract clearer and more robust
patterns (Ward, 1999). The correlation of the rotated EOF2 of humidity (REOF2) with the westerly index
Figure 3. Leading modes of tropospheric variability over High Asia. Spatial loadings of (a) EOF2 of geopotential thickness of
the 500–200 hPa layer, (b) rotated EOF2 of vertically integrated specific humidity of the 500–300 hPa layer, (c) EOF1 of
zonal wind at 300 hPa, and (d) EOF1 of meridional wind at 300 hPa in JAS, 1948–2014. Explained total-field variance is given
in parentheses, while rshows the pattern’s time series correlation with the westerly index (Figure 1). The contour lines
depict the topography (equidistance 1,000 m), with the 2,000 m contour emphasized.
Mode
12345
Explained variance (%)
0
20
40
60
80
thickness
Mode
12345
0
10
20
30
40
50
humidity
Mode
12345
0
10
20
30
40
50
u-wind
Mode
12345
0
10
20
30
40
50
reference
extended
v-wind
Figure 4. Robustness of the EOFs. Explained variance of the five leading EOF modes for the same variables as in Figure 3.
The error bars are the result from the North et al. (1982) rule, considering autocorrelation. The black color applies to the
High Asia reference domain of Figure 3 (60–100°E; 25–45°N) and the blue color to an extended domain (plus 5° at every
margin).
Journal of Geophysical Research: Atmospheres 10.1002/2017JD027414
MÖLG ET AL. CIRCULATION INFLUENCES ON HIGH ASIA 12,706
thereby increases, and the pattern shows that W+ promotes drying in the southwest and moistening in the
northeast sector over High Asia (Figure 3b). This result aligns nicely with the interpretation from Figure 2b.
The key features of the EOF results are not sensitive to the delineation of the High Asia domain. Taking geo-
potential thickness as an example, expanding the domain by 5° at every margin increases the number of grid
points by 72%, yet the three leading EOF patterns remain stable. EOF2 continues to (i) explain a significant
part of the total variance (20%), (ii) strongly correlate with the westerly index (r= 0.80), and (iii) be unique
in North’s sense; further, its spatial pattern still embodies the characteristic dipole influence on the layer
temperature evident in Figure 3a. The other EOFs of Figure 3 are similarly insensitive to the domain delinea-
tion, as illustrated in Figure 4 by the presence of the same uniqueness or degeneracy features for the
modes computed using the extended domain (blue entries). The EOF analyses, therefore, corroborate that
the W+-related atmospheric anomalies in Figure 2 are a significant part of the total tropospheric variability
over High Asia from a statistical perspective.
The next essential question is to what extent are tropospheric anomalies manifested at the surface, where
proxy phenomena are sampled? A general hurdle to answering this question is the lack of spatially complete,
multidecadal surface data at an appropriate resolution (a few tens of kilometers or less); therefore, prior stu-
dies of tropospheric variability that made a link to High Asia’s surface climate typically had to use data on the
order of 10
2
km resolution (Bothe et al., 2011; Liu et al., 2015), which enable a basic assessment but face
inherent limitations regarding spatial details (Collier & Immerzeel, 2015; Curio et al., 2015). To detect a
Figure 5. Effects of southward-penetrating westerlies on the surface climate in boreal summer. Differences between W+ and Wcomposites in JAS, 1980–2014, in
the (left) MERRA and (right) WRF data sets for (a and b) 2 m air temperature, (c and d) surface precipitation, and (e and f) 2 m specific humidity. The stippled
areas denote significant differences at the FDR control level of 0.01 and the dashed bold line again the 2,000 m elevation contour; the thin grey lines show national
borders. Station data are included in the temperature and precipitation plots. Note that the station’s W+ and Wcomposites may consist of sample years outside of
the 1980–2014 period (see section 2.3 and Table S3).
Journal of Geophysical Research: Atmospheres 10.1002/2017JD027414
MÖLG ET AL. CIRCULATION INFLUENCES ON HIGH ASIA 12,707
robust W+ signature over at least the length of a climate normal period (≥30 years) and with regional to local
skill, we bring together the following suite of different data (section 2): the MERRA reanalysis with a relatively
high spatial resolution up to 0.5°, available meteorological station data, and results from the high-resolution
atmospheric modeling forced by the Era-Interim reanalysis. The synthesis of these data in Figure 5 yields one
salient feature, which is a dipole response of the surface-air temperature to W+ influence with strong cooling
in the west but warming in the northeast sector of High Asia (Figures 5a and 5b). The patterns of the preci-
pitation response are less cohesive, yet a general tendency toward wetter conditions is a robust outcome,
mainly in the east and northwest of High Asia (Figures 5c and 5b). This tendency coincides with a prominent
humidity increase in these regions, while reductions of air humidity only develop in the southwestern rim
mountains as a part of widespread drying in Pakistan and northwest India (Figures 5e and 5f). Taken together,
these results strongly suggest that (i) significant westerly related anomalies are not confined to the upper
troposphere and (ii) the anomaly characteristics of the troposphere are well expressed at the surface too.
We must finally address potential physical mechanisms that drive the tropospheric and surface response
patterns described above. Cooling in the west and drying along the southwestern margin of High Asia is most
certainly related to the large-scale advection properties. A cyclonic anomaly centered east of the Caspian Sea
(Figure 2c) seems to be a major factor, as has been emphasized in connection with surface cooling of
West Central-Asia in summer (Krishnan & Sugi, 2001). Moreover, the local surface energy budget changes
(Figure S3) do not exhibit obvious west-east dipoles, which also points to the influence of large-scale
advection. In this regard, the position of the upper-tropospheric westerly jet is key since the jet acts as
waveguide (Ding & Wang, 2005). During W+ conditions in JAS, a southward shift of the jet is indicated by
a significant correlation between the jet position index (see section 2) and the westerly index of Figure 1b
(r= 0.52; p<0.01). Figure 6 illustrates that this shift is 2°–4° on average during the summer months in the
jet entrance region over High Asia, which increases the mean zonal wind south of 40°N coherently. The
magnitude of the shift is significant considering the ~10° amplitude of the seasonal cycle (Schiemann
et al., 2009). The importance of the jet position was also emphasized for climate variability in neighboring
regions of High Asia (Zhao et al., 2014) and for East Asian monsoon seasonality (Chiang et al., 2015).
Nonetheless, our results demonstrate that jet variability cannot simply be associated with homogenous
climatic anomalies over High Asia, which is a consequence of the stationary wave structure of the westerlies.
In particular, tendencies for wet anomalies in the northwest and for wet/warm anomalies in regions of East
High Asia appear in addition.
Regarding the northwest of High Asia, an obvious influence concerns reduction of the mean zonal flow in this
region (Figures 3c and 6). A weaker upper-tropospheric flow is known to facilitate the development of regio-
nal convection, which is supported by observations (Mölg et al., 2009; Uyeda et al., 2001) and modeling
(Collier & Immerzeel, 2015). Cold-air advection in the upper troposphere (Figure 2a) likely contributes further,
since it tends to decrease the potential stability of the atmosphere. Signs of significantly enhanced convec-
tion indeed appear in the moister air over northwest High Asia: surface longwave radiation and evaporation
increase, while top-of-atmosphere outgoing longwave radiation decreases (Figure S3). If we turn to the wet
and warm anomalies in the eastern half (Figures 2, 3, and 5), there is one noteworthy feature. These anomalies
are only observed east of the anomalous cyclonic circulation “B”(Figure 2c). Consistent local developments
June
Speed (m s
-1
)
0102030
Latitude (°N)
30
35
40
45
50
July
Speed (m s
-1
)
0102030
August
Speed (m s
-1
)
0102030
September
Speed (m s
-1
)
0102030
Figure 6. Jet stream shift. Monthly meridional profiles (averaged over 75°–80°E) of zonal wind velocity at 200 hPa for every
month in the June–September season of a W+ year (light red) and Wyear (light grey) in the 1980–2014 period (MERRA
reanalysis). The thick red (dark gray) line is the mean W+ (W) profile.
Journal of Geophysical Research: Atmospheres 10.1002/2017JD027414
MÖLG ET AL. CIRCULATION INFLUENCES ON HIGH ASIA 12,708
include enhanced heating of the atmospheric surface layer due to tur-
bulence and longwave radiation and moistening as a result of intensi-
fied evaporation (Figure S3).
A closer look at the large-scale wave structure helps to frame the
mechanisms of the varying anomaly patterns. The meridional wind in
the upper troposphere demonstrates how the westerly wave stretches
all from the Mediterranean Sea to North China in W+ seasons, passing
over High Asia (Figure 7a). Yet in Wcases, the wave signatures end
west of High Asia (Figure 7b). Figure 7c further illustrates the dimin-
ished wave activity over High Asia during weak westerlies along the
35°N latitude circle (which runs through the cooling/warming zone of
West/East High Asia) and down to 500 hPa, hence near the surface of
High Asia. Associated with this feature, the geopotential thickness field
shows that the South Asian High retreats southward over West High
Asia and advances poleward north of 30°N in East High Asia (Figure S4);
this movement is consistent with the patterns in Figures 2a and 3a.
Interrelated shifts in the jet stream and the South Asian High also
emerged in another recent study as drivers of opposite climate anoma-
lies in regions of midlatitude Asia (Wei et al., 2017). Hence, contrasting
temperature anomalies in High Asia during W+ despite continuous
southerly flow anomalies (Figure 7c) can be explained plausibly by
these shifts as well. Southerlies contribute to cooling in the west where
their linkage to the strong zonal component of the anomalous cyclonic
circulation “B”(Figure 2c) occurs. Yet with increasing distance from the
center of “B”in the east, where the southerlies meet the expansion
zone of the South Asian High, temperature advection would promote
warming. Wei et al. (2017) also underline that warm-air advection from
the south characterizes the eastern flank of cyclonic circulation anoma-
lies (like “B”), which favors air ascent and precipitation. The tendency of
precipitation increases in East High Asia during W+, therefore is most
likely triggered by both the local moistening (Figure S3) and advective
processes including water transport (e.g., Curio et al., 2015).
The main anomaly patterns and mechanisms for JAS presented above
already appear in June, except that the surface-troposphere interac-
tions are most likely less developed (Text S3 and Figures S5 and S6).
Hence, climate anomalies at the surface are less distinct in June, and
capturing the potential westerly influence in the early monsoon season
by proxy data may seem less obvious. However, proxies do possess this potential if the May–June season
accounts for a significant fraction of their annual variability (Li et al., 2016; Mölg et al., 2014) and the under-
lying sensitivity shows a linkage to the upper-tropospheric climate (e.g., Mölg et al., 2014).
4. A Note on the Forcings of Westerly and ISM Influences
The results of our study could be construed as meaning the anomaly patterns in Figures 2, 3, and 5 are forced
by midlatitude climate dynamics. However, assuming that these patterns are entirely separate from the ISM
state is not possible, as westerlies and ISM are interactive systems (Ding & Wang, 2005; Krishnan et al., 2009;
Wei et al., 2017). A certain degree of interdependence is therefore unavoidable and manifested as statistical
covariability: the westerly index and the standard ISM index (section 2.1) correlate at r=0.59 (0.63) in JAS
(June) over 1948–2014 (both significant at the 1% level). This relation indicates that W+ coincides with a weak
ISM. In this context, the importance of cyclonic circulation anomalies in the upper troposphere during W+
(Figure 2c) was also shown for weak ISM phases (Krishnan et al., 2009). It is none of our goals to decompose
the factors of the westerlies-ISM interplay; we still provide a short synopsis of arguments below, suggesting
that a southward shift of the westerlies is unlikely to be a fully passive response to ISM forcing.
Figure 7. Tracks and strength of the westerly wave train. Mean meridional wind
in JAS, 1948–2014, for (a) W+ and (b) Wphases at 300 hPa and (c) along 35°N
for both phases and vertically integrated for the 500–300 hPa layer. As before,
the dashed bold line in the maps is the 2,000 m elevation contour, which roughly
delineates High Asia. In the bottom plot, the blue rectangle indicates the region
of tropospheric and surface cooling in West High Asia during W+, while the
red one signifies the region of warming in East High Asia (e.g., Figure 5, top).
Journal of Geophysical Research: Atmospheres 10.1002/2017JD027414
MÖLG ET AL. CIRCULATION INFLUENCES ON HIGH ASIA 12,709
(i) Ding and Wang (2005) already concluded that a clear driver role in
the interaction of westerlies and ISM cannot be established and that
the most promising approach is to think in terms of mutual feedbacks.
Later, the results of Krishnan et al. (2009) supported this viewpoint,
showing that anomalous cold-air advection by the westerlies
decreases ISM convection; this reduced convection feeds back into
the midlatitudes through Rossby wave dispersion, which reinforces
the anomalous midlatitude circulation pattern. (ii) Some relevant
mechanistic elements are clearly outside of the ISM influence.
This concerns the circulation anomaly center “A”in Figure 2c, which
is also evident at 850 hPa and therefore is of dynamic nature. Also,
the wave activity that originates from the North Atlantic and western
Europe can be considered external, since Rossby dispersion acts
eastward (Ding & Wang, 2005). And (iii) correlations of the major
EOF patterns of climate variability over High Asia with the ISM index
are not systematically higher than the correlations with the westerly
index (Table S1).
In synthesis, the surface climate anomalies during weak ISM should
therefore not be a mirror image of Figure 5. To test this notion, we
reproduced Figure 5, but this time, as the difference between weak
ISM composites (n= 10; 1982, 1985–1987, 1991, 2000–2002, 2004,
and 2009) and strong ISM composites (n= 8; 1983, 1988, 1990, 1994,
2005, 2006, 2007, and 2010). The +0.75 and 0.75 units of standard
deviation (cf. Figure 1) again serve as the criterion for selecting strong
and weak years, respectively. WRF data are not included, since the
simulations were designed for westerly phases (section 2.2) and do
not contain all relevant ISM years. The resulting patterns (Figure 8),
however, are unmistakably robust and clearly differ from Figure 5. In
particular, the westerly related dipole response in surface-air tempera-
ture does not evolve in case of the weak ISM signal (Figure 8a). Also,
regions of increased precipitation in High Asia are more localized and
confined to the southeast sector in the weak ISM signature (Figure 8b).
Air humidity anomalies, in turn, seem to be mostly negative (Figure 8c).
The greatest similarities in the response patterns of Figures 5 and 8
are outside of High Asia in the lowlands to the southwest. Although
not the focus of our study, the last point reminds us that responses
to large-scale circulation variability can differ dramatically between the lowlands and the high-altitude land-
scapes of High Asia (e.g., Vellore et al., 2014).
5. Concluding Remarks and Outlook
The results of our study substantiate the idea that the midlatitude westerly circulation leaves a distinct signa-
ture in the surface climate of High Asia during boreal summer. Merging three major and widely used reana-
lysis products, in situ observations, and atmospheric modeling increases the confidence in and robustness of
the present findings. What stands out is the heterogeneous influence of strong westerly phases, in particular
cold anomalies in the west and regionally confined warm/wet anomalies in the east of High Asia. It is beyond
the scope of the paper to attribute the associated meridional shift of the upper-tropospheric jet to individual
drivers in the ISM-westerly interaction and its teleconnections (Ding & Wang, 2005). As discussed above, how-
ever, the westerlies most likely contribute actively to the interplay of the two circulations, which also emerges
from the large-scale dynamics viewpoint (e.g., Ding & Wang, 2005; Krishnan et al., 2009).
Our study extends on previous dynamical research, by solidifying a link between the large-scale westerlies
and regional-to-local climate variability in High Asia. Moreover, these results provide perspectives for future
research on climate proxies. First, the demonstrated role of the westerlies creates opportunities for explaining
Figure 8. Effects of weak monsoon on the surface climate during boreal
summer. Differences between weak ISM and strong ISM composites in JAS,
1980–2014, in the MERRA data set for (a) 2 m air temperature, (b) surface
precipitation, and (c) 2 m specific humidity. The stippled areas denote significant
differences at the FDR control level of 0.01 and the dashed bold line the 2,000 m
elevation contour; the thin grey lines show national borders. Station data are
included in the temperature and precipitation plots (as in Figure 5, the station
composites may consist of sample years outside 1980–2014).
Journal of Geophysical Research: Atmospheres 10.1002/2017JD027414
MÖLG ET AL. CIRCULATION INFLUENCES ON HIGH ASIA 12,710
additional variance in recent climate proxy properties; in this vein, a westerly related predictor in addition to
ISM indices should also enhance the identification of atmospheric signals. Second, and regarding the
paleoclimate side, reconstructed patterns of summertime moisture and temperature (e.g., Wang et al.,
2016) could be used to infer past characteristics of the combined ISM-westerly influence on High Asia. This
relevance exists since dominant high-frequency variability in the climate system is prone to emerge in the
patterns of lower-frequency variability (Krishnan et al., 2009) or long-term changes (Vecchi and Soden,
2007) as well. The maps we present (Figure 5) and the associated inclusion of a well-founded index of the
westerly waves (Ding & Wang, 2005) provide a useful basis for both avenues.
It is clear that linking our results to proxy data would be a worthwhile next step for future studies. Research
along this line will also help to excite further ideas and efforts to learn about the interaction of two major
circulation systems on the planet, by integrating climate dynamics knowledge and proxy data. This integra-
tion has been emphasized before as a pivotal step forward in climate research (Trenberth & Otto-Bliesner,
2003) and will benefit both disciplines.
References
Benn, D., & Owen, L. A. (1998). The role of the Indian summer monsoon and the mid-latitude westerlies in Himalayan glaciation: Review and
speculative discussion. Journal of the Geological Society of London,155(2), 353–363. https://doi.org/10.1144/gsjgs.155.2.0353
Bolch, T., Kulkarni, A., Kääb, A., Huggel, C., Paul, F., Cogley, J. G., …Stoffel, M. (2012). The state and fate of Himalayan glaciers. Science,336,
310–314. https://doi.org/10.1126/science.1215828
Bosilovich, M. G., Lucchesi, R., & Suarez, M. (2015). MERRA-2: File specification, GMAO Office Note No. 9 (Version1.0).
Bothe, O., Fraedrich, K., & Zhu, X. (2011). Large-scale circulations and Tibetan Plateau summer drought and wetness in a high-resolution
climate model. International Journal of Climatology,31, 832–846. https://doi.org/10.1002/joc.2124
Chiang, J. C. H., Fung, I. Y., Wu, C. H., Cai, Y., Edman, J. P., Liu, Y., …Labrousse, C. A. (2015). Role of seasonal transitions and westerly jets in East
Asian paleoclimate. Quaternary Science Reviews,108, 111–129. https://doi.org/10.1016/j.quascirev.2014.11.009
Collier, E., & Immerzeel, W. W. (2015). High-resolution modeling of atmospheric dynamics in the Nepalese Himalaya. Journal of Geophysical
Research: Atmospheres,120, 9882–9896. https://doi.org/10.1002/2015JD023266
Curio, J., Maussion, F., & Scherer, D. (2015). A 12-year high-resolution climatology of atmospheric water transport over the Tibetan Plateau.
Earth System Dynamics,6(1), 109–124. https://doi.org/10.5194/esd-6-109-2015
Dee, D. P., Uppala, S. M., Simmons, A. J., Berrisford, P., Poli, P., Kobayashi, S., …Vitart, F. (2011). The ERA-Interim reanalysis: Configuration and
performance of the data assimilation system. Quarterly Journal of the Royal Meteorological Society,137, 553–597. https://doi.org/10.1002/
qj.828
Ding, Q., & Wang, B. (2005). Circumglobal teleconnection in the Northern Hemisphere summer. Journal of Climate,18(17), 3483–3505. https://
doi.org/10.1175/JCLI3473.1
Grießinger, J., Bräuning, A., Helle, G., Hochreuther, P., & Schleser, G. (2016). Late Holocene relative humidity history on the southeastern
Tibetan Plateau inferred from a tree-ring δ
18
O record: Recent decrease and conditions during the last 1500 years. Quaternary
International,430,52–59. https://doi.org/10.1016/j.quaint.2016.02.011
Hochreuther, P., Wernicke, J., Grießinger, J., Mölg, T., Zhu, H., Liang, E., …Bräuning, A. (2016). Influence of the Indian Ocean Dipole on
tree-ring δ18O of monsoonal southeast Tibet. Climatic Change,137(1-2), 217–230. https://doi.org/10.1007/s10584-016-1663-8
Joswiak, D. R., Yao, T., Wu, G., Tian, L., & Xu, B. (2013). Ice-core evidence of westerly and monsoon moisture contributions in the central
Tibetan Plateau. Journal of Glaciology,59(213), 56–66. https://doi.org/10.3189/2013JoG12J035
Kalnay, E., Kanamitsu, M., Kistler, R., Collins, W., Deaven, D., Gandin, L., …Joseph, D. (1996). The NCEP/NCAR 40-year reanalysis project. Bulletin
of the American Meteorological Society,77(3), 437–471. https://doi.org/10.1175/1520-0477(1996)077%3C0437:TNYRP%3E2.0.CO;2
Krishnan, R., Kumar, V., Sugi, M., & Yoshimura, J. (2009). Internal feedbacks from monsoon-midlatitude interactions during droughts in the
Indian summer monsoon. Journal of the Atmospheric Sciences,66, 553–578. https://doi.org/10.1175/2008JAS2723.1
Krishnan, R., & Sugi, M. (2001). Baiu rainfall variability and associated monsoon teleconnections. Journal of the Meteorological Society of Japan,
79(3), 851–860. https://doi.org/10.2151/jmsj.79.851
Kropáček, J., Braun, A., Kang, S., Feng, C., Ye, Q., & Hochschild, V. (2012). Analysis of lake level changes in Nam Co in central Tibet utilizing
synergistic satellite altimetry and optical imagery. International Journal of Applied Earth Observation,17,3–11. https://doi.org/10.1016/
j.jag.2011.10.001
Li, R., Luo, T., Mölg, T., Zhao, J., Li, X., Cui, X., …Tang, Y. (2016). Leaf unfolding of Tibetan alpine meadows captures the arrival of monsoon
rainfall. Scientific Reports,6(1), 20,985. https://doi.org/10.1038/srep20985
Liang, H., Lyu, L., & Wahab, M. (2016). A 382-year reconstruction of August mean minimum temperature from tree-ring maximum
latewood density on the southeastern Tibetan Plateau, China. Dendrochronologia,37,1–8. https://doi.org/10.1016/j.dendro.
2015.11.001
Liu, H., Duan, K., Li, M., Shi, P., Yang, J., Zhang, X., & Sun, J. (2015). Impact of the North Atlantic Oscillation on the dipole oscillation of summer
precipitation over the central and eastern Tibetan Plateau. International Journal of Climatology,35(15), 4539–4546. https://doi.org/
10.1002/joc.4304
Maussion, F., Scherer, D., Mölg, T., Collier, E., Curio, J., & Finkelnburg, R. (2014). Precipitation seasonality and variability over the Tibetan
Plateau as resolved by the High Asia reanalysis. Journal of Climate,27(5), 1910–1927. https://doi.org/10.1175/JCLI-D-13-00282.1
Mölg, T., Chiang, J. C. H., Gohm, A., & Cullen, N. J. (2009). Temporal precipitation variability versus altitude on a tropical high mountain:
Observations and mesoscale atmospheric modeling. Quarterly Journal of the Royal Meteorological Society,135, 1439–1455. https://doi.org/
10.1002/qj.461
Mölg, T., Maussion, F., & Scherer, D. (2014). Mid-latitude westerlies as a driver of glacier variability in monsoonal High Asia. Nature Climate
Change,4(1), 68–73. https://doi.org/10.1038/NCLIMATE2055
Mölg, T., Maussion, F., Yang, W., & Scherer, D. (2012). The footprint of Asian monsoon dynamics in the mass and energy balance of a Tibetan
glacier. The Cryosphere,6, 1445–1461. https://doi.org/10.5194/tc-6-1445-2012
Journal of Geophysical Research: Atmospheres 10.1002/2017JD027414
MÖLG ET AL. CIRCULATION INFLUENCES ON HIGH ASIA 12,711
Acknowledgments
This work was supported by the German
Research Foundation (DFG) grant MO
2869/1-1 and the National Science
Foundation grant AGS-1405479. We
thank Yanhong Gao for the discussion
on WRF data for Tibet and Kai Uwe Eiselt
and Elena Kropačfor assisting with
the station data analysis. The WRF
modeling was conducted at the High-
Performance Computing Center (HPC)
at the University of Erlangen-Nürnberg’s
Regional Computation Center (RRZE).
The resultant data are only available
upon request due to the very large data
size. All other employed data sets are
freely available (see section 2) and were
obtained from the Earth System
Research Laboratory (ESRL), National
Aeronautics and Space Agency (NASA),
European Centre for Medium-Range
Forecast (ECMWF), and Dutch
Meteorological Office (KNMI) platforms.
The constructive comments of three
anonymous reviewers improved
the paper.
North, G. R., Bell, T. L., Cahalan, R. F., & Moeng, F. J. (1982). Sampling errors in the estimation of empirical orthogonal functions. Monthly
Weather Review,110(7), 699–706. https://doi.org/10.1175/1520-0493(1982)110%3C0699:SEITEO%3E2.0.CO;2
Ramaswamy, C. (1962). Breaks in the Indian summer monsoon as a phenomenon of interaction between the easterly and the sub-tropical
westerly jet streams. Tellus,14(3), 337–349.
Saeed, S., Müller, W. A., Hagemann, S., & Jacob, D. (2011). Circumglobal wave train and the summer monsoon over northwestern India and
Pakistan: The explicit role of the surface heat low. Climate Dynamics,37, 1045–1060. https://doi.org/10.1007/s00382-010-0888-x
Schiemann, R., Lüthi, D., & Schär, C. (2009). Seasonality and interannual variability of the westerly jet in the Tibetan Plateau region. Journal of
Climate,22(11), 2940–2957. https://doi.org/10.1175/2008JCLI2625.1
Skamarock, W. C., & Klemp, J. B. (2008). A time-2split nonhydrostatic atmospheric model for weather research and forecasting applications.
Journal of Computational Physics,227(7), 3465–3485. https://doi.org/10.1016/j.jcp.2007.01.037
Trenberth, K. E., & Otto-Bliesner, B. L. (2003). Toward integrated reconstruction of past climates. Science,300(5619), 589–590. https://doi.org/
10.1126/science.1083122
Uyeda, H., Yamada, H., Horikomi, J., Shirooka, R., Shimizu, S., Liping, L., …Koike, T. (2001). Characteristics of convective clouds observed by a
Doppler radar at Naqu on Tibetan Plateau during the GAME-Tibet IOP. Journal of the Meteorological Society of Japan,79(1B), 463–474.
https://doi.org/10.2151/jmsj.79.463
Vecchi, G. A., & Soden, B. J. (2007). Global warming and the weakening of the tropical circulation. Journal of Climate,20(17), 4316–4340.
https://doi.org/10.1175/JCLI4258.1
Vellore, R. K., Krishnan, R., Pendharkar, J., Choudhury, A. D., & Sabin, T. P. (2014). On the anomalous precipitation enhancement over the
Himalayan foothills during monsoon breaks. Climate Dynamics,43(7-8), 2009–2031. https://doi.org/10.1007/s00382-013-2024-1
Wang, M., Liang, J., Hou, J., & Hu, L. (2016). Distribution of GDGTs in lake surface sediments on the Tibetan Plateau and its influencing factors.
Science China Earth Sciences,59(5), 961–974. https://doi.org/10.1007/s11430-015-5214-3
Ward, M. N. (1999). Analyzing the boreal summer relationship between worldwide sea-surface temperature and atmospheric variability. In
H. Von Storch & A. Navarra (Eds.), Analysis of Climate Variability (pp. 95–118). Berlin, Heidelberg and New York: Springer. https://doi.org/
10.1007/978-3-662-03744-7_6
Wei, W., Zhang, R., Wen, M., & Yang, S. (2017). Relationship between the Asian westerly jet stream and summer rainfall over Central Asia and
North China: Roles of the Indian monsoon and the South Asian High. Journal of Climate,30(2), 537–552. https://doi.org/10.1175/JCLI-D-15-
0814.1
Wilks, D. S. (2016). “The stippling shows statistically significant grid points”. How research results are routinely overstated and
overinterpreted, and what to do about it. Bulletin of the American Meteorological Society,97(12), 2263–2273. https://doi.org/10.1175/
BAMS-D-15-00267.1
Yao, T., Thompson, L. G., Yang, W., Yu, W., Gao, Y., Guo, X., …Joswiak, D. (2012). Different glacier status with atmospheric circulations in
Tibetan Plateau and surroundings. Nature Climate Change,2, 663–667. https://doi.org/10.1038/NCLIMATE1580
Zhang, L., Su, F., Yang, D., Hao, Z., & Tong, K. (2013). Discharge regime and simulation for the upstream of major rivers over Tibetan Plateau.
Journal of Geophysical Research: Atmospheres,118, 8500–8518. https://doi.org/10.1002/jgrd.50665
Zhang, Y., Zhang, J., & Dong, X. X. (2013). Green-up dates in the Tibetan Plateau have continuously advanced from 1982 to 2011. Proceedings
of the National Academy of Sciences of the United States of America,110(11), 4309–4314. https://doi.org/10.1073/pnas.1210423110
Zhao, Y., Huang, A., Zhou, Y., Huang, D., Yang, Q., Ma, Y., …Wei, G. (2014). Impact of the middle and upper tropospheric cooling over central
Asia on the summer rainfall in the Tarim Basin, China. Journal of Climate,27(12), 4721–4732. https://doi.org/10.1175/JCLI-D-13-00456.1
Zhu, H. F., Shao, X. M., Yin, Z. Y., Xu, P., Xu, Y., & Tian, H. (2011). August temperature variability in the southeastern Tibetan Plateau since
AD 1385 inferred from tree rings. Palaeogeography Palaeoclimatology Palaeoecology,305,84–92. https://doi.org/10.1016/
j.palaeo.2011.02.017
Journal of Geophysical Research: Atmospheres 10.1002/2017JD027414
MÖLG ET AL. CIRCULATION INFLUENCES ON HIGH ASIA 12,712
