Impact of the Madden-Julian Oscillation on summer rainfall in Southeast China
Impact of the Madden–Julian Oscillation on Summer Rainfall in Southeast China
L INA Z HANG*
International Center for Climate and Environment Sciences,Institute of Atmospheric Physics,Chinese Academy of Sciences,and Graduate School of the Chinese Academy of Sciences,Beijing,China
B IZHENG W ANG AND Q INGCUN Z ENG
International Center for Climate and Environment Sciences,Institute of Atmospheric Physics,
Chinese Academy of Sciences,Beijing,China
(Manuscript received19March2007,in?nal form8July2008)
ABSTRACT
The impact of the Madden–Julian oscillation(MJO)on summer rainfall in Southeast China is investigated
using the Real-time Multivariate MJO(RMM)index and the observational rainfall data.A marked tran-
sition of rainfall patterns from being enhanced to being suppressed is found in Southeast China(east of1058E
and south of358N)on intraseasonal time scales as the MJO convective center moves from the Indian Ocean
to the western Paci?c Ocean.The maximum positive and negative anomalies of regional mean rainfall are in
excess of10%relative to the climatological regional mean.Such different rainfall regimes are associated with
the corresponding changes in physical?elds such as the western Paci?c subtropical high(WPSH),moisture,
and vertical motions.When the MJO is mainly over the Indian Ocean,the WPSH shifts farther westward,
and the moisture and upward motions in Southeast China are increased.In contrast,when the MJO enters
the western Paci?c,the WPSH retreats eastward,and the moisture and upward motions in Southeast China
are decreased.
It is suggested that the MJO may in?uence summer rainfall in Southeast China through remote and local
dynamical mechanisms,which correspond to the rainfall enhancement and suppression,respectively.The
remote role is the energy propagation of the Rossby wave forced by the MJO-related heating over the Indian
Ocean through the low-level westerly waveguide from the tropical Indian Ocean to Southeast China.The
local role is the northward shift of the upward branch of the anomalous meridional circulation when the MJO
is over the western Paci?c,which causes eastward retreat of the WPSH and suppressed moisture transport
toward Southeast China.
1.Introduction
Since?rst documented by Madden and Julian(1971, 1972),the Madden–Julian oscillation(MJO)has been increasingly recognized as an important atmospheric phenomenon in the tropics(Madden and Julian1994; Zhang2005).It propagates eastward at about an aver-aged speed of5m s21across the equatorial Indian and western-central Paci?c Oceans,with a local period of 30–90days.Associated with eastward propagation,it also propagates northward in boreal summer(hereafter summer;Wang and Rui1990).Although the MJO exists all year-round,its strength undergoes a strong seasonal cycle,which is strongest in winter and weakest in sum-mer(Madden and Julian1994).No matter summer or winter,many studies have revealed that the MJO in?u-ences rainfall not only in the tropics,but also in the ex-tratropics(Mo and Higgins1998;Jones2000;Paegle et al. 2000;Higgins and Shi2001;Carvalho et al.2004;Jones et al.2004;Barlow et al.2005;Lorenz and Hartmann 2006;Jeong et al.2008).
China lies within the prevailing region of East Asian summer monsoon.The summer rainfall in China involves multiscale variations.Among them,the intraseasonal variation is prominent,which often causes severe?ood or drought in the eastern part of China and greatly
*Current af?liation:China Meteorological Administration
Training Center,Beijing,China.
Corresponding author address:Dr.Lina Zhang,China Meteo-
rological Administration Training Center,No.46,Zhongguancun
South St.,Haidian District,Beijing100081,China.
E-mail:zhangln@https://www.sodocs.net/doc/8c12432515.html,
V OLUME22J O U R N A L O F C L I M A T E15J ANUARY2009
DOI:10.1175/2008JCLI1959.1
ó2009American Meteorological Society201
impacts people,as well as the development of economy and agriculture.It is shown that such intraseasonal variation is under the in?uence of the MJO(Hsu2005; Waliser2006;Donald et al.2006,hereafter D06),such as the oscillations of the rain belt and the mei-yu con-vergence zone(Lau et al.1988;Chen and Murakami 1988;Chen et al.1988).
In previous studies,?ltering methods are often ap-plied to extract the intraseasonal signals of the MJO. However,there is no consistent method or criterion to denote the MJO because different?ltering periods or regions are adopted.A new MJO index without?l-tering was advanced by Wheeler and Hendon(2004, hereafter WH04),which provides real-time MJO in-formation(position and strength).Using this index,the effect of the MJO on rainfall can be directly investi-gated(WH04;D06;Jeong et al.2008).The near-global effect of the MJO on rainfall is shown in D06.Though they did not mention directly,it can be seen from their Fig.3that the summer rainfall in China,especially the southeastern part,is under the in?uence of the MJO. However,how the evolution of the MJO leads to the variation of rainfall is not well documented and needs further study.
With regard to the mechanism of the intraseasonal variation of summer rainfall in China,the northward or northwestward propagation of the MJO are considered important factors(Lau and Chan1986;Wang and Rui 1990;Hsu2005;Wang2006).The characteristics of rainfall and other physical?elds in suppressed phases (when the MJO is over the Indian Ocean)are thought to be the mirror image of that in active phases(when the MJO is over the western Paci?c).And the mechanism in suppressed phases is considered to be just contrary to that in active phases.However,since the MJO-related energy moves from the Indian Ocean to the western Paci?c associated with the propagation of the MJO, different mechanisms should be proposed according to the evolution of the MJO.
For the above issues,we will further investigate the impact of the MJO on summer rainfall in Southeast China using the MJO index proposed by WH04and the observational rainfall data.The structure and evolu-tion of the MJO-related large-scale physical?elds such as horizontal winds,the western Paci?c subtropical high(WPSH),moisture transport,and vertical motions are also explored.Based on the diagnostic results, some new insights linking the MJO and summer rain-fall in Southeast China are provided from both the remote and local points of view.Section2describes the data and methods utilized in the study.Then the var-iations of rainfall and the corresponding large-scale physical?elds affected by the MJO are presented in sections3and4,respectively.Section5proposes the possible mechanism responsible for the impact of the MJO on summer rainfall in Southeast China.Finally,a summary is given in section6.
2.Data and methods
a.Data
Daily rainfall data at730stations in China are used in this study.This dataset was collected and compiled by the National Meteorological Information Centre of the China Meteorological Administration(CMA).Only 625out of730stations with less missing data and cov-ering the time period1980–2000are included in the analysis here.These625stations are not evenly dis-tributed.More are located in the eastern part of China and less in the western and northwestern part of China (Fig.1a).The summer season in this paper is de?ned as the period from1May to31August.
Considering the geographical limitation of station data(especially over oceans),outgoing longwave ra-diation(OLR)from the National Oceanic and Atmospheric Administration(NOAA)polar-orbiting satellites(Liebmann and Smith1996)are used as a proxy for large-scale convection and rainfall in the tropics.However,in the extratropics and regions with inhomogeneous surface conditions,the cloud data produced by International Satellite Cloud Climatology Project(ISCCP)can represent large-scale rainfall variability better than OLR(Kang et al.1999). Because the large-scale rainfall in East Asia during boreal summer is generally related with deep convec-tion(Hsu2005),we adopt daily deep convective cloud amount averaged from ISCCP D13-h data(Rossow and Schiffer1999)during the period from1984to2004 (whose starting time is July1983).In addition,daily horizontal winds,geopotential height,and vertical velocity are taken from National Centers for Environ-mental Prediction(NCEP)Department of Energy (DOE)Atmospheric Model Intercomparison Project (AMIP-II)reanalysis data(hereafter as NCEP2; Kanamitsu et al.2002).
The MJO index used here is the Real-time Multi-variate MJO(RMM;the daily RMM data are available online at https://www.sodocs.net/doc/8c12432515.html,.au/bmrc/clfor/cfstaff/matw/ maproom/RMM/RMM1RMM2.74toRealtime.txt)index of WH04.This index is the pair of normalized projec-tion time series of the?rst two leading empirical or-thogonal functions(EOFs)of the combined?elds, including158S–158N averaged850-hPa zonal wind, 200-hPa zonal wind,and OLR data(called RMM1and RMM2).Although no time bandpass?ltering is
202J O U R N A L O F C L I M A T E V OLUME22
applied,the index can strongly discriminates the 30–80-day signal.Based on this RMM index,the eastward propagation of the MJO can be categorized into eight phases,each corresponding to the geographical posi-tion of its active convective center (see Fig.7in WH04).These phases make up of a full MJO cycle originating from the western Indian Ocean and de-caying over the central Paci?c.On average,each phase lasts for about 6days.
b.Methods
While the MJO is the most important mode of trop-ical intraseasonal variability,its strength varies from event to event and from year to year.According to the magnitude and propagation of RMM index,strong MJO can be distinguished from weak MJO.MJO cycles with ????????????????????????????????????????
RMM121RMM22p .1and characterized by a clear and continuous eastward propagation (anticlockwise rotation in the WH04diagram)are denoted as ‘‘strong’’MJO.Besides,a few days (about 1–3days)before/after the start/end of the strong MJO event with
????????????????????????????????????????
RMM121RMM22p .0:9are permitted in the event.Here,0.9is the climatological mean MJO amplitude (Pohl and Matthews 2007).‘‘Weak’’MJO is de?ned as ????????????????????????????????????????
RMM121RMM22p ,0:9,which appears as somewhat random movement near the origin in the WH04dia-gram.The MJO phases mentioned in the following all refer to the eight phases of the strong MJO event.
Daily anomalies of rainfall and other variables were calculated by subtracting the daily climatological means from the original https://www.sodocs.net/doc/8c12432515.html,posites were made for these daily anomalies according to RMM index with the same MJO phase in boreal summer during 1980–2000(1984–2004for deep convective cloud).Although the period of deep convective cloud is dif-ferent,the nonoverlapping period of 2000–04does not distort the main features,since the composites of 1984–2000and 1984–2004are almost consistent (not shown).
Statistical signi?cance for the composites is judged using a local t test applied to the difference between two sample means.For horizontal wind,geopotential height,and moisture ?ux,which are normally dis-tributed,the composites are estimated to be signi?-cantly different from zero at the 5%(10%)level,the formula is
t 5 x
à y j j s 1m 111m 2
1=2;
e1T
where s 25?P m 1i 51ex i
à x T21P m 2
i 51ey i à y T2 em 191m 2
9à2T, x and y are the sample means for certain phase x i ei 51;m 1Tand for all data y i ei 51;m 2T,respectively.Considering the autocorrelation of daily data,the ef-fective degree of freedom is estimated.Here m 1
9is the number of strong MJO events at certain phase,and m 2
9is estimated as (Wilks 2006)
m 295m 2
e1àr T
e11r T
;
e2T
where r is the autocorrelation coef?cient and m 2is the sample size for y i .
Rainfall has very skewed probability distribution and the direct t test is not valid.Before the t test is used,we ?rst transform the original daily rainfall data to make them less skewed.Since there are lots of zeros in daily rainfall data,the traditional Box–Cox method,which requests that the data be positive,is not applicable.We adopt the method introduced by Yeo and Johnson (2000),which can transform the zero and negative data directly.The transformation
is
F I
G .1.(a)Locations of 625meteorological stations in China.(b)Averaged daily rainfall (mm day 21)in China during May–August 1980–2000from station data.Only those greater than 3mm day 21are shaded.The rectangle indicates the location of Southeast China (east of 1058E and south of 358N,231stations).
15J ANUARY 2009Z H A N G E T A L.
203
ex 11Tl à1
l
;x $0
àeàx 11T
2àl à1h i 2àl
;
x ,0
8>><
>>:;e3T
where l is set to be the best theoretical value 0.555.After the transformation,the data have approximately normal distribution and the t test can be used.
In view of the uneven spatial distribution of the sta-tions,the ?eld test is required after the t test.For each MJO phase,the null hypothesis of the ?eld test is that there are no signi?cant differences between the com-posite anomalous ?eld and the zero ?eld (Shiryaev 1996).The power function of the test is
b ep T5P eX .m T51àP eX #m T
51àX
m i 50
C i K p i e1àp TK ài ;
e4T
where p is the signi?cance level of the t test,K is the
total station numbers,and m is the number of the t tests that are signi?cant at the p level.Let a be the level of the ?eld signi?cance,the null hypothesis is rejected if b ep T,a .
Set K 5625,p 50:05,a 50:05,and the critical value m c can be achieved according to b ep T5a 50:05[i.e.,P eX #m c T51àa 50:95].If m .m c ,then the null hypothesis is rejected.
Based on the De Moivre–Laplace theorem,we can get
X m c i 50
C i K p i e1àp TK ài ’F
m c
àKp 10:5??????????Kpq p
àF
àKp à0:5??????????Kpq
p 50:95;e5T
where F x eTis the standard normal distribution function,q 51àp 50:95.If K 5625,then m c 539.If K 5231,then m c 516.
3.Impact of the MJO on summer rainfall in Southeast China The summer rainfall is the largest among the four seasons in China and exhibits obvious regional features (Fig.1b).The daily rainfall amounts are larger than 3mm in the southern and eastern part of China and less than 3mm in the western,northern,and northwestern part of China.In this study,we focus on Southeast China (south of 358N and east of 1058E)where most rainfall amounts are greater than 6mm day 21.
Using the MJO index,composites of rainfall anoma-lies for the eight MJO phases are obtained (Fig.2).Under the in?uence of the MJO,rainfall patterns dis-
play obvious changes in Southeast China.In the ma-jority of regions,the rainfall variations are in excess of 0.5mm day 21(about 10%relative to the climatological mean)and even 2–3mm day 21in some regions and phases.Although not all stations experience the same rainfall variations in certain MJO phase (Fig.3a),in general,rainfall in Southeast China is enhanced in the ?rst four phases (phases 1–4)and suppressed in the second four phases (phases 5–8).
Such characteristics are more clearly shown in Fig.3b.The regional mean rainfall anomalies in the composites are all positive in phases 1–4and negative in phases 5–8.That means there is an abrupt switch from phases 4to 5.The variations are more signi?cant in phases 4and 7when the number of stations with positive rainfall anom-alies are much larger (smaller)than those with negative rainfall anomalies.The maximum positive anomaly is 1.1mm day 21in phase 4and the maximum negative anomaly is 0.7mm day 21in phase 7,which occupy 15%and 13%relative to the climatological mean,respectively.
On the whole,the variation of rainfall in Southeast China is prominent associated with the evolution of the MJO.In the ?rst (second)four MJO phases,the rainfall in Southeast China is enhanced (suppressed)when there are more (less)stations with increased rainfall.Because the intensity and scope of rainfall anomalies affected by weak MJO (Fig.2i)are much weaker (,1mm day 21)and smaller than that by strong MJO,we just focus on the impact of strong MJO on summer rainfall in Southeast China in the following.
It has been pointed out in section 2a that the deep convective cloud amount from ISCCP can well represent the large-scale rainfall beyond the coverage of station data.The variation of deep convective cloud amount is consistent with that of station rainfall in Southeast China (Fig.4;i.e.,an abrupt switch between the ?rst four phases and the second four phases).Moreover,the deep convective cloud extends northeastward from South-eastern China to the East China Sea,Korea Peninsula,and southern Japan,and southward to the South China Sea.Hence,the rainfall variation is not a land-bound phenomenon,it occurs across larger areas including oceans.Such variation may be related to the changes in the large-scale circulation.
4.Evolution of large-scale physical ?elds associated with the MJO There are mainly two necessary large-scale conditions favorable for the development of rainfall.One is abun-dant moisture and the other is strong upward motion.It is found that both conditions are satis?ed in Southeast China in summer from the climatological mean.However,
204
J O U R N A L O F C L I M A T E V OLUME 22
15J ANUARY2009Z H A N G E T A L.205
F https://www.sodocs.net/doc/8c12432515.html,posites of rainfall anomalies(mm day21)in each of the MJO phases and the weak
MJO.Black open and closed circles denote the stations that pass the90%and95%signi?cant
tests,respectively,based on the Student’s t test.The numbers of stations that pass the90%and
95%signi?cant tests are given at the bottom-left corner.The number is greater than the critical
value16for each phase at95%signi?cance,so each passes the?eld signi?cance test and clas-
sifying the rainfall anomalies into eight categories is reasonable.
the moisture and vertical motions experience variations on intraseasonal time scales.In the following,the evo-lutions of moisture and vertical motions associated with the MJO and the corresponding large-scale ?elds will be studied.
a.MJO-related convection and circulation
To investigate how the MJO affects the variation of rainfall in Southeast China,it is necessary to explore the evolution of MJO-related convection and circulation ?rst (Fig.5).In phase 1,the convective is located over the tropical eastern Paci?c and western Indian Ocean.From phases 2to 4,the convection is mainly over the tropical Indian Ocean.It gradually moves northward and extends eastward into the equatorial western Paci?c (Figs.5b–d).The MJO-related convection leaves the Indian Ocean in phase 5and enters the Maritime Con-tinent and the tropical Paci?c Ocean east of 1208E.Then it strengthens and moves northward,centered at 158N and 1008–1508E (Figs.5f,g).In phase 8,the con-vection decays and moves to the eastern Paci?c Ocean again (Fig.5h).
Combining the intraseasonal characteristics of rain-fall patterns in section 3,the rainfall is enhanced when the MJO is located over the Indian Ocean and sup-pressed when it is located over the western Paci?c.That
is,the rainfall in Southeast China experiences variation associated with the eastward propagation of the MJO from the Indian Ocean to the Paci?c Ocean.Corre-spondingly,wind anomalies also change with the prop-agation of the MJO.The details will be described in section 4c due to the similar patterns between 850-hPa horizontal winds and moisture ?ux.
b.Western Paci?c subtropical high
The geopotential heights at 850and 500hPa can both be used to describe the WPSH.Here,850-hPa geo-potential height is chosen for analysis in order to connect it with the water vapor transport at low levels conve-niently.The WPSH ridgeline is located at about 258N in summer.The contours of 148and 150dagpm lie over the coast of South China and the Philippines,respectively (the ?gure is not shown,but is similar to Fig.6d).Gen-erally,when these two contours extend farther west (east)than normal,it indicates that the WPSH propa-gates westward (eastward).
The variation of the WPSH from MJO phases 1–8is explored in the following (Fig.6).In phase 1,the WPSH shifts westward and there is an east–west-oriented positive anomaly of geopotential height over the north-ern Philippines.In the following two phases,the WPSH shifts farther westward and the positive anomaly changes to be northeast–southwest oriented over the western North Paci?c (Figs.6b,c).In phase 4,the WPSH approaches the climatological mean.Since phase 5,the contour of 150dagpm retreats eastward and extends east of the Philippines,which indicates that the mon-soon trough is intensifying (Figs.6e–g).Correspondingly,the WPSH shifts eastward,and a negative anomaly of geopotential height is located over the South China Sea and the western North Paci?c.In phase 8,the monsoon trough disappears and the WPSH approaches the cli-matological mean state again.
The above analysis indicated that the intraseasonal variation of WPSH may also be regulated by the MJO.When the MJO is located over the Indian Ocean,the WPSH shifts westward and geopotential height in-creases over the western North Paci?c.When the MJO propagates eastward and northward into the subtropical western Paci?c,the monsoon trough deepens and the geopotential height decreases.The variation of the WPSH will directly impact the water vapor transport to its northwest.
c.Water vapor transport and precipitable water From the climatological mean,there are three main branches of northward water vapor transport converg-ing in Southeast China in summer.They are the south-westerly ?ow from the Bay of Bengal,the southerly
?ow
F I
G .3.(a)Number of stations with rainfall anomalies .0.5mm day 21and ,20.5mm day 21among 231stations for the eight MJO phases.(b)Composite regional mean rainfall anomalies (mm day 21)of Southeast China for the eight MJO phases.
206
J O U R N A L O F C L I M A T E V OLUME 22
from the South China Sea,and the southeasterly ?ow from the northwestern part of the WPSH.The con-vergence of moisture makes the abundant rainfall in Southeast China possible.
However,on intraseasonal time scales,the moisture transport switches from being increased to being de-creased associated with the variation of the WPSH (Fig.7).In phase 1,corresponding to the east–west-oriented positive anomaly of geopotential height and anomalous anticyclonic circulation over the Philippine,abnormal southwesterlies at the northwestern part of anticyclone carry more moisture to Southeast China.In the following two phases,when the positive geo-potential height and anticyclonic anomalies move to the western North Paci?c and change to be northeast–southwest oriented,the water vapor transport turns to be northward or northeastward.The moisture transport approaches to the climatological mean in phase 4.
When
F I
G https://www.sodocs.net/doc/8c12432515.html,posites of deep convective cloudiness anomalies (%)in MJO phases (a)1,(b)2,(c)3,(d)4,(e)5,(f)6,
(g)7,and (h)8.
15J ANUARY 2009Z H A N G E T A L.
207
208J O U R N A L O F C L I M A T E V OLUME22
F IG.5.As in Fig.4but for composites of OLR and850-hPa wind vector anomalies in MJO Dark and light
shaded regions denote OLR anomalies,25W m22and.5W m22,respectively.Heavy vectors indicate that
the wind anomalies are signi?cantly different from zero at the95%level(based on the t test)in at least one of
the wind components(zonal or meridional).
15J ANUARY2009Z H A N G E T A L.209
F IG.6.As in Fig.4but for composites of850-hPa geopotential height and its anomalies in MJO.The solid lines are
geopotential height.Dark and light shaded areas are its positive and negative anomalies,respectively.
210J O U R N A L O F C L I M A T E V OLUME22
F IG.7.As in Fig.4but for composites anomalies of moisture?ux(kg m21s21),precipitable water,and850-hPa geopotential height (geopotential meters)in MJO.Heavy arrows indicate moisture?ux where they are deemed to be signi?cantly different from zero at the 90%level based on the t test.Shaded areas represent increased(red)or decreased(blue)precipitable water.Black solid lines depict geopotential heights that are signi?cantly different from zero at the90%level based on the t test.
the MJO propagates eastward and northward into the subtropical western Paci?c,there are strong negative geopotential height anomaly and an anomalous cyclonic circulation over the western North Paci?c.Hence,pre-dominant northeasterly or easterly anomalies prevail over Southeast China(Figs.7e–g).In phase8,moisture transport again approaches to the climatological mean. By comparison,the magnitudes of northeasterly or easterly anomalies in phases5–7are obviously larger than that of southerly or southwesterly anomalies in phases1–3,which suppress the moisture transport into Southeast China to a great extent.Corresponding to the variation of moisture transport,the precipitable water in Southeast China also turns from being enhanced to being reduced(Fig.7).The increase/decrease of rainfall is in conjunction with enhancement/suppressment of precipitable water.
d.Vertical motions
In summer,there are upward motions in the entire troposphere associated with active deep convections and strong rainfall in Southeast China(?gure is not shown).Such rising motions exhibit variations associ-ated with the MJO propagation(Fig.8).The ascent in Southeast China(about208–358N)gradually enhances from phases1to4(Figs.8a–d),and the anomalous centers are just located at the latitudes with increased deep convective cloud and rainfall(Figs.4a–d and2a–d). Then the ascent in Southeast China weakens since phase5(Figs.8e–g),and the anomalous centers are just located at the regions with decreased deep convective cloud and rainfall(Figs.4e–g and2e–g).In phase8,the ascent around258N strengthens,corresponding to the increasement of clouds and rainfall(Figs.8h,4h,and 2h).In general,the enhanced(suppressed)deep con-vective cloud and rainfall are related to anomalous upward(downward)vertical motion.
5.Possible mechanisms of the impact of the MJO on
summer rainfall in Southeast China
It is indicated in section3that the rainfall in South-east China is enhanced in the?rst four phases(when the MJO is over the Indian Ocean)and suppressed in the second four phases(when the MJO is over the western Paci?c).In the following,two different
mechanisms
F IG.8.As in Fig.4but for meridional–height cross section of composite vertical velocity
anomalies(0.005Pa s21)in MJO,from108S to408N meridionally and from1000to100hPa
vertically.The zonal average of the cross section is taken from1058to1208E.The shaded area
means upward motions.The contour interval(CI)is0.005hPa s21.
15J ANUARY2009Z H A N G E T A L.211
212J O U R N A L O F C L I M A T E V OLUME22
F IG.9.As in Fig.4but for composites of OLR anomalies and850-hPa U wind in MJO.The solid and dashed line
denote the westerly and easterly,respectively.The heavy black line is the boundary of the easterly and westerly.
The gray areas are terrain and negative OLR anomalies(thick gray).
(i.e.,remote and local effects)are suggested on how the MJO affects rainfall in Southeast China when the MJO convective centers are located over two places.
a.The MJO acts as a Rossby wave source and
in?uences the rainfall through the westerly
waveguide
The tropical MJO heating can act as the source of Rossby waves.According to the theoretical and model studies on Rossby waves(Hoskins and Karoly1981;Kuo et al.2001),the energy of Rossby waves only propagates in the westerly(hereafter the westerly waveguide)and is trapped in the easterly.Hence,the energy of tropical Rossby waves forced by the MJO can propagate north-eastward into Southeast China and then affect the rainfall if there is a westerly waveguide and if it is lo-cated upstream of China.
When the MJO is located over the Indian Ocean (phases1–4),two westerly centers are located over the Indian Ocean and the North Paci?c,respectively.These two centers are connected with each other by the westerly over South Asia and East Asia(Figs.9a–d). Meanwhile,the Rossby wave inspired by the MJO is located upstream of China.The wave propagation can be clearly seen in previous studies(e.g.,Fig.5.9in Mo and Paegle2005).Thus,the wave energy can reach Southeast China through the westerly waveguide and contribute to the enhancement of rainfall.
After the MJO moves to the Maritime Continent,the westerly waveguide is interrupted or Rossby wave en-ergy is trapped by the easterly(phases5and6;Figs. 9e,f).When the MJO propagates into the western Pa-ci?c,it is not located upstream of China(phases7and8; Figs.9g,h).In these phases,the energy of the tropical Rossby wave cannot reach Southeast China.
b.The MJO arouses the variation of meridional
circulation and in?uences the rainfall through
the retreat of WPSH
Aforementioned analysis indicates that the energy of Rossby waves cannot affect the summer rainfall in Southeast China when the MJO is located over the western Paci?c.Thus,there should be another process in?uencing the rainfall in Southeast China. Associated with the eastward and northward propa-gation of the MJO,the anomalous meridional
circulation
F IG.10.As in Fig.4but for meridional–height cross section of composite vertical velocity
anomalies(0.005Pa s21)in MJO,from108S to408N meridionally and from1000to100hPa
vertically.The zonal average of the cross section is taken from1208to1408E.The shaded area
means upward motions.The CI is0.005hPa s21.
15J ANUARY2009Z H A N G E T A L.213
(local Hadley circulation)over the Paci?c Ocean evolves correspondingly.When the MJO enters the western Paci?c,the upward branch of anomalous local Hadley circulation in the tropics moves northward into the sub-tropics and its maximum is between 300and 400hPa (Figs.10).According to the relationship between ver-tical velocity (v )and relative vorticity (z )in pressure coordinates,
d z dt 5?z ?t 1eV á=Tz 1v ?z ?p 5àf =áV 5f ?v ?p
;e6T
where t stands for time,V is the horizontal wind vector,and f is the Coriolis parameter (f .0in the Northern Hemisphere).Here ?v =?p .0in the lower levels.Therefore,d z =dt .0,which arouses abnormal cyclonic circulation and the subsequent deepening of the mon-soon trough and eastward shift of the WPSH.The re-treat of WPSH results in less moisture transported into Southeast China and the rainfall is suppressed.
6.Summary
Based on the RMM index proposed by WH04,it is con?rmed in this study that the MJO has considerable in?uence on the rainfall in Southeast China.Rainfall
patterns in Southeast China undergo substantial changes (i.e.,from enhancement to suppression)as the convec-tive center of the MJO moves eastward from the trop-ical Indian Ocean to the western Paci?c.A sudden change is particularly obvious between MJO phases 4and 5when the regional mean rainfall anomaly switches from being positive to being negative.The rainfall in-creases by about 15%in phase 4and decreases by about 13%in phase 7relative to the climatological mean.The large-scale ?elds also change during the life cycle of the MJO.The longitudinal position of the WPSH,the characteristics of moisture,and vertical motions in Southeast China are almost distinct in the ?rst and the second four MJO phases.When the MJO is located over the Indian Ocean,the WPSH shifts westward.More moisture is transported to Southeast China and the precipitable water is increased.Meanwhile,upward motions are also enhanced in Southeast China.After the MJO enters the western Paci?c,the WPSH shifts eastward.Less water vapor is transported into South-east China and the precipitable water is decreased.Local ascent is suppressed at the same time.
Two mechanisms are suggested for the impact of the MJO on summer rainfall in Southeast China.One is a remotely forcing mechanism (Fig.11a).According
to
F I
G .11.The schematic diagrams of the two possible mechanisms:(a)mechanism 1and
(b)mechanism 2.
214
J O U R N A L O F C L I M A T E V OLUME 22
the theoretical and model studies,the energy of Rossby waves can only propagate in the westerlies.Considering that a strong low-level westerly prevailed from the In-dian Ocean to Southeast China in the?rst four phases, we suggest that when the MJO is located over the Indian Ocean,the energy of Rossby waves forced by the MJO-related convective heating can propagate northeastward through the westerly waveguide and reach Southeast China.The arrived energy contributes to rainfall en-hancement.
The other mechanism is the local effect(Fig.11b). When the MJO enters the western Paci?c,the Rossby wave energy cannot reach China.The convective center of the MJO moves about108latitude poleward com-pared to that over the Indian Ocean.Associated with the northward propagation,the meridional circulation with anomalous ascent near the equatorial western Paci?c and descent in the subtropical western North Paci?c in the ?rst four phases is reversed in the second four phases as the tropical upward branch of anomalous local Hadley circulation moves northward into the subtropics.Such abnormal meridional circulation leads to the deepening of the monsoon trough and the eastward shift of the WPSH. Consequently,the moisture transport toward Southeast China is decreased and,hence,rainfall is suppressed. Considering the complexity of the mechanisms,our results are primary and more in-depth study is needed in the future.
Acknowledgments.This study was supported by the Knowledge Innovation Program of the Chinese Acad-emy of Sciences Grants KZCX2-YW-217and IAP07116 and by the public welfare research foundation of China Grant GYHY200706042.The authors thank Dr. Chiding Zhang and the anonymous reviewers for pro-viding useful comments to this paper.Discussions with Matthew Wheeler,Mathew Barlow,Alexie Donald, Hai Lin,and Huang-Hsiung Hsu were also helpful.
REFERENCES
Barlow,M.,M.Wheeler,B.Lyon,and H.Cullen,2005:Modula-tion of daily precipitation over Southwest Asia by the Madden–Julian oscillation.Mon.Wea.Rev.,133,3579–3594. Carvalho,L.M.V.,C.Jones,and B.Liebmann,2004:The South Atlantic convergence zone:Intensity,form,persistence,re-lationships with intraseasonal to interannual activity,and extreme rainfall.J.Climate,17,88–108.
Chen,T.C.,and M.Murakami,1988:The30–50-day variation of convective activity over the western Paci?c Ocean with an emphasis on the northwestern region.Mon.Wea.Rev.,116, 892–906.
——,M. C.Yen,and M.Murakami,1988:The water vapor transport associated with the30–50-day oscillation over the Asian monsoon regions during1979summer.Mon.Wea.Rev., 116,1983–2002.Donald,A.,H.Meinke,B.Power,A.de H.N.Maia,M.C.
Wheeler,N.White,R.C.Stone,and J.Ribbe,2006:Near-global impact of the Madden-Julian Oscillation on rainfall.
Geophys.Res.Lett.,33,L09704,doi:10.1029/2005GL025155. Higgins,R.W.,and W.Shi,2001:Intercomparison of the principal modes of interannual and intraseasonal variability of the North American Monsoon System.J.Climate,14,403–417. Hoskins,B.J.,and D.J.Karoly,1981:The steady linear response of a spherical atmosphere to themal and orograhpic forcing.
J.Atmos.Sci.,38,1179–1196.
Hsu,H.H.,2005:East Asian monsoon.Intraseasonal Variability in the Atmosphere–Ocean Climate System,https://www.sodocs.net/doc/8c12432515.html,u and D.
E.Waliser,Eds.,Springer,63–94.
Jeong,J.H.,B.-M.Kim,C.-H.Ho,and Y.-H.Noh,2008:Systematic variation in wintertime precipitation in East Asia by MJO-induced extratropical vertical motion.J.Climate,21,788–801. Jones,C.,2000:Occurrence of extreme precipitation events in California and relationships with the Madden–Julian oscilla-tion.J.Climate,13,3576–3587.
——,D.E.Waliser,https://www.sodocs.net/doc/8c12432515.html,u,and W.Stern,2004:Global oc-currences of extreme precipitation events and the Madden–Julian oscillation:Observations and predictability.J.Climate, 17,4575–4589.
Kanamitsu,M.,W.Ebisuzaki,J.Woollen,S.-K.Yang,J.J.Hnilo, M.Fiorino,and G.L.Potter,2002:NCEP–DOE AMIP-II Reanalysis(R-2).Bull.Amer.Meteor.Soc.,83,1631–1643. Kang,I.S.,C.H.Ho,Y.K.Lim,and https://www.sodocs.net/doc/8c12432515.html,u,1999:Principal modes of climatological seasonal and intraseasonal variations of the Asian summer monsoon.Mon.Wea.Rev.,127,322–340. Kuo,H.C.,J.-H.Chen,R.T.Williams,and C.-P.Chang,2001: Rossby waves in zonally opposing mean?ow:Behavior in Northwest Paci?c summer monsoon.J.Atmos.Sci.,58,1035–1050.
Lau,K.M.,and P.H.Chan,1986:Aspects of the40-50-day os-cillation during the northern summer as inferred from out-going longwave radiation.Mon.Wea.Rev.,114,1354–1367.——,G.J.Yang,and S.H.Shen,1988:Seasonal and intraseasonal climatology of summer monsoon rainfall over East Asia.Mon.
Wea.Rev.,116,18–37.
Liebmann,B.,and C.A.Smith,1996:Description of a complete (interpolated)OLR dataset.Bull.Amer.Meteor.Soc.,77, 1275–1277.
Lorenz,D.J.,and D.L.Hartmann,2006:The effect of the MJO on the North American Monsoon.J.Climate,19,333–343. Madden,R.A.,and P.R.Julian,1971:Detection of a40-50-day oscillation in the zonal wind in the tropical Paci?c.J.Atmos.
Sci.,28,702–708.
——,and——,1972:Description of global scale circulation cells in the Tropics with a40-50-day period.J.Atmos.Sci.,29,1109–1123.——,and——,1994:Observations of the40–50-day tropical oscillation—A review.Mon.Wea.Rev.,122,814–837.
Mo,K.C.,and R.W.Higgins,1998:Tropical in?uences on Cal-ifornia precipitation.J.Climate,11,412–430.
——,and J.Paegle,2005:Pan-America.Intraseasonal Variability in the Atmosphere–Ocean Climate System,https://www.sodocs.net/doc/8c12432515.html,u and
D.E.Waliser,Eds.,Springer,95–124.
Paegle,J.N.,L.A.Byerle,and K.Mo,2000:Intraseasonal mod-ulation of South American summer precipitation.Mon.Wea.
Rev.,128,837–850.
Pohl,B.,and A.J.Matthews,2007:Observed changes in the lifetime and amplitude of the Madden–Julian oscillation associated with interannual ENSO sea surface temperature anomalies.J.Cli-mate,20,2659–2674.
15J ANUARY2009Z H A N G E T A L.215
Rossow,W.B.,and R.A.Schiffer,1999:Advances in understanding clouds from ISCCP.Bull.Amer.Meteor.Soc.,80,2261–2288. Shiryaev,A.N.,1996:Probability:No.95in Graduate Texts in Mathematics.2nd ed.Springer-Verlag,621pp.
Waliser,D.E.,2006:Intraseasonal variations.The Asian Mon-soon,B.Wang,Ed.,Springer,203–257.
Wang,B.,2006:The Asian Monsoon.Springer,787pp.
——,and H.Rui,1990:Synoptic climatology of transient tropical intraseasonal convection anomalies:1975-1985.Meteor.At-mos.Phys.,44,43–61.Wheeler,M.,and H.H.Hendon,2004:An all-season real-time multivariate MJO index:Development of an index for mon-itoring and prediction.Mon.Wea.Rev.,132,1917–1932. Wilks,D.S.,2006:Statistical Methods in the Atmospheric Sciences.
2nd ed.Academic Press,627pp.
Yeo,I.K.,and R.A.Johnson,2000:A new family of power transformation to improve normality or symmetry.Biometrika, 87,954–959.
Zhang,C.,2005:Madden-Julian Oscillation.Rev.Geophys.,43, RG2003,doi:10.1029/2004RG000158.
216J O U R N A L O F C L I M A T E V OLUME22