[ 回到主選單 ] INTERANNUAL VARIABILITY OF THE ONSET TIMING OF THE SOUTH CHINA SEA SUMMER MONSOON AND ITS POSSIBLE MECHANISMS Mong-Ming Lu*, Yun-Lan Chen and Meng-Shih Chen Central Weather Bureau, Taipei, Taiwan ROC 1. INTRODUCTION
INTERANNUAL VARIABILITY OF THE ONSET TIMING OF
THE SOUTH CHINA SEA SUMMER MONSOON AND ITS POSSIBLE MECHANISMS
Mong-Ming Lu*, Yun-Lan Chen and Meng-Shih Chen
Central Weather Bureau, Taipei, Taiwan ROC
The onset of the South China Sea (SCS) summer monsoon is characterized by a sudden shift of the ITCZ from the equatorial region to north of 10°N (Lau et al. 1999). After the onset, the low-level atmospheric circulation changes from anticyclonic to cyclonic. The onset is remarkable in the aspect that it clearly marks the commencement of the summer regime of the East Asian Monsoon (EAM) system.
The evolution of the EAM system can be represented by the north-south movement of the western Pacific subtropical high (WPSH). The persistent ridge of the WPSH is the foot of the subsidence leg of the ITCZ. Therefore, the seasonal latitudinal movement of the WPSH is primarily controlled by the annual march of the sun. However, we find that in some years the seasonal transition can substantially deviate from the normal cycles. To understand such deviation is essential for understanding and predicting the interannual variability of the EAM system.
The interannual variability of the EAM system can be affected by "fast" (less than 1 week), "intermediate" (weeks to months) and "slow" (months to years) subsystems (Lau et al. 1999). In this study we attempt to identify the possible "slow" subsystems.
2. DATA AND ONSET DEFINITION
The data used includes 42-year (1958-1999) NCEP/NCAR Reanalysis, NCEP SST and 25-year (1974-1999) OLR data sets. The summer and winter regimes of the EAM system are separated by a critical time determined by the characteristics of the WPSH. The onset time of the SCS summer monsoon, denoted as TMs, is the time when the anticyclonic circulation withdraws from the SCS.
TMs is defined by three criteria using the index VOR850, the average 850hPa vorticity over the area of 10°N-20°N, 110°E-120°E. First, the onset day must be later than April 1. Second, the onset day is the first day of the first continuous 5 days of positive VOR850. Third, within the 45-day period after the onset, the number of days with negative VOR850 is less than 30. Our results show that 78% (33/42) of the onset dates are in May and the early half of June. We assume that the interannual variation within this period of time is affected by "fast" or "intermediate" subsystems. As to the 10 years that their TMs occur either before May 5th or after June 15th, we assume that the extremely early or late onsets are resulted from the effects of "slow" (seasonal) subsystems. The five early onset years (TMs before May 5th) are 1966, 1971, 1985, 1996 and 1999. The five late onset years (TMs after June 15th) include 1959, 1968, 1977, 1992 and 1993.
3.1 Early vs. Late Onset Years
The contrasts between early and late onset years can be clearly seen in the maps of the composite low-level (850hPa) and upper-level (200hPa) winds and tropical convection. We find that in the early onset years, the SCS, Bay of Bengal, and the Philippine Sea are dominated by persistent low-level cyclonic anomalies from March to May. In the late onset years, the similar areas are dominated by anticyclonic anomalies from April to June. Accompanying the cyclonic anomalies in the early onset years, the monthly anomalies of the upper-level zonal flow are easterly from March to May, while in the late onset years, the anomalies are westerly from April to June. In the late onset years, the associate anomalous upper-level meridional flow clearly flows from the Southern to the Northern Hemisphere. The contrast between the monthly convection anomalies, particularly in April is remarkable. In the early onset years, deep convection over the SCS and Indonesia area is more active than normal. In the late onset years, the convection over the area extending from India Peninsula to western Pacific is much weaker than normal. Evolution of the dry anomalies resembles the seasonal march of tropical convection. Therefore, the late onset may reflect the weak annual cycle of tropical convection.
3.2 Possible Mechanisms
Surface forcing is a fundamental component of the "slow" subsystem. In this study, we have examined the contrast in surface forcing, circulation anomalies, interhemispheric interactions and tropical convection. The purpose is to establish a hypothesis for the cause of the extreme interannual variation of SCS summer monsoon onset.
The global air temperature at 1000hPa is first examined. From March to July, the equatorial eastern Pacific is colder than normal in the early onset years. The cold anomalies are strongest in May. Accompanying the cold anomalies, warm anomalies are observed over the western north Pacific. Over the SCS, it is warmer than normal in March and April, but becomes colder in May and June. Siberia is extremely cold in March. In the late onset years, distinct warm anomalies are observed over the eastern north Pacific. Siberia and northeast China are much warmer than normal. It should be noted that in March the anomalies have opposite signs in many places. The contrast is less obvious in other months. The onset timing may be particularly sensitive to the surface condition in March.
The monthly SST anomalies are also investigated with the usage of EOF analysis to separate the annual average anomalies and the principal components of seasonal variations. The SST in the surrounding area of the land bridge from Australia to Southeast Asia is warmer (colder) than normal in the early (late) onset years. In addition, the most distinct contrast of the SST anomalies is found in their meridional gradient in the western north Pacific depicted by the EOF modes. In early onset years, before May we find cold anomalies in the equatorial central Pacific, sandwiched by the warm anomalies over the mid-latitude Pacific in both North and South Hemispheres. The maximum gradient appears in the longitudinal belt of 150°E -150°W. In SCS we see clear zonal SST gradient with warm anomalies toward the Asia continent. The gradient becomes with warm anomalies toward the Philippine Sea after summer. The meridional gradient over the mid-latitude western Pacific is sharper in the late onset years than in the early ones. The gradient of the late onset years is particularly pronounced in the west of 150°E from Indonesia to Japan. Before May the warm anomalies are in the east of Japan (between Japan and Aleutian islands) with the maximum gradient between 30°N and 40°N. The gradient changes sign in May and become opposite in July.
The low-level convergence, upper-level divergence and convection are consistent with the surface forcing. In other words, over the area with warm surface temperature, the expected low-level convergence, upper-level divergence and more active convection are observed. In addition, two factors are found likely to be of critical importance to the extreme variations of the onset timing: the mid-latitude land sea contrast and interhemispheric interaction. In the early onset years, the land sea contrast near the area of eastern China, Korea and Japan is sharper than in the late onset years. Land is colder but ocean is warmer than normal. Probably with the help of more active interhemispheric interaction modulated by equatorial SST anomalies, the transient eddies become more active in that area. In the late onset years, the interhemispheric interaction is weak. The SCS is in covered by large-scale subsidence until late June.
4. CONCLUDING REMARKS
Our results suggest that surface forcing can be the primary cause for the extreme variations of the monsoon onset timing. The zonal gradient of the surface temperature between the western Pacific and the eastern Indian Ocean or the Asia Continent, from Australia to northeast China, and the meridional gradient between Indonesia and Japan is of particular importance. The anomaly patterns of surface temperature can modulate tropical convection, interhemispheric interaction over the Pacific and transient eddy activity in East Asia. Consequently, the seasonal transition of the EAM system can occur earlier or be delayed.
Ding, Yihui and Chongyin Li, 1999: Onset and evolution of the South China Sea Monsoon and its interaction with the ocean. China Meteorological Press. 423pp.
Lau, K.-M., K.-M. Kim and S. Yang, 1999: Dynamical and boundary forcing characteristics of regional components of the Asian summer monsoon. J. Climate (submitted).
* Corresponding author address: Mong-Ming Lu, Central Weather Bureau, Taipei, Taiwan ROC 100 e-mail: email@example.com