Decadal Transition of the Leading Mode of Interannual Moisture Circulation overEast Asia–Western North Pacific: Bonding to Different Evolution of ENSO

XIUZHEN LI

Center for Monsoon and Environment Research and Department of Atmospheric Sciences, Guangdong Province Key

Laboratory for Climate Change and Natural Disaster Studies, Sun Yat-sen University, Guangzhou, and Jiangsu

Collaborative Innovation Center for Climate Change, Nanjing, China

ZHIPING WEN

Institute of Atmospheric Sciences, Fudan University, Shanghai, and Center for Monsoon and Environment Research

and School of Atmospheric Sciences, Sun Yat-sen University, Guangzhou, and Jiangsu Collaborative Innovation

Center for Climate Change, Nanjing, China

DELIANG CHEN

Regional Climate Group, Department of Earth Science, University of Gothenburg, Gothenburg, Sweden

ZESHENG CHEN

State Key Laboratory of Tropical Oceanography, South China Sea Institute of Oceanology,

Chinese Academy of Sciences, Guangzhou, China

(Manuscript received 6 June 2018, in final form 26 October 2018)

ABSTRACT

The El Niño–Southern Oscillation (ENSO) cycle has a great impact on the summer moisture circulation

over East Asia (EA) and the western North Pacific [WNP (EA-WNP)] on an interannual time scale, and its

modulation is mainly embedded in the leadingmode. In contrast to the stable influence of themature phase of

ENSO, the impact of synchronous eastern Pacific sea surface temperature anomalies (SSTAs) on summer

moisture circulation is negligible during the 1970s–80s, while it intensifies after 1991. In response, the in-

terannual variation of moisture circulation exhibits a muchmore widespread anticyclonic/cyclonic pattern over

the subtropical WNP and a weaker counterpart to the north after 1991. Abnormal moisture moves farther

northwardwith the enhancedmoisture convergence, and thus precipitation shifts from theYangtze River to the

Huai River valley. The decadal shift in the modulation of ENSO on moisture circulation arises from a more

rapid evolution of the bonding ENSO cycle and its stronger coupling with circulation over the Indian Ocean

after 1991. The rapid development of cooling SSTAs over the central-eastern Pacific, andwarming SSTAs to the

west over the eastern Indian Ocean–Maritime Continent (EIO-MC) in summer, stimulates abnormal de-

scending motion over the western-central Pacific and ascending motion over the EIO-MC. The former excites

an anticyclone over theWNP as a Rossby wave response, sustaining and intensifying theWNP anticyclone; the

latter helps anchor the anticyclone over the tropical–subtropical WNP via an abnormal southwest–northeast

vertical circulation between EIO-MC and WNP.

1. Introduction

Summer precipitation over East Asia is extremely

complex. In the past few decades, summer precipitation

over East China has shown a dramatic interdecadal

variation accompanied by a weak East Asian summer

monsoon, with an anomalous rain belt retreating from

North and northeastern China to the Yangtze River

valley in the late 1970s and to South China in the early

1990s (Ding et al. 2008, 2009; Gong andHo 2003; Li et al.

2011). After the late 1990s, the mei-yu rain belt tends

to move northward again from south of the Yangtze

River valley to the Huai River valley (Liu et al. 2012; Si

et al. 2009). Concurrent with this decadal shift of the rain

belt, an interdecadal transition in the leading mode ofCorresponding author: Dr. Xiuzhen Li, lixiuzhen@mail.sysu.

edu.cn

15 JANUARY 2019 L I E T AL . 289

DOI: 10.1175/JCLI-D-18-0356.1

� 2018 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS CopyrightPolicy (www.ametsoc.org/PUBSReuseLicenses).

precipitation also occurs, shifting from a meridional

‘‘2 1 2’’ structure to a dipole ‘‘1 2’’ pattern. This

indicates a possible transition in the precipitation regime

during the last few decades (Ding et al. 2008; Sun and

Wang 2015; Si and Ding 2013). Although the decadal

southward shift of the rain belt has been intensively in-

vestigated and attributed mainly to the weakness of the

Asian monsoon, the cause of the transition of the in-

terannual mode of precipitation has not been well

addressed.

Precipitation variation over East Asia is closely as-

sociated with moisture circulation over East Asia (EA)

and the western North Pacific [WNP (EA-WNP); Li

et al. 2011, 2012; Sun and Wang 2015], which is influ-

enced jointly by three Asian summer monsoon subsys-

tems and the western Pacific subtropical high (WPSH;

e.g., Chen et al. 1992; Murakami and Matsumoto 1994;

Ueda and Yasunari 1996). By performing an EOF anal-

ysis based on vertically integrated water vapor flux

anomalies in summer, Li and Zhou (2012) found that

variation in moisture circulation over EA-WNP is char-

acterized by an ‘‘anticyclone–cyclone’’ dipole structure,

which is responsible for the abnormal elongated con-

vergence from the Yangtze River valley to southern

Japan. This illustrates the vigorous variation of the

mei-yu–baiu rain belt. A recent study by Sun and Wang

(2015) shows that this anticyclone–cyclone dipole-

structured moisture circulation is replaced by an an-

ticyclonic monopole structure with a much larger

spatial domain over the WNP after 2000, as the main

rain belt centered along the Yangtze River valley shifts

northward to the region between the Yangtze and

Yellow Rivers. Thus, concurrent with the interdecadal

transition in the leading mode of summer precipitation,

the moisture circulation pattern over EA-WNP is seen

to change.

Many external factors, including El Niño–SouthernOscillation (ENSO), tropical Indian Ocean SST, snow

cover/depth on the Tibetan Plateau, Arctic sea ice,

Arctic Oscillation (AO), and Antarctic Oscillation, are

known to have impacts on the variability of circulation

over East Asia (Wang et al. 2000; Xie et al. 2009;Wu and

Qian 2003; Wu and Kirtman 2007; Duan et al. 2013;

Huang et al. 1995; Gong and Ho 2003; Nan et al. 2009),

and the impact of ENSO is emphasized on interannual

time scale (e.g., Huang and Wu 1989; Chang et al. 2000;

Wang et al. 2000, 2001; Chen 2002; Wu and Wang 2002;

Wu et al. 2003; Li et al. 2014, 2015). In the case of El

Niño, an anomalous cyclone is triggered in the lower

troposphere over theWNP in the developing summer by

enhanced convection over the tropical central Pacific via

the Gill–Matsuno mechanism (Gill 1980; Matsuno 1966;

Wu et al. 2003), while an abnormal anticyclone is formed

[western North Pacific anticyclone (WNPAC)] in the

decaying summer through the Indian Ocean capacitor

mechanism (Xie et al. 2009) and local air–sea interaction

(Wang et al. 2000; Wang and Zhang 2002) when the

positive sea surface temperature anomalies (SSTAs) in

the tropical central-eastern Pacific diminish (Wang et al.

2003; Wang and Wu 2012). The anticyclone–cyclone

dipole-structured moisture circulation tends to appear

over EA-WNP in the transitional summer between a

decaying El Niño and a developing La Niña (Li and

Zhou 2012). Moreover, recent studies have proposed

that the speed at which El Niño transforms to La Niñacould modulate the summer circulation over East Asia

(Fan et al. 2013; Chen et al. 2016).When it shifts quickly,

as cooling develops over the central-eastern Pacific from

spring to summer, it starts to influence the mainte-

nance and intensification of the WNPAC in summer

(Chen et al. 2016).

The modulation of these external signals may change

on decadal time scales. In the past few decades, a shift

in the relationship between ENSO and the East Asian

summer monsoon and thus China rainfall has been de-

tected (Wu and Wang 2002). As for the cause, Wu and

Wang (2002) emphasize the role of variability of the

summer SSTAs over the Indian Ocean and WNP.

During the 1970s–90s, when summer SST increased over

the Philippine Sea, convection over theWNP intensified

and moved northward, and the cyclonic anomaly over

Japan shifted eastward. This resulted in an anticyclone–

cyclone dipole-structured anomalous moisture circula-

tion. After the 1990s, the interannual variability of

tropical SSTs over both the Indian and Pacific Oceans

diminishes and fails to excite the dipole-structured

anomalous moisture circulation (Sun and Wang 2015).

Modulation of the relationship by the Pacific decadal

oscillation (PDO) is also proposed (Feng et al. 2014).

However, only the synchronous summer SSTAs are

emphasized in their studies (Wu and Wang 2002; Sun

and Wang 2015).

Questions remain regarding how the interannual

variations of moisture circulation and thus precipitation

have changed in the past few decades and whether the

ENSO signal has played a role. If so, what characteristics

of ENSO are responsible for such variations, the am-

plitude, evolution, or its coherence with the SSTAs over

other regions? This paper addresses these questions.

Sections are organized as follows: The data used in this

study are introduced in section 2. The decadal shift of

the interannual variation of moisture circulation mod-

ulated by SSTAs is investigated in section 3. The role

played by the ENSO signal, especially its evolution

speed, is examined in detail in section 4. The underlying

physical processes responsible for the decadal change in

290 JOURNAL OF CL IMATE VOLUME 32

the modulation of the ENSO revolution on moisture

circulation over EA-WNP are studied in section 5, and a

discussion and conclusion are provided in section 6.

2. Data and methodology

Daily precipitation data from 756 stations in China,

collected and subjected to quality-control procedures by

the China Meteorological Administration (Bao 2007),

are employed. The key dataset applied in this study is

the Japanese 55-year Reanalysis (JRA-55), the second

global reanalysis constructed by the Japan Meteoro-

logical Agency (JMA; Kobayashi et al. 2015; Harada

et al. 2016). It is the first to apply four-dimensional

variational analysis in the last half-century. Its main

objectives are to address issues found in the previous

reanalysis and to produce a comprehensive atmospheric

dataset suitable for studying multidecadal variability

and climate change (Kobayashi et al. 2015; Harada et al.

2016). Improvements are found in JRA-55 in its repre-

sentations of the phenomena on a wide range of space–

time scales, such as equatorial waves and transient

eddies in storm track regions (Kobayashi and Iwasaki

2016). Moreover, temporal consistency is improved in

JRA-55 compared with the older reanalysis (JRA-25)

throughout the reanalysis period. It is similar to ERA-40

in its spatial patterns, as well as in the magnitude of its

moisture convergence and divergence, which are com-

parable to those of the Special Sensor Microwave Im-

ager (SSM/I; Park et al. 2007). It covers the period

starting in 1958 when regular radiosonde observations

begin on a global basis. The variables employed are

monthly wind fields, vertical velocity, and vertically in-

tegrated water vapor flux. The monthly Hadley Centre

Sea Ice and Sea Surface Temperature dataset (HadISST;

Rayner et al. 2003), with a resolution of 18 latitude 318 longitude, is also included to study the variation in

SSTAs and their impact on the moisture circulation. The

period of all datasets spans 1961–2012. Themature winter

(WIN) and decaying spring (SPR) and summer (SUM) of

ENSO are represented as WIN(0), SPR(11), SUM(11)

for short.

TheCommunityAtmosphereModel, version 4 (CAM4),

the atmospheric component of the Community Climate

System Model, version 4 (CCSM4; Gent et al. 2011), de-

veloped with significant community collaboration at the

National Center for Atmospheric Research (Neale et al.

2013), is employed in this study to explore the atmospheric

response over EA-WNP to the SSTA evolution over

the tropical eastern Pacific. It reasonably captures the

spatial features of the summer climate in observations,

though with some infidelity (e.g., Neale et al. 2013;

Chen et al. 2014). Bymodifyingdeep-convectionprocesses,

CAM4 reduces many of the model biases, resulting in

significant improvement in the Hadley circulation. It

also enhances estimations of the phase, amplitude, and

spatial anomaly patterns of the modeled El Niño. Amore comprehensive and detailed description about

CAM4 can be found in Neale et al. (2010). In this study,

the CAM4 model uses a finite-volume dynamic core

and is run for a period of 20 years at a horizontal res-

olution roughly equivalent to 1.98 latitude 3 1.258longitude, with 26 vertical levels in a hybrid sigma–

pressure coordinate system extending from the surface

to approximately 3.5 hPa.

3. Decadal shift of the interannual variation ofmoisture circulation

a. Leading mode of moisture circulation and itsrelationship with ENSO

EOF analysis is applied to the summer vertically in-

tegrated water vapor flux during 1962–2013 to examine

the interannual variation of moisture circulation over

EA-WNP. The leading mode, which explains 36.8% of

the total variance, nearly triple that of the second mode,

is depicted in Fig. 1 together with the principal compo-

nent (PC1). An anticyclone–cyclone dipole-structured pat-

tern features the moisture circulation over EA-WNP. In

positive-phase years, a widespread anticyclonic moisture

circulation prevails over the subtropical WNP; abnormal

moisture diverging over the tropical–subtropical WNP is

advected by the peripheral flow of the anticyclone to

East China. To the south, abnormal easterly moisture

transport dominates the tropical western Pacific,

implying a weaker-than-normal climatological west-

erly moisture transport from the Indian Ocean. To the

north, a cyclonic moisture circulation controls the mid-

latitudes, with abnormal northerly transport west to Japan.

In comparisonwith its anticyclonic counterpart, the cyclonic

moisture circulation is much weaker in both intensity and

spatial extent. In this mode, the abnormal moisture trans-

port over East Asia converges at around 308N over the

Yangtze River valley instead of moving northward. This

leading mode represents variations mainly on an in-

terannual time scale. The interannual component of the

corresponding PC1 explains a majority of the variance

(.80%), together with a weak interdecadal variation and a

nearly negligible long-term trend.With this consideration in

mind, only the interannual component is investigated in this

study. All variables are filtered by applying a 9-yr high-pass

Lanczos filter before further analysis.

This leading mode of moisture circulation was pro-

posed in our previous studies (Li and Zhou 2012) to be

modulated by the ENSO signal to a great extent, and its

positive phase was apparent in the transitional summer

15 JANUARY 2019 L I E T AL . 291

from an El Niño to a La Niña event. To examine

whether such modulation is stable over the past few

decades, a 13-yr running correlation between PC1 and

the Niño-3.4 index is examined (Fig. 2). The Niño-3.4index is calculated based on SSTAs in the previous

winter, the synchronous summer, and the following

winter to observe the possible impact of ENSO evolu-

tion. Several dramatic features are evident in Fig. 2.

First, the PC1 of the moisture circulation over EA-WNP

is positively correlated with the prewinter Niño-3.4 in-

dex and negatively with the postwinter Niño-3.4 index.

This supports the preference of the anticyclone–cyclone

dipole-structuredmoisture circulation in the transitional

summer between an El Niño and a La Niña. Second, andinterestingly, the influence of eastern Pacific SSTAs in

the synchronous summer shows a strong decadal varia-

tion. The correlation between PC1 and the synchronous

Niño-3.4 index is weak and insignificant during the

1970s–80s, while it intensifies and remains strongly

negative after 1991, with a correlation coefficient even

higher than that of the prewinter Niño-3.4 index. The

causes of such a decadal shift in this relationship are still

unknown. It has been widely accepted that the direct

influence of eastern Pacific summer SSTAs on the cli-

mate over East Asia is weak as the ENSO signal di-

minishes. However, a recent study by Chen et al. (2016)

argues that because an El Niño event can transition into

either a LaNiña event or a neutral or persistent warming

condition in the succeeding summer, composite analysis

by previous studies might conceal the impact of syn-

chronous SSTAs over the eastern Pacific. The rapid

development of synchronous cooling over the central-

eastern Pacificmay sustain theWNPACby stimulating a

Rossby wave response to its northwest. A similar en-

hanced decadal relationship also appears between PC1

and the postwinter Niño-3.4 index. In sum, compared

to the stable influence from the previous winter, the

influence of synchronous eastern Pacific SSTAs on

summer moisture circulation over EA-WNP intensifies

after 1991.

FIG. 1. (a) The regressed vertically integrated summer water vapor fluxes (vectors;

kgm21 s21) and their divergence (shading; 1025 kgm22 s21) based on the standardized PC1 of

the vertically integrated summer water vapor fluxes over EA-WNP (58–508N, 908–1508E)during 1962–2013. (b) Time series of the PC1 (blue column). The 9-yr running mean of PC1

(black line) and the remaining interannual component (orange line) are superimposed.

292 JOURNAL OF CL IMATE VOLUME 32

b. Shift in the interannual coherence of moisturecirculation and ENSO

As themodulation of synchronous tropical SSTAs over

the eastern Pacific changes, decadal variation might also

appear in the corresponding moisture circulation. To

address this question, the interannual variation of the

moisture circulation and its coupling with tropical SSTAs

are examined for the periods of 1971–90 and 1991–2013,

when the correlation coefficient between the PC1 of

moisture circulation and the synchronous Niño-3.4 indexis insignificant and significantly negative, respectively.

The coupling between the ENSO signal and the mois-

ture circulation over the EA-WNP may be embedded in

not just the first leading mode. A singular value de-

composition (SVD) is applied to the summer moisture

circulation over the EA-WNP and the tropical SSTAs in

the previous winter (Fig. 3) in order to address this issue.

The fields ofmoisture circulation (Quy) and SSTAs in the

first SVD mode (SVD1) are referred to as SVD1(Quy)

and SVD1(SST), respectively. The squared covariance

fractions (SCF) accounted for by SVD1, the correlation

coefficients of the corresponding expansion coefficients

(ECs) between Quy and SSTAs (EC_R), and the spatial

correlation of Quy/SST between EOF1 and SVD1

(Spa_REOF,SVD_Quy/Spa_REOF,SVD_SST) are exhibited

inTable 1. It is interesting to note that the SCFs for SVD1

are quite high during the two periods (95.3% and 85.5%).

In addition, the spatial patterns of the outputs from SVD

show high similarities with that of EOF (correlation

coefficients . 0.7), with a widespread anticyclonic mois-

ture circulation lying over the EA-WNP and an El Niñosignal dominating the winter SSTAs (Fig. 3). These re-

sults illustrate that the ENSO signal exerts a strong in-

fluence on moisture circulation over the EA-WNP, and

its modulation is embedded mainly in the leading mode

of the moisture circulation. In other words, interannual

variation of summer moisture circulation over the

EA-WNP is dominated by the ENSO signal.

The SVD1(SST)s in the two periods are characterized

by a typical El Niño pattern with maximum warming

over the eastern Pacific, cooling of the ‘‘U shape’’ to the

west, basin-wide warming over the Indian Ocean, and

a zonal ‘‘warm–cool’’ contrast over the South China

Sea (SCS) and WNP (Figs. 3d,e). In contrast to the

small decadal variation in the prewinter SSTA pattern

(Fig. 3f), a decadal shift occurs in the leading mode of

moisture circulation over EA-WNP (Fig. 3c). In the

positive phase, dramatic discrepancies can be seen in the

location, spatial extent of the anticyclonic moisture cir-

culation over the subtropical EA-WNP, and its con-

currence with the cyclonic moisture circulation to the

north over Japan. Before 1990, the abnormal anticy-

clonic moisture circulation over the subtropical WNP is

much smaller in its meridional extent, and the cyclonic

moisture circulation to the north is strong. Abnormal

northerly transport to the west of the cyclonic anomaly

dominates the midlatitude WNP, and southerly trans-

port anomaly to the west of the anticyclonic anomaly

turns eastward at around 308N instead of penetrating

further. This results in abnormal moisture convergence

over South China, the Yangtze River valley, and the

region south to Japan, whereas abnormal moisture di-

vergence occurs to the north over the Korean Peninsula

and Japan (Fig. 3a). As a result, the corresponding

precipitation anomalies are enhanced mainly over the

Yangtze River valley and the western part of south-

eastern China (Fig. 4a). The correlation between the

corresponding expansion coefficient of the SVD1(Quy)

(EC1_Quy) and the Yangtze River valley summer pre-

cipitation during 1971–90 reaches 0.66, significant at the

99.9% confidence level (Fig. 4c). Hence, the ENSO

signal in the previous winter exerted a great influence

on precipitation over the Yangtze River valley during

1971–90 by modulating moisture circulation.

After 1991, the anticyclonic moisture circulation over

the subtropical WNP shows a larger meridional expan-

sion compared to a zonally elongated shape during

1971–90, while the cyclonic moisture circulation to the

north is weak. As a result, moisture from the lower lat-

itudes invades northward, and the stronger moisture

convergence shifts northward to the Huai River valley,

extending northeastward from the upper reaches of

the Yangtze River valley to the Korean Peninsula and

Japan, with enhanced moisture divergence to the south

over South China (Fig. 3b). In response to this abnor-

mal moisture circulation pattern, the precipitation

intensifies and shifts northward from the Yangtze

River valley to the Huai River valley and weakens over

most parts of South China (Fig. 4b). The correlation

between EC1_Quy and summer precipitation increases

from 0.13 to 0.51 (significant at the 95% confidence

FIG. 2. The 13-yr running correlation between the interannual

components of PC1 and Niño-3.4 calculated based on high-pass

(,9 yr) SSTAs in the previous winter (blue), simultaneous summer

(red), and post winter (green). The dashed line indicates the 90%

confidence level.

15 JANUARY 2019 L I E T AL . 293

level) over the Huai River valley after 1991, while it

decreases from 0.66 to 0.44 over the Yangtze River

valley (Figs. 4c,d).

The tropical Indian Ocean and western Pacific

experience a decadal shift not only in the moisture cir-

culation over EA-WNP but also in covariance with

moisture circulation. Before 1990, the anticyclonic mois-

ture circulation over the subtropical WNP was concurrent

with weakened westerly moisture circulation to its south

over the eastern Indian Ocean–Philippine Sea, but the

resultingmoisture divergence isweakand scattered (Fig. 3a).

The leading mode of moisture circulation over EA-WNP

during 1991–2013 shows a stronger coupling with moisture

circulation over tropical regions. Abnormal easterly mois-

ture transport dominates the tropical central-western Pacific

with abnormal moisture divergence over the tropical central

TABLE 1. SCF accounted for by SVD1, the correlation coefficients between EC_R of moisture circulation and tropical SSTAs, and the

spatial correlations of Quy/SSTAs between EOF1 and SVD1 (Spa_REOF,SVD_Quy/Spa_REOF,SVD_SST) during 1971–90 and 1991–2013;

** indicates significance at the 99% confidence level.

SCF EC_RQuy,SST Spa_REOF,SVD_Quy Spa_REOF,SVD_SST

1971–90 95.3% 0.71 (0.91**, 0.70**) 0.97**

1991–2013 85.5% 0.71 (0.97**, 0.88**) 0.98**

FIG. 3. The regressed (a)–(c) vertically integrated summer water vapor fluxes (vectors; kgm21 s21) and their

divergence (shading; 1025 kgm22 s21) based on EC1_Quy between the moisture circulation and prewinter SSTAs

during (a) 1971–90 and (b) 1991–2013 and (c) their difference. Only the regressed results significant at the 90%

confidence level are shown in (a) and (b). (d)–(f) As in (a)–(c), but for the SSTAs (shading; K) regressed based

on the corresponding expansion coefficient of EC1_SSTAs. Results above the 90% confidence level are marked

in (d)–(e).

294 JOURNAL OF CL IMATE VOLUME 32

FIG. 4. Regression of summer precipitation (mm) based on the EC1_Quy during (a) 1971–90 and (b) 1991–2013.

Results above the 90% confidence level are marked. The blue boxes in (a) and (b) represent the Yangtze River

valley (288–328N, 1078–1228E) and the Huai River valley (318–358N, 1148–1208E), respectively. Time series of re-

gional summer precipitation over (c) the Yangtze River valley (288–328N, 1078–1228E) and (d) the Huai River

valley (318–358N, 1148–1208E) are shown together with EC1_Quy during 1971–90 and 1991–2013.

15 JANUARY 2019 L I E T AL . 295

Pacific and convergence over the Maritime Continent

(Fig. 3b). In addition, a C-shaped moisture circulation

occurs in the tropical Indian Ocean, with abnormal mois-

ture converging over the eastern IndianOcean south to the

equator. This C-shaped circulation in the surfacewinds has

been described by Xie et al. (2009) as a typical response to

antisymmetric convective heating about the equator, even

though the SSTAs are more or less symmetrical. This

C-shaped moisture circulation is hardly ever found

before 1990, illustrating the enhancement of the cou-

pling between the leading mode of moisture circulation

over EA-WNP and the circulation over the Indian

Ocean after 1991. This will be investigated further in next

section. In the differential panel (Fig. 3c), the afore-

mentioned weakness of the cyclonic moisture circulation

north to 308N, the intensification of the easterly transport

from the central Pacific to the Maritime Continent, and

the existence of the C-shaped moisture circulation over

the Indian Ocean after 1991 are even more robust. In

other words, the influence of tropical SSTAs on the in-

terannual variation of moisture circulation lies more

southward over tropical–subtropical regions rather than

in midlatitude regions after 1991.

4. Modulation from the evolution of ENSO

Given the similarity in the prewinter SVD1(SST)

between the two periods and the stable relationship

between the prewinter El Niño index and the leading

mode of the moisture circulation (Figs. 2, 3), the decadal

shift of the moisture circulation over EA-WNP bonding

to the ENSO signal might be caused by the change in

the SSTA evolution and the intensification of the mod-

ulation from synchronous SSTAs. The evolution of

SSTAs may act to amend the impact of the prewinter

ENSO signal.

a. Faster evolution of ENSO

To verify this hypothesis, SSTAs and 850-hPa stream-

function and 500-hPa vertical velocity in the decaying

phases are regressed by the corresponding expansion

coefficients of EC_1(SST), and their spatial pattern

and temporal evolution (58S–58N mean) are displayed

in Figs. 5 and 6. In the mature winter, responding to

warming SSTAs over the eastern Pacific, an abnor-

mal Walker circulation dominates the tropical Pacific

(Fig. 7). Abnormal ascending motion appears over

the tropical central-eastern Pacific, whereas descend-

ing motion occurs to the west over the tropical western

Pacific. They stimulate a pair of cyclones and anti-

cyclones on their northern and southern sides, which

is nearly symmetrical about the equator (Figs. 6a,d).

Consistent with SSTAs, common features are also found in

the corresponding circulations in WIN(0) during the

two periods with the exception of a slight difference in

the anticyclone over the WNP, which is referred to as

the Philippine Sea anticyclone (PSAC). The PSAC has

aroused much attention for its key role in bridging

the ENSO signal and the abnormal climate over East

Asia from the mature winter to the decaying summer.

It is characterized by a more westward displacement

after 1991, with its center over the Philippines shifting

to the SCS.

In contrast to the similarity in prewinter SSTAs, the

evolution of El Niño during the two periods shows

dramatic differences. Before 1990, the abnormal SSTAs

over the tropical Pacific start to decay in the spring,

develop into a neutral condition in the late summer, and

evolve further into a La Niña in the following winter.

This is concurrent with the persistence of Indian Ocean

basin-wide warming (IOBW) from the mature winter to

the decaying summer (Figs. 5a–d). The prewinter strong

abnormal rising motion over the tropical eastern Pacific

weakens concurrently in early spring and decays rapidly

into the neutral condition afterward. Its counterpart

over the western Pacific lies asymmetrically about the

equator, over the WNP, and to the east of Australia in

the decaying spring. The descending motion over the

WNP couples well with the WNP anticyclone, which

extends eastward in SPR(11) and shifts northward and

diminishes in SUM(11).

After 1991, the development of La Niña is seen to

be much faster. The tropical eastern Pacific warming

evolves into abnormal cooling instead of a neutral con-

dition in SUM(11). This is concurrent with the rever-

sal of the SSTAs over the tropical western Pacific from

negative to significantly positive (Figs. 5f,h). The

IOBW persists to SPR(11) and begins to diminish in

SUM(11) as the warming SSTAs develop over

the eastern Indian Ocean, Maritime Continent, and the

WNP. The atmospheric response over the Pacific in the

decaying spring is similar to that before 1990, while it

shows a distinct seasonal reversal in the decaying

summer consistent with the SSTAs. This will be dis-

cussed further in section 4c.

b. Intensified effect of IOBW in the mature winter anddecaying spring of ENSO

As previouslymentioned, concurrent with prewinter El

Niño, IOBW dominates the Indian Ocean and persists to

late summer before the 1990s, while it disappears faster in

early summer after the 1990s. Though the tropical In-

dian Ocean warms up with a similar pattern and intensity

in the mature winter and decaying spring during the two

periods, a dramatic atmospheric response can be seen

only after 1991 (Figs. 6, 7). Abnormal upward motion

296 JOURNAL OF CL IMATE VOLUME 32

FIG. 5. (a)–(j) Regression of seasonal SSTAs (K) from the mature winter to the following winter

based on the EC1_SSTAs. (k),(l) The temporal evolution of the regressed tropical (58S–58N aver-

aged) SSTAs.Results above the 90%confidence level aremarked. The panels (a)–(e) and (k) are for

the period of 1971–90 and (f)–(j) and (l) are for the period of 1991–2013.

15 JANUARY 2019 L I E T AL . 297

features the tropical western and central Indian Ocean,

accompanying the abnormal downward motion over the

western Pacific (Fig. 6d) and lower-level easterly and

upper-level westerly anomalies, forming an anti-Walker

circulation over the Indian Ocean–western Pacific

(Fig. 7d). This is counter to the scattered and insignificant

signals before 1990. The descending motion over the

western Pacific, a common downward branch of Walker

circulation over the tropical Pacific and anti-Walker cir-

culation over the tropical Indian Ocean, intensifies and

FIG. 6. Regression of seasonal 850-hPa streamfunction (contour; 106m2 s21) and 500-hPa vertical velocity

(shading; hPa s21) from the mature winter to the following summer based on the EC1_SSTAs. (g),(h) Temporal

evolution of regressed tropical (58S–58N averaged) 500-hPa vertical velocity. Results significant at the 90% con-

fidence level are marked. The panels (a)–(c) and (g) are for the period of 1971–90 and (d)–(f) and (h) are for the

period of 1991–2013.

298 JOURNAL OF CL IMATE VOLUME 32

extends southwestward after 1991 (Fig. 7d) and so does

the PSAC (Fig. 6d). In the decaying spring, abnormal

upwardmotion over the Indian Oceanmigrates eastward

and lies south to the equator (Fig. 6e), with the bonding

lower-tropospheric circulation characterized by C-shaped

wind anomalies (figure not shown). The anti-Walker

circulation over the Indian Ocean remains, though its

counterpart over the Pacific weakens to be insignificant

(Fig. 7e). However, this teleconnection between the trop-

ical Indian Ocean and the western Pacific could hardly

be found in the regressed zonal circulation before 1990

(Figs. 7a,b).

To further reveal the decadal change in atmospheric

response of the IOBW and its teleconnection with the

FIG. 7. Regression of seasonal tropical (108S–108N averaged) zonal circulation (vector) and 500-hPa vertical

velocity (shading, 102 hPa s21) in the (a),(d) mature winter, (b),(e) decaying spring, and (c),(f) summer based on

the EC1_SSTAs. Only the results of vertical velocity significant at the 90% confidence level are shown. The panels

(a)–(c) are for the period of 1971–90 and (d)–(f) are for the period of 1991–2013.

15 JANUARY 2019 L I E T AL . 299

circulation over the WNP, divergent and rotational

winds in both the lower and upper troposphere are

regressed based on EC_1(SST) (Figs. 8, 9). Before 1990,

descending motion accompanied by lower-level di-

vergence and upper-level convergence occurs in the

circulation over the WNP, with the opposite over

Southeast China and the central Pacific in WIN(0) and

SPR(11), indicating a possible linkage between the

WNP anticyclone and precipitation over Southeast

China. The atmospheric teleconnection between the

WNP and the Indian Ocean is weak and insignificant,

except in the decaying spring when the circulation de-

parts from the southeastern Indian Ocean to the WNP

in the upper level (Fig. 8d). In contrast, the airflow

ascends over the western Indian Ocean, goes eastward

to the WNP in the upper troposphere, sinks over the

WNP, and returns to the western Indian Ocean in the

lower troposphere in WIN(0) after 1991 (Figs. 9a,b).

The signal over the Indian Ocean shifts southwestward

to the central Indian Ocean south of the equator in

the following spring, forming a southwest–northeast

teleconnection between the Indian Ocean and WNP

(Figs. 9c,d). It helps anchor the abnormal widespread

anticyclone over the tropical–subtropical WNP. The

intensification of this teleconnection between the In-

dian Ocean and WNP is proposed as one of the causes

of the widespread anticyclonic moisture circulation

after 1991.

FIG. 8. Regression of 500-hPa vertical motion (shading; hPa s21), divergent wind (vectors; m s21), and velocity

potential (contours; 105m2 s21) at (a),(c),(e) 850 hPa and (b),(d),(f) 200 hPa from previous winter to summer based

on the EC1_SSTAs during 1971–90. Only the results of divergent wind and vertical motion significant at the 90%

confidence level are shown.

300 JOURNAL OF CL IMATE VOLUME 32

This decadal intensification of the IOBW effect has

also been proposed by other studies. In recent decades,

air–sea interaction over the tropical Indian Ocean has

enhanced, including an ENSO-induced downwelling

Rossby wave, southwestern Indian Ocean warming, the

antisymmetric C-shaped wind pattern, and its coupling

with the WNPAC (Xie et al. 2010; Huang et al. 2010;

Zheng et al. 2011). As pointed out by Luffman et al.

(2010), the SST over the Indian Ocean has increased at a

rate of 0.78C during the past century, and the western

IndianOcean especially haswarmed faster than any other

tropical ocean region. Ding et al. (2010) argues that the

summer SSTAs superposed on a higher mean SST con-

tribute to a stronger response of the atmospheric circu-

lation to SSTA forcing. By using numerical modeling,

Zheng et al. (2011) proposes that the ‘‘capacitor effect’’ of

IOBW enhances in response to global warming even

thoughENSO itself weakened. Further work is needed to

illustrate how the air–sea interaction and thus the atmo-

spheric response change under global warming given

the comparable SSTAs over the Indian Ocean in both

periods.

c. Role of simultaneous eastern Pacific cooling

In summer, responding to the rapid establishment of

cooling over the eastern Pacific and warming over the

eastern Indian Ocean–Maritime Continent (EIO-MC),

abnormal descending motion begins to dominate the

western-central Pacific in early summer. It is concurrent

with abnormal ascending motion to the west over the

EIO-MC (Figs. 6f,h). This indicates an establishment of

an anti-Walker circulation between the EIO-MC and

FIG. 9. As in Fig. 8, but for the period of 1991–2013.

15 JANUARY 2019 L I E T AL . 301

central Pacific, the atmospheric counterpart of rapidly

developing La Niña in summer after 1991 (Fig. 7f). The

suppressed convection over the tropical western-central

Pacific excites a pair of anticyclones to its northwest over

the WNP and southwest to the east of northeast Aus-

tralia apart from the equator as a Rossby wave response

via the Gill–Matsuno mechanism (Fig. 6f). The north-

west anticyclone tends to be superimposed upon and

thus enhances the WNPAC excited by the preceding El

Niño. A similar result is found in Chen et al. (2016) via

numerical experiments forced by tropical eastern Pacific

cooling. However, these regressed circulation anomalies

cannot be found before 1990 accompanying neutral

SSTAs in the tropical eastern Pacific in the decaying

summer of El Niño (Figs. 6c,g, 7c).

The aforementioned abnormal vertical circulation

can be found not only in a west–east direction between

EIO-MC and the tropical western Pacific, but also in a

southwest–northeast direction between the EIO-MC

and the subtropical WNP in the rotational and di-

vergent wind fields after 1991 (Figs. 9e,f). The airflow

ascends over the EIO-MC, and diverges northeast-

ward in the upper level to the WNP, where it descends

to the lower level and returns to the EIO-MC. This

is consistent with He and Wu (2014), in which such

southwest–northwest vertical circulation is proposed as

an alternative mechanism for the Indian Ocean SST

influence on the SCS climate other than the Kelvin

wave–induced Ekman divergence mechanism (Wu et al.

2012; He and Wu 2014).

Hence, the rapid development of cooling SSTAs

over the central and eastern Pacific and warming

SSTAs over EIO-MC appear to act in concert to

stimulate a west–east teleconnection with abnormal

downward motion over the western-central Pacific

and upward motion over the EIO-MC. The former

excites an anticyclone over the WNP, sustaining and

intensifying the WNPAC; the latter helps anchor

the WNPAC via the abnormal southwest–northeast

vertical circulation between the EIO-MC and WNP.

In response, a more widespread anticyclonic mois-

ture circulation prevails in the tropical–subtropical

western Pacific, with the northward shift of abnormal

moisture convergence over East Asia and precipitation

enhancement over the Huai River valley instead of the

YangtzeRiver valley inChina. That is, the different speed

of evolution from the prewinter El Niño to La Niña tendsto amend the leading mode of moisture circulation over

EA-WNP.

The decadal change of the covariance between the

moisture circulation over WNP and the SSTAs over the

eastern Pacific and EIO-MC is also supported by the cases

studied. The regional SSTAs in abnormal years when

the EC1_Quy is significantly anomalous during two

subperiods are listed in Table 2. All positive/negative

anomalous cases are preceded by tropical eastern Pa-

cific warming/cooling. However, in contrast to the higher

possibility of reversal instead of maintenance of the east-

ern Pacific SSTAs from winter to summer during 1991–

2013, the possibility is evenly distributed during 1971–90.

That is why the leading mode of moisture circulation

concurs with the reversal of ENSO signal in summer

after 1991, while with neutral conditions in summer

before 1990 in the regression analysis. For the SSTAs

over the EIO-MC, comparing to the 3 out of 4 cases in

the pre-1990 period, all cases in the post-1991 period

occur in line with the regression results. Hence, the modu-

lation from the summer SSTAs is not stable in the pre-1990

period, while it is relatively stable in the post-1990 period.

d. Numerical experiments

To evaluate the impact of a different transforming speed

of El Niño to La Niña on the circulation over EA-WNP,

three idealized numerical experiments are designed and

TABLE 2. Relationship between the SVD1(Quy) and SSTAs over the tropical eastern Pacific in previous winter [EP_WIN(0)], spring

(EP_SPR), simultaneous summer (EP_SUM), and summer SSTAs over the EIO-MC for the periods of 1971–90 and 1991–2013. The

symbols1/2 represent positive/negative regional SSTAs, and * indicates the SSTAs are significant based on the criterion of.0.8 standard

deviation. EP overturn with O denotes the years with prewinter positive/negative SSTAs over the tropical eastern Pacific transforming to

the opposite in summer, and cooling/warming SSTAs dominating eastern Pacific in summer when the EC1_Quy is positive/negative.

EIO-MCrolewithOdenotes the yearswithwarming/cooling SSTAsdominatingEIO-MCwhen theEC1_Quy is positive/negative. The3 represents

the opposite condition to O.

1971–90 1991–2013

Negative EC1_Quy Positive EC1_Quy Negative EC1_Quy Positive EC1_Quy

71 74 76 89 73 77 83 88 97 99 06 08 11 95 98 07 10

EP_WIN (0) 2* 2* 2* 2* 1* 1 1* 1 2 2* 2 2* 2* 1* 1* 1* 1*

EP_SPR 2* 2* 2 2* 1* 1 1* 2 1 2* 2 2* 2* 1 1* 1 1*

EP_SUM 2* 1 1* 2 2* 1* 1 2* 1* 2* 1 2 2 2 2* 2 2*

EIO-MC 2 2 2* 1 1* 2 1 1 2* 2 2 2* 2* 1* 1* 1 1*

EP overturn 3 O O 3 O 3 3 O O 3 O 3 3 O O O OEIO-MC role O O O 3 O 3 O O O O O O O O O O O

302 JOURNAL OF CL IMATE VOLUME 32

conducted. One is a control (CTL) run in which the model

is forced by repeating climatological annual cycles of

SST and sea ice. The other two idealized experiments

are designed with climatological SST superposed by the

SSTAs over the tropical central-eastern Pacific (208S–208N, 1608E–808W) regressed during the period of 1971–

90 when the transforming of El Niño is slow (EP_pre90)

and during the period of 1991–2013 when the transforming

of El Niño is fast (EP_post90). Only the regressed SSTAs

significant at the 90% confidence level are superposed

(Figs. 6, 11). The last 19 ensembles’ mean results from

winter to summer are analyzed.

The difference of the simulated lower-level circula-

tion and midlevel vertical motion between the EP_

pre90/EP_post90 and CTL runs are shown in Fig. 10.

Similar atmospheric response is found in boreal winter

in both simulations, as the enforcing SSTAs in WIN(0)

are more or less alike (Figs. 10a,d). When it comes to

spring, in the EP_pre90 run, as the eastern Pacific

warming is sustained, upward motion dominates the

central-eastern Pacific with its counterpart over the

western Pacific; a weak anticyclonic circulation appears

over the western Pacific and the SCS (Fig. 10b). Re-

sponding to the fast decay of El Niño in the EP_post90

run, both the upward and downward motions over the

tropical Pacific weaken to a large extent and so does the

anticyclone over the SCS-WNP (Fig. 11d). In boreal

summer, as the eastern Pacific SST turns to be neutral in

the EP_pre90 run, the atmospheric response over the

WNP dissipates (Fig. 11c); whereas in the EP_post90 run,

an anticyclonic circulation is stimulated over theWNP and

SCS as the cooling SSTAs over the tropical eastern Pacific

and abnormal descent to its west are established (Fig. 11f).

These results validate the importance of the early estab-

lishment of eastern Pacific cooling in intensifying the si-

multaneous WNPAC in summer after 1991.

The upward motion over the EIO-MC is proposed to

be directly triggered by the local simultaneous warming

SSTAs (He andWu2014). To evaluate the individual role

of the EIO-MC warming and also its combined effect

with the eastern Pacific cooling, two additional experi-

ments are conducted with the SSTAs specified around

the EIO-MC region (EIO-MC_post91; 158S–58N, 908E–1608W) and both the EIO-MC and eastern Pacific region

(EP&EIO-MC_post91) during 1991–2013, and the cir-

culation in summer is shown in Fig. 11. In the EIO-MC_

post91 run, though the stimulated low-level circulation is

not significant, an anticyclonic circulation could still be

found over the WNP. The airflow ascends over the EIO-

MC, diverges northward to the SCS-WNP in the upper

level, then descends and returns to the EIO-MC in the

lower level, forming an abnormal meridional circulation.

With the SSTAs over the eastern Pacific superposed, the

WNP anticyclone is enhanced and enlarged; the meridi-

onal circulation shifts southwest–northeastward, and the

west–east teleconnection between the EIO-MC and mid-

Pacific is found. Hence, with the EIO-MC warming and

the eastern Pacific cooling acting in concert, theWNPAC

FIG. 10. Difference of the simulated 850-hPa wind (vector; m s21) and 500-hPa omega (shading; 1022 hPa s21)

between the (a)–(c) EP_pre90 run, (d)–(f) EP_post90 run, and the CTL run in the (a),(d) previous winter, (b),(e)

previous spring, and (c),(f) summer. Only the results significant at the 90% confidence level are shown for omega

and in black vector for wind. The red/blue contours with the interval of 0.3 K indicate the SSTAs over the tropical

eastern Pacific imposed in the stimulation.

15 JANUARY 2019 L I E T AL . 303

is enhanced and enlarged in the transitional summer be-

tween El Niño and La Niña.

5. Conclusions and discussion

The leading mode of summer moisture circulation over

EA-WNP, characterized by an ‘‘anticyclone–cyclone’’ di-

pole structure, is likely to take place in the transitional

summer from an El Niño to a La Niña event. In the past

few decades, its modulation by the prewinter tropical

eastern Pacific SSTAs was stable, while the influence of

synchronous eastern Pacific SSTAs intensified after 1991.

As a result, the zonally elongated anticyclonic mois-

ture circulation over the subtropical WNP enhances and

shows a much larger meridional expansion, while the

cyclonic moisture circulation to the north weakens after

1991. As a result, stronger moisture invades northward

and the associated convergence shifts from the Yangtze

River valley to the Huai River valley, as does the

precipitation.

The possible shift in the modulation of ENSO might

arise in its evolution. After 1991, the prewinter El Niñodecays and reverses to be cooling, concurrent with

the replacement of IOBW by the EIO-MC warming in

the decaying summer. Such rapid development of La

Niña stimulates abnormal descending motion over the

western-central Pacific and ascending motion over the

EIO-MC. The former excites a pair of anticyclones

over the western Pacific, sustaining and intensifying the

WNP anticyclone. The latter helps in anchoring the

WNP anticyclone via the southwest–northeast circu-

lation between the EIO-MC and WNP. Numerical

experiments using CAM4 enforced by the regressed

SSTAs over the tropical central-eastern Pacific and

EIO-MC verify these results. In addition, the influence

of the IOBW shows a decadal variation. Though the

FIG. 11. Difference of the simulated 500-hPa omega (shading; 1022 hPa s21) and (a),(d) 850-hPa wind (vector;

m s21), (b),(c),(e),(f) divergent wind (vectors; m s21) and velocity potential (contours; 105m2 s21) wind between the

(a)–(c) EIO-MC_post90 run, (d)–(f) EP&EIO-MC_post90 run, and the CTL run in summer. Only the results

significant at the 90% confidence level are shown for omega and in black vector for wind. The red/blue contours

with the interval of 0.1K indicate the SSTAs imposed in the stimulation.

304 JOURNAL OF CL IMATE VOLUME 32

tropical Indian Ocean warms up with a similar pattern

and intensity in the mature winter and decaying spring

of El Niño during the two periods, apparent atmo-

spheric response could be found only after 1991.

Besides the wider expansion of anticyclonic moisture

circulation over the WNP, decadal change is also man-

ifested in the weakness of the cyclonic moisture circu-

lation to the north in the positive phase of the leading

mode. The cyclonic moisture circulation is strong before

1990, lying east to Japan, together with the anticyclonic

circulation over the WNP, implying the modulation

of the ‘‘East Asia–Pacific’’ teleconnection or ‘‘Pacific–

Japan pattern’’ on the moisture circulation (Nitta 1987).

However, it weakens to a large extent after 1991. Huang

and Sun (1992) suggested that this East Asia–Pacific

teleconnection, characterized by the meridional prop-

agation of quasi-stationary planetary waves, was forced

by heat sources around the Philippines (Hoskins and

Karoly 1981). To examine the possible causes of the

weakness of the cyclone over Japan, the wave activity

fluxes, and the quasigeostrophic streamfunction (Takaya

and Nakamura 2001) in the lower and upper tropo-

sphere in the two periods are exhibited in Fig. 12.

The cyclone is significant before 1990 and shows a

quasi-barotropic vertical structure, with its location

shifting slightly northeastward in the upper level. In

the lower troposphere, its variation is closely associ-

ated with the northward propagation of wave activity

fluxes from the subtropical WNP, implying possible

forcing from the lower latitudes (Huang and Sun 1992;

Kosaka and Nakamura 2006). The wave activity flux

ascends over the region where the cyclone lies. In ad-

dition, it couples well with a barotropic wave train from

the high latitudes, which could be traced back to the

Arctic Ocean. The wave activity flux from the high lat-

itudes propagates along the wave train in the upper

troposphere and descends to the lower level over the

Sea ofOkhotsk north to themidlatitude cyclone. That is,

the cyclone over Japan is the joint result of wave prop-

agation from lower latitudes in the lower troposphere

and from higher latitudes in the upper troposphere. In

contrast, after 1991, the wave activity forced by the

convection over the tropical–subtropical western Pacific

is constrained in the lower latitudes. Meanwhile, the

wave train from higher latitudes cannot be found. The

underlying processes and mechanisms for such decadal

variation remain unknown at this stage. Other signals

besides tropical SSTAs could also play a role in the

decadal variation of the atmosphere and thus moisture

circulation over EA-WNP.

In this study, the decadal shift in 1990/91 is focused on

as the relationship between the moisture circulation

over the EA-WNP and the eastern equatorial Pacific

SSTAs in summer shifts to be significantly negative. It is

proposed to be caused by the rapid transforming of ENSO,

which starts to play roles in summer. In cases studied

(Table 2), we found that such rapid transforming also ap-

pears before 1990; however, it takes place with equal

chance as slow transforming cases, while it appears much

more often than the slow cases after 1990. For the causes,

many recent studies propose that the Pacific decadal os-

cillation (PDO) could modulate the decaying speed of

FIG. 12. Regression of the quasigeostrophic streamfunction (shading; 106m2 s21), horizontal (vectors; m2 s22)

and vertical (contours; 1024 Pam s22) wave activity flux proposed by Takaya and Nakamura (2001) averaged

between (a),(b) 500 and 200 hPa and (c),(d) 1000 and 550 hPa for the periods of (a),(c) 1971–90 and (b),(d) 1991–

2013. Only the results significant at the 90% confidence level are shown.

15 JANUARY 2019 L I E T AL . 305

ENSO and its impact on the global climate (e.g.,

Gershunov and Barnett 1998; Chen et al. 2013; Feng et al.

2014). The climate response of El Niño is likely to be

stronger when the PDO is highly positive; in contrast, the

climate response of La Niña is stronger when the PDO is

highly negative (Gershunov and Barnett 1998). During

high-PDO phases, El Niño decays slowly and has a strong

anchor in the north IndianOceanwarming; comparatively,

during low-PDO phases, El Niño decays rapidly and La

Niña develops in summer. As a result, the anomalous

WNPAC has a zonal band structure spanning southern

China–WNP during the positive PDO phases, while it

has a much larger spatial domain with meridional elon-

gated structure from the tropics to the midlatitudes during

the negative PDO phases (Chen et al. 2013; Feng et al.

2014). Hence, the more rapid evolution speed of ENSO

and its impact on the moisture circulation over EA-WNP

after 1991 proposed in this study might also be related to

the change of the PDO phases. However, it is found that a

decadal shift of the PDO from positive to negative phase

tends to appear in late 1990 instead of early 1990, though

two short terms of negative PDO phase appear in 1989–91

and mid-1990s (figure not shown). Whether the decadal

shift of the relationship between the ENSO and moisture

circulation could be attributed to the PDO signal requires

further investigation.

Acknowledgments. This work is jointly supported by

the National Key Research and Development Program

of China (Project Number 2016YFA0600601), National

Natural Science Foundation of China (Project Number

41775043, 41805057), and the Guangdong Natural

Science Foundation (2018A030310023). Support from

Swedish VR, VINOVVA, STINT, BECC, MERGE,

and SNIC through S-CMIP is also acknowledged.

REFERENCES

Bao,M., 2007: The statistical analysis of the persistent heavy rain in

the last 50 years over China and their backgrounds on the

large-scale circulation (in Chinese). Chin. J. Atmos. Sci., 31,

779–792.

Chang, C.-P., Y. Zhang, and T. Li, 2000: Interannual and interdecadal

variations of the East Asian summer monsoon and tropical

Pacific SSTs. Part I: Roles of the subtropical ridge. J. Climate, 13,

4310–4325, https://doi.org/10.1175/1520-0442(2000)013,4310:

IAIVOT.2.0.CO;2.

Chen, L., M. Dong, and Y. Shao, 1992: The characteristics of

interannual variations on the East Asian monsoon.

J. Meteor. Soc. Japan, 70, 397–421, https://doi.org/10.2151/jmsj1965.70.1B_397.

Chen,W., 2002: Impacts of El Niño and La Niña on the cycle of the

East Asian winter and summer monsoon (in Chinese). Chin.

J. Atmos. Sci., 26, 595–610.

——, J. Feng, and R. Wu, 2013: Roles of ENSO and PDO in the

link of the East Asian winter monsoon to the following

summer monsoon. J. Climate, 26, 622–635, https://doi.org/

10.1175/JCLI-D-12-00021.1.

Chen, Z., Z. Wen, R. Wu, P. Zhao, and J. Cao, 2014: Influence of

two types of El Niños on the East Asian climate during boreal

summer: a numerical study.Climate Dyn., 43, 469–481, https://

doi.org/10.1007/s00382-013-1943-1.

——,——,——,X. Lin, and J.Wang, 2016: Relative importance of

tropical SST anomalies in maintaining the western North Pa-

cific anomalous anticyclone during El Niño to La Niña tran-

sition years. Climate Dyn., 46, 1027–1041, https://doi.org/

10.1007/s00382-015-2630-1.

Ding, R., K.-J. Ha, and J. Li, 2010: Interdecadal shift in the re-

lationship between the East Asian summer monsoon and the

tropical Indian Ocean. Climate Dyn., 34, 1059–1071, https://

doi.org/10.1007/s00382-009-0555-2.

Ding, Y., Z. Wang, and Y. Sun, 2008: Inter-decadal variation of the

summer precipitation in East China and its association with

decreasing Asian summer monsoon. Part I: Observed evi-

dences. Int. J. Climatol., 28, 1139–1161, https://doi.org/10.1002/

joc.1615.

——, Y. Sun, Z. Wang, Y. Zhu, and Y. Song, 2009: Inter-decadal

variation of the summer precipitation in China and its asso-

ciation with decreasing Asian summer monsoon. Part II:

Possible causes. Int. J. Climatol., 29, 1926–1944, https://doi.org/

10.1002/joc.1759.

Duan, A., M. Wang, Y. Lei, and Y. Cui, 2013: Trends in summer

rainfall over China associated with the Tibetan Plateau sen-

sible heat source during 1980–2008. J. Climate, 26, 261–275,

https://doi.org/10.1175/JCLI-D-11-00669.1.

Fan, L., S.-I. Shin, Q. Liu, and Z. Liu, 2013: Relative importance of

tropical SST anomalies in forcing East Asian summer mon-

soon circulation. Geophys. Res. Lett., 40, 2471–2477, https://

doi.org/10.1002/grl.50494.

Feng, J., L. Wang, and W. Chen, 2014: How does the East Asian

summer monsoon behave in the decaying phase of El Niñoduring different PDO phases? J. Climate, 27, 2682–2698,

https://doi.org/10.1175/JCLI-D-13-00015.1.

Gent, P. R., and Coauthors, 2011: The Community Climate System

Model Version 4. J. Climate, 24, 4973–4991, https://doi.org/

10.1175/2011JCLI4083.1.

Gershunov, A., and T. P. Barnett, 1998: Interdecadal modulation

of ENSO teleconnections. Bull. Amer. Meteor. Soc., 79,

2715–2725, https://doi.org/10.1175/1520-0477(1998)079,2715:

IMOET.2.0.CO;2.

Gill, A. E., 1980: Some simple solutions for heat-induced tropical

circulation. Quart. J. Roy. Meteor. Soc., 106, 447–462, https://

doi.org/10.1002/qj.49710644905.

Gong, D.-Y., and C.-H. Ho, 2003: Arctic oscillation signals in the

East Asian summer monsoon. J. Geophys. Res., 108, 4066,

https://doi.org/10.1029/2002JD002193.

Harada, Y., and Coauthors, 2016: The JRA-55 reanalysis: Repre-

sentation of atmospheric circulation and climate variability.

J. Meteor. Soc. Japan, 94, 269–302, https://doi.org/10.2151/

jmsj.2016-015.

He, Z., and R. Wu, 2014: Indo-Pacific remote forcing in summer

rainfall variability over the South China Sea.Climate Dyn., 42,

2323–2337, https://doi.org/10.1007/s00382-014-2123-7.

Hoskins, B. J., and D. J. Karoly, 1981: The steady linear response of a

spherical atmosphere to thermal andorographic forcing. J.Atmos.

Sci., 38, 1179–1196, https://doi.org/10.1175/1520-0469(1981)

038,1179:TSLROA.2.0.CO;2.

Huang, G., K. Hu, and S.-P. Xie, 2010: Strengthening of tropical

Indian Ocean teleconnection to the northwest Pacific since

306 JOURNAL OF CL IMATE VOLUME 32

the mid-1970s: An atmospheric GCM study. J. Climate, 23,

5294–5304, https://doi.org/10.1175/2010JCLI3577.1.

Huang, R., and Y. Wu, 1989: The influence of ENSO on the sum-

mer climate change in China and its mechanism. Adv. Atmos.

Sci., 6, 21–32, https://doi.org/10.1007/BF02656915.

——, and F. Sun, 1992: Impacts of the tropical western Pacific on

the East Asian summer monsoon. J. Meteor. Soc. Japan, 70,

243–256, https://doi.org/10.2151/jmsj1965.70.1B_243.

Huang, S. S., X. Q. Yang, Q. R. Jiang, M. M. Tang, Z. M. Wang,

Q. Xie, and Y. C. Zhu, 1995: The effects of the polar sea ice on

climate (in Chinese). J. Meteor. Sci., 15, 46–56.Kobayashi, C., and T. Iwasaki, 2016: Brewer-Dobson circulation di-

agnosed from JRA-55. J. Geophys. Res. Atmos., 121, 1493–1510,

https://doi.org/10.1002/2015JD023476.

Kobayashi, S., andCoauthors, 2015: The JRA-55 reanalysis: General

specifications andbasic characteristics. J.Meteor. Soc. Japan, 93,

5–48, https://doi.org/10.2151/jmsj.2015-001.

Kosaka, Y., and H. Nakamura, 2006: Structure and dynamics of

the summertime Pacific Japan teleconnection pattern. Quart.

J. Roy. Meteor. Soc., 132, 2009–2030, https://doi.org/10.1256/

qj.05.204.

Li, X., and W. Zhou, 2012: Quasi-4-yr coupling between El Niño–Southern Oscillation and water vapor transport over East

Asia–WNP. J. Climate, 25, 5879–5891, https://doi.org/10.1175/

JCLI-D-11-00433.1.

——, Z. Wen, and W. Zhou, 2011: Long-term change in summer

water vapor transport over South China in recent decades.

J. Meteor. Soc. Japan, 89A, 271–282, https://doi.org/10.2151/

jmsj.2011-A17.

——, ——, ——, and D. Wang, 2012: Atmospheric water vapor

transport associated with two decadal rainfall shifts over East

China. J. Meteor. Soc. Japan, 90, 587–602, https://doi.org/

10.2151/jmsj.2012-501.

——, W. Zhou, D. Chen, C. Li, and J. Song, 2014: Water vapor

transport and moisture budget over eastern China: Remote

forcing from the two types of El Niño. J. Climate, 27, 8778–8792,

https://doi.org/10.1175/JCLI-D-14-00049.1.

——, ——, and Y. D. Chen, 2015: Assessment of regional drought

trend and risk over China: A drought climate division per-

spective. J. Climate, 28, 7025–7037, https://doi.org/10.1175/

JCLI-D-14-00403.1.

Liu, H., T. J. Zhou, Y. Zhu, and Y. Lin, 2012: The strength-

ening East Asia summer monsoon since the early 1990s.

Chin. Sci. Bull., 57, 1553–1558, https://doi.org/10.1007/

s11434-012-4991-8.

Luffman, J. J., A. S. Taschetto, and M. H. England, 2010: Global

and regional climate response to late twentieth-century

warming over the Indian Ocean. J. Climate, 23, 1660–1674,

https://doi.org/10.1175/2009JCLI3086.1.

Matsuno, T., 1966: Quasi-geostrophic motions in the equatorial

area. J. Meteor. Soc. Japan, 44, 25–43, https://doi.org/10.2151/

jmsj1965.44.1_25.

Murakami, T., and J. Matsumoto, 1994: Summermonsoon over the

Asian continent and western North Pacific. J. Meteor. Soc.

Japan, 70, 597–611, https://doi.org/10.2151/jmsj1965.72.5_719.

Nan, S., J. Li, X. Yuan, and P. Zhao, 2009: Boreal spring Southern

Hemisphere annular mode, Indian Ocean sea surface tem-

perature, and East Asian summer monsoon. J. Geophys. Res.,

114, D02103, https://doi.org/10.1029/2008JD010045.

Neale, R. B., and Coauthors, 2010: Description of the NCAR

Community Atmosphere Model (CAM 4.0). NCAR Tech.

Note NCAR/TN-4851STR, 212 pp., www.cesm.ucar.edu/

models/ccsm4.0/cam/docs/description/cam4_desc.pdf.

——, J. Richter, S. Park, P. H. Lauritzen, S. J. Vavrus, P. J.

Rasch, and M. Zhang, 2013: The mean climate of the

Community AtmosphereModel (CAM4) in forced SST and

fully coupled experiments. J. Climate, 26, 5150–5168,

https://doi.org/10.1175/JCLI-D-12-00236.1.

Nitta, T., 1987: Convective activities in the tropical western Pacific

and their impact on the Northern Hemisphere summer cir-

culation. J. Meteor. Soc. Japan, 65, 373–390, https://doi.org/

10.2151/jmsj1965.65.3_373.

Park, S.-C., B.-J. Sohn, and B. Wang, 2007: Satellite assessment of

divergent water vapor transport from NCEP, ERA40, and

JRA25 reanalyses over the Asian summer monsoon region.

J. Meteor. Soc. Japan, 85, 615–632, https://doi.org/10.2151/

jmsj.85.615.

Rayner, N. A., D. E. Parker, E. B. Horton, C. K. Folland,

L. V. Alexander, D. P. Rowell, E. C. Kent, and A. Kaplan, 2003:

Global analyses of sea surface temperature, sea ice, and night

marine air temperature since the late nineteenth century.

J. Geophys. Res., 108, 4407, https://doi.org/10.1029/2002JD002670.

Si, D., and Y. Ding, 2013: Decadal change in the correlation pattern

between the Tibetan Plateau winter snow and the East Asian

summer precipitation during 1979–2011. J.Climate, 26, 7622–7634,

https://doi.org/10.1175/JCLI-D-12-00587.1.

——,——, and Y. Liu, 2009: Decadal northward shift of theMeiyu

belt and the possible cause. Chin. Sci. Bull., 54, 4742–4748,

https://doi.org/10.1007/s11434-009-0385-y.

Sun, B., and H. Wang, 2015: Analysis of the major atmo-

spheric moisture sources affecting three sub-regions of East

China. Int. J. Climatol., 35, 2243–2257, https://doi.org/10.1002/

joc.4145.

Takaya, K., and H. Nakamura, 2001: A formulation of a phase-

independent wave-activity flux for stationary and migratory

quasigeostrophic eddies on a zonally varying basic flow.

J. Atmos. Sci., 58, 608–627, https://doi.org/10.1175/1520-0469

(2001)058,0608:AFOAPI.2.0.CO;2.

Ueda, H., and T. Yasunari, 1996: Maturing process of summer

monsoon over the western North Pacific: A coupled ocean/

atmosphere system. J. Meteor. Soc. Japan, 74, 493–508, https://

doi.org/10.2151/jmsj1965.74.4_493.

Wang, B., and Q. Zhang, 2002: Pacific-East Asian teleconnection.

Part II: How the Philippine Sea anomalous anticyclone is

established during El Niño development. J. Climate, 15,

3252–3265, https://doi.org/10.1175/1520-0442(2002)015,3252:

PEATPI.2.0.CO;2.

——, R. Wu, and X. Fu, 2000: Pacific–East Asian teleconnection:

How does ENSO affect East Asian climate? J. Climate, 13,

1517–1536, https://doi.org/10.1175/1520-0442(2000)013,1517:

PEATHD.2.0.CO;2.

——, ——, and K.-M. Lau, 2001: Interannual variability of

Asian summer monsoon: Contrast between the Indian and

western North Pacific–East Asian monsoons. J. Climate,

14, 4073–4090, https://doi.org/10.1175/1520-0442(2001)014,4073:

IVOTAS.2.0.CO;2.

——,——, and T. Li, 2003: Atmosphere–warm ocean interaction and

its impacts on Asian–Australian monsoon variation. J. Climate,

16, 1195–1211, https://doi.org/10.1175/1520-0442(2003)16,1195:

AOIAII.2.0.CO;2.

Wang, L., and R. Wu, 2012: In-phase transition from the winter

monsoon to the summer monsoon over East Asia: Role of the

Indian Ocean. J. Geophys. Res., 117, D11112, https://doi.org/

10.1029/2012JD017509.

Wu, R., and B. Wang, 2002: A contrast of the East Asian summer

monsoon–ENSO relationship between 1962–77 and 1978–93.

15 JANUARY 2019 L I E T AL . 307

J. Climate, 15, 3266–3279, https://doi.org/10.1175/1520-0442(2002)

015,3266:ACOTEA.2.0.CO;2.

——, and B. P. Kirtman, 2007: Observed relationship of spring

and summer East Asian rainfall with winter and spring Eur-

asian snow. J. Climate, 20, 1285–1304, https://doi.org/10.1175/

JCLI4068.1.

——, Z.-Z. Hu, and B. P. Kirtman, 2003: Evolution of ENSO

related rainfall anomalies in East Asia. J. Climate, 16,3742–3758, https://doi.org/10.1175/1520-0442(2003)016,3742:

EOERAI.2.0.CO;2.

——, S. Yang, Z. Wen, G. Huang, and K. Hu, 2012: Interdecadal

change in the relationship of southern China summer rainfall

with tropical Indo-Pacific SST. Theor. Appl. Climatol., 108,

119–133, https://doi.org/10.1007/s00704-011-0519-4.

Wu, T.-W., and Z.-A. Qian, 2003: The relation between the Ti-

betan winter snow and the Asian summer monsoon and

rainfall: An observational investigation. J. Climate, 16,

2038–2051, https://doi.org/10.1175/1520-0442(2003)016,2038:

TRBTTW.2.0.CO;2.

Xie, S.-P., K. Hu, J. Hafner, H. Tokinaga, Y. Du, G. Huang, and

T. Sampe, 2009: Indian Ocean capacitor effect on Indo–

western Pacific climate during the summer following

El Niño. J. Climate, 22, 730–747, https://doi.org/10.1175/

2008JCLI2544.1.

——, Y. Du, G. Huang, X.-T. Zheng, H. Tokinaga, K. Hu, and

Q. Liu, 2010: Decadal shift in El Niño influences on Indo–

western Pacific and East Asian climate in the 1970s. J. Climate,

23, 3352–3368, https://doi.org/10.1175/2010JCLI3429.1.Zheng, X.-T., S.-P. Xie, and Q. Liu, 2011: Response of the In-

dian Ocean basin mode and its capacitor effect to global

warming. J. Climate, 24, 6146–6164, https://doi.org/10.1175/2011JCLI4169.1.

308 JOURNAL OF CL IMATE VOLUME 32