結合北太平洋海溫模態與熱帶地區海洋熱含量探討ENSO的演變

Abstract

先前的研究指出考慮熱帶地區次表層海洋熱含量的暖水體 (例如：Warm water volume，簡稱WWV) 以及北太平洋中緯度地區海平面氣壓變異所引起的海溫模態 (例如：Victoria mode，簡稱VM) 有利於發展聖嬰/反聖嬰事件 (El Niño Southern Oscillation，簡稱ENSO)。此研究利用觀測資料與數值模擬來探討結合北太平洋海溫變異與熱帶次表層熱含量對ENSO演變的影響。首先透過經驗正交函數的分析來檢驗次表層熱含量的氣候變異，其中第一與第二個模態為ENSO模態與它的相位轉換過程 (WWV)。交互相關分析指出這兩者氣候模態之間有著中度相關且約七個月的相位轉換，並且WWV的形成是透過中部赤道太平洋地區附近的風應力旋度所引起的經向副熱帶海洋環流(Subtropical cells)的變異，進而將水體注入至赤道地區逐漸形成WWV。透過區分南北側近赤道區域的風應力旋度所造成經向水體運輸的分析，指出南北側的信風(Trade winds)距平將會對赤道地區注入熱含量，讓WWV更有利於發展ENSO。這樣的結果能夠確認在ENSO的週期下改變熱帶地區熱含量的變異，不僅是由近赤道地區的海洋與大氣耦合動力下的貢獻，還有來自副熱帶地區信風變異的影響。 VM 是北太平洋中緯度海溫第二個氣候模態，並且有著典型的海溫足跡來影響熱帶地區的氣候動力。合成分析指出結合正(負)相位的VM與正(負)相位的WWV事件將有利於聖嬰(反聖嬰)事件的形成。主要的動力機制是VM的海溫變異能改變赤道地區的大氣環流且造成西風距平的生成，這樣的風場變異將能產生海洋凱爾文波(Kelvin waves)，並且將原先赤道地區的WWV向東傳遞來觸發ENSO事件。 本研究最後透過地球系統模式來驗證VM的貢獻對於WWV的改變所引起ENSO的發展。當VM與WWV指標皆是同個相位時，模式的表現將傾向於隨後幾個月發生ENSO事件。敏感度測試實驗指出如果持續地添加VM海溫變異至模式中，VM與Niño3.4指標之間的關聯性將會提高，指出ENSO事件的發生裡VM是其中可控制的氣候因素。這樣的結果能夠確認VM與WWV是不同的氣候變異，並且皆是ENSO重要的預測指標。儘管最近的研究指出WWV在2000年以後不再是一個良好預測ENSO的指標，但是2015-16年的聖嬰事件能夠驗證結合VM與WWV的氣候變異能發展ENSO事件，並且如果考慮VM的變異能夠讓WWV更具有ENSO的可預報性。

Previous studies suggested that both tropical subsurface heat content (i.e., Warm Water Volume, WWV) and the North Pacific sea surface temperature pattern (specifically Victoria Mode, VM), linked to the North Pacific sea level pressures, may tend to develop El Niño Southern Oscillation (ENSO) events. Here, we study the linkages between the VM and tropical ocean heat content within the ENSO evolution using the observation and model experiments. The first two dominant Empirical Orthogonal Function (EOF) modes of tropical ocean heat content represent the typical ENSO variation and its phase transition. Cross-correlation analysis indicates a moderate correlation of a 7-month phase lag between these two modes. The formation of WWV is directly caused by the meridional transport of Subtropical Cell driven by the wind stress curl. Separation of Sverdrup transport induced by the off-equatorial wind stress curl between northern and southern regions suggests that anomalous trade winds from both hemispheres may provide an additional charging of the subsurface heat content, leading to the WWV be conducive to develop the ENSO. These results confirm that not only coupled air-sea interactions near the tropics but also anomalous trade winds from the subtropics can modulate tropical subsurface heat content during the ENSO cycle. The VM is defined as the second mode of sea surface temperature in the North Pacific, typical footprinting of the extratropical forcing on the tropical dynamics. Composite analysis confirms that the joint impacts of positive (negative) VM and positive (negative) WWV favor the development of El Niño (La Niña) events. The VM may favor the occurrence of anomalous westerlies in the tropics that drives oceanic Kelvin waves, causing the eastward propagation of WWV to trigger the ENSO events. The main contribution of VM on the WWV change leading to the ENSO development is verified using the Community Earth System Model. When the VM and WWV indices have the same sign, the model tends to develop ENSO events a few months later. An additional sensitivity experiment shows that the correlation between VM and Niño3.4 can be increased if the VM pattern is consistently imposed on the model, suggesting the controlling role of VM on the ENSO events. These results confirm the distinct roles of VM and WWV as the critical predictors of ENSO variability. Finally, although recent studies indicate that WWV is not a good predictor of ENSO events after 2000, the El Niño event in 2015-16 can verify that the joint impacts of VM and WWV variability on ENSO evolution, making the WWV more predictable about the ENSO event in consideration of the VM variability.