Westward shedding of anticyclones from the Tibetan anticyclone is indeed evident in NCEP reanalysis data. These eddies are trapped near the tropopause. Cutoff potential vorticity features are confined to within about 20 K of the tropopause; in geopotential, they extend somewhat further, but not below about 400 hPa. 

The following observations have been made in support of this hypothesis. First, the summer precipitation regime is dominated by a continental-scale precipitation pattern characterized by an our-of-phase relationship between precipitation in the southwestern United States and that in the Great Plains of the United States. Second, interannual fluctuations in the onset date of the monsoon in the southwestern United States are significantly correlated with interannual fluctuations in the intensity of summer rainfall in this region such that early monsoons are often very wet and late monsoons tend to be dry. Third, wet (dry) monsoons in the southwestern United States often follow winters characterized by dry (wet) conditions in the southwestern United States and wet (dry) conditions in the northwestern United States. Finally, interannual variability of the summer monsoon in the southwestern United States is modulated by long term (decade scale) fluctuations in the North Pacific SSTs. The mechanism relating the North Pacific SST pattern to interannual variability in the summer monsoon appears to be via the impact of variations in the Pacific jet on West Coast precipitation regimes during the preceding winter Multiyear fluctuations in the North Pacific SST pattern are consistent with multiyear fluctuations in the atmospheric circulation and in the West Coast precipitation regimes during Northern Hemisphere (NH) winter. hence with multiyear variability in the summer monsoon state. Influences on the summer monsoon during the preceding winter and spring are tied together using appropriate SST indices that capture decade-scale variability in the North Pacific during NH winter and interannual variability in the eastern tropical Pacific during NH spring. The results suggest that decadal variability in the North Pacific SSTs may be an important factor in determining long-term periods of summertime drought or rainy conditions both in the southwestern United States and in the Great Plains of the United States. 

The physical link through which ENSO is related to decreased monsoon rainfall on both interannual and interdecadal timescales has been investigated using National Centers for Environmental Prediction-National Center for Atmospheric Research reanalysis products. The decrease in the Indian monsoon rainfall associated with the warm phases of ENSO is due to an anomalous regional Hadley circulation with descending motion over the Indian continent and ascending motion near the equator sustained by the ascending phase of the anomalous Walker circulation in the equatorial Indian Ocean. It is shown that, to a large extent, both the regional Hadley circulation anomalies and Walker circulation anomalies over the monsoon region associated with the strong (weak) phases of the interdecadal oscillation are similar to those associated with the strong (weak) phases of the interannual variability. However, within a particular phase of the interdecadal oscillation, there are several strong and weak phases of the interannual variation. During a warm eastern Pacific phase of the interdecadal variation, the regional Hadley circulation associated with El Nino reinforces the prevailing anomalous interdecadal Hadley circulation while that associated with La Nina opposes the prevailing interdecadal Hadley circulation. During the warm phase of the interdecadal oscillation, El Nino events are expected to be strongly related to monsoon droughts while La Nina events may not have significant relation: On the other hand, during the cold eastern Pacific phase of the interdecadal SST oscillation, La Nina events are more likely tobe strongly related to monsoon floods while El Nino events are unlikely to have a significant relation with the Indian monsoon. This picture explains the observation that the correlations between IMR and ENSO indices on the interannual timescale do not follow the interdecadal oscillation as neither phase of the interdecadal oscillation favors a stronger (or weaker) correlation between monsoon and ENSO indices. 

The numerical experiments are done with an atmospheric general circulation model over an aquaplanet with zonally uniform sea surface temperature. The existence of the two flow regimes, the two "forces," and the abrupt transition are all demonstrated in the experiments. Experimental results show high dependence on the choice of cumulus parameterization scheme. especially during the equatorial trough circulation regime. Although the proposed interpretation is more suitable for explaining the monsoon trough onset in the western Pacific, it is hypothesized that the same basic mechanism is also at the core of monsoon onset in other parts of the Tropics. 

It is shown that the summer rainfall in the eastern China is closely associated with the winter snow cover over QXP not only in the interannual variation but also in the decadal variation. A clear relationship exists in the quasi-biennial oscillation between the summer rainfall in the northern part of North China and the southern China and the winter snow cover over QXP. Furthermore, the summer rainfall in the four climate divisions of Qinling-Daba Mountains, the Yangtze-Huaihe River Plain, the upper and lower reaches of the Yangtze River showed a remarkable transition from drought period to rainy period in the end of 1970's, in good correspondence with the decadal transition of the winter snow cover over QXP. 

In the second part of the paper, the interannual variability of EAM and its relationship with sea surface temperature (SST) in the 200 year simulation were studied by using the composite method, wavelet transformation, and the moving correlation coefficient method. The summer EAMI is negatively correlated with ENSO (El Nine and Southern Oscillation) cycle represented by the NINO3 sea surface temperature anomaly (SSTA) in the preceding April and January, while the winter EAM is closely correlated with the succeeding spring SST over the Pacific in the coupled model. The general differences of EAM between El Nine and La Nina cases were studied in the model through composite analysis. It was also revealed that the dominating timescales of EAM variability may change in the long-term variation and the strength may also change. The anomalous winter EAM may have some correlation with the succeeding summer EAM, but this relationship may disappear sometimes in the long-term climate variation. Such time-dependence was found in the relationship between EAM and SST in the long-term climate simulation as well. 

First, a 50-year atmospheric general-circulation model simulation with climatological SST is examined to determine the tropical Pacific wind-stress anomalies that are associated with a variable monsoon but that are also independent of SST variability in the tropical Pacific. Using simple statistical techniques, it is found that a weak (strong) monsoon results in a weakening (strengthening) of the trade winds over the tropical Pacific. To examine how these 'monsoon-forced wind-stress anomalies' in the tropical Pacific affect ENSO, simulations were made with a simple coupled model that does not include the effects of a variable monsoon. The effects of the monsoon are then added in the coupled model by either specifying the strength of the monsoon or by parametrizing the strength of the monsoon in terms of the coupled-model simulated SSTA in the east Pacific. Based on these coupled simulations, a variable monsoon enhances the ENSO variability, particularly three to six months after the monsoon ends, and can also serve as a trigger mechanism for ENSO. It is found that an ongoing warm (cold) ENSO event is made even warmer (colder) by a weak (strong) monsoon. Similarly, warm (cold) events are weakened by a strong (weak) monsoon. These results also reproduce the observed lag/lead ENSO-monsoon relation where the maximum negative correlation between the monsoon and the SSTA in the east Pacific occurs 3-6 months after the monsoon season. 

A multiple linear regression equation is developed to predict the Indian summer monsoon rainfall using these indexes and the empirical relations are verified on independent data.

Good results were obtained in forecasting the summer monsoon rainfall for the whole of India. The forecast of summer monsoon rainfall for west-central India and all-India rainfall for July also appears to be encouraging. The indexes, thus, seem to be useful in long-range forecasting of the Indian summer monsoon rainfall. 

The CGCM reasonably well reproduces the observed ENSO-related interannual variability of the tropical circulation system. Its main deficit is associated with a westward displacement of simulated SST variability. There is an underestimation of precipitation around the Philippines. Differences are found between the CGCM and the AGCM in the variability over the Indian monsoon region. The AGCM responds well to the prescribed SST anomaly in the Pacific. It behaves erroneously over the Indian Ocean. This may be related to the fact that the AGCM is only responding to the prescribed SST fields, while the CGCM includes two-way atmosphere-ocean interactions.

The CGCM results show that the simulated Indian summer monsoon rainfall anomalies are negatively correlated with the equatorial Pacific SST anomalies. It is consistent with the observed one, i.e., good monsoon is associated with La Nina. As a precursory signal, ground temperature is significantly warmer in spring in central Asia preceding a good monsoon. It is noted that the snow cover anomalies are negative in the above region, but its significance is marginal. 

In the strong phase of the TBO, the area of relatively strong monsoon convective maximum over South Asia in the spring to summer season moves southeastward to Indonesia in the autumn to winter season. This movement superimposes on its climatological seasonal cycle. It suggests that the northern winter monsoon convection tends to be strong around Indonesia to northern Australia when the summer Asian monsoon is strong. The anomalous state of the air-sea coupled system in the Pacific sector which forms in the summer season, seems to dissipate from its eastern edge. This occurs by a local atmosphere-ocean coupling process through a large scale Walker circulation.

The convection anomalies persist during the entire monsoon season over Indonesia and northern Australia. As a response to this equatorial monsoon convection anomaly, a Matsuno-Gill type stationary Rossby wave is established over the South Asian region. The appearance of upper level anticyclonic circulation and lower level cyclonic circulation anomaly in a strong monsoon year is a favorable condition for bringing the cold air advection over the Eurasian continent. Cold air advection after the strong monsoon persists through the whole winter to spring season to form the cold tropospheric temperature around Central to South Asia. Then reduced land-sea, or north-south temperature contrasts sets up the following weak South Asian summer monsoon. The simulated TBO of the South Asian monsoon is tightly phase locked with a seasonal cycle. The phase of the TBO changes in northern spring, which suggests that the extratropical-tropical interaction be realized mainly during winter to spring through the onset of South Asian monsoon. Our results imply that the TBO is an inherent feature in the land-monsoon-ocean coupled system, and emphasize a more active role of monsoon-extratropical interaction in the Indian sector in winter to spring season for regulating the TBO cycle. 

A well defined east-west oriented trough system, extending from about 28 degrees N Latitude/65 degrees E Longitude to 20 degrees N Latitude/90 degrees E Longitude, in the lower levels, has been the main feature associated with the strong phase of the monsoon corresponding to PC I. The trough in the lower levels is more marked in the eastern half compared to the western half in both the sets of charts associated with strong phases of the monsoon related to the PC II and PC III. With PC II, the position of the troughs in the lower levels is further north of its location in PC III. The east-west trough system associated with the strong phase of PC IV has a large southward tilt with height. The charts corresponding to the weak phases of these PCs have synoptic features, such as the position of the trough close to the foothills of the Himalayas, and the shifting of middle and upper tropospheric anticyclones to the south.

The study suggests an objective method of interpretation of principal components by utilising synoptic data. In addition, synoptic models and data sets corresponding to different phases of the monsoon can also be prepared in an objective manner by such PCA. 

An inquiry into circulation and precipitation data indicates that it was the intraseasonal variability of the monsoon system that brought the above-normal rainfall over India. Furthermore, it is shown that the 1997 El Nino not only suppressed the large-scale Asian monsoon circulations, but also produced a convectively unstable area off the east coast of Somalia through the modifications in sea surface and lower tropospheric conditions. Anomalous convection was triggered, amplified over this area, then intruded northward to the Indian subcontinent as it propagated eastward.

A moisture budget analysis using 41 years (1958-98) of the NCEP/NCAR reanalysis data set confirmed that the transient part of moisture transport anomaly in the 1997 summer was larger than the stationary one, contributing positively to the small positive rainfall anomalies over India. The 1997 summer appears to be one of the singular cases in the context of the general positive correlation between the monsoon circulation and precipitation indices over the recent 41 years. 

Analysis of the model output shows that some of the systematic errors in the model in the monsoon simulation are reduced with incorporation of downdrafts. The systematic error by which the monsoon low level westerlies extend deep into the west Pacific ocean is substantially reduced by the inclusion of downdrafts. The lower tropospheric monsoon flow over the north Indian ocean is strengthened with downdrafts. TEJ simulated with downdrafts is closer to NCEP reanalysis. Inclusion of downdraft improves the precipitation simulation over land areas of India, and also simulates the correct position of the monsoon trough. When downdrafts are not included there is excessive precipitation over the west equatorial Indian ocean and less precipitation over India. Surface latent heat Aux increases with the downdrafts. The systematic error of the model which simulates stronger low level circulation in El Nino years than in La Nina years is reduced, though not completely eliminated. The interannual variability of simulated monsoon precipitation is more realistic with downdrafts. Spells of active and 'break,' or weak monsoon situations, are more prominently simulated when downdrafts are included. Larger number of strong synoptic disturbances are simulated over India with the inclusion of downdrafts. 

Comparing the difference of OLR, wind fields, vertical circulations and SST anomalies in the strong and weak STH, we investigate the characteristics of global circulations and the SST distributions related to the anomalous STH at the western Pacific both in winter and summer. Much attention has been paid to the study of the air-sea interaction and the relationship between the East Asian monsoon and the STH in the western Pacific. A special vertical circulation, related to the STH anomalies is found, which connects the monsoon current to the west and the vertical flow influenced by the SST anomaly in the tropical eastern Pacific. 

Part II investigates these relations further. Time-lag correlations are calculated between the key indices and atmospheric variables over the equatorial Indo-Pacific Oceans. If type II years, derived by the WAI, are removed from the 34-year time series, correlations between the TOI and these variables increase appreciably, now showing clearly the biennial character. The analysis identifies a sequence of events involving biennial oscillation of the ENSO-monsoon system from approximately JAS(-1) to JAS(0), followed by intensification of the ENSO from JAS(0) to November-December-January(+1). The ENSO-monsoon oscillation system is not sinusoidal but skewed.

To show the geographical patterns associated with the above sequence of events, planar maps are presented of the lag correlation between the observed TOI(JAS) and (i) Vertical velocity at the 500 hPa level, (ii) precipitation, (iii) sea surface temperature (SST), and (iv) atmospheric sea level pressure. Distinct geographical distributions of the ENSO-monsoon oscillation emerge in both the observations and model data. One pattern is characterized by a horseshoe shape over the Pacific, which is generally symmetric around the equator, but with geographical differences depending on location in the lag sequence. The other pattern is a see-saw shape, primarily a standing oscillation located in the eastern South Pacific and the Indian Oceans, resembling the sea-level-pressure pattern found by Trenberth and Shea. Applying the lead-lag relationship, it is demonstrated that the SST over the central Pacific four months ahead can be projected, based on the TOI(JAS). Conversely, the intensity of the Indian monsoon rainfall for non-type II years can be projected 15 months ahead by the SST over the eastern Pacific Ocean. This indicates that the ENSO-monsoon oscillation system is quasi-periodic, as opposed to irregular, with a two-year cycle; this is clearly revealed with the removal of type II years. 

The results of composite and individual analyses of TBB anomalies show that the interannual variability of the convective activities associated with the summer monsoon in East Asia is large and has a close relation to the thermal distribution of SST anomalies in the tropical Pacific, especially in the western Pacific warm pool. In the sumner with ENSO-like distribution of SST anomalies in the tropical Pacific, the convective activities are weak around the Philippines, then the convective activities are intensified and the summer monsoon rainfall is strong in the area from the Yangtze River basin and the Huaihe River basin in China to Republic of Korea and Japan. On the contrary, in the summer with anti-ENSO-like distribution of SST anomalies in the tropical Pacific, the convective activities are strong around the Philippines, then the convective activities are weakened and the summer monsoon rainfall is weak in the area from the Yangtze River basin and the Huaihe River basin to Republic of Korea and Japan.

It may be also found either from the composite analysis or from the individual analysis of TBB anomalies that the convective activities associated with the summer monsoon in East Asia have a good negative relation to that around the Philippines and a positive relation to that over the equatorial central Pacific. 

It is found, somewhat surprisingly, that western Eurasia is the only geographical region for which a significant inverse correlation exists between winter snow cover and subsequent summer monsoon rainfall. However, composites for high and low snow cover over Eurasia show spatially homogeneous large-scale patterns of snow cover and surface temperature anomalies. Winters of high and low snow cover for Eurasia are found to be associated with colder and warmer than normal temperatures, respectively, for large regions of the Eurasian continent. The inverse snow-monsoon relationship holds especially in those years when snow is anomalously high or low for both the winter as well as the consecutive spring season. Contrary to previous findings, no significant relation is found between the Himalayan seasonal snow cover and subsequent monsoon rainfall. 

The moisture balance over a domain bounded by 20-40 degrees N and 110-125 degrees E (Area-C, main part of China) indicates that the moisture flux convergence within this domain is mainly due to the southerly moisture inflow across the southern boundary. The precipitation in this domain increases in the early June with the abrupt increase of the southerly moisture inflow. The precipitation within this domain is highly correlated with the southerly moisture inflow crossing 20 degrees N.

To relate the southerly moisture flow into Area-C with the moisture balance in the tropical zone, the moisture balance in a domain bounded by 0-20 degrees N and 110=125 degrees E (Area-S, the South China Sea including the monsoon trough region) is further studied. The variation of the cross-equatorial moisture inflow in Area-S is significantly smaller than that of the moisture efflux across the northern boundary of Area-S. In Area-S, the strong correlation is found between the zonal moisture flux convergence and the southerly moisture flux across the 20 degrees N latitude circle. This zonal moisture flux convergence is seen within the monsoon trough region between the monsoon westerly and the easterly flux in the southern rim of the Pacific subtropical anticyclone. The monsoon trough region plays the role of the channel to transport the moisture from 0-20 degrees N zone toward north, although Area-S itself is not a moisture source region. The west-east intraseasonal oscillation of the monsoon trough, monsoon westerly and the North Pacific subtropical anticyclone have strong influence on the precipitation and westerly moisture transport over China.

The features of precipitation over China in early July are characterized by a narrow intense rainfall zone over the Yangtze River Valley. The moisture flux field during this period indicates the strong moisture flux convergence within the precipitation zone. 

The annual-cycle (1 yr) variance time series of Nino3 SST and Indian rainfall is negatively correlated with the interannual ENSO signal. The 1-yr variance is larger during 1935-60, suggesting a negative correlation between annual-cycle variance and ENSO variance on interdecadal timescales.

The method of wavelet coherency is applied to the ENSO and monsoon indexes. The Nino3 SST and Indian rainfall are found to be highly coherent, especially during intervals of high variance. The Nino3 SST and Indian rainfall are approximately 180 degrees out of phase and show a gradual increase in phase difference versus Fourier period. All of the results are shown to be robust with respect to different datasets and analysis methods. 

In agreement with observations, the simulated intraseasonal monsoon activity is mainly described by irregular alternations of active spells and break spells associated with fluctuations of the Tropical Convergence Zone (TCZ) between a continental and an oceanic regime. In the model, the spatial characteristic of the intraseasonal monsoon variability is a robust feature which is primarily related to an internal mode of variability of the system, rather than to a response to land-surface feedbacks. Experimentation indicates that the simulation of northward propagating events, related to transitions in the regime, does not require the inclusion of interactive surface hydrological processes. This suggests that the transitions are also mainly related to internal atmospheric dynamics.

The temporal characteristics of the fluctuations between the two TCZ regimes, however, are influenced by an interactive surface. The low-frequency intraseasonal monsoon variability is enhanced by hydrological surface feedbacks. When the surface interacts with the atmosphere, the active and break regimes of the monsoon are equally likely. In the absence of surface feedbacks, the probability distribution is modified and the changes depend on the land-surface conditions imposed. The results show that the probability of a monsoon break exceeds that of an active phase when the imposed land-surface conditions are based on climatological values for July. This asymmetry in the probability distribution affects intraseasonal monsoon variability. In turn, the time-mean monsoon circulation, depending as it does on the statistics of the intraseasonal oscillations (such as frequency of occurrence and mean amplitude), is also modified by the surface feedbacks. It follows that the surface conditions play a role in the interannual predictability of the time-mean monsoon. 

The impact of the ISO on the development of the low-level southwesterlies is both local and remote, and depends on the strength and phasing of the ISO with the seasonal cycle. Of the 14 yr examined here, 6 showed a strong contribution to the northeastward progression and onset of the monsoon rains over India. In these cases, the ISO is initially (about 2 weeks prior to onset of rains over India) out of phase with, and therefore suppresses, the seasonal development of the regions of large-scale rising and sinking motion. As the ISO moves to the northeast, the rising branch enters the Indian Ocean and acts to enhance the latent heating in the region of the emerging Somali jet. At low levels the response rakes the form of an anticyclonic circulation anomaly over the Arabian Sea, and a cyclonic circulation anomaly to the south, which nets to inhibit the eastward progression of the Somali jet. As the ISO moves in phase with and enhances the seasonal mean upper-level divergent circulation, there is an abrupt and intense development of the southwesterly winds leading to an unusually rapid northeast shift and intensification of the monsoon rains over India and the Bay of Bengal. The general northeast progression of the anomalies may be viewed a!, an initial suppression and then acceleration of the "normal" seasonal cycle of the monsoon. 

1) Interannual variation. Based upon the SST averaged over the National Oceanic and Atmospheric Administration NINO3 region (150 degrees-90 degrees W, 5 degrees S-5 degrees N), the summers of 1982, 1983, 1987, and 1991 and 1981, 1984, 1985, 1988, 1989, and 1994 are defined as warm and cold, respectively. A clear interannual variation can be seen in the frequency of monsoon depressions in the Bay of Bengal: an enhancement (reduction) of monsoon depression activity occurs during cold (warm) summers. This interannual variation of monsoon depression activity is traceable to the corresponding variation of the combined tropical cyclone and 12-24-day monsoon low frequency in the south China Sea. The latter interannual variation results from the development of an;anomalous anticyclonic (cyclonic) circulation between 15 degrees and 30 degrees N in the WTP-SCS region in response to the warm (cold) SST anomalies in the eastern tropical Pacific.

2) Intraseasonal variation. There is an intraseasonal variability in the occurrence of tropical cyclones and of 12-24-day monsoon lows over the south China Sea, which is followed by a corresponding variability of monsoon depressions over the Bay of Bengal. The formation frequency of these depressions is dependent on the penetration role of the residual lows of these two types of disturbances across Indochina. These residual lows lead to an intraseasonal change in monsoon depression formation in connection with a deepening/filling of the monsoon trough over northern India and the Bay of Bengal. 

It has been found that during normal, flood and drought years, the first four (most dominated) principal component with 'significantly positive' correlated stations explains 73%, 77% and 100% of the variance, while the remainder 'weakly correlated' stations explains 58%, 66% and 52% of the variance with all India seasonal mean rainfall. 

During winter (November-February), the equatorward-flowing East India Coastal Current in the western Bay of Bengal and westward-flowing North Equatorial Current in the southern Bay bring low-saline waters into the south-eastern Arabian Sea, causing a haline stratification within the near-surface isothermal layer. During December-April, the positive surface-wind-stress curl and the associated Ekman divergence shoals the pycnocline. A south-westward propagating mode-2 Rossby wave from off south-west India seen in satellite-derived mean sea level and model solutions also modulates the underlying pycnocline. During the pre-summer monsoon season, under clear skies and light wind conditions, the radiative heat input overwhelms turbulent heat losses at the air-sea interface, and the net surplus heat energy is absorbed in a shallow haline stratified near-surface layer, resulting in the formation of the observed mini-warm pool.

An examination of historical data on the genesis of monsoon onset vortices reveals that on most occasions the genesis has occurred over this mini-warm pool region. Evidence for the geographic coincidence in the occurrence of the genesis of onset vortex and the sea surface temperature maxima during individual years of a three decade period (1961-90) is presented. 

The change in the Bay of Bengal convection (the ISM) has planetary-scale implications, whereas the change in Philippine convection has primarily a regional impact including a linkage with the east Asia subtropical monsoon. The equatorial western Pacific winds exhibit a considerably higher correlation with the ISM convection than with the Philippine convection. During the summers when a major Pacific warm episode occurs (e.g., 1982-83, 1986-87, 1991-92, and 1997), the convection and circulation indices describing the ISM often diverge considerably, causing inconsistency among various normally coherent monsoon indices. This poses a primary difficulty for using a single monsoon index to characterize the interannual variability of a regional monsoon. The cause of the breakdown of the coherence between various convection and circulation indices during ENSO warm phase needs to be understood. 

It is an important fact that the seasonal variation of the moisture flux crossing the boundary of a large domain, which is bounded by 10 degrees S, 40 degrees N, 35 degrees E and 140 degrees E, and the variation of cross equatorial moisture transport from the southern hemisphere, are significantly smaller than the variations of the moisture transport within the domain. The moisture balance calculation also indicates that the essential feature of the summer monsoon is the formation of a pair of moisture source region, and the adjacent moisture sink region, within the large domain. The moisture source regions form under the subsidence and convective stable stratification. The moisture sink regions appear under the ascent motion and convective unstable stratification. Several parameters such as "moisture influx ratio" and "rainfall production ratio" are defined to discern the characteristics of the moisture balance. The temporal and spatial variations of these parameters depict well the features of moisture balance within the Asia summer monsoon region. 

It is argued that the interannual variability of the monsoon circulation is primarily driven by gradients of diabatic heating associated with variations of the EIMR, and that the regional monsoon Hadley circulation is a manifestation of this heating. An index of the monsoon Hadley (MH) circulation is defined as the meridional wind-shear anomaly (between 850 hPa and 200 hPa) averaged over the same domain as the EIMR. Using circulation data from two independent reanalysis products, namely the National Centers for Environmental Prediction/National Center for Atmospheric Research reanalysis and the European Centre for Medium-Range Weather Forecasts reanalysis, it is shown that the MH index is significantly correlated with the EIMR. Also it is shown that both the EIMR and MH indices have a dominant quasi-biennial variability, consistent with previous studies of IMR. Teleconnections of IMR, EIMR and MH indices with summer sea surface temperature (SST) have also been investigated. There are indications that the south equatorial Indian Ocean SST has a strong positive correlation with the EIMR. Also it is noted that the correlation of the monsoon indices with the eastern Pacific SST was weak during the period under consideration primarily due to almost a reverse relationship between monsoon and El Nino and Southern Oscillation during the latest eight years. 

Results show the close relationship between the fluctuations of atmospheric circulation, heating, and surface condition. It is suggested that the atmospheric circulation abruptly changes during the transition owing to the interaction between convection, large-scale circulation, and lower-boundary forcing that includes topographically lifting ocean and land surface heating. 

The first mode is characterized by an increase in large-scale cloud over India and the subtropical western Pacific until mid-August. The second mode depicts large-scale cloud variations associated with the East Asian rainband referred to as Mei Yu and Baiu. This mode is associated with the development of summer monsoon circulation: a low pressure system over the Asian continent and a subtropical high over the Pacific. The third eigenmode is characterized by zonal cloud hands from northern India crossing the Korean peninsula to Japan, aad dryness over the oceans in the south of cloud bands. This mode is related to the mature phase of Changma rainy season in Korea associated with the northward movement of cloud bands and circulation systems from the subtropical western Pacific. This mode appears as a first principal mode of climatological intraseasonal oscillation (CISO) over the entire Asian monsoon region. The CISO mode has a timescale of about 2 months.

The northward moving CISO also appears in the 850- and 200-mb geopotential height fields as a first mode of each. dataset. Based on the height variations of the CISO mode, it is suggested that the extratropical CISO during summer is related to a regional index cycle associated with the variation of north-south temperature gradient in East Asia. 

Normalized daily precipitation time series are analyzed between 1951 and 1990 for 85 observation stations to develop criteria that describe general rainband characteristics throughout China. Rainfall is defined to be "heavy" if the daily value at a given location is greater than 1.5% of the annual mean total. Heavy precipitation is then shown to be "persistent" and is thus identified with the rainband when the 1.5% threshold is exceeded at least 6 times in a 25-day period. Finally, rainband initial (final) dates are defined to immediately follow (precede) a minimum period of 5 consecutive days with no measurable precipitation. A semiobjective analysis based on the above definitions and rainband climatology is then applied to the time series to determine annual initial and final dates.

Analysis application produces results that closely correspond to the systematic pattern observed across China, where the rainband arrives in the south during May, advances to the Yangtze River valley in June, and then to the north in July. Rainband duration (i.e., final - initial + 1)is approximately 30-40 days while total rainfall decreases from south to north. A significant positive correlation is found between total rainfall and duration interannual variability, where increased rainband precipitation corresponds to initial (final) dates that are anomalously early (late). No clear trends are identified except over north China, where both duration and total rainfall decrease substantially after 1967.

The Eurasian sea level pressure and 500-hPa height fields are then correlated with total rainfall over south China, the Yangtze River valley, and north China to identify statistically significant relationships. Results indicate that precipitation amount is influenced by the interaction of several circulation features. Total rainfall increases over south China when the surface Siberian high ridges to the south and is overrun by warm moist air aloft. Yangtze River valley precipitation intensifies when westward expansion of the subtropical high along with strengthening of the Siberian high and monsoon low cause moisture advection, upward motion, and the thermal gradient along the Mei-Yu front to increase. North China total rainfall increases in response to intense heating over the landmass, westward ridging of the subtropical high, and greater moisture transport over the region. 

The most remarkable change in the onset period of the Asian summer rainy season over the Indian Ocean and the Southeastern Asia is the rapid displacement of the major precipitation areas from the equatorial zone to the Indochina Peninsula and Indian subcontinent. The notable change in the onset phase of the rainy season over East Asia is the formation of a precipitation belt, which corresponds to the Meiyu and Baiu frontal zone.

In order to realize the relation between the variations of the precipitation with the those of the low-level circulation, several major circulation systems (CSs) in the lower troposphere are defined. Among them, the variations of the clockwise circulation centered over the Indian Ocean (CS-3) and a cyclonic circulation over the northern part of the Indian subcontinent (CS-6) are closely related with variations of the Indian monsoon rainfalls, while the cross-equatorial flow from the Australia anticyclone (CS-4) and the circulation around the North Pacific subtropical anticyclone (CS-5) are the major systems related with the East Asia rainfalls.

The variation of the mixing ratio of water vapor, equivalent potential temperature, vertical stability and the vertical motion are also described in relation with the variations of the precipitation and CSs. During the Asian summer monsoon season, the gradual decrease of the mixing ratio and the increase of the vertical stability, in association with the subsidence, are significant features over the western tropical Indian Ocean, while the increase of the specific humidity and the vertical instability are the notable features over the Indian sub continent, Indochina Peninsula and the southern part of China. 

Due to attenuated Walker circulation in response to a warm episode, convection is suppressed over the northern tropical Indian Ocean and the maritime continent from the preceding winter to spring. The suppressed tropical convection in the preceding spring generates anomalous cyclonic circulation to the west of the Tibetan Plateau as a result of the Rossby-type response to convective heating off the equator. The convection-induced anomalous cyclonic circulation accompanied by large-scale ascending atmospheric motion contributes substantially to increased rainfall and greater soil moisture, thus resulting in decreased land-surface temperature over central Asia to the northwest of the Indian subcontinent. On the other hand, warm SST anomalies are initially introduced over the tropical Indian Ocean in late spring prior to the onset of the monsoon due to the changes in the surface heat flux and/or dynamic response of the ocean to wind forcing, in intimately association with pronounced in situ low-level northeasterly wind anomalies and less cloud cover. Both these different physical processes in the land and ocean areas are crucially responsible for reduced land-ocean thermal contrast (or reduced meridional tropospheric temperature gradient), eventually bringing about the weakening of the Asian summer monsoon. The reverse situation is quite true for strong monsoon years. Once the summer monsoon becomes weak (strong) at its early stage due to these processes, the initially induced warm (cool) SST anomalies over the tropical Indian Ocean are further intensified.

The mechanism proposed here is valid during the period from the late 1970s to the early 1990s when weak and strong monsoon years are categorized. During that period, the unusual Nino-3 SST anomalies tend to persist from the preceding winter until summer, hence serving as a bridge between the ENSO prevailing in the preceding winter and anomalous summer monsoon. However, regardless of when the monsoon-ENSO coupling is prominent, both the springtime outgoing longwave radiation and low-level wind anomalies dominating over the tropical Indian Ocean, associated with anomalous Walker circulation, are still crucial factors in terms of the potential predictability of the Asian summer monsoon. 

(1) During the Asian summer monsoon season, the prescribed deep heat sources in the southern part of Asia form the Tibetan High, the monsoon trough, the low-level circulation over South Asia, and furthermore, the downward motion in the western part of the Eurasian Continent. The heat sources near the surface over central Asia also induce downward motions aloft.

(2) In early summer (June), the deep heat sources in the southern part of Asia tend to form southwesterly low-level flows and upward motion southeast of Japan. Those are considered to be the background for the Baiu formation in East Asia as well as heat lows produced in the southern part of Asia. The mid-latitude heat sources associated with the Baiu precipitation produce a low-level jet south of that.

(3) Climatological seasonal change from early summer (June) to mid-summer (July) is characterized by an air temperature increase in the whole Northern Hemisphere and a northward shift of a weakened westerly jet. When in the model a zonal mean field in June is replaced by that in July, the major characteristics of the seasonal change are obtained qualitatively; low-level jets and upward motion areas in South Asia and East Asia shift from the ocean side of the coasts toward the land side. This change of vertical motion is consistent with the seasonal change of deep heat sources from June to July.

(4) The climatological seasonal change from mid-summer (July) to late summer (August) is characterized by enhanced convective activity in the extended area of the subtropical western Pacific. When deep heat sources in July are replaced by those in August over the western Pacific only, the major characteristics of the seasonal change over the Pacific and the Indian Ocean are obtained. The expansion of the Tibetan High at the upper-level and the Pacific High at low-level over Japan is also simulated by the seasonal change of the western Pacific heat sources only.

(5) The model simulation with the combination of the diabatic heat source for August and the zonal mean field for June is compared with the climatological August simulation. It is indicated that the zonal mean field delayed from its seasonal migration could be related to weak monsoon circulation and the associated precipitation anomalies in the mid-latitudes and the subtropics. 

Results indicate that ocean basin-scale SST anomalies exert a stronger control on the interannual variability of the monsoon compared to GW anomalies. The impact of SST anomalies on the monsoon appears nonlinear with respect to warm and cold events. The monsoon is weakened during the warm events but changes less noticeably during the cold events. The diminution of monsoon circulation associated with the warm SST anomalies is accompanied by a broad-scale reduction in water vapor convergence and monsoon rainfall.

Results also indicate that, following wet land surface conditions (enhanced snow and soil moisture) in the Asian continent during previous cold seasons, the summer monsoon becomes moderately weaker. Antecedent land surface processes mainly influence the early part of the monsoon. Wetter and colder conditions occur in the Asian continent during warm SST events. This results in reduced land-sea thermal contrast, which reinforces the weak monsoon anomalies produced initially by warm SST forcing. These interactive forcings are also responsible for the changes in the winter-spring westerlies over subtropical Asia, which are key precursory signals for the subsequent summer monsoon.

It should be pointed out that this study is conducted for the climate decade of 1979-88 only. The general robustness of the results needs to be explored by further investigations. In addition, chaotic features may have affected the results because of sampling errors. 

Variations in lapse rate or vertical stratification through the depth of the troposphere are found to occur mainly between active and break periods, rather than on a day-to-day basis. This is interpreted as being due to mid-tropospheric temperature being adjusted by dynamical processes over large scales rather than in situ response to localized convection. Between active and break periods large changes occurred in midtropospheric moisture. Variations in convective activity are well related to variations in lower and middle tropospheric moisture content. The break coincided with a drying due to large-scale horizontal advection.

Convective activity is weakly but inversely related to convective available potential energy variations. Day-to-day variations in CAPE are dominated by variations in equivalent potential temperature of the source level (boundary layer) air. The physical effect is one of changing the moist adiabat along which the air parcel rises. In temperature-log pressure space, moist adiabats diverge in the upper half of the troposphere. Since CAPE variations are dominated by changes in the moist adiabat of the rising parcel, the day-to-day CAPE changes occur almost totally in positive area variations above the 600-hPa level.

The above results are discussed in the context of other studies in the literature. It is proposed that stabilization of the atmosphere in response to deep convection occurs almost entirely through the modification of CAPE through decreasing theta(e) of the source air in the boundary layer. This occurs over relatively small spatial scales, whereas variations in lapse rate through the deep troposphere are hypothesized to occur over the relatively large scales associated with monsoon active and break events. 

Results suggest the necessity of including a detailed land surface parameterization in the realistic short-range weather numerical predictions. An enhanced short-range prediction of hydrological cycle including precipitation was produced by the model, with land surface processes parameterized. This parameterization appears to simulate all the main circulation features associated with the summer monsoon in a realistic manner. 

We have compared our results of monsoon studies at a resolution T255 with those at resolution T62. The transform grid separation at the resolution T255 is approximately 50 km and at the resolution T62, it is approximately 200 km. We find that the model at the higher resolution (T255) performs better and has more realistic energy conversions for the convectively driven synoptic scale monsoon.

An organization of convection, at the synoptic scales, is not seen in the forecasts at lower resolutions, T62, where the rainfall patterns are generally much broader and tend to be more zonal. Such organization appears more realistic at the resolution T255. Variances of the energy conversion, calculated in the two-dimensional spectral space, from physically initialized short range forecasts at the higher resolution are seen to be lamest on the scales of the monsoon. Similar calculations for the reanalyzed fields at lower resolutions show the spectral distribution of variances to be biased towards local Hadley scale overturnings. 

There were statistically significant differences in the composite of temperature anomaly patterns between excess and deficient monsoon years over north Europe, central Asia and north America during January and May, over NW India during May, over central parts of Africa during May and July and over Indian sub-continent and eastern parts of Asia during July. It has been also found that temperature anomalies over NW Europe, central parts of Africa and NW India during January and May were positively correlated with Indian summer monsoon rainfall. Similarly temperature anomalies over central Asia during January and temperature anomalies over central Africa and Indian region during July were negatively correlated. There were secular variations in the strength of relationships between temperature anomalies and Indian summer monsoon rainfall. In general, temperature anomalies over NW Europe and NW India showed stronger correlations during the recent years. It has been also found that during excess (deficient) monsoon years temperature gradient over Eurasian land mass from sub-tropics to higher latitudes was directed equatowards (polewards) indicating strong (weak) zonal flow. This temperature anomaly gradient index was found to be a useful predictor for long range forecasting of Indian summer monsoon rainfall. 

The Florida State University Global Spectral Model at the resolution T170 (170 waves triangular truncation) was run to carry out several experiments to investigate the issue of the maintenance of the Asian monsoon. In this context the authors examined the issue of the maintenance of the monsoon over a south Asian domain. The computations show that differential heating (i.e., the covariance of heating and temperature) lends to the growth of available potential energy, which is next passed on to the divergent motions via the covariance of vertical velocity and temperature. The final link in this scenario is the transfer of energy from the divergent to the rotational part of the motion field that describes the monsoon. These are largely described by psi-chi interaction via the covariance of del psi and del chi (where psi is a streamfunction and chi is a velocity potential). It is noted that lateral boundary fluxes are also important for the maintenance of the monsoon. Lateral coupling with the monsoon of southeast Asia and with the Southern Hemispheric circulation are also some of the crucial elements of the overall energetics. 

Monsoon transitions in the model, as depicted by the abrupt meridional movement of the axis of maximum vertical motion from equator to northern latitudes, occur during 16-20 May for the East Asian Monsoon (EAM) and 1-5 June for the South Asian Monsoon (SAM) regions. The necessary stability criterion for dry (moist) SI over the EAM and SAM regions reveals a sudden cross equatorial advection of negative dry potential vorticity (DPV) and moist potential vorticity (MPV) into the summer hemisphere five to ten days preceding the model monsoon transition. This causes dry and moist SI. Maximum shift of the zero line of DPV and MPV (dry and moist symmetrically unstable regions) happens subsequent to monsoon transition. Simplified analysis of the potential vorticity (PV) budget equation reveals that the lower tropospheric negative PV advection into the summer hemisphere is largely governed by the dominance of vertical differential diabatic heating over horizontal differential diabatic heating.

The diabatic heating also shows an abrupt increase from 2-3 K day(-1) before the transition, to 12-14 K day(-1) at the time of monsoon transition. The genesis of pre-monsoon weak heat source arises primarily due to unstable SI and CSI of the pre-monsoon basic states, which consequently produce moderately large scale lower (upper) tropospheric convergence (divergence) patterns slightly poleward of the zero line of DPV and MPV. Lower tropospheric conditionally unstable tropical atmosphere, in the presence of off equatorial large scale lower (upper) tropospheric convergence (divergence), is conducive to exciting CISK-like processes, which may eventually release large amounts of latent heat and develop a strong heat source at the time of monsoon transition. We have noted that a fully established model meridional circulation originates only when the diabatic forcing magnitude exceeds some threshold value of around 5 K day(-1) at the time of monsoon transition. The model transition is more pronounced over the EAM region than over the SAM region. The linear steady-state dynamical response of a zonally symmetric atmosphere as a consequence of varying the location and magnitude of an idealized asymmetric thermal forcing reveals that the most intense meridional circulation (maximum efficiency of vertical motion) is accomplished when the thermal forcing is located around 10 degrees N. The interrelationship between the location of zero DPV/MPV contour, lower tropospheric maximum convergence versus maximum vertical velocity of the monsoonal circulation in the summer hemisphere, clearly suggests that SI (CSI) of zonal monsoon flows is a causal mechanism for the onset of monsoon transition. 

Thus, the models fall into two distinct classes on the basis of the seasonal variation of the major rainbelt over the Asia West Pacific sector, the first (class I) are models with a realistic simulation of the seasonal migration and the major rainbelt over the continent in the boreal summer; and the second (class II) are models with a smaller amplitude of seasonal migration than observed. The mean rainfall pattern over the Indian region for July-August (the peak monsoon months) is even more complex because, in addition to the primary rainbelt over the Indian monsoon zone (the monsoon rainbelt) and the secondary one over the equatorial Indian ocean, another zone with significant rainfall occurs over the foothills of Himalayas just north of the monsoon zone. Eleven models simulate the monsoon rainbelt reasonably realistically. Of these, in the simulations of five belonging to class I, the monsoon rainbelt over India in the summer is a manifestation of the seasonal migration of the planetary scale system. However in those belonging to class II it is associated with a more localised system. In several models, the oceanic rainbelt dominates the continental one. On the whole, the skill in simulation of excess/deficit summer monsoon rainfall over the Indian region is found to be much larger for models of class I than II, particularly for the ENSO associated seasons. Thus, the classification based on seasonal mean patterns is found to be useful for interpreting the simulation of interannual variation. The mean rainfall pattern of models of class I is closer to the observed and has a higher pattern correlation coefficient than that of class II. This supports Sperber and Palmer's (1996) result of the association of better simulation of interannual variability with better simulation of the mean rainfall pattern. The hypothesis, that the skill of simulation of the interannual variation of the all-India monsoon rainfall in association with ENSO depends upon the skill of simulation of the seasonal variation over the Asia West Pacific sector, is supported by a case in which we have two versions of the model where NCEP1 is in class II and NCEP2 is in class I. The simulation of the interannual variation of the local response over the central Pacific as well as the all-India monsoon rainfall are good for NCEP2 and poor for NCEP1. Our results suggest that when the model climatology is reasonably close to observations, to achieve a realistic simulation of the interannual variation of all-India monsoon rainfall associated with ENSO, the focus should be on improvement of the simulation of the seasonal variation over the Asia West Pacific sector rather than further improvement of the simulation of the mean rainfall pattern over the Indian region. 

In excess monsoon years, the average percentage contribution of each monsoon month to the long term mean (1871-1994) seasonal rainfall (June - September) is more than that of the normal while in the deficient years it is less than normal. This is noticed in all 29 meteorological sub-divisions. From the probability analysis, it is seen that there, is a rare possibility of occurrence of seasonal rainfall to be excess/deficient when the monthly rainfall of any month is deficient/excess. 

Temporal variations of kinetic energy of wavenumber 2 over Region 1 and Region 2 are almost identical. The correlation matrix of different time series shows that (i) wavenumber 2 over Regions 1 and 2 might have the same energy source and (ii) there is a possibility of an exchange of kinetic energy between wavenumber 1 over Region 1 and short waves over Region 2. Wave to wave interactions indicate that short waves over Region 2 are the common source of kinetic energy to wavenumber 2 over Regions 1 and 2 and wavenumber 1 over Region i. Time spectral analysis of kinetic energy of zonal waves indicates that wavenumber 1 is dominated by 30-45 day and bi-weekly oscillations while short waves are dominated by weekly and bi-weekly oscillations.

The correlation matrix, wave to wave interaction and time spectral analysis together suggest that short period oscillations of kinetic energy of wavenumber I might be one of the factors causing dominant weekly (5-9 day) and bi-weekly (10-18 day) oscillations in the kinetic energy of short waves. 

The results indicate that the previous May's tropospheric temperature anomaly has a strong and direct relationship with southwest monsoon rainfall, suggesting that warmer/colder tropospheric temperature in May leads to good/bad southwest monsoon rainfall over India. The correlations are stronger for AIR and NWR followed by PIR. Its stability analysis displays best correlation during 1967-1981 and 1968-1982 and again during the recent years 1975-1989 and 1976-1990 for all the regions. Significant direct correlations are also obtained with the tropospheric temperature anomaly of the previous April and spring season. The result also suggests that antecedent May tropospheric temperature anomaly may be useful in the long range prediction of the following southwest monsoon rainfall over India. (C) 1998 Royal Meteorological Society. 

Time series of area-averaged SSTs for the NINO3 region in the eastern equatorial Pacific show that the CSM is producing about 60% of the amplitude of the observed variability in that region, consistent with most other global coupled models. Global correlations between NINO3 rim; series, global surface temperatures, and sea level pressure (SLP) show that the CSM qualitatively reproduces the major spatial patterns associated with the Southern Oscillation (lower SLP in the central and eastern tropical Pacific when NINO3 SSTs are relatively warmer and higher SLP over the far western Pacific and Indian Oceans, with colder water in the northwest and southwest Pacific). Indices of Asian-Australian monsoon strength are negatively correlated with NINO3 SSTs as in the observations. Spectra of time series of Indian monsoon, Australian monsoon, and NINO3 SST indices from the CSM show amplitude peaks in the Southern Oscillation and tropospheric biennial oscillation frequencies (3-6 yr and about 2.3 yr, respectively) as observed. Lag correlations between the NINO3 SST index and upper-ocean heat content along the equator show eastward propagation of heat content anomalies with a phase speed of about 0.3 m s(-1), compared to observed values of roughly 0.7 m s(-1). Composites of El Nino (La Nina) events in the CSM show similar seasonal evolution to composites of observed events with warming (cooling) of greater than several tenths of a degree beginning early in northern spring of year 0 and diminishing around northern spring of year +1, but with a secondary resurgence in the CSM events later in northern spring of year +1. The CSM also shows the largest amplitude ENSO SST and low-level wind anomalies in the western tropical Pacific, with enhanced interannual variability of SSTs extending northeastward and southeastward toward the subtropics, compared to largest interannual SST variability in the central and eastern tropical Pacific in the observations. 

A case study of 1989-90 South American summer monsoon (SASM) reveals the following characteristics.

1) In late spring, the onset of SASM is signaled by an abrupt merging of the upper-tropospheric double westerly jets, one in the subtropics and the other in the subpolar region, into a single jet in the midlatitudes. This is followed by the establishment of a vortex to the southeast of Altiplano and occurrence of heavy precipitation over subtropical eastern Brazil.

2) During the mature phase of SASM, the heavy rainfall zone moves over the Altiplano Plateau and the southernmost Brazilian highland. The fully established SASM features are the following: (a) an enhancement of equatorial North Atlantic trade wind, which emanates from the Sahara high and crosses the equator over the South American continent; (b) a buildup of strong northwesterlies along the eastern side of the tropical Andes; and (c) development of the South Atlantic convergence zone in the southernmost position with strong convective activity. Meanwhile, the upper-tropospheric return flow emerges from an anticyclone formed over the Altiplano Plateau, crosses the equator, and sinks over northwestern Africa.

3) The withdrawal of SASM in late summer is signaled by the resplitting of the midlatitude westerly jet. At the same time, the low-level northwest monsoon how diminishes, reducing the moisture supply and leading to the termination of heavy precipitation over the subtropical highland.

Results also show that the above-mentioned characteristics of SASM are clearly linked to the tropospheric temperature changes over the central South American highland. Sensible versus latent heating over the highland are bound to play an important role in the evolution of SASM.

To provide further support of presence of a monsoon climate over South America, SASM is compared and contrasted to the "classic" east Asian summer monsoon (EASM). Many similar features, including evolution characteristics between the two systems, have been identified. Contrasting aspects of the SASM from the EASM are also discussed. It is pointed out that a number of monsoonal characteristics of the climate of South America, such as the seasonal reversal of the low-level wind, become apparent only when the strong annual mean wind is removed. Based on the characteristic features and their evolution, the authors conclude that a monsoon climate does exist over South America. 

Furthermore, distinct precursory signals, including the Eurasian snow in winter and soil moisture anomalies in spring, have been found in the pre-monsoon seasons of weak and strong monsoon years. There is a sharp contrast between weak and strong monsoon years; excessive snow over Eurasia south of 50 degrees N in winter and the increased soil moisture in spring are found prior to weak summer monsoon. These results are consistent with evidence found in observational data analyses and some model experiments. A detailed analysis of surface heat budget shows that snow-albedo feedback dominates over the Tibetan Plateau. On the other hand, to its west in the central Asia, the relatively lower land, the effective cloud albedo anomalies due to excessive rainfall and surface evaporation influence the surface conditions.

A numerical experiment with the Eurasian land surface initial conditions in spring, interchanged between weak and strong monsoon years, indicates positive roles played by the land surface processes in influencing the subsequent summer monsoon circulations during the 10-year period. However, such land surface feedbacks are not strong enough to change the sign of the monsoon circulation anomalies. The direct influence of the El Nino/Southern Oscillation through the changes in Walker circulation appears to predominate. 

The results of our analysis clearly shaw a consistent dramatic reversal in wind direction over the western Arabian Sea three weeks in advance of the onset of monsoon. The wind speed shows a large increase coinciding with the onset of monsoon. These findings together show the dominant role of sea surface winds in establishing the monsoon circulation. The study confirms that the cross equatorial current phenomenon becomes more important after the onset of monsoon. 

The synoptic structure of the 30-60 day mode is similar in all years and is shown to be intimately related to the strong ('active') or weak ('break') phases of the Indian summer monsoon circulation. The peak (trough) phase of the mode in the north Bay of Bengal corresponds to the 'active' ('break') phase of monsoon strengthening (weakening) the entire large scale monsoon circulation. The ISOs modulate synoptic activity through the intensification or weakening of the large scale monsoon flow (monsoon trough). The peak wind anomalies associated with these ISOs could be as large as 30% of the seasonal mean winds in many regions. The vorticity pattern associated with the 30-60 day mode has a bi-modal meridional structure similar to the one associated with the seasonal mean winds but with a smaller meridional scale. The spatial structure of the 30-60 day mode is consistent with fluctuations of the tropical convergence zone (TCZ) between one continental and an equatorial Indian Ocean position. The 10-20 day mode has maximum amplitude in the north Bay of Bengal, where it is comparable to that of the 30-60 day mode. Elsewhere in the Indian Ocean, this mode is almost always weaker than the 30-60 day mode. In the Bay of Bengal region, the wind curl anomalies associated with the peak phases of the ISOs could be as large as 50% of the seasonal mean wind curl. Hence, ISOs in this region could drive significant ISOs in the ocean and might influence the seasonal mean currents in the Bay.

On the interannual time scale, the NCEP/NCAR reanalysed wind stress is compared with the Florida State University monthly mean stress. The seasonal mean stress as well a interannual standard deviation of monthly stress from the two analyses agree well, indicating absence of any serious systematic bias in the NCEP/NCAR reanalysed winds. It is also found that the composite structure of the 30-60 day mode is strikingly similar to the dominant mode of interannual variability of the seasonal mean winds indicating a strong link between the ISOs and the seasonal mean. The ISO influences the seasonal mean and its interannual variability either through increased/decreased residence time of the TCZ in the continental position or through occurrence of stronger/weaker active/break spells. Thus, the ISOs seem to modulate all variability in this region from synoptic to interannual scales. 

Based on the criteria defined in this paper for the SCS summer monsoon onset, the average onset date over the SCS from 1979 to 1995 is around the fourth pentad of May. The airflow and general circulation over the SCS changes dramatically after the onset. The ridge of the subtropical high in the western Pacific in the lower troposphere weakens and retreats eastward from the SCS region with an establishment of westerly winds over the whole region. During the SCS monsoon onset, the most direct impact in the vicinity of the SCS are the equatorial westerlies in the Bay of Bengal through their eastward extension and northward movement. An indirect influence on the SCS onset is also caused by the enhancement of the Somali cross-equatorial flow and the vanishing Arabian High over the sea; the latter may be a signal for the SCS onset.

There are quite significant interannual variations in the SCS onset. In the years of a delayed onset, the most profound feature is that the easterly winds stay longer in the SCS than on average. Deep convection activities are suppressed. The direct cause is the abnormal existence of the western Pacific subtropical high over the SCS region. Moreover, compared to the average, the equatorial westerlies in the Bay of Bengal are also weaker in the years of a delayed onset. No significant changes for the cross-equatorial flow at 105 degrees E are observed for these years. It has also been found that the interannual variations of the SCS onset are closely related with the ENSO events. In the years of a delay, the Walker circulation is weaker, and the sea surface temperature (SST) anomalies in the western Pacific are negative. 

A measure of internal variability of the model summer monsoon is obtained from a 20-yr integration of the same model with fixed annual cycle SST as boundary conditions but with predicted soil moisture and snow cover. A comparison of summer monsoon indexes between this run and the observed SST run shows that the internal oscillations can account for a large fraction of the simulated monsoon variability. The regional-scale oscillations in the observed SST run seems to arise from these internal oscillations. It is discovered that most of the interannual internal variability is due to an internal quasi-biennial oscillation (QBO) of the model atmosphere. Such a QBO is also found in the author's third 18-yr simulation in which fixed annual cycle of SST as well as soil moisture and snow cover are prescribed. This shows that the model QBO is not due to land-surface-atmosphere interaction. It is proposed that the model QBO arises due to an interaction between nonlinear intraseasonal oscillations and the annual cycle. Spatial structure of the QBO and its role in limiting the predictability of the Indian summer monsoon is discussed. 

The warming over the Tibetan Plateau from spring to summer is found in the 200-500 hPa thickness data on about 15-day intervals. Of importance is the observational evidence that the warming phase over the Tibetan Plateau around mid-May is concurrent with the early onset of the SEAM. Thus, the thermal contrast between the Tibetan Plateau and the adjacent ocean is likely to induce the acceleration and eastward extension of the low-level monsoon flow, causing the abrupt commencement of the SEAM including onset of the South China Sea monsoon (SCSM). This relationship between low-level wind over the key region (10 degrees-20 degrees N, 80 degrees-120 degrees E) and 200-500 hPa thickness over the Tibetan Plateau is also confirmed based on the correlation analysis in the interannual variabilities.

An influence for the mid-latitude atmosphere, stationary Rossby waves are generated over the South China Sea and propagate in a northeastward direction toward Japan because of the cyclonic vorticity and the tropical heat source associated with the onset of the SCSM. As a result of this wave propagation, a high pressure anomaly appears over Japan, which is consistent with a singularity of clear skies around Japan in mid-May (Kawamura and Tian, 1992). 

Finally, the timing of the onset of the Asian monsoon in 1989 was explored. It was shown that the onset of the Asian monsoon occurs when the warm or rising phase of different low-frequency oscillations reach the "East Asian monsoon area" (EAMA) concurrently. These include the warm phase of the eastward propagating two- to three-week oscillation (TTO) of the upper-layer temperature in middle latitudes, the rising phase of the northward propagating Madden-Julian oscillation of the southern tropical divergence, and the rising phase of the westward propagating TTO of the western Pacific divergence. It was concluded that the timing of the Asian monsoon onset is determined when the favorable phases of different low-frequency oscillations are locked over the EAMA. 

(1) In the northern winter, the SCS SSTA are quite sensitive to the longitudinal shift of global wind anomalies associated with the equatorial Pacific SSTA. This fact is related to that the SCS SSTA and the neighbor SSTA have strong biennial oscillation.

(2) When the global wind anomalies are shifted eastward in the winter (BO-type years), the tropical eastern Pacific SSTA tend to change in the following spring. On the other hand, when those wind anomalies are shifted westward (LF-type years), the eastern Pacific SSTA tend to be maintained through the year. The associated differences between the BO and LF-type years are found in the seasonal change of the low-level tropical wind anomalies from the preceding summer through winter.

(3) The northern summer SCS SSTA seem to be controlled by in-situ low-level wind anomalies. Further more, easterly anomalies over South Asia and the tropical western Pacific and westerly anomalies over East Asia are found in the lower atmosphere for the positive SCS SSTA. It is also shown that the summer SCS SSTA have a statistical relationship with the equatorial central Pacific SSTA in the preceding winter. This fact suggests a relationship between the summer Asian monsoon and the winter phase of ENSO. 

Based on the instrumental data available and on an objective criteria, regional rainfall anomaly time series for contiguous regions over Thailand, Malaysia, Singapore, Brunei, Indonesia and Philippines are prepared. Results reveal that although there are year-to-year random fluctuations, there are certain epochs of the above-and below-normal rainfall over each region. These epochs are not forced by the El Nino/La Nina frequencies. Near the equatorial regions the epochs tend to last for about a decade, whereas over the tropical regions, away from the Equator, epochs last for about three decades. There is no systematic climate change or trend in any of the series. Further, the impact of El Nino (La Nina) on the rainfall regimes is more severe during the below (above) normal epochs than during the above (below) normal epochs. Extreme drought/flood situations tend to occur when the epochal behaviour and the El Nino/La Nina events are phase-locked. (C) 1997 by the Royal Meteorological Society. 

A global atmospheric GCM in perpetual January mode is run with observed SSTs with specified convective heating anomalies to demonstrate that convective heating anomalies elsewhere in the Tropics associated with the coupled ocean-atmosphere biennial mechanism can contribute to altering seasonal midlatitude circulation. These changes in the midlatitude longwave pattern, forced by a combination of tropical convective heating anomalies over East Africa, Southeast Asia, and the western Pacific (in association with SST anomalies), are then able to maintain temperature anomalies over south Asia via advection through winter and spring to set up the land-sea meridional tropospheric temperature contrast for the subsequent monsoon. The role of the Indian Ocean, then, is to provide a moisture source and a low-amplitude coupled response component for meridional temperature contrast to help drive the south Asian monsoon. The role of the Pacific is to produce shifts in regionally coupled convection-SST anomalies. These regions are tied together and mutually interact via the large-scale east-west circulation in the atmosphere and contribute to altering midlatitude circulations as well. The coupled model results, and experiments with an atmospheric GCM that includes specified convective heating anomalies, suggest that the influence of south Asian snow cover in the monsoon is not a driving force by itself, but is symptomatic of the larger-scale shift in the midlatitude longwave pattern associated with tropical SST and convective heating anomalies. 

The CISO results from a phase-locking of transient intraseasonal oscillation to annual cycle. It exhibits a dynamically coherent structure between enhanced convection and low-level convergent (upper-level divergent) cyclonic (anticyclonic) circulation. Its phase propagates primarily northward from the equator to the northern Philippines during early summer (May-July), and westward along 15 degrees N from 170 degrees E to the Bay of Bengal during August and September.

The propagation of CISO links monsoon singularities occurring in different regions. Four CISO cycles are identified from May to October. The first cycle has a peak wet phase in mid-May that starts the monsoon over the South China Sea and Philippines. Its dry phase in late May and early June brings the premonsoon dry weather over the regions of western North Pacific summer monsoon (WNPSM), Meiyu/Baiu, and Indian summer monsoon (ISM). The wet phase of Cycle II peaking in mid-June marks the onsets of WNPSM, continental ISM, and Meiyu, whereas the dry phase in early to mid-July corresponds to the first major breaks in WNPSM and ISM, and the end of Meiyu. The wet phase of Cycle III peaking in mid-August benchmarks the height of WNPSM, which was followed by a conspicuous dry phase propagating westward and causing the second breaks of WNPSM (in early September) and ISM (in mid-September). The wet phase of Cycle IV represents the last active WNPSM and withdrawal of ISM in mid-October.

The relationships among ISM, WNPSM, and East Asian Subtropical Monsoon (EASM) are season dependent. During Cycle II, convective activities in the three monsoon regions are nearly in phase. During Cycle III, however, the convective activities are out of phase between ISM and WNPSM; meanwhile, little linkage exists between WNPSM and EASM. The causes of unstable relationships and the phase propagation of CISO are discussed. 

The analysis identifies centers over the central and eastern Indian Ocean, and western Pacific Ocean, along the equator, and along 15 degrees N, where seasonal mean cloud amounts range from 40% to 80% with cloud tops mostly in the middle and upper troposphere. Intraseasonal variability of clouds is also large over these centers (% variances >25%). Consistently, seasonal mean cooling rates are at a maximum (3 degrees-5 degrees C day(-1)) in the upper troposphere between 300 and 400 mb, related to cloud-top cooling. The cooling rates below 400 mb are between 1 degrees and 3 degrees C day(-1). The cooling rates exhibit intraseasonal amplitudes of 1.0 degrees-1.5 degrees C day(-1). The largest amplitudes are found between 300 and 500 mb, indicating that cooling rate variability is directly related to intraseasonal variability of convective clouds. Spatial distributions of clouds and cooling rates remain similar during the 1984 and 1987 summer seasons. However, during 1987, intraseasonal amplitudes of deep convective cloud amount and cooling rate over the Indian Ocean are 10%-15% larger than in 1984.

It is shown that intraseasonal variability of cooling rates over the Indian Ocean can perturb convective heating by 10%-30% in the upper and lower troposphere. Based on a one-dimensional radiative-convective equilibrium model, it is estimated that the radiative damping timescale over the Indian Ocean region is similar to 3 days. Based on this damping timescale and in conjunction with a model of equatorial Kelvin waves with first baroclinic mode, it is hypothesized that the variable cloud-radiative cooling rates can alter phase speeds of Kelvin waves by up to 60%. This helps explain why the frequency range of intraseasonal oscillations is so broad. 

The main features of the relationships Free from such bias are: (a) Indian monsoon rainfall and SST tendency from DJF to MAM before as well as after monsoon are significantly related except within 1904-1940 in respect of relationship with tendency before monsoon, (b) Indian monsoon rainfall and SOI tendency before and after monsoon are significantly related over some non-overlapping component periods only, (c) though the best SST-SOI tendency coupling is for DJF to MAM tendency, no coupling is observed between these tendencies within 1904-1940, (d) linkage of SST tendency from DJF to MAM with the preceding Indian monsoon rainfall appears to be stronger than that with the concurrent SOI tendency and continues even during the period of no coupling between the tendencies, thus bringing out the dominating active role played by the Indian monsoon. 

A Mobile Wave CISK (MWC) form of interaction between the large scale monsoon and the transient circulations associated with the Madden Julian Oscillation (MJO) is projected as a viable physical mechanism for the northward movement of low frequency modes. It is demonstrated that the effective low level convergence, following such an interaction, tends to shift northward relative to the site of interaction. This enables the heating perturbations to be displaced northward which in turn causes the secondary circulations and wind perturbations to follow. The essential criterion for the occurrence of a prolonged northward propagation of the low frequency modes is that the heating perturbations should phase lead the wind perturbations at all times.

An examination of the psi-chi interactions on the 30-50 day time scale reveals that the conversion from the transient divergent motions to rotational motions is quite intense (feeble) in the strong (weak) monsoon differential heating experiments. Because of the closer proximity to the monsoon heat source and also due to the latitudinal variation of earth's rotational effects, the psi-chi interactions tend to be more pronounced to the north of 15 degrees N while they are less robust in the near equatorial latitudes.

The regularity of the monsoonal modes is found to depend on the strength of the monsoon differential heating and also on the periodic behaviour of the equatorial intraseasonal oscillations. The monsoonal modes are quite steady and exhibit extreme regularity in the presence of a weak north-south differential heating provided the equatorial forcing due to the MJO varies in a periodic manner. This result supports the findings of Mehta and Krishnamurti (1988) who found greater regularity of the 30-50 day modes during bad monsoon years.

The low frequency monsoonal modes are found to be quite sensitive to the moisture availability factor (m) and the vertical profile of heating used in the MWC parameterization. A small increase in the value of (m) is found to significantly intensify the amplitude of the monsoonal oscillations while there is no considerable shift in the spectral frequency within the 30-50 day band as such. The 30-50 day motions show significant enhancement, with a relatively sharp spectral peak around 45 days, when the vertical profile of MWC heating has a maximum in the lower troposphere. However an upward displacement of the heating maximum tends to weaken the low frequency oscillations. 

The results of the experimentation suggest that, at least for the test cases of 1983 and 1984, the modulation of the Walker circulation, with implied additional subsidence over the eastern hemisphere, is the dominant mechanism whereby the Asian summer monsoon is weakened during El Nino years. However, the late onset during El Nino years may also be associated with the complementary cord SST anomalies in the west Pacific which delay the northwards transition of the TCM. During La Nina, the modulation of the Walker circulation appears not to be the controlling factor which determines the stronger monsoons. The UGCM results suggest that the complementary warm SST anomalies in the west Pacific enhance the TCM, and it is this in situ response by the TCM that leads to an early onset and stronger monsoon. The importance of warm anomalies in the west Pacific in the development of a strong monsoon has been investigated further through a case-study of the 1994 season. The year 1994 was an El Nino year in which the monsoon was unexpectedly active, but which was also marked by warmer than normal SSTs in the west Pacific.

The sensitivity experiments have also elucidated the role of El Nino in influencing the precursory signature of stronger subtropical westerlies over India and south-east Asia during the winter and spring preceding weak monsoons. The results suggested that the equatorwards shift of the subtropical jet is a remote response to the warm SST anomalies in the central and east Pacific associated with El Nino.