The Variability of the Asian Summer Monsoon

Journal of the Meteorological Society of Japan, Vol. 85B, pp. 21--54, 2007 21

The Variability of the Asian Summer Monsoon

Yihui DING

National Climate Center, China Meteorological Administration, Beijing, P.R. China

(Manuscript received 8 September 2006, in final form 26 February 2007)

Abstract

The Asian monsoon is the most significant component of the global climate system. During recent two decades, a more and more efforts have been made to study the Asian monsoon. A substantial achieve- ment has been made in basic physical processes, predictability and prediction since the MONEX of 1978–1979. The major advance in our new understanding of the variability of the Asian summer mon- soon has been highlighted in this paper. The present paper is structured with four parts. The first part is the introduction, indicating the new regional division of the Asian monsoon system and significant events of the history of the Asian monsoon research. The second part discusses the annual cycle and sea- sonal march of the Asian monsoon as the mean state, with a special emphasis on the onset, propagation, active-break cycle and withdrawal of the Asian summer monsoon. The process and mechanism of the earliest onset of the Asian summer monsoon which takes place in the near-equatorial East Indian ocean-central and southern Indochina Peninsula have been well documented. The third part deals with the multiple scale variability of the Asian summer monsoon, including the intraseasonal, interannual and inter-decadal variability. Their dominant modes such as 10–20 day and 30–60 day oscillations for the intraseasonal variability, the Tropospheric Biennial Oscillation (TBO), the Indian Ocean Dippole Mode (IODM) and teleconnection patterns for interannual variability and the 60-year oscillation for the inter-decadal variability, as well as related SST-monsoon relationship and land-monsoon relationship have been discussed in more details. The fourth part is the conclusion, summarizing the major findings and proposing future work.

1. Introduction was just thought of as the northward extension of the Indian monsoon, on the one hand, and The Asian monsoon is characterized with a on the other hand, the summer monsoon over distinct seasonal reversal of wind and rapid al- the western North Pacific summer monsoon ternation of dry and wet or rainy season in the (WNPSM) was fully taken to be the eastward annual cycle, which is concert with the seasonal extension of the Indian monsoon (Ding 1994). reversal of the large-scale atmospheric heating However, now it has been increasingly realized and steady circulation features (Webster et al. that the Asian monsoon system should further 1998; Ding and Chan 2005; Trenberth et al. extend to incorporate these two regions due to 2006). About two decades ago, the Asian mon- similar features of monsoon climate, as a large soon was mainly viewed as the Indian or the amount of literatures has been contributed to South Asian monsoon (ISM) in the English lit- the study of the East Asian summer monsoon erature. The summer monsoon over East Asia (EASM) (Chang 2004) and the WNPSM in re- cent two decades (Murakami and Matsumoto Corresponding author: Yihui Ding, National Cli- 1994; Ueda and Yasunari 1996; Wu and Wang mate Center, China Meteorological Administra- 2001; Wu 2002; Wang and Lin 2002; Wang tion, No. 46, Zhongguancun Nan Da Jie, Haidian et al. 2005a; Wang et al. 2005c). Thus, the District, Beijing 100081, P.R. China. E-mail: [email protected] Asian-Pacific monsoon is demarcated into three ( 2007, Meteorological Society of Japan sub-systems: the Indian summer monsoon, the 22 Journal of the Meteorological Society of Japan Vol. 85B

Fig. 1. western North Pacific summer monsoon and the EASM region. Wang and Lin (2002) believe the East Asian summer monsoon (Fig. 1a). The that the ISM and WNPSM are tropical mon- EASM domain defined by Wang and Lin (2002) soons in which the low level winds reverse includes the region of 20–45N and 110–140E, from winter easterlies to summer westerelies, covering eastern China, Korea, Japan and the whereas the EASM is a sub-tropical monsoon adjacent marginal seas. This definition does in which the low-level winds reverse primarily not fully agree with the conventional notion from winter northerlies to summer southerlies. used by Chinese meteorologists (Tao and Chen However, if the SCS region is included in the 1987; Ding 1994; Ding and Chan 2005), who EASM, the EASM should be a hybrid type of usually includes the South China Sea (SCS) in tropical and subtropical monsoon. July 2007 Y.H. DING 23

Fig. 1. (a) Climatological July–August mean precipitation rates (color shading in mm/day) and 925 hPa wind vectors (arrow) in the Asian-Australian monsoon region. The precipitation and wind climatology are derived from CMAP (Xie and Arkin 1997) (1979–2000) and NCEP/NCAR reanalysis (1951–2000), respectively. The three boxes define major summer precipitation areas of the Indian tropical monsoon (5N–27.5N, 65E–105E), western North Pacific tropical monsoon (5N–22.5N, 105E–150E), and the East Asian subtropical monsoon (Wang et al. 2005). (b) Cli- matological mean tropical easterly jet (150–100 hPa layer) over the Asian-African monsoon region averaged for summer (JJA) of 1958–2002. Isolines (Dashed: easterly; and solid: westerly wind) de- note mean wind speed for 150–100 hPa layer. CD and CID are jet axis. AB is the demarcation line of entrance (right) and exit (left) zone. Div: divergence, Con: convergence, Up: upward motion, Dn: downward motion, R: rainfall region, and D: dry region. Unit: ms1. (c) Same as (b), but for rainfall and divergence (Unit: 106 s1). Shaded areas: >4 mm day1 (Chen et al. 2006)

In the ISM region, the monsoon or seasonal south of the huge South Asian high. The major changes of winds and rainfall could be inter- monsoon rainfalls are located in the right sec- preted as a result of northward seasonal migra- tor of the jet entrance zone where the upper- tion of the east-west oriented precipitation belt level divergence and low-level convergence, i.e. accompanied by the inter-tropical convergence upward motion is observed (Fig. 1b, c) (Hoskins zone (ITCZ or TCZ), while in the EASM region and Wang 2006; Chen et al. 2006). In short, the seasonal march of the monsoon is displayed the Asian monsoon system is a huge monsoon in the northward excursion of the planetary system and constitutes an essential part of frontal zone or Meiyu-Baiu frontal system. So, the Asian-Australian monsoon system. On the one of main differences between these two other hand, it is closely related to the African monsoon sub-regions is the different effect of monsoon system through the TEJ. mid-and high latitude events. In the WNPSM The research of the Asian monsoon has a region, the ITCZ is the dominating weather long history (Webster 2006), but the substan- system which is the major birthplace of ty- tial progress has been made since 1960’s phoons and tropical convective systems. At the through a number of international and regional upper level, the Asian monsoon region is domi- monsoon projects and field experiments such nated by the tropical easterly jet (TEJ) to the as the International Indian Ocean Expedition 24 Journal of the Meteorological Society of Japan Vol. 85B

(IIOE) in the mid-1960s and in 1975–1976, the the energy and water cycle of the Asian mon- FGGE Monsoon Experiments in 1978–1979 soon; (5) synoptic systems, mesoscale process (MONEX) (Krishnamurti 1985), the Coupled and diurnal variations associated with the Ocean-Atmosphere Response Experiment in Asian monsoon; (6) large-scale physical pro- 1992–1993 (TOGA-COARE) (Webster and Lu- cesses and dynamics of the Asian monsoon, kas 1992), the GEWEX Asian Monsoon Ex- including dynamics of tropical waves, coupled periment (GAME) in 1996–2000 (GAME ISP, monsoon system and SST-monsoon relation- 1998), the South China Sea Monsoon Experi- ship, ecosystem-monsoon relationship and land- ment (SCSMEX) in 1996–2000 (Lau et al. monsoon interaction, snow-monsoon relation- 2000; Ding and Liu 2001; Johnson et al. 2004), ship, Rossby wave-teleconnection theory and the Bay of Bengal Monsoon Experiment (BO- dynamical and thermal effects of the Tibetan MEX) in 1999 (Bhat et al. 2001) and the Joint Plateau; (7) predictability and prediction of the Air-Sea Monsoon Interactive Experiment (JAS- Asian monsoon; (8) the monsoon evolution in MINE) in 1999 and the Arabian Sea Monsoon paleo-climate including centennial and millen- Experiment (ARMEX) in 2002 (Webster et al. nial variability, and (9) impacts of the Asian 2002). At the same time, a number of books, monsoon on the socio-economic sectors, espe- monographs and review papers written in En- cially on the agriculture and water resources. glish summarizing the major scientific achieve- The present paper will make a review to high- ments of the Asian monsoon in different periods light major achievements and new findings con- have been published, including Monsoon Mete- cerning the above-described items (2) and (3) orology (Ramage 1971), Southwest Monsoon which are central issues of the Asian monsoon. (Rao 1976), Monsoons (Fein and Stephens Due to the space limitation, the review will be 1987), Monsoon Meteorology (Chang and Krish- confined in the aspect of the summer monsoon. namurti 1987), Monsoon over China (Ding 2. Annual cycle and seasonal march of 1994), The East Asian Monsoon (Chang 2004), the Asian monsoon The Global Monsoon System (Chang et al. 2005) and The Asian Monsoon (Wang 2006). 2.1 The onset and the seasonal march of the They provide international scientific commu- Asian summer monsoon nities with continuously updated input of The Asian monsoon shows a strong annual knowledge of observations, processes, dy- cycle which distinguishes it from other mon- namics, prediction and socio-economic impact soon systems over the world that have much of the Asian monsoon. Now, it has been realized weaker annual cycles. Observations indicate that the Asian monsoon can not only have sig- that the largest amplitude of the annual cycle nificant regional implications for occurrence of of precipitation, zonal and meridional winds at floods and droughts and other significant disas- low-level occurs in the Asian monsoon regions trous events, but also can exert very important where there is the strongest atmospheric heat- influences on the global climate system and ing or heat sources driving the monsoon. In the the global climate prediction through, e.g., the process of the annual cycle, the onset of the tropical-extratopical interaction, the monsoon- Asian summer monsoon is a key indicator char- ENSO relationship and the global teleconnec- acterzing the abrupt transition from the dry tion (Webster et al. 2005b; Ding and Wang season to the rainy season and subsequent sea- 2005; Wang et al. 2005a). sonal change (Lin and Wang 2002; Qian et al. During recent two decades, a more and more 2002; Wang and Ding 2006). It also marks the attention has been paid to study the Asian beginning of the summer monsoon season and monsoon. Recent studies on the Asian monsoon arrival of monsoon rainfalls. Numerous investi- have been devoted to the following aspects: (1) gators have studied this problem from the re- the global perspectives of the Asian monsoon; gional perspectives. It is to some extent difficult (2) the seasonal march and annual cycle of to obtain a unified and consistent picture of the the Asian monsoon including the onset, active- climatological onset of the Asian summer mon- break cycle and withdrawal; (3) multiple-scale soon in different regions due to differences in i.e. intraseasonal, interannual and inter- data, monsoon indices and definitions of mon- decadal variability of the Asian monsoon; (4) soon onset used in these investigations. Ding July 2007 Y.H. DING 25

(2004) has summarized the climatological dates of the onset of the Asian summer monsoon in different monsoon regions based on various sources, with dividing the whole onset process into four stages: (1) Stage 1 (late in April or early in May): the earliest onset in the conti- nental Asia is often observed in the central Indochina Peninsula late in April and early in May. Over the ocean, one may trace earlier on- set in the near-equatorial East Indian Ocean in late April. (2) Stage 2 (from mid- to late May): Fig. 2. Climatological onset dates of the this stage is characterized by the areal extend- Asian summer monsoon for 1961–2000. ing of the summer monsoon, advancing north- Unit: pentad. The monsoon onset is ward up to the Bay of Bengal and eastward defined by (1) 5 day-averaged 850 hPa down to the SCS. (3) Stage 3 (from the early to zonal wind > 0, (2) OLR < 230 wm2 the middle June): this stage is well known for and (3) 5 day-averaged precipitation the onset of the Indian summer monsoon rate > 6 mm day1. Wind and precipi- and the arrival of the East Asian rainy season tation data are taken from NCEP/ such as the Meiyu over the Yangtze River Ba- NCAR Reanalysis (1961–2000), and sin and the Baiu season in Japan. At the same OLR data are taken from NOAA /CDC time, the summer monsoon over the western (1974–2004). (Ding and Liu 2006a) North Pacific extends from the SCS to the southwestern Phillipine Sea, accompanied by the increase in convective cloudness and the is established across the Somali jet into the eastward shift of the ITCZ (Wu and Wang near-equatorial Arabian Sea and a cyclonic vor- 2001; Wu 2002). (4) Stage 4 (the early and tex, so-called onset vortex (Krishnamurti et al. middle July): the summer monsoon at this 1981) is often observed. Then the Indian sum- stage can advance up to Northwest India, mer monsoon gradually advances northward North China, the Korean Peninsula (so-called across the western Indian subcontinent and Changma rainy season) and even Central Ja- eventually merges with the summer monsoon pan. After mid-July, the WNPSM and associ- simultaneously propagating northwestward ated ITCZ further marches northeastward as from the Bay of Bengal. By the middle of July monsoon rainfall and active convection abruptly the whole of the Indian subcontinent comes un- penetrate to the region of 25N, 150 –160Ein der the grip of summer monsoon. Therefore, the Pentad 42 (July 25–29) (Ueda et al. 1995). This onset of the Asian summer monsoon is one of is the maturing process of the WNPSM which most important events, as a singularity of an- may maintain till late September, the major nual cycle of the monsoon. Furthermore, the period of the typhoon season of the western onset process of the summer monsoon in a cer- North Pacific (Ueda and Yasunari 1996; Wu tain location or region is very rapid or even and Wang 2001; Wang et al. 2005c, 2006). abrupt, with taking a couple of days or one Figure 2 presents a climatological illustra- week to complete the dramatic change from tion of the onset process (Ding and Liu 2006a). dry season to rainy season. Figure 3 clearly The earliest onset occurs in the end of April shows abrupt changes in rainfalls, OLR and (the 24th pentad) in the near-equatorial East 850 hPa zonal wind around the onset dates for Indian Ocean and Sumatra. Then the onset the South China Sea, the Indochina Peninsula process propagates northeastward and north- and the Indian subcontinent, respectively. The westward, respectively. In western coast of bursts of convective activity during the onset the Indian subcontinent, the onset begins first process are very significant (see the curves of across the Kerala coast, normally by the end of OLR). The regional onset dates defined from May or early June (Ananthakrishnan and So- Fig. 3 are quite consistent with those seen in man 1988) when heavy rains lash the coastal Fig. 2 (the 4th pentad of May for the SCS, the state after the cross-equatorial low-level jet 2nd pentad of May for the Indochina Peninsula 26 Journal of the Meteorological Society of Japan Vol. 85B

and the 1st –2nd pentad of June for the Indian Peninsula). The abrupt increase in precipita- tion after the summer onset was previously indicated by various investigators, e.g., Anan- thakrishnan and Soman (1989) for south Ker- ala, Matsumoto (1997) for the central Indo- china Peninsula and Tao and Chen (1987) for the SCS. The onset of the Asian summer monsoon brings with it a dramatic change in large-scale circulation features and weather situation. Numerous investigators have examined this problem from climatological and synoptic per- spectives based on the large-scale wind, geopo- tential height, precipitation and OLR patterns (Lau and Yang 1997; Matsumoto 1997; Fong and Wang 2001; Ding and Liu 2001; Ding and Sun 2001; Wang and Lin 2002; Liu et al. 2002; Goswami 2005c; Zhang et al. 2004; Ding and Liu 2006a). Their detailed description will not be given, and instead, based on these studies, the sequence chain of significant events during the onset of the Asian summer wind may be identified below:

The development of a cross-equatiorial cur- rent in the equatorial East Indian Ocean (80–90E) and off the Somali coast, and the rapid seasonal enhancement of heat sources over the Indochina Peninsula, South China, Tibetan Plateau, and neighboring oceanic areas, thus leading to seasonal reversal of sign of the tropospheric meridional tempera- ture gradient in the Asian monsoon region. The acceleration of low-level westerly wind in the tropical East Indian ocean; Formation of a mid-tropospheric shear zone across the central Bay of Bengal to the South- east Arabian Sea in which may be embedded a cyclonic vortex in most of cases, which may even intensify into a cyclonic storm either in Fig. 3. Annual changes in 850 hPa zonal the Bay of Bengal or the Southeast Arabian 1 2 wind (Unit: ms ), OLR (Unit: Wm ) Sea (the onset vortex). and precipitation (Unit: mm day1) The eastward and northward expansion of averaged for the SCS (a), the Indochina tropical monsoon from the tropical East In- Peninsula (b) and the Indian Peninsula (c). Data sources are same as Fig. 2. dian Ocean, initiating the arrival of the sum- (Liu and Ding 2006a) mer monsoon and associated rainy season in the regions of the Bay of Bengal and Indo- china Peninsula with involvement of impacts from mid-latitudes; The significant weakening, breaking around the Bay of Bengal and eastward retreat of July 2007 Y.H. DING 27

the main body of the subtropical high, and bility (Tomas and Webster 1997; Krishna Ku- eventual onset of the SCS summer monsoon mar and Lau 1998). The change in the sign of with convective clouds, rainfall, low-level meridional gradient of tropospheric tempera- southwesterly wind and upper-level north- ture thus makes the circulation conducive for easterly wind suddenly developing in this re- symmetric instability that forces frictional gion. boundary layer convergence, overcomes the in- The ITCZ or TCZ usually used in the ISM hibition from the subsidence above the plane- region and accompanying zonally oriented tary boundary layer over the Indian continent precipitation belt rapidly moves from a mean and the Bay of Bengal caused by compensating position about 5S in winter and early spring subsiding air of the convection over the Mari- to about 20N in northern summer with the time Continent during April and May and cap- progress of the monsoon onset in the Indian ping effect of southward flow of dry and colder subcontinent. air on pre-existing moist and warm air, and Significant development and northwestward leads to explosive development of off-equatorial movement of the South Asian high at convection over India and the Bay of Bengal 200 hPa, thus leading to establishment of the and the subsequent onset of the Indian summer tropical easterly jet on its southern flank and monsoon. rapid northward jump of the upper-level The reversal of the meridional temperature westerly jet on its northern flank over the gradient has a significant regional difference, northern Tibetan Plateau. with the earliest reversal taking place in the re- gion from the eastern part of the Bay of Bengal A number of investigators have studied the to the Indochina Peninsula (Fig. 4) in the early problem on generation and maintenance of May (Wu and Zhang 1998; Zhang et al. 2004; deep tropospheric heat source in the north Ding and Liu 2006a). And this reversal process which is a central issue of the onset of the was first found in the upper troposphere and Asian summer monsoon. They have found that then it rapidly propagated downward, mainly the heating over the Tibetan Plateau plays an through vigorous convective vertical transport important role in the seasonal evolution of the of strong sensible and latent heat through (Ya- meridional gradient of heating and in trigger- nai et al. 1973; He et al. 1987; Yanai et al. ing the onset of the Asian summer monsoon 1992; Li and Yanai 1996; He et al. 2006). The (Murakami and Ding 1982; Ding 1992a; Yanai sensible heat driven air-pumping effect also et al. 1992; Wu and Zhang 1998; Ueda and Ya- plays a considerable role in vertical heat trans- sunari 1998; Zhang et al. 2004; Yanai and Wu port (Yanai and Wu 2006). Then, the reversal 2006). Goswami (2005c) used the evolution of process extends eastward to the SCS (the 4th the mean temperature of the tropospheric layer pentad of May) and westward to the Indian between 200 hPa and 700 hPa averaged be- Peninsula (the 1st –2nd pentad of June). Timing tween 30E and 110E to illustrate the sea- of the large-scale reversal of the meridional sonal evolution of the meridional gradient of temperature gradient is fully consistent with tropospheric heating. The change in the sign onset dates of the summer monsoon in these re- of meridional gradient of the tropospheric tem- gions. The cause why the earliest reversal of perature ushers in the setting up of an off- the meridional gradient occurs in the Indochina equatorial large-scale deep heat source. The Peninsula rather than the Indian Peninsula is monsoon onset itself is a problem of internal at- attributed to earlier occurrence of sensible and mospheric dynamic involving the instability of latent heating and vigorous convective activity the basic flow in tropics. However, the change over the land of Indochina Peninsula under in the large-scale heat source is a necessary favorable effect of upward motion ahead of the prerequisite. The atmospheric response to such Indian-Burma trough. In contrast, at this time a heat source leads to cross-equatorial flow and the Indian Peninsula is located behind the strengthening of the low-level westerlies. This trough where downward motion suppress the causes the zero absolute vorticity line in lower development of convection (Wu and Zhang atmosphere to move north of the equator to 1998; Zhang and Qian 2002; He et al. 2006; about 5N facilitating symmetric inertial insta- Ding and Liu 2006a). 28 Journal of the Meteorological Society of Japan Vol. 85B

over the WNP seems to follow the northeast- ward extension of the warm SST tongure (Wu and Wang 2001) which is created by the cloud-radiation and wind-evaporation feedback processes. Monsoon-induced changes in cloudi- ness and surface wind produce contrasting changes in surface short-wave radiation, and latent heat flux between the convection and pre-convection region. The resulting SST ten- dency difference turns around the SST gradient east of the convection region in about one month, and induces the northeastward shift of the highest SST center. Under this condition, the convective instability and low-level mois- ture convergence increase, thus triggering con- vection and atmospheric destabilization and subsequent monsoon onset. After the onset of the Asian summer mon- soon, the monsoonal rainfall, the TCZ or ITCZ and other properties or variables (e.g., yse) as- sume significant northward advance, but with quite marked regional differences. In India, the seasonal advance of the monsoon gradually proceeds northwards. However, the northward progress is not a fully smooth affair as well and takes place often in surge or in stages, in- terspersed by periods of weakening or stagna- tion of monsoon activity (Gadgil and Kumar Fig. 4. Longitude-time cross-section of 2006). For each surge or stage, the advance the latitudinal difference (20N–5N) of thickness between 500 hPa and 200 hPa process is accompanied by a synoptic-scale averaged for 1961–2000. Ordinate: pen- disturbance pushing the leading boundary of tad. Unit: gpm (Ding and Liu 2006a) monsoon either northward or westward. Along the Indo-Ganetic Plains the advance of the monsoon occurs from east to west and is associ- The summer monsoon progress over the ated with the formation of 2 or 3 monsoon Western North Pacific (WNP) cannot be ex- depressions/lows (Ding and Sikka 2006). The plained by the change in the tropospheric heat monsoon advances takes nearly 3 to 4 weeks to source that reflects the land-sea thermal con- cover the entire Gangetic Plains. Thus, the ad- trast. The results obtained by most of investi- vance of the monsoon over the entire subconti- gators suggest that the summer monsoon ad- nent is a rather slow process taking on average vance over the WNP may result from air-sea nearly 40–45 days from its start off Kerala interaction. The key is to understand the role around 31 May to its culmination by mid-July of SST change and SST gradient as well as re- over central Pakistan. The ITCZ or monsoon lated feedback processes. The zonal asymmetry trough and accompanying major rain belt in in SST between western and eastern Pacific this region move farther northward compared along 10 –20N is thought to be a leading to the ITCZ in the WNPSM region and reaches factor for the derelopment of the WNP sum- the northernmost latitude (@25N) in mid- mer monsoon (e.g., Murakami and Matsumoto July. The climatological pentad-mean maxi- 1994). Ueda and Yasunari (1996) indicated the mum rainfall rate (10–12 mm day1) occurs in importance of the development of a warm SST early June. Note also that another rain belt is tongue around 20N and 150–160E in early located at 5S, and the Indian monsoon rain July. The northeastward monsoon advance has a close linkage with the equatorial convec- July 2007 Y.H. DING 29 tion throughout the entire summer. On the 2.2 Active-break cycle and retreat of the Asian other hand, in the East Asian-West Pacific re- summer monsoon gions, the ITCZ or the tropical rainfall belt The active and break periods of the monsoon is located in the latitudinal range of 5@25N, are characterized by precipitation maxima and after the onset of the summer monsoon. There minima over South Asia, depending on the is no such near-equatorial convective band as season (Webster et al. 1998). These periods are in the Indian region. The maximum intensity thought to be associated with shifts in the and the northernmost location (@25N) are ob- location of the monsoon trough in India (Krish- served in August when the subtropical high namurti and Ardanuy 1980; Ding and Sikka over the western North Pacific considerably 2006). During the active period of the monsoon, moves northward. North of the ITCZ, there is the monsoon trough is generally located in the another significant seasonal rain belt which is central and northern Indian Peninsula. The separate from the ITCZ and propagates north- break monsoon phenomenon is a reverse of the ward in a stepwise way, not in continuous way. active monsoon spell over central and northern Numerous studies have demonstrated that it India. During the monsoon break the monsoon generally undergoes three standing stages and trough hugs the Himalayan rim, the low-level two stages of abrupt northward shifts (Ding easterlies disappear entirely along the Indo- 1992b). These stepwise northward jumps are Gangetic plains and are replaced by west- closely related to seasonal changes in the gen- northwest flow along the periphery of the Hi- eral circulation in East Asia, mainly evolution malayas, resulting in decrease of rainfall over of the planetary frontal zone or the Meiyu- much of India, but enhanced rainfall in the far Baiu frontal zone, the upper-level westerly jet north and south. These anomalies are large- steam and the subtropical high over the west- scale and extend across the entire South Asia. ern North Pacific. The peak rainy seasons tend Active and break cycles vary in duration and to occur primarily in three standing stages may last between a few days and weeks. They i.e., the first standing period from May to are significant variations on the intraseasonal mid-June for the presummer rainy season in scale to cause alterations between wet spells South China and Taiwan, the second standing and dry spells. The phenomenon of break mon- period from mid-June to mid-July for the soon has been of interest because prolonged Meiyu-Baiu-Changma rainy season, and the droughts often occur during intense break. For third standing period from mid-July to mid- example, a prolonged break situation in the August for the rainy season in North and peak monsoon of July in 2002 not only resulted Northeast China and tropical western North in rainfall deficit of the month of July, but also Pacific. From the end of August to early Sep- caused a seasonal-scale drought over the whole tember the monsoon rain belt rapidly moves country (Gadgil and Joseph 2003; Gadgil and back to South China, with most of the eastern Kumar 2006). The break and active periods part of China dominated by a dry spell, which clearly demonstrate the two modes of the South symbolizes the termination of the East Asian Asian monsoon system at the intraseasonal summer monsoon. Therefore, in the EASM re- scale (ISO). Figure 5a shows active minus gion, the monsoon rainy season has a shorter break composite of intraseasonal anomalies duration and weaker precipitation intensity (10–90-day filted) of precipitation and 850 hPa than the ISM and the WNPSM which generally wind (Goswami 2005c). It can be seen that the end in the late September and the late October, meridional shear of the low-level zonal wind respectively (Li et al. 2005; Wang et al. 2005a, and cyclonic vorticity at 850 hPa are signifi- c). In addition, in the EASM region, three sepa- cantly enhanced (weakened) during an active rate low-level westerly wind systems originat- (break) phase of the ISO. Hence, conditions for ing in the mid-latitude, subtropical and tropical cyclogenesis are much more favorable during regions, respectively, actually exist, and their an active phase compared to a break phase. interaction sets a large-scale stage for occur- Based on the analysis of genesisi and tracks of rence of many significant weather and climate low pressure systems (LPS) for 40 years (1954– events in this region that is very unique in the 1993), Goswami et al. (2003) show that genesis world monsoon climate zone. of the LPS is nearly 3.5 times more favorable 30 Journal of the Meteorological Society of Japan Vol. 85B during an active condition compared to a break Surgi 1987; Goswami 2005c). Their observa- condition of the monsoon (Fig. 5b). They also tions have revealed that the active and break show that the LPS are spatially strongly clus- period of the Indian summer monsoon or the tered to be along the monsoon trough region wet and dry spells over the Indian continent during an active phase. are manifestation of repeated northward propa- The above result has shown that there is an gation of the TCZ or the monsoon trough from association of active and break periods of the the equatorial position to the continental posi- monsoon with the ISO. Numerous investigators tion and results from superposition of 10–20- have studied this problem, with a special em- day and 30–60-day oscillations. Krishnamurti phasis on the role of 30–50 day oscillation and Ardanuy (1980) examined the 10 to 20-day (MJO mode) and 10–20 day oscillation in mod- waves at 20N, utilizing 30 years of sea level ulating the active-break cycle of the summer pressure. They noted that alternate passage of monsoon (Yasunari 1979; Sikka and Gadgil troughs and ridges of these west-propagating 1980; Krishnamurti 1985; Krishnamurti and oscillations modulated the pressure of the mon- soon trough over northern India, and there is a strong relationship between the arrival of the ridges of these waves over the central India and occurrence of breaks in the monsoon. On the other hand, the arrival of ridges of north- ward and eastward propagating 30–50-day mode over the region of a climatological mon- soon trough has been known to coincide with the periods of occurrence of breaks in the sum- mer monsoon (Yasunari 1980; Krishnamurti 1985). The nearly simultaneous arrival (phase- locking) of ridges of these two ISO modes ap- peared to modulate the pressure of the mon- soon trough considerably. Recently, a number of investigators have found similarities in the ISO and the interannual oscillation of mon- soon system (Krishnamurthy and Shukla 2000;

Fig. 5. (a) Active minus break composite of intrseasonal anomalies (10–90-day filtered) of precipitation and 850 hPa winds averaged for the summer (from June 1 to September 30 of 1979–2002). Positive (negative) anomalies are repre- sented by full (dashed) isolines (Gos- wami, 2004) and (b) tracks of LPS for the period of 1954–1983 during active ISO phase (top panel) and break phase (bottom panel). Dark dots represent the genesis point and the lines their tracks. Large number of LPS during active phase are strongly clustered to be along the monsoon trough. The few LPS that form during breaks clearly avoid the monsoon trough region and form either near the foothills of Hima- layas or off the western coast and move westward (Goswami et al. 2003). July 2007 Y.H. DING 31

Lawrence and Webster 2001). This is not monsoon, respectively, the first peak occurring surprising as it is due to the dominance of before the onset of the summer monsoon in two modes of active and break of monsoon East Asia (the 25th pentad) is produced by fron- activity in the monsoon season (Ding and Sikka tal systems (the spring rainy season). It seems 2006). that the active-break cycle of the monsoon is The breaks in the monsoon also occurs under mainly observed in the region to the south of mid-latitude interaction in which case a west- the Yangtze River where the rainfall events erly trough in the upper-middle troposphere are greatly affected by tropical and subtropical and the associated western disturbance in the summer monsoon. Farther northward, only sin- lower troposphere greatly influence the activity gle peak of rainfalls is observed in North China, of the monsoon trough over the Indo-Pakistan implying that the active-break cycle is not very region. The western disturbances from mid- marked. For Changma rainy season in Korea, latitude can shift the monsoon trough to the the peak rainfall occurs in the period of late foothills of the Himalayas, thus leading to June-mid-July. Afterwards, a short break is the break in the monsoon when the monsoon generally observed in the late July. Starting has already remained active over central and from mid- and late August, the revival of mon- northern India for 10–15 days and it is time soon rainy period is also observed (Chen et al. for intraseasonal reversal of the monsoon activ- 2004). This second rain spell is not long and ity in terms of rainfall as the ridge phase of the it maintains until early September, forming ISO is approached over the northern Bay of the autumn rainy season in Korea (Qian et al. Bengal (Ding and Sikka 2006). 2002). In Japan, the autumn rainy season or In the East Asian monsoon region, the simi- the Shurin season (Akisama) generally takes lar active-break cycle of the summer mon- place during the period of September 6–26 af- soon is climatologically observed. After the pre- ter the Baiu season (e.g., Matsumoto 1988). summer rainy season in May and early June in Previously, the Shurin frontal zone is regarded South China and the Meiyu-Baiu rainy season as the polar frontal zone with strong tempera- from the early June to mid-July in the Yangtze ture gradient in both Japan and China, but ac- River basin and Japan are terminated, breaks cording to the later study made by Matsumoto of the monsoon rainy period occur, respectively. (1988), the Shurin frontal zone is characterized Breaks of different spans are observed in South by a weak thermal gradient (@4C/1000 Km) China, Taiwan, central East China, Northeast and a strong moisture gradient (@5g/Kg/ China and Korea (Chen et al. 2004; Wang and 1000 Km) to the west of 130E along the south- Ding 2006). Therefore, the monsoon rainfall ern coast of Japan like the Baiu frontal zone, variation during the warm season in East Asia with little sunshine observed (Inoue and Mat- is generally characterized by two active rainfall sumoto 2003). However, the stagnating nature periods separated by a break spell. The north- of the Shurin frontal zone was less pronounced ward passage of the Meiyu-Baiu rain band than the Baiu frontal zone. During the Shurin is followed by a break spell (monsoon break) season, the mid-latitude westerlies shows a which also propagates northward. Then the blocking pattern over Eastern Eurasia. In monsoon rainfall revival after the break is China, the autumn rainy season begins in early clearly observed. Chen et al. (2004) has shown September and ends in mid-October, with pro- that the monsoon revival in East Asia is caused nunceed occurrence over western China and by different mechanisms associated with the the Yangtze River delta (Kao and Kuo 1958). development of other monsoon circulation com- This rainy season causes the second and third ponents including the ITCZ, typhoons, and peaks of precipitation in southwest China and weather systems in mid-latitudes. Therefore, the Yangtze River basin, respectively. The the seasonal variations with the dual peak in end of the autumn season, accompanied by the South China and Southwest China and the southward shift of the subtropical jet stream triple peak in the Yangtze River basin can be over the Tibetan Plateau, is nearly simulatane- observed. In the Yangtze River basin, in addi- ous with the end of the Indian summer mon- tion to two peaks associated with northward soon (Yeh et al. 1958). advance and southward retreat of the summer The retreat of the Asian summer monsoon 32 Journal of the Meteorological Society of Japan Vol. 85B has been earliest observed in East Asia, which evolution of the monsoon system as well as generally starts from the 44th pentad (6–10 Au- related singularities. However, other modes of gust). The process of the retreat is very rapid, the monsoon variability such as intraseasonal, taking only one month or even less to retreat interannual and inter-decadal are of the most from northern to southern China. Two pentads interesting and important due to their close as- later, the low-level southwesterlies disappear sociation with anomalous, significant weather in the region to the south of the Yangtze River and climate events. basin. Early in September, the leading zone of the summer monsoon quickly withdraws south- 3.1 Intraseasonal variability (ISV) ward to the northern part of the South China The character of the intraseasonal variability Sea and then is stationary there, marking the in the Asian monsoon region is especially prom- end of the summer monsoon in East Asia. Mat- inent and unique, with their well-organized sumoto (1997) also indicated that the monsoon structures, preferred propagation paths and westerlies are already replaced by easterlies significant regional or local modulating effect in the northern part of the South China Sea on the monsoonal rainy seasons. The initial in early September, while monsoonal westerlies recognition of the intraseasonal variability or are still dominant over the Indochina Penin- oscillation came in the early 1970’s with the sula, until late October when the summer in discovery of the 30–60 day or 40–50 day intra- East Asia fully ends over Southeast Asia. So, seasonal oscillations in the tropics named later the life cycle of the activity of the East Asian as the Madden-Julian Oscillation (MJO) (Mad- summer monsoon is about four months from den and Julian 1971), and a very active re- May to early September. The retreat of the In- search period was followed due to the thrust dian summer monsoon begins in the western of the MONEX/FGGE of 1979 (Krishnamurti parts of the Northwest Indian state of Rajas- 1985). The space-time structures and propaga- than in early September (Gadgil and Kumar tion of the ISV have been studied by a number 2006). Thus, the effective duration of the south- of investigators (e.g., Goswami 2005a; Waliser west monsoon rains over this area is only about 2005) on the global basis through the analysis one and a half month. The southward retreat of of wind, rainfall and OLR fields. They have re- the summer monsoon rains continues rapidly vealed the ISV eastward propagation and equa- until about the middle of October, by which torially trapped character, baroclinic nature of time it withdraws completely for the northern the wind anomalies, the global scale character- half of the Indian Peninsula. During the period ized primarily by wave number 1 for wind and of October and December, major parts of South wave number 2 for rainfalls and its connections Asia is generally dry, only with the south and to mid-latitudes and the boreal summer mon- southeast peninsular of India and Sri Lamka soon. In particular, the relationship between receiving significant amounts of rainfall. By the global ISV and the Asian summer monsoon the end of December when the northeast has been extensively studied, and many new monsoon blows with its full strength over the findings have been made of the strong depen- northern Indian Ocean, the rainy season has dence of the ISV periodicity, amplitude and practically ended (Matsumoto 1990; Ding and propagation on geographical regions and sea- Sikka 2006). In the WNPSM region, the sum- sons (Li et al. 2005) and its modulating effect mer monsoon retreats latest, generally in the on the monsoon activity (active and break peri- end of October. Therefore, the summer mon- ods) and dry and wet spells. The typical ISV or soon season has a life cycle of about five months ISO structure of summer rainfall anomalies in from June to October in this region (Li et al. the entire Asian monsoon region is character- 2005; Wang et al. 2005c). ized by the fact that positive rainfall anomalies in the western and central Indian Ocean occur 3. Multi-scale variability of the Asian in conjunction with negative rainfall anomalies summer monsoon over a region extending between India and the The annual cycle of the monsoon is a major equatorial western Pacific. This system appears component of its variability, which is usually to propagate in both an eastward and north- used to define typical seasonal character and ward direction, especially in the longitudinal July 2007 Y.H. DING 33 range between 80E and 130E. In this longi- day disturbance, with comparable variances tudinal sector the very evident northward- (Wang et al. 2005a). However, the 10–20 day propagating phenomenon is studied by many disturbances have prominent westward propa- investigators (e.g., Yasunari 1979; Lau and gation from the Philippine Sea and appears Chan 1986a). Two dominant modes of the ISV to be more active in the Meiyu region, the in the Asian monsoon region have been identi- South China Sea, and the Indochina Peninsula fied during the Asian summer monsoon: the than in the Indian monsoon region (Murakami 10–20 day and 30–60 day oscillations. In ad- 1980; Lau 1992). The synoptic disturbances dition to the above-described ‘‘synoptic’’ ISO with period shorter than 10 days have the larg- events, ‘‘the Climatological ISO’’ (CISO) (Wang est variance everywhere in the monsoon do- and Xu 1997) have been detected and examined main. by a number of studies. This CISO is produced Overall the synoptic variability in the East due to the enough year-to-year similarity in the Asian sector is larger than that in the Indian Asian summer monsoon. sector. Wang et al. (2005a) have pointed out Figure 6 presents the pattern of the intra- that the synoptic scale disturbances in the seasonal variability (30–90 days) and its com- East Asian summer monsoon have its unique parison with annual rainfall variability and subtropical features, which are associated with interannual variability for the Northern Hemi- the Meiyu-Baiu front system, ranging from sphere summer seasons (Waliser 2005). The meso-g scale to synoptic scale (e.g., Ninomiya annual standard deviation exhibits strong vari- and Murakami 1987; Ninomiya 2004; Johnson ability in either side of the equator, which is a and Ding 2005). The ISO life cycle is character- depiction of the annual meridional migration ized by its initiation and eastward propagation of the tropical rainfall band. As indicated in of convective anomaly over the Indian Ocean, Section 2, this is a fundermental manifestation followed by poleward propagation, with the of the monsoon. The map of interannual vari- northward-moving branch having greater am- ability emphasize the connection to monsoon- plitude than the southward-moving branch related SST variability in the tropical Pacific (Kemball-Cook and Wang 2001; Wang et al. Ocean (See Subsection 3.2). It can be seen that 2005a). The transition of convection from the its pattern is rather similar to that of the ISV Indian Ocean to the western Pacific occurred which implies a close link of the latter with the next, followed by dissipation of the current cy- interannual variability. The map of the ISV cle and initiation of subsequent cycle. Further- illustrates two important features. First, the more, the ISO life cycle may be further divided intraseasonal rainfall variability is as large or into two periods of the early (May–June) and larger than the variability associated with the late (August–October) summer which have other time scales illustrated. Second, it tends several different significant characteristics. to be relatively most prominent in the Asian The ISO in the early summer shows strong and Australian monsoon sectors, with the max- eastward movement of convection along the imum variance observed in the tropical East equator in both the Indian and the western Indian Ocean and the Bay of Bengal, and the Pacific Ocean, while the convection in the late South China Sea and tropical western North summer has a weaker eastward propagation Pacific. Wang et al. (2005a) have obtained the signal along the equator, and displays a discon- similar results by using the OLR data. In early tinuous jump from the Indian Ocean to the summer (MJJ), the largest 20–60 day variance western Pacific and then strong northwestward is located in the equatorial eastern Indian propagation of convection in the western Pa- Ocean and the Bay of Bengal. In mid- and late cific. This difference of the ISO life cycle be- summer (ASO), the 20–60 day variability in the tween the early and late summer reflects the Indian sector weakens significantly while the seasonal shift of atmospheric conditions (in- strongest variability is shifted to the western creased easterly vertical shear and low-level North Pacific between 5N and 20N with ex- specific humidity) in which the westward- tending from 110E to 170E. The 10–20 day propagating Rossby wave is more favorably intraseasonal variability in general has a simi- emanated. The amount and character of ISO lar subseasonal variation as that of the 20–60 associated with the Asia monsoon system 34 Journal of the Meteorological Society of Japan Vol. 85B

Fig. 6. Rainfall variability maps for the global tropics. Rainfall data is based on pentad values of the satellite and in-situ merged CMAP product of Xie and Arkin (1997) from 1979 to 1999. Upper panel: Annual cycle. In this case, the mean 73-pentad annual cycle was constructed from the data and the variance was computed about the annual mean; values shown in terms of standard devia- tion. Middle panel: Interannual variability. In this case, the data were low-pass filtered, retaining periods longer than 90 days. The variance of these interannual anomalies was computed for the June–September periods separately; values shown in terms of standard deviation. Lower panel: ISV. In this case, the data were band-pass filtered, retaining periods between 30 and 90 days. The variance of these intraseasonal anomalies was computed for the June–September periods sepa- rately; values shown in terms of standard deviation (Waliser 2005).

varies considerably from year to year. In terms summer monsoon, with interannual variations of variance, the typical change from year to producing a rectified signal onto the seasonal year is on the order of 30%. Thus, the varia- mean state and variability. A number of obser- tions are considerable and have a profound vational and modeling studies on interannual impact on the year-to-year character of the variations of the ISO have tended to fall into July 2007 Y.H. DING 35 mechanisms responsible for producing observed on inter-decadal time scales between Indian interannual variability of the ISO and the ISO Ocean SSTs and both the summer mean low- modulating effect of the interannual character level zonal winds and the boreal ISO intensity. of the boreal monsoon (Waliser 2005). In these In addition, it was found that the 10–20 day two aspects, major results showed that: (1) the intraseasonal variability showed evidence of observational and model predictability studies increasing in intensity over much of the Asian found a very weak negative correlation between monsoon region. tropical interannual SST variability and MJO The intraseasonal oscillation can play a very activity (Slingo et al. 1999), but with signifi- important role in modulation of the seasonal cantly enhanced predictability during El Nino variation or the annual march of the Asian events (a decrease in the ISO activity). But, monsoon, thus leading to occurrence of singu- Lawrence and Webster (2001) found very little larities and significant events as well as relationship between ENSO and boreal sum- quasi-periodic phenomena. At least, the ISO is mer ISO activity, suggestive of at most inter- closely related to such intraseasonal variations mittent predictability. On the other hand, or events as the onset and active-break cycle of Teng and Wang (2003) indicate a much more the monsoon, establishment and break /revival robust relationship of significant positive rela- cycle of rainy period, synoptic disturbances tionship with interannual eastern Pacific SST and tropical cyclones, cold surges and moisture variability for the westward propagating com- transport of the monsoon. In Section 2, the as- ponent of the ISO; (2) the intraseasonal ‘‘cloud- sociation of the ISO with the onset and the iness fluctuations’’ exhibited longer periods active-break cycle of the summer monsoon during El Nino years (Yasunari 1980). The has been discussed on the basis of synoptic- character of the monsoon may be influenced climatological events. For the CISO, its cycles via the ENSO-related sensitivity of the north- regulate significantly the onset, duration, break westward propagating component of the ISO and retreat of rainy periods in the Asia mon- and variations of the boreal summer ISO soon region (Wang et al. 2005a). Lin and Wang (Teng and Wang 2003; Flatau et al. 2003). But, (2002) have found three principal CISO cycles there is still a need to sort out more clearly the in the East Asian summer monsoon. The first effects from ENSO; and (3) the interannual and cycle (cycle I) is associated with the northward intraseasonal fluctuations share a common migration of subtropical front (wet phase) and mode of spatial variability as indicated in Sec- subtropical high (dry phase) in June. The onset tion 2. The ISO activity exhibits a strong in- of the summer monsoon may be viewed as verse relationship with the seasonal mean the beginning of the occurrence of the first Indian summer monsoon strength (Krishna- major intraseasonal oscillation accompanied murthy and Shukla 2000; Lawrence and Web- by a northward propagation of the precipi- ster 2001). Strong monsoon tended to have few tation maximum and monsoonal flow (e.g., intraseasonal variability whereas weak mon- u-component). The wet phase of the cycle-I in soon tended to have more. The common pat- mid-June marks the grand onset of the Asian terns shared by principal intraseasonal and in- monsoon, i.e., the continental Indian summer terannual variability may be explained by the monsoon, Meiyu/Baiu and western Philippine conclusion that the ISO of the monsoon was Sea, while the dry phase of this cycle in early the fundamental building block of interannual to mid-July corresponds to a grand CMB (the variability. climatological monsoon break). The second and The inter-decadal variability of the MJO third CISO cycles, occurring after the CMB, are activity is characterized by a positive ‘‘trend’’, characterized mainly by westerward propaga- possibly with a transition to greater activity tion from 150Eto80E of wet and dry phases, around the late 1970’s and/or a very low fre- peaking in mid-August and last active phase in quency (@30 years) of variability (Slingo et al. mid-October of the western North Pacific sum- 1999). This trend may be possibly associated mer monsoon. Kang et al. (1999) pointed out with the low-frequency SST warming over the that another northward propagation of the Indian Ocean. For the ISO, Zveryaev (2002) CISO in East Asia may be observed around further revealed a strong positive correlation early August (in the cycle-II). The extreme 36 Journal of the Meteorological Society of Japan Vol. 85B phases of the CISO indicate wet or dry events numerous investigators, which were reviewed occurring in a fixed pentad on a regular basis by Wang (2004), Hoskins and Wang (2006), Wal- i.e., monsoon singularities (Wang and Xu 1997). iser (2005, 2006), and Webster (2005, 2006). The propagation of the CISO links monsoon Here only a very concise summary is presented singularities occurring in different regions. The to highlight major achievements in this subject. above studies on the CISO may provide a The major scientific problems involved in the framework to understand, and even predict, re- physical processes and theory of the ISO gional monsoon onset and climatological fea- have been identified below: (1) plantetary-scale tures of regional rainy seasons. wave, with core large-scale system and convec- The ‘‘active’’ and ‘‘break’’ phase of the ISO tive activity, (2) the slow eastward and north- has the close relationship with the genesis and ward propagation in the near-equatorial region, tracks of the low pressure system (LPS) in the (3) the off-equatorial westward propagation in Indian summer monsoon region which is indi- the boreal summer, (4) coupling to the ocean, cated in Section 2. Based on recent climatologi- clouds, radiation, mean state circulation, land cal study by Wang et al. (2006), the MJO has a surface conditions, and mid-latitude variability, significant effect on the genesis of typhoons and (5) instability/maintenance. The earlier ef- over the western North Pacific. During the forts were directed to instability mechanisms ‘‘wet’’ and westerly wind phase of the MJO, the of the MJO, with dealing primarily with the at- genesis of typhoons tended to be more frequent mosphere alone, which include the CISK and whereas during the dry phase the genesis of Wave-CISK. In these studies, low level conver- typhoons tended to be less frequent. gent (divergent) areas of equatorial wave mo- The ISO activity may also play an important tions were tied to atmospheric heating (cooling) role in triggering and abrupt terminating of El to examine the influence on instability growth Nino events through changing the large-scale and propagation speed. The notable shortcom- wind and convection fields. In the triggering ings of earlier contributions in this area in- process, the westerly wind bursts and enhanced volved a tendency to amplify short rather than convective activity were caused by the ISO or planetary-scale wave and to have phase speeds MJO passages (Lau and Chan 1986; Nitta for the excited waves propagating too fast (Wal- et al. 1992; McPhaden 1999). The abrupt termi- iser 2005). Later, the ‘‘frictional wave-CISK’’ nation of the 1997–1998 El Nino event was and the ‘‘evaporation-wind-feedback’’ theories associated with an abrupt intensification of the were put forward to illustrate the propagation easterly trade over the eastern equatorial Pa- and growth/maintenance of the MJO, with in- cific Ocean, excited by a exceptionally strong clusion of boundary layer friction and coupling MJO traveling around the equator in May between moist Kelvin and Rossby modes for 1998 (Takayabu et al. 1999). the former (Wang and Rui 1990) and the The monsoon rainfall and the activity of positive surface evaporation anomalies for the rainfall-generating synoptic systems in the latter, thus leading to the increase in the the Asian monsoon region are generated/ low-level moist static energy to the east of modulated by the coupling of the global intra- the convection region and unstable eastward seasonal mode with the regional 30–60-day propagating mode of MJO-like convection monsoon trough/ridge, and the westward prop- region (Yano and Emanuel 1991). The above agation of the 12–24-day monsoon mode. This suggested and other similar mechanisms downscaling interaction can clearly be reflected (e.g., discharge-recharge hypothesis (Blade and in effect of the global intraseasonal modes on Hartman 1993)) all emphasized the importance intraseasonals of the hemispheric and regional of local feedback. Although the MJO/ISO is not moisture transport and water vapor budget consistent with a stationary oscillation that (Chen 2006). For example, water vapor trans- might ensure when considering only local feed- port is fluctuated by the 30–60-day oscillation, back, it is possible that the above sorts of mech- especially in the Meiyu rainfall belt in East anisms play complimentary roles to those that Asia. are based on a wave-like infrastructure, includ- The physical processes and theory for the in- ing establishing instability, setting time scale traseasonal oscillations have been studied by and propagating the signal. July 2007 Y.H. DING 37

Wang (2004) has recently made an attempt frequency perturbations (Hoskins and Wang to provide a comprehensive view of MJO/ISO 2006). The basic state of the warm pool is con- theory of ‘‘frictional CID’’ with combining of ducive to the occurrence of coupled unstable frictional wave-CISK and ‘‘Convective Interac- mode in intraseasonal timescales (Wang and tion with Dynamics’’. He takes into account Xie 1998). The wind-evaporation-SST feedback most essential processes and feedbacks, with is central to this coupled instability. Recently, basics of instability deriving from low-frequency Stephens et al. (2004) and Wang (2004) have equatorial waves, convective latent heating, provided evidence of tropical intraseasonal vari- boundary-layer friction, and the spatial distri- ability as a self-regulating coupled system. bution of atmospheric moisture. The frictional They divided intraseasonal oscillation cycle CID model is able to simulate the ISO in into three phases or eight phases. For the case the Asian monsoon region during the boreal of the study made by Stephens et al. (2004), the summer, which is characterized by prominent destabilization phase, convective phase and northward propagation and off-equatorial west- restoring phase consititute the cycling self- ward propagation (Wang and Xie 1998; Hos- regulating system in which the corresponding kins and Wang 2006). The model results sug- variations of the SST plays an essential role. gest that the northward propagation in the As indicated by Waliser (2006), while there is Asia-Pacific summer region is due to north- yet to be overall agreement on a given theory, it westward emanation of Rossby waves from appears that most of the essential issues that the equatorial disturbances when the latter need to be considered in developing and refin- decay over Indonesia and near the Date Line. ing the theories for the ISO have at least been The vertical shear mechanism (strong easterly identified. The theoretical development of the shear) induces a northward component from ISO is a challenging issue due mainly to the the westward propagating Rossby waves in the fact that the MJO/ISO involves a multi-scale northern summer monsoon region. The major interaction, namely, it is itself a planetary- drawback of the frictional CID theory lies in scale phenomenon, yet it is apparent that it its simple representation of diabatic heating does not directly organize convection and the between the large-scale equatorial waves and convective latent heat release is largely con- convection. sumed directly by mesoscale and synoptic-scale In addition to the atmospheric dynamics de- disturbances. Therefore, it involves in complex scribed above, important effect of the air-sea in- multi-scale interaction. Besides, in the process teraction on the MJO was well realized due to of this energy cascade, complex air-sea interac- the recurrent and pronounced signature of the tion and cloud-radiation feedback process may MJO in the tropical climate. At the early time, also play an important role in sustaining the a wealth of studies documented the relation- ISO. ships between the MJO, air-sea heat, momen- tum and mass fluxes, and SST. Later, a num- 3.2 Interannual variability ber of numerical studies have addressed the The interannual variability of the Asian role of SST coupling in the MJO due to the fact monsoon is defined as the yearly deviation of that the instability of the MJO only exists upon seasonal transition from the mean annual cycle coupling. The interactive SST and surface heat (Yang and Lau 2006). Furthermore, substantial fluxes may also contribute to the eastward spatial variability exists in the Asian monsoon propagation of the MJO and the northward system, so that several major regional monsoon propagation of the summer monsoon ISO. The components or sub-regions are usually divided coupled model results suggest that, while atmo- in order to classify monsoon variability within spheric internal dynamics are essential in gen- Asia. Therefore, the interannual variability of erating the ISO, interaction between the atmo- the Asian monsoon is characterized by a set sphere and ocean mixed layer may further of seasonally and spatially varying characteris- enhance and better organize the eastward tic features. The interannual variability of the propagating MJO and northward propagating Asian monsoon has a significant impact on the monsoon ISO through an additional coupled in- agriculture and economy for many countries in stability that amplified moist atmospheric low- this region. Especially, the interannual vari- 38 Journal of the Meteorological Society of Japan Vol. 85B

Fig. 7. Schematic illustrating key anomalies of convection, SST, surface winds, Walker circulation, extratropical circulation, land surface changes and thermocline tendencies in the Indian Ocean and the western Pacific associated with the TBO (Meehl and Arblaster 2002). ability of seasonal rainfalls has a strong link to 7-yr modes of the interannual variability are the yield, areas under cultivation and the pro- also evident, which were extensively studied duction of agricultural crops through occur- previously. The TBO signal appears to be a fun- rence of the large deficent and excess seasonal damental characteristic of the Asian monsoon rainfalls or extreme events of droughts and rainfall, with increasing rainfall in one summer floods (Gadgil and Kumar 2006). and decreasing it in the next. But, it is also Dominant modes or spatial patterns of the found in the relationship between the Asian interannual variability of the Asian monsoon monsoon and other climate phenomena such as have been extensively studied based on the ENSO (Lau et al. 2004; Webster 2005b). The fields of wind or circulation, precipitation, tem- TBO possesses a characteristics spatial struc- perature, OLR, monsoon indices and other ture, seasonablity and evolutive sequence variables, with using different methods includ- (Ropelewski et al. 1992; Meehl 1994; Meehl ing estimates of seasonal standard deviations, and Arblaster 2002). In DJF (Fig. 7a), one can EOF analysis, regression analysis and so on. see above-normal SST in the Indian Ocean and Three dominating modes have been identified: the central and eastern Pacific Ocean and the TBO (the Tropospheric Biennial Oscilla- below-normal SST in the region to the north tion), IODM (Indian Ocean Dipole Mode) and of Australia. The SST distribution, related dipole-like variation of precipitation between Walker circulations and convective activity re- South Asian and Southeast Asia (Lau et al. semble those associated with the warm phase 2004). In addition to the above modes, the 4– of ENSO event. The Eurasian continent is July 2007 Y.H. DING 39 warmer and has less snow, in connection with interactive processes by the climatological sea- the establishment of an anomalous high pres- sonal cycle. Especially, the biennial tendency sure ridge that signals an atmospheric Rossby of the Asian monsoon perhaps is one of im- wave response to tropical heating. In JJA portant consequences of the ENSO-monsoon (Fig. 7b) the warmest SST anomalies shift system interaction (Lau and Wang 2005, 2006). westward over the equatorial western Pacific Recently, a plausible mechanism through which and the eastern Indian Ocean which represents ocean-atmosphere coupling leads to a TBO has La Nina phase. As a consequence, the summer been proposed with coupled ocean-atmosphere monsoon rainfall increases over South Asia, in- GCMs (Wu and Kirtman 2004). A strong Asian dicating occurrence of a stronger Indian mon- summer monsoon during JJA can enhance soon. In the following DJF, the SST in the west- surface easterlies in the equatorial central Pa- ern Pacific reaches it maximum, while the cific, induces an eastward propagating upwell- Indian Ocean cooling is most extensive (Fig. ing Kelvin wave and give rise to negative SST 7c). The increased convection over the western anomalies in the eastern Pacific that amplifies Pacific and northern Australia and weak con- through air-sea interaction. Cold SST in the vection over the Indian Ocean provide forcing eastern Pacific is also associated with warmer to a Rossby wave response characterized by an SST in the western Pacific. A strong Asian anomalous trough and a colder, more snow- summer monsoon also cools the Indian Ocean covered subtropical Eurasia. Thus, in about the through enhanced evaporation and upwelling. time period of one year, the signs of the anoma- Associated intensification of the Walker circu- lous monsoon in coupled ocean-atmosphere- lation leads to divergence of moisture supply land system have completely reversed. The in the western Indian Ocean. Reduced mois- cycle is repeated in the next twelve months ture supply at low levels together with upper with the opposite polarities completing a full level subsidence leads to a weaker Asian sum- biennial cycle, but with a weaker Indian mon- mer monsoon during the next summer. A weak soon observed in the second JJA. Asian summer monsoon induces opposite ef- During the life cycle of the TBO, a prominent fects and can lead to a stronger monsoon next feature is the biennial alternation between year. Thus, the ocean-atmosphere interaction successive strong and weak Indian summer generates interannual variability of the Asian monsoon. Numerous investigators have studied summer monsoon via generation of the TBO spatial distributions of strong and weak mon- signal. Lau and Wang (2005) have also indi- soon year and their shifts based on observed cated that the interaction of ENSO and sea- analysis. Among them, Fasullo (2004) found sonal anomalies could contribute to biennial that for the strongest monsoon seasons, statis- tendency of monsoon system by modulating tically significant association with the preced- the coupled atmosphere-ocean environment in ing summer rains exist that are consistent the warm Indo-western Pacific region. Never- with the biennial hypothesis. In contrast, for theless, still more studies are required to years following the strongest monsoon seasons, clarify the origion of the biennial oscillation of and for years that both precede and follow the monsoon-ENSO system. weakest monsoon seasons, significant biennial In addition to the remote ENSO-related associations are largely absent. It is clear that ocean-atmosphere interaction, local warm more work needs to be done regarding biennial ocean-atmosphere interaction over the Indian variability in the monsoon in order to assert the Ocean and western North Pacific can also give biennial hypothesis. rise to the interannual variability of monsoon A number of mechanisms have been proposed annual cycle. The Indian Ocean dipole or zonal to explain the TBO (Clarke et al. 1998; Chang mode (IODM) is a prominent manifestation of and Li 2000; Kim and Lau 2001; Meehl and such ocean-atmosphere interaction (Saji et al. Arblaster 2002; Webster 2005b; Yang and Lau 1999; Webster et al. 1999). This mode is not 2006). Overall, the TBO is likely related to the an equatorially confined zonal mode. The SST interaction between the Asian monsoon, the dipole is coupled with the south Indian Ocean ENSO cycle, and myriad of oceanic and land anticyclonic anomalies (SIO AC). This ocean- surface processes, as well as regulation of the atmosphere interaction also contributes to a 40 Journal of the Meteorological Society of Japan Vol. 85B

to following winter depending on the monsoon basic state. The SIO AC peaks in fall. However, in the WNP, the positive feedback between Rossby waves and SST maintains the Phil- lppine Sea anticyclonic anomaly and SST di- pole after mature El Nino, providing a delayed impacts on the East Asian summer monsoon. The above mechanism implies that the Asian monsoon variations are determined not only by remote ENSO forcing, the local monsoon warm pool ocean interaction also plays a critical role. In this process, the annual reversal of monsoon circulation is another hidden factor that con- tributes to interannual variability of the Asian monsoon. Recently, Kawamura et al. (2001) and Anna- malai et al. (2005) have examined the local effect and remote influence of Indian Ocean SST variability on Asian monsoons by using an ocean general circulation model (OGCM) Fig. 8. Schematic diagrams showing the and an atmospheric general circulation model essential dynamics of the coupled (AGCM), respectively. Kawamura et al. (2001) monsoon-ocean modes over (a) the studied the effect of the transition of circulation South Indian Ocean (SIO) and (b) the regime in the tropical Indian ocean from spring western North Pacific (WNP). A posi- to early summer on the inter-annual atmo- tive feedback exists between the anom- sphere-ocean variabilities in the WNP, includ- alous atmospheric anticyclone and SST dipole [warm (cold) to the east (west) ing the SCS. They found that in spring before of the anticyclone center] in the pres- strong Asian summer monsoon, an equatorially ence of the mean circulation during asymmetric air-sea coupled mode in the tropi- local winter and spring (solid double cal Indian Ocean led to the subsequent cool arrows). (Wang et al. 2000) SST in the northern Indian Ocean and the SCS and warm anomalies in the warm pool east of the Philippines. With establishment of quasi-biennial signal of the monsoon (Losch- this east-west gradient of SST anomalies, the nigg et al. 2003; Li et al. 2003). But, the direct convection over the warm pool region of the contribution of the IODM to the interannual WNP was enhanced. Then, equatorially asym- variability of the South Asian monsoon is un- metric atmospheric waves were excited to the clear at this time. The warm atmosphere-ocean west of anomalously enhanced convection and interaction is also found in the South China related heating area, which further induced Sea and the western North Pacific (WNP) low-level westerly wind anomalies. This dy- where the establishment of Philippine Sea anti- namic process facilitates the localization of cyclonic anomaly is followed by oceanic warm- intense convection through the change in sur- ing to its west and cooling to its east (Wang face latent heat and SST and maintains a et al. 2000; Wang and Zhang 2002; Wang et al. positive local atmospheric-ocean feedback sys- 2003). Figure 8 highlights the common and dif- tem. The model solutions by Annamalai et al. ferent atmosphere ocean interaction between (2005) demonstrate that precipitation varia- the South Indian Ocean and Western North Pa- tions over the southwest Indian are tied to local cific (Wang et al. 2003), based on off-equatorial SST anomalies and are highly reproductive. Rossby waves (RW)-SST feedback mechanism Changes in the Indian Ocean-Walker circula- (Wang et al. 2000). In the SIO, there is the pos- tion suppress precipitation over the tropical itive feedback from summer (o) to fall (o), but West Pacific-Maritime Continent, contributing it turn into the negative feedback from fall (o) to the development of the Philippine and South July 2007 Y.H. DING 41

China Sea low-level anticyclone. This remote tropical central and eastern Pacific SST and forcing from the Indian Ocean SST anomalies ‘‘local effect’’ by the regional SST of tropical- offers an additional mechanism for develop- extratropical oceans near the Asian monsoon. ment of the Philippine Sea anticyclone apart The relationship between the central and east- from Pacific SST. This mechanism also causes ern Pacific SST and the monsoon is generally a significant delay in the Indian summer onset regarded as a problem of ENSO-monsoon rela- in June by 6–7 day through the anomalous sub- tionship. During the past decades, extensive sidence over the Maritime Continent. studies have been carried out to understand The seasonal migration of the monsoonal the various aspects of this relationship and re- rain bands is one of the central issue of the sponsible physical mechanism. One of the most monsoon-ENSO and monsoon-TBO relation- important results is that the Indian summer ships. The most dominant modes of the climato- monsoon is weaker (stronger) than normal be- logical variations of summer monsoon rainfall fore (after) the peak of an El Nin˜ o in winter, are characterized by various unique regional and that the relationship is opposite for the rain bands such as Meiyu in China, Baiu in monsoon and La Nin˜ a. However, the impact of Japan, Changma in Korea, and the monsoon ENSO on the East Asian monsoon is usually trough rain band in South Asia (Ninomiya and more complex, greatly depending on different Murakami 1987; Ninomiya 2004). The most stages of ENSO cycle and is often characterized dominant mode of the interannual variability by strong regional features (Huang et al. 2004). of summer monsoon rainfall is associated with During a summer at the developing stage of an a dipole-like variation of precipitation between El Nin˜ o event, the summer monsoon rainfall South Asia and Southeast Asia (Li and Zeng is above-normal, often with floods occurring in 2002). The change in precipitation over South the Yangtze River and Huaihe River valleys of Asia (India and the Bay of Bengal) is usually China, while the summer monsoon rainfall is accompanied by an opposite signal over the weak in North China, with drought frequently equatorial Indian Ocean. In East Asia, the pre- occurring. In contrast, during a summer at cipitation variability over northern China and the decaying stage of an El Nino event, the change in Meiyu over the Yangtze River valley below-normal summer rainfall and associated also assumes opposite dipole-like tendency (Lau droughts tends to occur in the Yangtze River and Weng 2002; Ding and Sikka 2006). But, and Huaihe River valleys, while the summer positive-negative-positive meridional mode of monsoon rainfall may be normal or above- precipitation anomaly is of equal importance to normal in North China, and large positive rain- the dipole variation (Ding et al. 2007). fall anomalies including excessively heavy rain- It has been realized that there are three ma- falls and floods may be observed in the region jor influencing factors affecting the interannual to the south of the Yangtze River. A better un- variability of the Asian monsoon: SST, land derstanding of the ENSO-monsoon relationship surface process and the large-scale climate phe- (Lau and Wang 2006) requires consideration nomena (Yang and Lau 2006). The first two fac- of monsoon-ocean interaction, including annual tors reflect the impacts of the slowly varying cycles of both ENSO and the monsoon. The lo- boundary (or external) forcing from the under- cal monsoon-ocean interaction is not only one lying oceans and land, while the third factor of fundamental driving mechanisms for the may itself be the result of atmospheric inter- biennial and 4–7-yr variability of the Asian nal dynamics, which is characterized by low- monsoon, but also influences various compo- frequency global variations outside the Asian nents of the Asian system through changes in monsoon system such as the North Atlantic surface heat and moisture fluxes, direct mois- Oscillation (NAO), the Arctic Oscillation (AO), ture supply, thermal difference between land and the Antarctic Oscillation (AAO). As pointed and oceans, and wavetrains generated by atmo- out by Yang and Lau (2006), among all the ex- spheric heat sources. One good example is the ternal causes of interannual variability of the correlative relationship between the Indian Asian monsoon, SST is perhaps the leading im- Ocean SST and the Asian monsoon, as indi- pacting factor. The effect of SST on the mon- cated previously. soon can be divided into ‘‘remote effect’’ by the Among land surface processes affecting the 42 Journal of the Meteorological Society of Japan Vol. 85B interannual variability of the Asian monsoon, turbs mid-latitude jet stream, which in turn the most important impacts come from varia- excited optimum downstream circulation anom- tions of soil moisture and snow cover. Snow aly mode (Yang et al. 2002; Ding and Wang cover may have an important effect on the in- 2005). The existence of teleconnection patterns terannual variability of the monsoon because associated with the Asian monsoon clearly indi- of its ability to alter the surface albedo and to cates remote or global linkage of the Asian regulate soil moisture (Yang and Lau 2006; monsoon and climate in other regions (Fig. 9). Yasunari 2006) which in turn can affect the monsoon condition as a persistence signal. The 3.3 Inter-decadal variability study on snow-monsoon relationship has a long The inter-decadal variability of the Asian history which is characterized by a negative monsoon is a long time component of variability relationship: the Indian summer monsoon is on the variety of timescales of the Asian mon- generally weaker (stronger) than normal when soon, generally ranging from 10 years to 100 more (less) extensive preceding winter and years. Due to the lack of availability of good spring snow cover occurs in Eurasia (Liu and quality data for a sufficiently long period, the Yanai 2002). However, such a relationships is temporal evolution and spatial patterns of the argued by Shinoda (2001) who, based on analy- interdecdal variability is less documented than sis of snow depths over central Eurasia, sug- the other shorter time scale variability, and gests that such a carryover on the summer the impacting factors and mechanisms for it monsoon circulation may not be produced di- are poorly understood (Goswami 2005b). The rectly through land-surface-atmosphere inter- instrumental records of surface observations action. The springtime soil moisture and snow such as near-surface air temperature, precipi- anomalies over central Eurasia are probably tation and surface pressure could be extended symptomatic of large-scale circulation varia- back to about 140 years in the Northern Hemi- tions which directly influence the Indian mon- sphere, but the upper air data are available for soon. In contrast, the preceding winter and only about 50 years. Therefore, the study of the spring snow cover or snow days/depths over monsoon inter-decadal variability needs a mix the Tibetan Plateau have a generally positive of the high-resolution proxy data (e.g., tree- relationship with the Asian summer monsoon ring, ice cores and historical documentary re- (Wu and Qian 2003; Zhang et al. 2004; Ding cords) with the instrumental data to produce a et al. 2007). Several GCM studies have found longer time-series and/or spatial patterns as that this region possibly has the greatest carry- accurately as possible. This is a challenging is- over influence on the Asian monsoon circula- sue. A better undersanding of the inter-decadal tion (Yasunari et al. 1991; Vernekar 1995). variability is very important in improving the The snow-monsoon relationship was earliest prediction of seasonal-interannual prediction of used in seasonal and interannual climate pre- the Asian monsoon, due to the fact that modu- diction (Bamzai and Shukla 1999). lation of the interannual variability by the Finally, we shall briefly discuss the telecon- inter-decadal variability influences predictabil- nection patterns which are usually excited by ity of the seasonal mean monsoon which is one the interannual and intraseasonal variability of the most important problems in the Asian of the Asian monsoon. A number of the telecon- monsoon region. In addition, the identification nection patterns have been revealed which are of regime shifts and evolution of the monsoon summarized in Table 1. It can be seen that the climate state may be used as the necessary origins of most of teleconnection patterns are background condition for prediction of the sea- normally regions of anomalous heat sources sonal and interannual variability, because the produced by significant monsoon rainfalls or dominant modes of the inter-decadal variability tropical convective activities that can generate and its shift of monsoon climate regimes may the Rossby wavetrain to propagate to North be manifestation of a global coupled ocean- America via North Pacific if these heat sources atmosphere-land mode of the inter-decadal are located north of 20N (Wang et al. 2001). variability. The decadal-inter-decadal variabil- The other possible mechanism is through heat- ity contributes to the total variability of the ing induced meridional circulation that per- ISO intensity with 20–35% (Zveryaev 2002). July 2007 Y.H. DING 43

Table 1 Summary of teleconnection modes in the Asian monsoon region* Investigators Names Origin regions and propagating routes Nitta (1987) Japan-Pacific pattern (JP) Anomalous convective activities over tropical Huang and Wu (1989) or East Asia/Pacific western North Pacific pattern (EAP) From Phillppine Sea across East Asia to Japan Wang et al. (2000) Pacific-East Asia ENSO events have a delayed impacts on the teleconnection pattern East Asian summer monsoon via development (PEA) of anticyclonic anomaly over Philippine Sea. Wang et al. (2001) WNP-North America Strong summer monsoon over Western North Li and Zhang (1999) teleconnection pattern Pacific (WNP) From WNP crossing North Pacific to North America Lau and Weng (2002) Tokyo-Chicago Express Rainfall anomalies over Japan and northeastern Lau et al. (2004) China From Japan and Northeast China, crossing North Pacific, to Western Canada, the northern Great Plains and Mid-west of US Lau et al. (2004) Shanghai-Kansas Express Fluctuations of heat sources and sinks in the Indo-Pacific monsoon region. Emanating from central East Asia across North Pacific to North America Guo and Wang (1988); India-North China pattern Summer precipitation over Indian Peninsula Kripalani and Kulkarni From Indian Peninsula across the Tibetan (1997, 2001) Plateau to North China and further to Japan Ding and Wang (2005) Circumglobal The heat source produced by abnormal Indian teleconnection pattern summer monsoon and the jet exit region of the North Atlantic From Northwest India/Pakistan, crossing East Asia and North Pacific to North America or From Western Europe, across European Russia, India and East Asia, to North America *Remarks: some teleconnection patterns having their origin regions outside the Asian monsoon region are not included in this table, e.g., the Silk Road pattern (Enomoto et al. 2003).

The inter-decadal variability of ISO activity Indian summer monsoon rainfall shows a lack over the Indian summer monsoon region has of trend or climate change signal, but contain a close correspondence with the inter-decadal coherent multidecadal variability with an variability of ENSO-monsoon mode (Krishna- approximate periodicity of 55–60 years. That murthy and Goswami 2000; Goswami 2006). may be one of dominant modes of the inter- The periods of 1955–1975 and 1980–1995 were decadal variability in this region (e.g., Goswami characterized by systematically higher and 2006). The tridecades between 1871 and 1900 lower than normal ISO activity over the Indian and between 1930 and 1960 generally saw monsoon region, respectively. The transition more above than below-normal rainfall over from higher than normal to lower than normal the country. Frequency of occurrence of large- ISO activity took place simultaneously with that scale floods related to the ISO and high- of the coupled inter-decadal mode in mid-1970s. frequency variability were also higher during Based on analysis of long records of reliable these periods. Similarly, the tridecades be- summer rainfall observations (JJAS) over the tween 1901 and 1930 and between 1971 and Indian continent between 1871 and 2000, the 2000 saw more below than above-normal 44 Journal of the Meteorological Society of Japan Vol. 85B

rainfall over the country. These periods were also characterized by a higher frequency of droughts. The equatorial eastern Pacific SST (Nino 3) also shows a similar inter-decadal variability, but is approximately out of phase with that of the Indian summer monsoon rain- fall. Another important method to characterize features of inter-decadal variability is to exam- ine the spatial coherence of the inter-decadal variability of summer monsoon rainfall. The results have indicated that the rainfall tends to be above-(below) normal over most of the country during these above-(1900–1925) (below 1940–1965) normal inter-decadal epochs, indi- cating a large spatial scale for the inter-decadal variability of the Indian summer monsoon. It is very interesting to note that the spatial pattern of the inter-decadal variability over the Indian subcontinent has a similarity with the domi- nant pattern of interannual variability over India (Krishnamurthy and Shukla 2001). From above evidences revealed in the South Asian monsoon region, the quasi-60-year inter- decadal mode is to a certain extent is a robust mode that may be a intrinsic mode of oscilla- tion of monsoon (Goswami 2006). It is very interesting to note that the interannual vari- ability of both the Indian monsoon and ENSO Fig. 9. Schematic diagram illustrating shows a quasi-60-year inter-decadal fluctua- the entire mechanism of circumglobal tion. Strong correlation ðr ¼ 0:82Þ between the teleconnection consisting of two scenar- two indicates that the quasi-60-year oscillation ios during the positive phase of CGTI is a global mode of variability that modulates (Circumglobal Teleconnection Index). activity of both the Indian monsoon and ENSO. The cloud denotes the strong Indian A quasi-50–60-year inter-decadal periodicity monsoon and the circles represent the is also evident in East Asia based on analy- circumglobal teleconnection at the up- sis of 1901–1998 historical rainfall records per level. For the first scenario (a), the (MJJA) (Goswami 2006). Roughly after 1920, Indian summer monsoon plays a more the phase of the inter-decadal variability of the active role in affecting the mid-latitude atmosphere. An enhanced Indian sum- East Asian summer monsoon rainfall also has a mer monsoon initially generates an similarity with that of the South Asian summer upper-level anomalous high to its monsoon. In East Asian monsoon region (main- northwest over west-central Asia, and ly in China) a 60–70-year periodicity of the sur- then excites successive downstream face air temperature is very marked (Ding et al. cells along the waveguide through 2006b). This signal of the 65–70-year periodic- Rossby wave dispersion and (b) the sec- ity was also found previously in the global tem- ond scenario is the means by which the perature record (Schlesinger and Ramankutty barotropic instability of the westerly jet 1994). Thus, the quasi 60–70-year oscillation dominates, while the Indian summer in both the South Asian summer monsoon and monsoon plays a more passive role in the East Asian summer monsoon may be a receiving influences from mid-latitude atmosphere. (Ding and Wang 2005) manifestation of a global mode of inter-decadal variability. In addition to this period of the inter-decadal variability, a 80-year periodcity July 2007 Y.H. DING 45 of the summer rainfall and the index of the Asian summer monsoon under the condition of East Asian summer monsoon in China may the obvious warming trend in the equatorial exist based on 1873–2000 instrumental obser- central and eastern Pacific in this period vation and 500-year reconstructed summer (Wang 2001; Chang et al. 2001; Huang et al. rainfall time series (Gong and Ho 2003; Wang 2004). Recently, Wang and Ding (2006) have in- et al. 2005d; Ding et al. 2007). dicated an overall weakening of the global land For the temperature and rainfalls in China, monsoon precipitation in last 56 years, primar- there are also obvious decadal (10–14 years) ily due to weakening of the summer monsoon and 30–40-year periodicity (Wang et al. 2005). rainfall in the Northern Hemisphere. This re- Furthermore, the East Asian summer monsoon sult is likely to be a reflection of the weakening goes through a major regime shift in the mid- of the Asian summer monsoon. and late 1970s which occurred concurrently The precise cause of the interdecdal vari- with the major climate transition in the tropi- ability are still unclear. Physical mechamisms cal Pacific. In addition, in mid-1960s and early responsible for quasi 60-year or near-century 1990s other two regime shifts were also noted (e.g., 80-year) periodicity are poorly under- (Ding et al. 2007). stood. Some studies (Mehta and Lau 1997; It is well known that the simultaneous corre- Agnihotri et al. 2002) indicate a possibility lation between the ENSO event (JJAS Nino 3 of a forcing of the quasi 60-year variability of SSTAs) and all India rainfalls has remained the monsoon by solar forcing variability over significantly negative over a long period of the same timescales. But the solar-monsoon time. But, this ENSO-monsoon relationship relationship at 50–60-year timescale has has decreased sharply over the last two decades not well established. Therefore, one plausible and is currently insignificantly small (Krishna mechanism for the inter-decadal variability Kumar et al. 1999). Another aspect of this of the monsoon is the result of a global relationship is that even when correlation was coupled ocean-atmosphere mode of variability strong, the largest correlation occurs with Nino (Goswami 2005c, 2006). Using long records of 3 SST following the monsoon season. During Indian monsoon rainfall, global SST, and sea- recent decades, however, this largest correla- level pressure data, it is shown that interdecdal tion takes place with SST one year prior to the variability of both all India rainfall and ENSO Indian monsoon. This indicates a significant (Nino 3 SST) are associated with almost identi- change in the ENSO-monsoon relationship dur- cal global patterns of SST and sea level pres- ing recent decades. The rapid weakening of sure (SLP). Thus, the inter-decadal variability the ENSO-monsoon relationship on interan- of the Indian monsoon and that of the ENSO nual timescales during the recent decades has are likely to be parts of a global-scale oscilla- been a subject of considerable attention. Gos- tion on quasi 60-year timescales. Recently, wami (2006) has explored the above changing inter-decadal changes in global scale circula- relation between the ENSO and Indian mon- tion patterns or changes in the circulation over soon in terms of modulation of the ENSO mon- North Atlantic (e.g., NAO or AO) and North soon relationship on interannual time scales by Pacific are suggested to be related to inter- the large-scale circulation changes (the mon- decadal variability of the monsoon (Krishna soon Hadley and Walker circulations) associ- Kumar et al. 1999; Chang et al. 2001). Possible ated with the inter-decadal variability. On the causative mechanisms for the inter-decadal other hand, the inter-decadal changes in the transition of the East Asian summer monsoon ENSO-monsoon relationship also have signifi- have been examined in several studies (Hu cantly affected the long-term change in summer 1997; Wu and Wang 2002; Yang and Lau 2004; precipitation in East Asia, with the decrease Huang et al. 2004; Ding et al. 2007). Three in- in summer rainfall in North China and the sig- fluencing factors: SST, snow cover in the Tibe- nificant increase in the Yangtze River Basin tau Plateau and changes in large-scale circula- and South China from the late 1970s and the tion pattern are identified to be associated with early 1990s. This change may be caused by the the inter-decadal variability in precipitation inter-decadal weakening of northward moisture pattern and the regime shift of the mid-1970s transport which is related to weakening of the in this region of the Asian monsoon. 46 Journal of the Meteorological Society of Japan Vol. 85B

4. Concluding remarks system with complex ocean-land-atmosphere interactions. In this coupled monsoon system, The major advances in our understanding the TBO is one of most salient features, and the of the annual cycle/seasonal change and the SST-monsoon relationship and snow-monsoon multi-scale variability of the Asian summer relationship are of most important relationship monsoon during recent two decades has been manifested in this coupled system resulting highlighted in the present paper. It can be from land-monsoon and ocean-monsoon interac- seen that the change and the variability of the tions. Asian summer monsoon is one of most active (5) Teleconnection patterns and their areas of research interest in the field of atmo- planetary-scale propagation routes originating spheric science. A substantial achievement has in the Asian monsoon region have been re- been made in many aspects since the MONEX vealed. Although the teleconnection modes and of 1978–1979. Compared to the MONEX era, steady Rossby wave propagation (e.g., PNA the major advances in this problem can be sum- mode) have been known for three decades, the marized in the following points: monsoon forced teleconnection patterns have (1) In the context of the global aspect, the been revealed only in recent 20 years. Its signi- climatology of the Asian monsoon system has ficance lies in that the Asian monsoon not only been widely studied, including mean circula- assumes regional importance in this part of tion patterns, moisture transport, and regional the world, but also it can exert an important features and differences in rainy season. Three effect upon the weather and climate over re- sub-systems (the Indian monsoon, the East mote regions (e.g., North China, Japan, North Asian monsoon and the Western North Pacific) America) which can provide useful precursors of the huge Asian monsoon system have been for long-range climate prediction in many re- identified which are independent of each other gions. and at the same time interact. Research and prediction of the complex Asian (2) The fact that the earliest onset of the monsoon system will continue to be a goal of Asian summer monsoon takes place in the trop- long-time endeavour. Future work will be at ical eastern Indian Ocean-Indochina Peninsula least directed toward the following five aspects: in the early or mid-May has been well docu- (1) Improvement and coordination of observa- mented. As the seasonal march progresses, tional networks and platforms in the Asian the onset process propagates northwestward, monsoon region. The CEOP (Coordinated En- northeastward and eastward to establish the hanced Observing Period) and WCRP/COPES monsoon rainy seasons in the Indian subconti- (Coordinated Observation and Prediction of nent, Southeast- and East Asia and western Earth System) will provide a great opportunity North Pacific, respectively. to enhance observational and predictive capa- (3) During recent two decades, many new bility of the Asian monsoon in the future. In findings of multiscale temoporal and spatial addition, the application of satellite data is variability of the Asian summer monsoon have very essential, especially for the Indian and been revealed. Among them our understand- Pacific oceans, and the Tibetan Plateau. Spe- ing of the intraseasonal variability has been cial regional monsoon field campaigns will be substantially improved. Their climatological as- useful to understand the synoptic and physical pects, evolutive processes, physical mechanism, processes associated with different monsoon predictability and prediction have been widely weather systems and verify parameterization studied with observed analysis and model sim- schemes of physical processes. ulations. The achievements made in these as- (2) Physical processes and mechanisms re- pects are one of most significant progress of sponsible for the monsoon onset, interactions the Asian monsoon research and forecasts. of multiple-scale monsoon variability, the (4) A better understanding of monsoon physi- monsoon-ocean-land relationship, the internal cal processes and dynamics has been made. The dynamics, and teleconnection patterns. The mean state and its variability of the Asian mon- mechanisms related to the Asian monsoon on- soon system can be well understood and ex- set are very complicated. The timing and loca- plained in the context of a coupled monsoon tion of the Asian monsoon onset is influenced July 2007 Y.H. DING 47 by many factors, not just the Tibetan Plateau lem is the subseasonal (15–30 days) prediction heating, such as the persisitently external forc- which associated with the intraseasonal vari- ing and low-frequency oscillation in the atmo- ability. Many uncertainties are included in sim- sphere (e.g., the MJO). Although the ocean is ulations and forecasts of the ISV. It is not well responding to forcing from the atmosphere, the understood which condition: initial or surface response is such that a strong feedback to the boundary condition is the main influencing fac- atmosphere is produced. Due to relatively small tor. Finally, the application of the monsoon pre- year-to-year variability of the South Asian diction to society, especially for agriculture and monsoon, the theory of regulation of monsoon water resources management, has been empha- rests on negative feedback between the ocean sized, although the predictive skill of the mon- and the atmosphere (Webster 2005b; Wang et soon is relatively low (Webster et al. 2005a; al. 2005b). This feedback governs the ampli- Gadgil and Kumar 2006). The ‘‘useful predic- tude and phase of the annual cycle and also tion’’ for various users may be obtained through modulates interannual variability. It should be downscaling techniques and thus help estimate clarified to what degree the interannual vari- risk of the occurrence of some event and make ability of the monsoon is coupled to the ocean. reasonable decisions. (3) The energy and water cycle in the Asian monsoon and their association with extreme Acknowledgments weather and climate events (e.g., droughts, In the preparation of this paper, frequent floods and heat wave). Hydrological processes references to the books of The East Asian Mon- can also provide a new way to validate Asian soon (Ed. by C.P. Chang) and The Asian Mon- monsoon simulation. soon (Ed. by Bin Wang) have been made. The (4) There is much need in better capability author would like express his sincere grati- for identification of establishment timing of re- tudes to those contributors of these two books. gional rainy seasons and the abrupt develop- The author also thanks Profs. B. Hoskins, Bin ment of meso-scale disturbances embedded in Wang and Tao Shiyan for their comments and summer monsoon (Johnson and Ding 2005; suggestions. Thanks also go to Ms.Wang Zunya Ding and Sikka 2006). and Ms. Song Yafang for their much assistance. (5) Improvement of monsoon weather and cli- Finally, the author gratefully acknowledges the mate prediction and its application to society. invitation of Meteorological Society of Japan Further efforts should be devoted to improve- and the responsible editor, Dr. H. Ueda. This ment of model initialization, resolution, physi- research is supported by the SCSMEX (South cal parameteration, and ensemble techniques China Sea Experiment) Project (1997–2003) (Wang et al. 2004; Sumi et al. 2005; Krishna- and 973 Project 2006 CB403604. murti et al. 1999, 2006; Kang and Shukla 2005, 2006). The multimodel ensemble is a References very promising approach for dynamical predic- Agnihotri, R., K. Dutta, R. Bhushan, and B. Somaya- tion of the monsoon. However, improvement julu, 2002: Evidence for solar forcing on the of initial conditions and model itself is a funda- Indian monsoon during the last millennium. mental precondition. During recent decade, Earth and Planetary Science Letters, 198, although the emphasis has been shifted to dy- 521–527. namic predictions of the monsoon, the empiri- Ananthakrishnam, R. and M.K. Soman, 1988: The cal prediction methods of the monsoon are still onset of SW monsoon over Kerala. Int. J. Cli- very important at present time and even for the matol., 8, 283–296. future. Therefore, in a considerably long time Ananthakrishnam, R. and M.K. Soman, 1989: The period in the future, the monsoon prediction dates of onset of the southwest monsoon over Kerala for the period 1870–1900. Int. J. Clima- will use multiple techniques of empirical and tol., 9, 321–322. dynamical nature. As a long-term background, Annamalai, H., P. Liu, and S.-P. Xie, 2005: South- future changes in the Asian monsoon system west Indian Ocean SST variability: Its local ef- under the global warming should be taken into fect and remote influence on Asian monsoons. account in the climate research and prediction J. Climate, 18, 4150–4167. of the monsoon (Kitoh 2006). A difficult prob- Bamzai, A.S. and J. Shukla, 2003: Relation between 48 Journal of the Meteorological Society of Japan Vol. 85B

Eurasian snow cover, snow depth, and the Ding, Y.H. and Y.J. Liu, 2001: Onset and the evolu- Indian summer monsoon: An observational tion of the summer monsoon over the South study. J. Climate, 12, 3117–3132. China Sea during SCSMEX Field Experiment Bhat, G.S., S. Gadgil, P.V. Harrish Kunmar, S.R. in 1998. J. Meteor. Soc. Japan, 79, 255–276. Kalsi, P. Madhusoodanan, V.S.N. Murty, Ding, Y.H. and Y. Sun, 2001: A study on anomalous C.V.K. Prasada Rao, V. Ramesh Babu, L.V.G. activities of East Asian summer monsoon Rao, R.R. Rao et al., 2001: BOBMEX: The Bay during 1999. J. Meteor. Soc. Japan, 79, 1119– of Bengal Monsoon Experiment. Bull. Amer. 1137. Meteor. Soc., 82, 2217–2243. Ding, Y.H., 2004: Seasonal march of the East Asian Blade, I. and D.L. Hartman, 1993: Tropical intrasea- summer monsoon, In: The East Asian Mon- sonal oscillation in a simple nonlinear model. soon. Ed. by C.P. Chang, World Scientific, Sin- J. Atmos. Sci., 50, 2922–2939. gapore, 564 pp. Chang, C.-P. and T.N. Krishnamurti (Eds), 1987: Ding, Y.H. and C.L. Chan, 2005: The East Asian Monsoon Meteorology (Oxford Monographs on summer monsoon: an overview. Meteor. Atmos. Geology and Geophys. No. 7). Oxford Univer- Phys., 89, 117–142. sity Press, 544 pp. Ding, Y.H. and D.R. Sikka, 2006: Synoptic systems Chang, C.-P. and T. Li, 2000: A theory for the tropo- and weather. In: The Asian Monsoon. Ed. by spheric biennial oscillation. J. Atmos. Sci., 57, Bin Wang, Praxis Publishing Ltd., Chichester, 2209–2224. UK, 131–202. Chang, C.-P., P. Harr, and J. Ju, 2001: Possible roles Ding, Y.H. and Y.J. Liu 2006a: Predictability of the of Atlantic circulations on the weakening South China Sea summer monsoon onset. Sym- Indian monsoon rainfall-ENSO relationship. posium on Asian Monsoon, Winter MONEX: A J. Climate, 14, 2376–2380. Quarter Century and Beyond. Kuala Lumpur, Chang, C.-P. (Ed), 2004: The East Asian Monsoon. Malaysia, 4–7 April, 2006. World Scientific, Singapore, 564 pp. Ding, Y.H., G.Y. Ren, G.Y. Shi, P. Gong, X.H. Zhen, Chang, C.-P., B. Wang, and N.-C.G. Lau (Eds), 2005: P.M. Zhai, D.E. Zhang, Z.C. Zhao, S.W. Wang The Global Monsoon System: Research and and co-authors, 2006b: National assessment Forecast. WMO/TD No. 1266 (TMPR Report report of climate change (I): Climate change in No. 70). 542 pp. China and its future trend. Adv. in Climate Chen, H., Y.H. Ding, and J.H. He, 2006: The structure Change Res. 2,3–8. and variations of tropical easterly jet and its Ding, Y.H., Y. Sun, and Z.Y. Wang, 2007: Inter- relationship with the monsoon rainfall in Asia decadal variation of summer precipitation in and Africa. Accepted by Chinese J. Atmos. Sci. China and its association with deceasing East Chen, T.C., S.Y. Wang, W.-R. Huang, and M.-C. Yen, Asian summer monsoon. Submitted to Int. J. 2004: Variation of the East Asian summer Climatol. monsoon rainfall. J. Climate, 17, 2271–2290. Enomoto, T., B.J. Hoskins, and Y. Matsuda, 2003: Chen, T.C., 2006: Variation of the Asian monsoon The formation mechanism of Bonin high in water vapor budget: Interaction with the August. Quart. J. Roy. Meteor. Soc., 129, 157– global-scale modes. In: The Asian Monsoon. 178. Ed. by B. Wang. Praxis Publishing Ltd., Chi- Fasullo, J.T., 2004: Biennial characteristics of Indian chester, UK, 787 pp. monsoon rainfall. J. Climate, 17, 2972–2982. Clarke, A.J., X. Liu, and S.V. Gorder, 1998: Dynam- Fein, J.S. and P.L. Stephens, (Eds), 1987: Monsoons. ics of the biennial oscillation in the equatorial Wiley Interscience Publication. John Wiley and Indian and far Western Pacific Ocean. J. Cli- Sons, New York, 632 pp. mate, 11, 987–1001. Flatau, M.K., P.J. Flatau, J. Schmidt, and G.N. Ding, Q.H. and B. Wang, 2005: Circumglobal telecon- Kiladis, 2003: Delayed onset of the 2002 In- nection in the Northern Hemisphere Summer. dian monsoon. Geophys. Res. Lett., 30, 1768, J. Climate, 18, 3483–3505. doi:10.1029/2003 GL01734. Ding, Y.H., 1992a: Effects of the Qinghai-Xizang (Ti- Fong, S.K. and A.Y. Wang (Eds.), 2001: Climatologi- bet) Plateau on the circulation features over cal Atlas for Asian Summer Monsoon. Macau the plateau and its surrounding areas. Adv. in Meteorological and Geophysical Bureau and Atmos. Sci, 9, 112–130. Macau Foundation, 318 pp. Ding, Y.H., 1992b: Summer monsoon rainfall in Gadgil, S. and P.V. Joseph, 2003: On breaks of the China. J. Meteor. Soc. Japan, 70, 373–396. Indian monsoon. Proc. Indian Acad. Sci.-Earth Ding, Y.H., 1994: Monsoons over China, Kluwer Aca- Planet. Sci., 112, 529–558. demic Publisher, Dordrecht/Boston/London, Gadgil, S. and K.R. Kumar, 2006: The Asian 419 pp. monsoon-agriculture and economy. In: The July 2007 Y.H. DING 49

Asian Monsoon. Ed. by Bin Wang, Praxis and its mechanisms. Adv. in Atmos. Sci., 6,21– Publishing Ltd., Chichester, UK, 651–682. 32. GAME International Science Panel, 1998: GEWEX Huang, R.H., G. Huang, and Z.G. Wei, 2004: Climate Asian Monsoon Experiment (GAME) Implemen- variations of the summer monsoon over China, tation Plan, 136 pp. In: The East Asian Monsoon. Ed. by C.-P. Gong, D.Y. and C.H. Ho, 2003: Arctic oscillation Chang, World Scientific, Singapore, 564 pp. signals in the East Asian summer monsoon. Inoue, T. and J. Matsumoto, 2003: Seasonal and J. Geophys. Res., 108 (D2), 40–60. secular variations of sunshine duration and Goswami, B.N., R.S. Ajayamohan, P.K. Xaver, and natural seasons in Japan. Int. J. Climatol., 23, D. Sengupta, 2003: Clustering of low pres- 1219–1234. sure systems during the Indian summer mon- Johnson, R.H., H.P. Ciesielski, and T.D. Keenan, soon by intraseasonal oscillations. Geophs. 2004: Oceanic East Asian monsoon convection: Res. Lett., 304, 8, 1431 doi: 10.1029/2002 GL Results from the 1998 SCSMEX. In: The East 016734. Asian Monsoon. Ed. by C.-P. Chang. World Sci- Goswami, B.N., 2005a: Intraseasonal variability entific, Singapore, 436–459. (ISV) of south Asian summer monsoon. In: Johnson, R. and Y.H. Ding, 2005: Mesoscale and syn- Intraseasonal variability of the Atmosphere- optic processes in monsoons. In: The Global Ocean Climate System. Ed. by K.-M. Lau and Monsoon System: Research and Forecast. Ed. D. Waliser. Springer, Heideberg, Germany, by C.-P. Chang, Bin Wang and N.-C.G. Lau. 474 pp. WMO/TD NO. 1266. (TMRP Report No. 70), Goswami, B.N., 2005b: The Asian monsoon: Inter- 472–511. decadal variability. In: The Global Monsoon Kang, I., S. An, C.-H. Ho, Y.-K. Lim, and K.-M. Lau, System: Research and Forecast. Eds. by C.P. 1999: Principal modes of climatological sea- Chang, Bin Wang and N.-C.G. Lau, WMO/TD sonal and intraseasonal variations of the Asian NO. 1266, 472–511. summer monsoon. Mon. Wea. Rev., 127, 322– Goswami, B.N., 2005c: South Asian summer mon- 340. soon: An overview in: The Global Monsoon Kang, I.-S. and J. Shukla, 2005: Dynamical seasonal System: Research and Forecast. Ed. by C.-P. prediction and predictability of monsoon. In: Chang, Bin Wang and N.-C.G. Lau, WMO/TD The Global Monsoon System: Research and No. 1266, 47–71. Forecast. Ed. by C.P. Chang, Bin Wang and Goswami, B.N., 2006: The Asian monsoon: Inter- N.-C.G. Lau, WMO/TD No. 126, 386–402. decadal variability. In: The Asian Monsoon. Kang, I.-S. and J. Shukla, 2006: Dynamic seasonal Ed. by Bin Wang, Praxis Publishing, Chiches- prediction and predictability of the monsoon. ter, UK, 295–328. In: The Asian Monsoon. Ed. by Bin Wang. Guo, Q.Y. and J.Q., Wang, 1988: A comparative Praxis Publishing Ltd., Chichester, UK, 585– study on summer monsoon in China and India 612. (in Chinese). J. Trop. Meteor. 4, 53–60. Kao, Y.-S. and C.Y. Kuo, 1958: On the autumn rain- He, H., J.W. Mc Ginnis, Z. Song and M. Yanai, 1987: ing area in China. Acta Meteoro. Sinica, 29 Onset of the Asian monsoon in 1979 and the i264–273. effect of the Tibetan Plateau. Mon. Wea. Rev., Kemball-Cook, S. and B. Wang, 2001: Equatorial 115, 1966–1999. waves and air-sea interaction in the boreal He, J.H., M. Wen, Y.H. Ding, and R.H. Zhang, 2006: summer intraseasonal oscillation. J. Climate, Possible mechanism of the effect of convection 14, 2923–2942. over Asian-Australian ‘‘land bridge’’ on the Kim, K.M. and K.-M. Lau, 2001: Dynamics of mon- East Asian summer monsoon onset. Science in soon induced biennial variability in ENSO. China, Series D: Earth Sciences, 49, 1223– Geophys. Res. Lett., 28, 315–318. 1232. Kitoh, A., 2006: Asian monsoon in the future, In: The Hoskins, B. and B. Wang, 2006: Large-scale atmo- Asian Monsoon. Ed. by Bin Wang, Praxis Pub- spheric dynamics. In: The Asian Monsoon. Ed. lishing Ltd., Chichester, UK, 631–649. by Bin Wang, Praxis Publishing Ltd, Chiches- Krishna Kumar, K. and K.M. Lau, 1998: Possible ter, UK, 787 pp. role of symmetric instability in the onset and Hu, Z.Z., 1997: Inter-decadal variability of summer abrupt transition of the Asian monsoon. J. Me- climate over East Asia and its association with teor. Soc. Japan, 76, 363–383. 500 hPa height and global sea surface temper- Krishna Kumar, K., B. Rajagopalan and M.A. Cane, ature. J. Geophy. Res. 102, 19430–19412. 1999: On the weakening relationship between Huang, R.H. and Y.F. Wu, 1989: The influence of the Indian monsoon and ENSO. Science, 284, ENSO on the summer climate change in China 2156–2159. 50 Journal of the Meteorological Society of Japan Vol. 85B

Kriplani, R.H. and A. Kulkarni, 1997: Rainfall vari- Lau, K.-M., Y.H. Ding, J.T. Wang, R. Johnson, T. ability over Southeast Asia-Connection with Keenan, R. Cifelli, J. Gerlach, O. Thiele, T. Indian monsoon and ENSO extremes: New Rickenbach, S.C. Tsay et al., 2000: A report perspective. Int. J. Climatol., 17, 1155–1168. of the field operation and early results of Kripalani, R.H. and A. Kulkarni, 2001: Monsoon the South China Sea Monsoon Experiment rainfall variations and teleconnections over (SCSMEX). Bull. Amer. Meteor. Soc. 81, 1261– South and East Asia. Int. J. Climatol., 21, 1270. 603–616. Lau, K.-M. and H.-Y. Weng, 2002: Recurrent telecon- Krishnamurthy, V. and B.N. Goswami, 2000: Indian- nection patterns linking summertime precipi- ENSO relationship on inter-decadal time scale. tation variability over East Asia and North J. Climate, 13, 579–595. America. J. Meteor. Soc. Japan, 80, 1309– Krishnamurthy, V. and J. Shukla, 2000: Intra- 1324. seasonal and inter-annual variations of rainfall Lau, K.-M., K.-M. Kim, and J.Y. Lee, 2004: Interan- over India. J. Climate, 13, 4366–4375. nual variability, global teleconnection and po- Krishnamurthy, V. and J. Shukla, 2001: Observed tential predictability associated with the Asian and model simulated interannual variability of summer monsoon. In: The East Asian Mon- the Indian monsoon. Mausam, 52, 133–150. soon. Ed. by C.P. Chang, World Scientific, Sin- Krishnamurti, T.N. and P. Ardanuy, 1980: The 10 to gapore, 564 pp. 20 day westward propagating mode and breaks Lau, N.-C. and B. Wang, 2005: Monsoon-ENSO in- in the monsoon. Tellus, 32, 15–26. teraction. In: The Global Monsoon System: Re- Krishnamurti, T.N., P. Ardanuy, Y. Ramanathan, search and Forecast. Ed. by C.-P. Chang, Bin and R. Pasch, 1981: On the onset vortex of the Wang and N.-C.G. Lau. WMO/TD NO. 1266. summer monsoon. Mon. Wea. Rev., 109, 344– (TMRP Report No. 70) 363. Lau, N.-C. and B. Wang, 2006: Interaction between Krishnamurti, T.N., 1985: Summer Monsoon the Asian monsoon and El Nin˜ o/Southern Os- Experiment-A review. Mon. Wea. Rev., 113, cillation. In: The Asian Monsoon. Ed. by Bin 1590–1626. Wang, Praxis Publishing Ltd., Chichester, UK, Krishnamurti, T.N. and N. Surgi, 1987: Observatio- 479–512. nal aspects of summer monsoon. In: Monsoon Lawrence, D.M. and P.J. Webster, 2001: Inter- Meteorology. Ed. by C.-P. Chang and T.N. annual variations of the intra-seasonal oscilla- Krishnamurti, (Oxford Monographs on Geology tion in the South Asian summer monsoon and Geophysics). Oxford University Press, 3–25. region. J. Climate, 14, 2910–2922. Krishnamurti, T.N., C.M. Kishtanai, T.E. LaRow, Li, C. and M. Yanai, 1996: The onset and interan- D.R. Bachiochi, Z. Zhang, C.E. Willfore, S. nual variability of the Asian summer monsoon Gadgil, and S. Surendran, 1999: Improved in relation to land-sea thermal contrast. J. Cli- weather and seasonal climate forecasts from mate, 9, 358–375. multimodel superensemble. Science, 285, Li, J.P. and Q.C. Zeng, 2002: A unified monsoon 1548–1550. index. Geophy. Res. Lett., 29, 10.1029/2001GL Krishnamurti, T.N., T.S.V. Vijaya Kumar, and A.K. 013874. Mitra, 2006: Seasonal climate prediction of Li, C.Y. and L.P. Zhang, 1999: Summer monsoon ac- Indian summer monsoon. In: The Asian Mon- tivities in the South China Sea and its impacts. soon. Ed. by Bin Wang. Praxis Publishing, Chi- Scientia Atmosphereica Sinica. 23, 257–266 (in chester, UK, 553–584. Chinese). Lau, K.-M. and P.H. Chan, 1986a: Aspects of the 40– Li, T., B. Wang, C.-P. Chang, and Y. Zhang, 2003: A 50 day oscillation during the northern summer theory for the Indian Ocean dipole-zonal mode. as inferred from outgoing longwave radiation. J. Atmos. Sci., 60, 2119–2135. Mon. Wea. Rev., 114, 1354–1367. Li, T., B. Wang, and R.H. Zhang, 2005: Western Lau, K.-M. and P.H. Chan, 1986b: The 40–50 day North Pacific monsoon: Its annual cycle and oscillation and El Nino/Southern Oscillation: A subseasonal-to-interannual variability. In: The new perspective. Bull. Amer. Meteor. Soc., 67, Global Monsoon System: Research and Fore- 533–534. cast. Ed. by C.-P. Chang, Bin Wang and Lau, K.-M., 1992: East Asian summer monsoon rain- N.-C.G. Lau. WMO/TD NO. 1266. (TMRP Re- fall variability and climate teleconnection. port No. 70), 115–135. J. Meteor. Soc. Japan, 70, 211–242. Lin, H. and B. Wang, 2002: The time-space structure Lau, K.-M. and Song Yang, 1997: Climatology and of the Asian–Pacific summer monsoon: A fast interannual variability of the Southeast Asian annual cycle view. J. Climate, 15, 2001–2019. monsoon. Adv. in Atmos. Sci, 14, 141–162. Liu, X.D. and Yanai, M. 2002: Influence of Eurasian July 2007 Y.H. DING 51

spring snow cover on Asian summer rainfall. falls. In: The East Asian Monsoon. Ed. by C.P. Int. J. Climatol. 25, 1075–1089. Chang, World Scientific. Singapore, 564 pp. Liu, Y.M., J.C.L. Chan, J.Y. Mao, and G.X. Wu, 2002: Nitta, T., 1987: Convective activities in the tropical The role of Bay of Bengal convection in the western Pacific and their impacts on the onset of the 1998 South China Sea summer Northern Hemisphere summer circulation. monsoon. Mon. Wea. Rev., 130, 2731–2744. J. Meteor. Soc. Japan, 65, 373–390. Loschnigg, J., G.A. Meehl, P.J. Webster, J.M. Nitta, Ts, T. Mizuno, and K. Takahashi, 1992: Multi- Arblaster, and G.P. Compo, 2003: The Asian scale convective systems during the initial monsoon, the trospheric biennial oscillation, phase of 19886/1987 El Nin˜ o. J. Meteor. Soc. and the Indian Ocean zonal mode in the Japan, 70, 317–335. NCAR GCM. J. Climate, 16, 1617–1642. Qian, W.H., H.-S. Kang, and D.-K. Lee, 2002: Distri- Madden, R.A. and P.R. Julian, 1971: Detection of a bution of seasonal rainfall in the East Asia 40–50 day oscillation in the zonal wind in the monsoon region. Theor. Appl. Climatol., 73, tropical Pacific. J. Atmos. Sci., 28, 702–708. 151–168. Matsumoto, J., 1988: Large-scale features associated Ramage, C.S., 1971: Monsoon Meteorology (Int. Geo- with the frontal zone over East Asia from late phys. Ser. Vol. 15), Academic Press, San Diego, summer to Autumn. J. Meteor. Soc. Japan, 66, California, 296 pp. 565–579. Rao, Y.P., 1976: Southwest Monsoon (Meteorological Matsumoto, J., 1990: Withdrawal of the Indian sum- Monograph, Synoptic Meteorology No. 1), In- mer monsoon and its relation to the seasonal dian Meteorological Department, New Dehli, transition from summer to autumn over East 367 pp. Asia. Mausaun, 41, 196–202. Ropelewski, C.F., M.S. Halpert, and X. Wang, 1992: Matsumoto, J., 1997: Seasonal transition of summer Observed tropospheric biennial vaiability and rainy season over Indo-China and adjacent its relationship to the Southern Oscillation. monsoon. Adv. in Atmos. Sci., 14, 231–245. J. Climate, 5, 594–614. Mc Phaden, M.J., 1999: Genesis and evolution of Saji, N.H., B.N. Goswami, P.N. Vinayachandran, and 1997–1978 El Nin˜ o. Science, 283, 950–954. T. Yamagata, 1999: A dipole mode in the tropi- Meehl, G.A., 1994: Coupled land-ocean-atmosphere cal Indian Ocean. Nature, 401, 360–363. processes and South Asian monsoon variabil- Schlesinger, M.E. and N. Ramankutty, 1994: An os- ity. Science, 266, 263–267. cillation in the global climate system of period Meehl, G.A. and J.M. Arblaster, 2002: The tro- 65–70 years. Nature, 367, 723–726. pospheric biennial oscillation and Asian- Shinoda, M., 2001: Climate memory of snow mass as Australian monsoon rainfall. J. Climate, 15, soil moisture over central Eurasia. J. Geophys. 722–744. Res., 106, 33393–33403. Mehta, V.M. and K.-M. Lau, 1997: Influence of solar Sikka, D.R. and S. Gadgil, 1980: On the maximum irradiance on the Indian monsoon: ENSO rela- cloud zone and the ITCZ over Indian longitude tionship at decadal-multidecadal time scales. during southwest monsoon. Mon. Wea. Rev., Geophys. Res. Lett., 24, 159–162. 108, 1840–1853. Murakami, T., 1980: Empirical orthogonal function Slingo, J.M., D.P. Rowell, K.R. Sperber, and E. analysis of satellite-observed outgoing long- Northey, 1999: On the predictability of inter- wave radiation during summer. Mon. Wea. annual behavior of the Madden-Julian oscilla- Rev., 108, 205–222. tion and its relationship with El Nin˜ o. Quart. Murakami, T. and Y.H. Ding, 1982: Wind and tem- J. Roy. Meteor. Soc., 125, 583–609. perature changes over Eurasia during the Stephens, G.L., P.J. Webster, R.H. Johnson, R. Eng- early summer of 1979. J. Meteor. Soc. Japan, len, and T. L’euler, 2004: Observational evi- 60, 183–196. dence for the mutual regulation of tropical Murakami, T. and J. Matsumoto, 1994: Summer hydrological cycle and tropical sea surface tem- monsoon over the Asian continent and western perature. J. Climate, 17, 2213–2224. North Pacific. J. Meteor. Soc. Japan, 70, 597– Sumi, A., N.-C. Lau, and W.-C. Wang, 2005: Present 611. status of Asian monsoon simulation. In: The Ninomiya, K. and T. Murakami, 1987: The early Global Monsoon System: Research and Fore- summer rainy season (Baiu) over Japan. In: cast. Ed. by C.-P. Chang, Bin Wang and The Monsoon Meteorology, Ed. by C.-P. Chang N.-C.G. Lau. WMO/TD No. 1266, 376–385. and T.N. Krishnamurti, Oxford University Takayabu, Y.N., T. Iguchi, M. Kachi, A. Shibata, and Press, 93–121. H. Kanzawa, 1999: Abrupt termination of the Ninomiya, K., 2004: Large- and mesoscale features of 1997–1998 El Nin˜ o in response to a Madden- Meiyu-Baiu front associated with intense rain Julian oscillation. Nature, 402, 279–282. 52 Journal of the Meteorological Society of Japan Vol. 85B

Tao, S.Y. and L.X. Chen, 1987: A review of recent Wang, B., R. Wu, and K.-M. Lau, 2001: Interannual research on the East Asian monsoon in China. variability of Asian summer monsoon: Con- In: Monsoon Meteorology. Ed. by C.-P. Chang trast between the Indian and western North and T.N. Krishnamurti, Oxford University Pacific-East Asian monsoons. J. Climate, 14, Press, London, 60–92. 4073–4090. Teng, H.Y. and B. Wang, 2003: Interannual varia- Wang, B. and Lin Ho, 2002: Rainy seasons of The tions of the boreal summer oscillation in the Asian-Pacific monsoon. J. Climate, 15, 386– Asian–Pacific region. J. Climate, 16, 3572– 396. 3584. Wang, B. and Q. Zhang, 2002: Pacific-East Asian tel- Tomas, R.A. and P.J. Webster, 1997: The role of econnection. Part II: How the Philippine Sea inertial instability in determing the location anticyclone established during development of and strength of the near-equatorial convection. El Nin˜ o. J. Climate, 15, 3252–3265. Quart. J. Roy. Meteor. Soc., 123, 1445–1482. Wang, B., R. Wu, and T. Li, 2003: Atmosphere-warm Trenberth, K.E., J.W. Hurrell, and D.P. Stepaniak, Ocean interaction and its impact on Asian- 2006: The Asian mosoon: Global perspectives. Australian Monsoon variation. J. Climate, 16, In: The Asian Monsoon, Ed. by Bin Wang, 1195–1211. Praxis Publishing Ltd., Chichester, UK, 781 Wang, B., 2004: The Asian monsoon intraseasonal pp. variability: Basic theories. In: Intraseasonal Ueda, H., T. Yasunari, and Kawamura, 1996: Abrupt Variability of the Atmosphere-Ocean Climate seasonal change of large-scale convective activ- System. Ed. by K.-M. Lau and D. Waliser, ity over the Western Pacific in the Northern Springer-Praxis, Chichester, UK. summer. J. Meteor. Soc. Japan, 73, 795–809. Wang, B., T. Li, Y.H. Ding, R.H. Zhang, and H.J. Ueda, H. and T. Yasunari, 1996: Maturing process Wang, 2005a: East Asian-Western North Pa- of summer monsoon over the western North cific monsoon: A distinctive component of the Pacific-A coupled Ocean/Atmosphere system. Asian-Australian monsoon system. In: The J. Meteor. Soc. Japan, 74, 493–508 (JMSJ Global Monsoon System: Research and Fore- award paper). cast. Ed. by C.-P. Chang, Bin Wang and Ueda, H. and T. Yasunari, 1998: Role of warming N.-C.G. Lau, WMO/TD No. 1266 (TMRP Re- over the Tibetan Plateau in early onset of the port No. 70), 72–94. summer monsoon over the Bay of Bengal and Wang, B., Q.H. Ding, X.H. Fu, I.-S. Kang, K. Jin, J. the South China Sea. J. Meteor. Soc. Japan, Shukla, and F. Doblas-Reyes, 2005b: Funder- 76, 1–12. mental challenge in simulation and prediction Waliser, D.E., 2005: Intraseasonal variability, In: of summer monsoon rainfall. Geophys. Res. the Global Monsoon System: Research and Lett., 32, L15711, doi: 10.1029/2005GL022734. Forecast. Ed. by C.-P. Chang, Bin Wang and Wang, B. (Ed.), 2006: The Asian Monsoon. Praxis N.-C.G. Lau. WMO/TD NO. 1266. (TMRP Re- Publishing Ltd., Chichester, UK, 787 pp. port No. 70), 403–439. Wang, B. and Q.H. Ding, 2006: Changes in global Vernekar, A.D., J. Zhou, and J. Shukla, 1995: The monsoon prediction over the past 56 years. effect of Eurasian snow cover on the Indian Geophys. Res. Lett., 33, L06711, doi: 10.1029/ monsoon. J. Climate, 8, 248–266. 2005GL025347, 2006. Waliser, D.E., 2006: Intraseasonal variability. In: Wang, H., Y.H. Ding, and J.H. He, 2005c: A climato- The Asian Monsoon. Ed. B. Wang. Praxis Pub- logical study of summer monsoon over the lishing Ltd, Chichester, UK, 787 pp. western North Pacific. Acta Meteorologica Wang, B. and H. Rui, 1990: Dynamics of the coupled Sinica. 63, 418–430 (in Chinese with English moist Kelvin-Rossby wave on an equatorial abstract). beta-plane. J. Atmos. Sci., 47, 397–413. Wang, H., Y.H. Ding, and J.H. He, 2006: The influ- Wang, B. and X., Xu, 1997: Northern Hemispheric ence of western North Pacific summer monsoon summer monsoon singularities and climatolog- changes on typhoon genesis. Acta Meteorolog- ical intraseasonal oscillation. J. Climate, 10, ica Sinica, 64, 345–356. 1071–1085. Wang, H.J., 2001, The weakening of the Asian mon- Wang, B. and X.S. Xie, 1998: A coupled model of the soon circulation after the end of 1970’s. Adv. warm pool climate system. Part I: The role in Atmos. Sci., 18, 376–386. of air-sea interaction on maintaining Madden- Wang, S.W., R.S. Wu, and X.Q. Yang, 2005: Climate Julian oscillation. J. Climate, 11, 2116–2135. change in China. In: Climate and Environment Wang, B., R. Wu, and X. Fu, 2000: Pacific-East Asia Change in China and Their Projections, China teleconnection: How does ENSO affect East Science Press, Beijing, 63–95. Asian Climate. J. Cliamte, 13, 1517–1536. Wang, Y.Q., L.R. Leung, J.L. Mc Gregor, D.K. Lee, July 2007 Y.H. DING 53

W.-C., Wang, Y. Ding, and F. Kimura, 2004: mer monsoon and rainfall: An observational in- Regional climate modeling: Progress, chal- vestigation. J. Climate. 16, 2038–2051. lenges, and prospects. J. Meteor. Soc., Japan, Wu, R.G. and B. Kirtman, 2004: The tropospheric 82, 1599–1628. biennial oscillation of the monsoon-ENSO Wang, Z.Y. and Y.H. Ding, 2006: Climatology of system in an Interactive Ensemble Coupled rainy seasons in China, 2006 Western Pacific GCM. J. Climate, 17, 1623–1640. Geophysics Meeting of AGU, 24–27 July, 2006, Yanai, M., S. Esbenson, and J. Chu, 1973: Detemina- Beijing, China. tion of a bulk properties of tropical cloud clus- Webster, P.J. and R. Lukas, 1992: The Coupled ter from large-scale heat and moisture bud- Ocean-Atmospheric Response Experiment. gets. J. Atmos. Sci., 20, 611–627. Bull. Amer. Meteor. Soc. 73, 1377–1416. Yanai, M., C. Li and Z. Song, 1992: Seasonal heating Webster, P.J., V.O. Magata, T.N. Palmer, J. Shukla, of the Tibetan Plateau and its effects on the R.A. Tomas, M. Yanai, and T. Yasunari, 1998: evolution of the Asian summer monsoon. J. Monsoons: Progresses, predictability, and the Meteor. Soc. Japan, 70, 319–351. prospects for prediction. J. Geophys. Res., 103, Yanai, M. and G.X. Wu, 2006: Effects of the Tibetan 144501–144510. Plateau. In: The Asian Monsoon. Ed. by Bin Webster, P.J., A.M. Moore, J.P. Loschnigg, and R.R. Wang, Praxing Publishing Ltd., UK, 513–549. Leben, 1999: Coupled ocean-atmosphere dy- Yang, F.L. and K.-M. Lau, 2004: Trend and variabil- namics in the Indian Ocean during 1997– ity of China precipitation in spring and sum- 1998. Nature, 401, 356–360. mer: Linkage to sea surface temperature. Int. Webster, P.J., E.F. Bradley, C.W. Fairall, J.S. God- J. Climatol., 24, 1625–1644. frey, P. Hacker, R.A. Houze Jr, R. Lukas, Y. Yang, S., K.-M. Lau, and K.M. Kim, 2002: Varia- Serra, J.M. Humman, T.D.M. Lawrence et al., tions of the East Asian jet stream and Asian- 2002: The JASMINE pilot study. Bull. Amer, Pacific-American winter climate anomalies. J. Meteorol. Soc, 83, 1603–1630. Climate, 15, 306–325. Webster, P.J., H.-R. Chang, T. Hopson, C. Hoyos, Yang, S. and K.-M. Lau, 2006: Interannual variabil- and A. Subbiah, 2005a: Bridging the gap ity of the Asian monsoon. In. The Asian Mon- between monsoon research and applications: soon. Ed. by Bin Wang, Praxis Publishing Development of overlapping three-tier predic- Ltd., Chichester, UK, 259–294. tion schemes to facilitate ‘‘useful’’ forecasts. In: Yano, J. and K. Emanuel, 1991: An improved model The Global Monsoon System: Research and of the equatorial troposphere and its coupling Forecast. Ed. by C.-P. Chang, Bin Wang and with the stratosphere. J. Atmos. Sci., 48, 377– N.-C.G. Lau. WMO/TD No. 1266, 3–13. 389. Webster, P.J., 2005b: Oceans and monsoons, In: The Yasunari, T., 1979: Cloudness fluctuation associated Global Monsoon System: Research and Fore- with the northern hemisphere summer mon- cast. Ed. by C.P. Chang, Bin Wang and N-C.G. soon. J. Meteor. Soc. Japan, 57, 227–242. Lau, WMO/TD NO. 1266, 253–298. Yasunari, T., 1980: A quasi-stationary appearance of Webster, P.J., 2006: The coupled monsoon system. In the 30–40 day period in the cloudiness fluctua- The Asian Monsoon. Ed. by Bin Wang, Praxis tions during the summer monsoon over India. Publishing Ltd., Chichester, UK, 3–66. J. Meteor. Soc. Japan, 59, 336–354. Wu, G.X. and Y.S. Zhang, 1998: Tibetan Plateau Yasunari, T., A. Kitoh, and T. Tokioka, 1991: Local forcing and the timing of the monsoon onset and remote responses to excessive snow mass over South Asia and the South China Sea. over Eurasia appearing in the Northern spring Mon. Wea. Rev., 126, 913–927. and summer climate-A study with the MRI. Wu, R. and B. Wang, 2001: Multi-stage onset of the GCM. J. Meteor. Soc. Japan, 69, 473–487. summer monsoon over the western North Pa- Yasunari, T., 2006: Land-atmosphere interaction. In: cific. Clim. Dyn., 17, 277–289. The Asian Monsoon. Ed. by Bin Wang, Praxis Wu, R., 2002: Processes for the northeastward ad- Publishing Ltd., Chichester, UK, 459–478. vance of the summer monsoon over the west- Yeh, T.C., S.Y. Dao, and M.T. Li, 1958, The abrupt ern North Pacific. J. Meteor. Soc. Japan, 80, change of circulation over northern hemisphere 67–83. during June and October. Acta Meteor. Sinica, Wu, R. and B. Wang, 2002: A contrast of the East 29, 249–263. Asian summer monsoon and ENSO relation- Zhang, Y., T. Li, and B. Wang, 2004: Decadal change ship between 1962–1977 and 1978–1993. J. of the spring snow depth over the Tibetan Pla- Climate, 15, 3266–3279. teau: The associated circulation and influence Wu., T.W. and Z.A. Qian, 2003: The relation between on the East Asian summer monsoon. J. Cli- the Tibetan winter snow and the Asian sum- mate, 17, 2780–2793. 54 Journal of the Meteorological Society of Japan Vol. 85B

Zhang, Y.C. and Y.F. Qian, 2002: Mechanism of ther- mer monsoon onset over Southeast Asia. Int. mal features over the Indochina Peninsula and J. Climatol., 24, 1461–1482. possible effects on the onsets of the South Zveryaev, I.I., 2002: Inter-decadal changes in the China Sea monsoon, Adv. in Atmos. Sci., 19, zonal wind and the intensity of intraseasonal 885–900. oscillation during boreal summer Asian mon- Zhang, Z., J.C. Chan, and Y. Ding, 2004: Character- soon. Tellus, 54A, 288–298. istics, evolution and mechanisms of the sum-