東亞地區夏季降水三極結構之模擬研究

Abstract

東亞地區夏季的年際降水變化，存在一個三極結構(tripole pattern)：當華中、日本一帶多雨時，華南及華北地區少雨，反之亦然。此三極結構為東亞地區夏季降水的年際變異的第一個EOF。本研究因此以其PC為指標，來定義三極結構的強弱。Hsu and Lin（2007）提到，此三極結構的正相位與負相位是不對稱的，其能量來源可能受南邊菲律賓海附近的異常加熱(Nitta 1987)，也可能受歐亞大陸、青藏高原的非絕熱加熱影響(Enomoto et al. 2003)，而此現象本身極可能是存在於大氣本質中的現象(intrinsic mode)，經由外界的擾動造成此現象的出現或幅度增強。 本篇研究運用模式資料驗證上述假說，所採用的模式分別為SINTEX-F與ECHAM，將模式資料分成大氣與海洋耦合的模式，以此模式的海溫做為未耦合模式的邊界條件，再分為年際變化與氣候值海溫的驅動模式。 將有大氣與海洋耦合作用的模式資料與觀測資料對照，可發現模式資料(SINTEX-F、ECHAM)在EOF1可清楚模擬出三極結構，但模式資料看不出觀測資料中正負相位之不對稱性。在低層表現較高層模擬佳，有耦合的資料表現上勝於沒有耦合，有年際變化勝於氣候值。因此可看出海洋對大氣的交互作用在三極結構上有一定程度的重要性。 在海氣耦合模式中，海溫與降雨呈現負相關，未耦合模式中，海溫與降雨則為正相關，顯見兩種模式中的對流降雨機制並不相同。推測在有海洋的模式中海溫是海氣交互作用下的產物。此海溫在沒有海洋的模式中，卻成為直接影響降雨的重要因子，這個現象可能是由大氣主導而影響海洋。

There is a tripole rainfall pattern in East Asia during the northern summer. The positive (negative) phase of the pattern is characterized by more (less) rainfall in central eastern China, Japan, and South Korea, and less (more) rainfall in northern and southern China. This rainfall pattern is the first EOF mode in East Asia interannual variability. Hsu and Lin (2003) pointed the positive and negative phases are asymmetric. It could be originated in the heating anomalies in the tropical Western Pacific, which in turn triggers a wave-like pattern emanating northward toward extra-tropical East Asia (Nitta 1987). It also could be originated in the heating over the eastern Tibetan Plateau, which induce the eastward-propagating wave-like structure (Enomoto et al. 2003). It is suggested that the tripole pattern is a result of the amplification of an intrinsic dynamic mode that can be triggered by various factors despite of their different origins. This study uses numerical simulation data (SINTEX-F, ECHAM) to identify the hypothesis. There are two types of simulations in this experiment. The first type coupled with atmosphere and ocean model. The second type is AGCM foreced by simulated SST, and AGCM forced by the simulated climatological SST. SINTEX and ECHAM model could produce realistically the tripole rainfall pattern, which is the dominant pattern in East Asia. But there is no significant asymmetric phenomenon in model simulation data. The simulation of the N-S wave, which is affected by tropical heating, is better than the simulation of the E-W wave, which is affected by the wave activity in the Eurasian continent. Un-coupled simulation produces the tripole pattern, but not as distinct as coupled model. The simulation forced by the simulated interannual monthly SST is better than the one forced by the simulated climatological monthly SST. Therefore the Ocean plays an important role in exciting the tripole pattern. The relationship between SST and precipitation in coupled and un-coupled model is different. In coupled model, they had negative correlation; in un-coupled model, they had positive correlation. It maybe SST is the product after air-sea interaction in coupled model, but SST effects directly the rainfall in un-coupled model. Atmospheric effect dominates this structure.