Baek-Jo Kim

Forecast Research Laboratory
Meteorological Research Institute (METRI)
Korea Meteorological Administration (KMA)
460-18, Shindaebang-dong, Dongjak-gu
Seoul, 156-720, Korea
Tel: +82 2 834 5920; Fax: +82 2 834 5922
E-Mail: bjkim@metri.re.kr


Education and Professional Career

Dissertation Research



ABSTRACT

Recently, East Asia frequently experienced large summer climatic anomalies, for example, the dry and hot summer of 1993 and the wet and cool summer of 1994 due to the large interannual and intraseasonal variabilities of the Asian monsoon circulation. The summer rainfall in Korea is largely affected by the Asian monsoon circulation. Thus, in order to search for the mechanism of large anomalous summer climate over East Asia associated with the interannual variation of summer rainfall in Korea, the summer rainfall variation in Korea and its association with Asian monsoon circulation are investigated by using composite analyses, singular value decomposition (SVD) analysis, empirical orthogonal function (EOF). The Asian monsoon circulations employed in present study consist of the Asian winter monsoon circulation, Indian summer monsoon circulation, and the East Asian blocking. Furthermore, the long-range predictability on the variability of summer rainfall in Korea is explored by using an advanced statistical technique called as canonical correlation analysis (CCA) for the contribution to a better long-range predictive skill of summer rainfall in Korea.

By analyzing the upper-air data at three stations (Nanning of China, Chichijima of Japan, and Pohang of Korea), the abrupt change of wind direction from westerly to easterly at the 100 hPa level is found to be a crucial process in predicting the onset of the Changma one month in advance. The abrupt change in wind direction appears earlier for dry summers than for wet summers. The withdrawal of the Changma is generally coincident with the first timing when the geopotential height of Pohang at the 500 hPa level is higher than that of both Nanning and Chichijima. This large-scale features are largely attributed by the seasonal march of the Asian summer monsoon circulation associated with the lower-level western Pacific high and upper-level Tibetan high.

The comparison of the atmospheric circulations over the northern hemisphere in preceding winter of the dry/wet summer in Korea shows that the East Asian cold surge and the extratropical upper tropospheric trough are stronger in preceding winter of dry summer than wet summer. This stronger cold surge and trough enhance convection over the South China Sea, intensifying the local Hadley circulation, which in turn interacts with the extratropics by accelerating the east Asian jet. Thus, it is indicated that two related interactions over Eastern hemisphere, namely the air-sea and tropical-extratropical interactions, are more dominant in preceding winter of dry summer than wet summer. In addition, the EU pattern is found to be appeared in composite anomaly of 500 hPa geopotential height in preceding winter of wet summer, which seems to be closely associated with the pattern for the snow cover anomaly over Eurasia. The transition of preceding atmospheric circulation for the dry/wet summer occurs in spring due to not only the anomalous transition of the westerly jet largely affected by the stronger east Asian jet in preceding winter of wet summer but also the effect of Eurasian snow coverrevealed in preceding winter of dry summer. Interestingly, it is also found that there exists the predominant oscillation with 20-25 day period in the 500 hPa height difference (the dry summer minus the wet summer) in preceding spring.

The summer rainfall in Korea is found to be negatively and significantly correlated (-0.49) with Indian summer monsoon rainfall especially over the northwestern and central parts of India. The variation of upper-level atmospheric circulation associated with the Tibetan high plays an important role in suggesting the plausible mechanism of this relationship : excessive rainfall over the northwestern and central parts of India in association with the stronger southwesterly at surface than normal may intensify the Tibetan high at 200 hPa. The intensified Tibetan high tends to shift northeastward, leading in the formation of the eastern Tibetan anticyclone over the eastern Tibetan Plateau and South China. This eastern Tibetan anticyclone may be accompanied with the abrupt northward jump of the subtropical jet at 200 hPa and the northwestward extension of the western Pacific high at 850 hPa, resulting in a dry summer over the Korean peninsula. It is also found that the variation of the eastern Tibetan anticyclone is largely related to the El Ni o events which is closely linked to the activity of the Indian sumemr monsoon.

Both the frequency and the number of persistent days of the East Asian blocking in summer largely affect the interannual variation of Mei-Yu/Baiu rainfall through the development of a baroclinic instability wave and a Rossy wave-like propagation. Due to the geographical location of Korea, these effects associated with the East Asian blocking on the interannual variation of the Changma rainfall are not clear. Thus, the East Asian blocking plays little role in influencing the interannual variation of summer rainfall in Korea. However, it is noted that the break-down of the East Asian blocking has a considerable influence on the withdrawal of the Changma. In addition, it is also found that the interannual variation of summer rainfall in Korea is indirectly affected by the SST anomalies over the tropical western Pacific through the EAP teleconnection which is similar to the composite distribution of geopotential height anomalies at 500 hPa in MBYs.

It is found that the summer rainfall in Korea has a positive and significant correlation (0.41) with JFM (January to March) precipitation, and a negative and significant correlation (-0.51) with April temperature. These parameters have potentially utilities in predicting the long-range prediction of summer rainfall in Korea. The statistical prediction of summer rainfall using canonical correlation analysis shows that high predictability (more than 0.70) of summer rainfall is appeared over the west-central part of Korea, while relatively low predictability (less than 0.50) is located over the eastern mountainous region of Korea. The CCA model yields the best Heidke skill of 53.8 at Kunsan. The predictive skill (correlation coefficient between the predictions and observations averaged for six stations) of summer rainfall in Korea is about 0.95, which indicates that the CCA model provides useful skill in predicting the variability of total summer rainfall in Korea.

From the above-mentioned results, it is concluded that the interannual variation of summer rainfall in Korea is largely influenced by the evolution of the Asian winter monsoon, the variation of the upper-level atmospheric circulation associated with Indian summer monsoon rainfall, and the SST anomaly in the tropical western Pacific via the EAP teleconnection. These results is give a useful insight into the causes of the interannual variation of summer rainfall in Korea. Furthermore, it is found that the temporal and spatial variability of summer rainfall in Korea using CCA model developed in this study can be predicted one month in advance.