亞澳季風關聯性之年代際變異

Abstract

夏季季風年際變異受到熱帶海洋的影響甚鉅，因此亞洲季風與澳洲季風雖間隔半年，相距萬里，卻因赤道東太平洋以及熱帶印度洋異常海溫的影響，讓兩者有顯著的相關。本文主要探討在年代際變異的架構下，熱帶印度洋、赤道東太平洋與季風交互作用所產生的變動，是如何影響亞澳季風之間的關連性。 根據降水所定義的季風指數顯示，亞澳季風在1990年過後關係突然生變，從正相關轉變為負相關。為找出影響兩者關聯性的主要變異，並顧及亞澳季風與澳亞季風的不對稱性，本文利用大尺度OLR場的經驗正交函數分析，分別將兩個季風盛行的時間點放在一起，以JJA、DJF代表亞澳季風關聯性，以前一年的DJF、當年JJA代表澳亞季風關聯性。結果顯示，經歷1990/91年劇變，澳洲季風對ENSO的反應大體維持原樣，但是亞洲季風卻呈現迥然不同的風貌。 90年代之前，影響亞澳季風關聯性分屬於ENSO發展年(ENSO developing year)與ENSO衰落年(ENSO decay year)兩個時期。以Nino為例，在發展的前一年冬天以及當年春天，印度洋整片洋面較冷，因此夏季可提供的水氣較少，造成亞洲季風降水偏少。Nino成熟時，透過反向的渥克環流，抑制澳洲季風降水，因此亞澳季風降水都減少，兩者為同相位。而在西北太平洋區上，會有一異常反氣旋環流在聖嬰成熟時發展，並且從冬天一直持續到隔年夏天，在衰落年時抑制季風槽的形成，造成亞洲尤其西北太平洋季風降水減少。所以在此階段，影響亞澳季風與澳亞季風關係的機制不同，亞洲季風對ENSO的反應也不同，甚至在衰落年比發展當年明顯，造成方向上的不對稱。本時期ENSO週期較長，因此ENSO對亞澳季風的影響會總和兩個效果，使得亞澳季風有著正相關的相連性。 然而到了90年代，印度洋的自主性突然增強，Indian Ocean Dipole(IOD)變得活躍，並且與北方的季風系統連成一氣(特別是西北太平洋季風)，成為海氣交互作用的主軸。在Nino發展年時，正相的IOD會隨著聖嬰在夏季一同出現，造成蘇門達臘西方海面上的大氣又乾又冷，提供北方的亞洲季風一沉降之處，因此與季風形成逆向的區域哈德里環流，彼此增強，導致Nino年亞洲季風的降水不減反增(Nina與Nino的情況正好相反)，因而扭轉了亞澳季風的關係，由正相關轉變為負相關，也讓ENSO對亞洲季風的影響集中在ENSO發展年。亞澳季風與澳亞季風也由於IOD與亞洲季風的交互作用而呈現反對稱的型態，帶有準雙年的週期。 季風指數也隱喻著印度洋在90年代轉為主動去影響太平洋。當IOD與ENSO耦合時，會透過改變赤道西太平洋的低層風場，藉由海洋的Kelvin波加快ENSO的轉相。在我們的研究中，此現象於90年代特別明顯：此時的IOD異常活躍，造成ENSO內在結構的根本改變，尤其在相位轉換上，變成急促而俐落,傾向雙週年週期。所以90年代，印度洋自主性增強，經由與亞洲季風的交互作用，回饋到整個氣候系統，不僅改變了ENSO的結構，也扭轉了亞澳季風的不對稱與關聯性。

Interannual variation of summer monsoons is strongly influenced by tropical oceans. Therefore, the Asian monsoon and the Australian monsoon, which separate by 6 months and by ten thousand kilograms, are well correlated, due to the impact of the SST anomaly in the tropical eastern Pacific and the tropical Indian Ocean. The monsoon index defined by precipitation shows that the Asian-Australian monsoon (the AA monsoon) linkage abruptly changed from positive to negative correlation after 1990. To identify the variation of the AA monsoon linkage and take the asymmetry of the Asian-Australian monsoon and the Australian-Asian monsoon into account, we used the large scale OLR field, by putting two monsoon prevailing period together, JJA and DJF as the Asian-Australian monsoon linkage, and D(-1)JF and JJA as the Australian- Asian monsoon linkage, to do EOF analysis. The result indicates that after the abrupt change in 1990/91, the Australian monsoon had similar reaction to ENSO, while the Asian monsoon had distinct response in two periods. Before 1990s, the AA monsoon was connected in ENSO developing and decay year. For example, in the preceding winter and the spring of the El Nino developing year, the whole tropical Indian Ocean became cool, which resulted in less water vapor, and leaded to dry Asian monsoon. In winter, El Nino matured, and suppressed the Australian monsoon rainfall by reversed Walker cell. As a result, the Asian monsoon and the Australian monsoon were both dry and in phase in the El Nino developing year. However, an anomalous anti-cyclonic circulation over Western North Pacific developed when El Nino matured, and persisted till the following summer to block the formation of the monsoon trough, which caused dry Asian monsoon, especially the Western North Pacific monsoon (WNPM).The Asian-Australian monsoon and the Australian-Asian monsoon were asymmetric owing to the impacts of ENSO to the Asian monsoon were different. Low frequency ENSO was dominate in this period, and made the AA monsoon in phase by combining two effects of ENSO. However, in 1990s, Indian Ocean dipole (IOD) turned active. IOD and the Asian monsoon, especially the WNPM, became the main part of the air-sea interaction in summer. In the El Nino developing year, the positive IOD, which accompanied El Nino in summer, made the low level atmosphere over the western ocean of Sumatra dry and cool, and convert this place into a subsidence region to support the Asian monsoon in the north. The Asian monsoon and the western ocean of Sumatra formed a reversed local Hadley cell and intensified by each other. Consequently, the Asian monsoon was wet rather than dry in the El Nino developing year (the El Nina year was opposite to the El Nino year) and thus the AA monsoon linkage changed from positive to negative correlation, and simply responded to ENSO in developing year. The Asian-Australian monsoon and the Australian-Asian monsoon were anti-symmetric and had TBO frequency because of the interaction between the Asian monsoon and IOD. The monsoon index also implies that Indian Ocean influenced the tropical eastern Pacific actively in 1990s. When IOD and ENSO are well-coupled, the low level wind over tropical western Pacific would be changed, and accelerates the phase turning of ENSO through the oceanic Kelvin waves. In 1990s, the Indian Ocean turned active, and had a feedback mechanism of the monsoon- tropical ocean system, the interaction with the Asian monsoon, which not only change the period of ENSO into TBO frequency, but also the AA monsoon linkage and the asymmetry of the monsoons.