| 摘要: | 亞洲季風是大尺度現象也是行星系統重要的部份，為影響東亞區域氣候的主要天氣系統之一，但夏季季風並非每年總是準時到達，探討夏季季風肇始年際差異機制一直是大氣科學界重要課題。造成季風肇始之年際差異因素雖然很多，但其中以海洋因素干擾最為顯著，本文藉由資料分析和數值實驗來探討熱帶海溫變化對於亞洲夏季季風的影響。
資料分析顯示，亞洲夏季季風最早建立於孟加拉灣，緊接著在南海以及阿拉伯海/印度半島西側，最後才在西北太平洋上建立。ASI季風建立與低層阿拉伯海季風槽和索馬利噴流有非常密切的關係，BOB季風是由於印度半島低層季風槽與青藏高原南側槽線合併而使季風爆發，SCS季風建立為印度季風槽持續發展和高壓脊東退的結果，WNP季風的建立主要受到北太平洋副高脊和季風槽形成的影響。
透過與季風區強、弱季風年的分析比較也可以發現，WNP強季風年的環流型態與聖嬰成熟前的夏季（Nino+0年）或反聖嬰成熟後之夏季（Nina+1年）類似，而WNP弱季風年的環流型態與聖嬰成熟後的夏季（Nino+1）或反聖嬰成熟前之夏季（Nina+0年）相近。ASI強、弱季風區的環流型態和WNP季風區相反，顯示赤道太平洋海溫年際變化對WNP和ASI夏季季風強弱有顯著之影響，其中台灣附近形成的反氣旋式（氣旋式）環流扮演著減弱（加強）WNP（ASI）夏季季風的關鍵角色。SCS和BOB季風區的環流變化型態與聖嬰（反聖嬰）年的分布出現相當大的不同，顯示聖嬰現象對於SCS和BOB季風區的影響並不明顯，此兩區的季風年際變化並不是由赤道東太平洋的海溫變化所主導。
經由上述之模式實驗結果可知，EEP實驗和WNP實驗對於西太平洋地區天氣型態的影響是一致的，但印度洋地區則相反。加入CNP海溫變化可以減緩太平洋副高的變化程度，EEP+CNP和WNP+CNP實驗於Nino+0年減弱太平洋副高強度，而隔年則增強副高環流。比較海溫模擬實驗與實際觀測結果（圖4.48），發現兩者之間的差異如下：在Nino+0年，WNP實驗在印度半島南端形成氣旋環流，使ASI和BOB西部降水增加。WNP+CNP實驗中阿拉伯海出現氣旋環流，使ASI的降水趨勢與實際情況相反。在Nino+1年，由於EEP實驗中的副高強度明顯減弱，台灣附近反而形成氣旋環流距平，造成WNP季風區的降水增加。WNP實驗的副高強度仍然不夠，同時南亞地區的西風增強，使SCS和WNP地區降水出現增加。EEP＋CNP實驗的西太平洋反氣旋強度仍然不夠，因此WNP降水仍呈現增加的趨勢。WNP+CNP實驗的降水趨勢和環流分布與實際觀測結果接近。由此可知，亞洲季風區的環流變化在Nino+0年主要受到赤道東太平洋加上中太平洋海溫變化所導致的結果，而Nino+1年則由西北太平洋配合上中太平洋的海溫變化所主導。
The tropical low-frequency oscillation involves complex interactions of ocean-atmosphere-land of multiple temporal and spatial scales. Its study has became a major focus of atmospheric sciences in recent decades. So far, two significant low-frequency signals are found in the tropics: one is the 30~60 days "Intra-Seasonal Oscillation", while the other is the so-called "Southern Oscillation". In spite of the fact that the above phenomena are resulted from different mechanisms, yet both phenomena involve significant convective latent heating and air-sea energy exchange. Thus examining the role played by large-scale circulation, tropical cumulus heating, and sea surface temperature variations becomes a key to understand the above tropical low-frequency oscillations. In this study, we utilize NCEP reanalysis data of atmosphere and SST, CMAP precipitation to examine the evolution and structure of atmospheric low-frequency oscillations, along with the relationship with SST variations. We also utilize UCLA AGCM to simulate the atmospheric response to SST variations in equatorial eastern Pacific, north western Pacific, and Indian ocean, respectively.
Observations indicate that the 30~60 days oscillation consists of wavenumber 1 or 2 precipitation signal developing in Tropical Indian ocean. The precipitation signal propagates eastward with a phase speed of 10~15 m/s along the equator. Its strength peaks in winter and spring. During the El Nino years, from September to the following May, the signal is weaker than normal in Indian ocean and western Pacific but stronger in eastern Pacific. Estimating the sea surface evaporation-wind feedback flux exhibits a 15 days lead time against the precipitation field which favors eastward propagation. The magnitude of transient evaporation flux shows that its not the major energy source for the maintenance of 30~60 days oscillation, but to play a secondary role in modulating the amplitudes and seasonal variations of tropical 30～60 days oscillations. Analysis using singular value decomposition further indicates that the evaporation-wind feedback flux interacts positively with surface westerly in the tropical western Pacific and Indian Ocean, while interacts negatively in the tropical eastern Pacific and Atlantic Ocean.
During the El Nino year , SST increases in the eastern Pacific between 10°S to 10°N , the maximum amplitude is about 4ºC. SST cooling occurs in the subtropical western Pacific at the same time. After mature phase, the negative SST anomalies switch to positive anomalies in the south western Pacific, while the negative SST anomalies in the north western Pacific can maintain several months later, showing asymmetric SST pattern in the western Pacific. An anti-cyclonic circulation is found over the north western Pacific which may increase moisture flux and precipitation in the regions of Taiwan and southeast China in February , March and April.
Numerical simulations indicate that the eastern Pacific SST variations is the major factor responsible for interannual variability while the Indian Ocean and Western Pacific SST play the secondary role in modulating the regional circulation and climate. The north western Pacific experiment shows that compensate cooling will increase local anti-cyclonic circulation and thermally indirect circulation over the western Pacific in the northern hemisphere. Indian ocean experiment shows that the convective activities are locked in equatorial which may delay the Indian monsoon onset. The atmospheric non-linear response occurs mostly in mid-latitude of winter hemisphere.
The relationship between Taiwan's spring precipitation and Nino3.4 winter SST anomaly exhibits high correlation in the period of 1979-2001, especially in strong El Nino years, and so does the relationship between Taiwan summer precipitation and the western North Pacific summer monsoon index (WNPSMI). In normal year, Strong (weak) WNPSM results in the increase (decrease) of Taiwan summer precipitation. During ENSO years, major rain bands move eastward to 130°E～160°E、10°N～30°N which decreases the summer precipitation in Taiwan. On the other hand, the WNPSM tends to be out of phase with the East Asian summer monsoon. |
