以長期衛星高度計資料探討南海北部黑潮套流所形成之反氣旋式渦旋

Abstract

過去的研究指出黑潮受東北季風與來自北太平洋內部的中尺度渦旋等影響，有機會進入南海北部，於臺灣西南外海形成套流，此套流可進一步剝離形成反氣旋式渦旋，本研究運用24年(2000–2023)的衛星高度計資料重新探討此一現象。夏季平均資料顯示黑潮蜿蜒進呂宋海峽，經南海邊緣流回到臺灣東岸外海；冬季平均資料顯示，黑潮進入南海形成套流與渦旋的複合結構，約盤據在臺灣西南外海200 × 200 km的區間，此區間內的平均相對渦度(ζ)時間序列顯示：套流/渦旋事件(ζ=0.3f-0.4f)主要發生在每年11月至隔年3月，且在近10年渦度有增強趨勢。通過時序列觀察發現，共發生 26 次渦旋事件，其中 12 次由季風驅動，3 次與西北太平洋渦旋撞擊黑潮有關，另有 11 次為季風與西北太平洋渦旋共同作用的結果。經由絕對動力高度之複數經驗正交函數分析結果指出，Mode-1 (30.5%) 代表選取區域的季節性變化，Mode-2 (25.7%) 代表黑潮擺動與套流的形成，Mode-3 (11.5%) 為反氣旋式渦旋的剝離與向西南行進。渦旋的發展與風場變化有關，本研究比較風應力（τ）、風速（Ｕ_10）分別與渦旋渦度（∣ζ/f∣）、面積（A）與環量（|ζ/f | A）之間的統計關聯性。結果顯示，風應力與三者之相關係數分別為 0.52、0.48 與 0.56，正相關不明顯；將風應力與渦度進行時間積分後，相關性可微幅提升至 0.64，顯示風場累積作用較能對應渦旋的發展歷程。風速的統計相關性較風應力明顯高，其分別與面積與環量相關性達 0.93 與 0.90。為瞭解渦旋的一般特性，本研究自 26 個事件中以 Okubo–Weiss 參數篩選出 10 個具有良好對稱形狀的渦旋作為觀察案例，結果顯示平均生存期約為 200 – 350 天，其中發展期約占 20%、成熟期約 30%、行進期則達 50%；最大渦旋動能可達 0.3 m² s⁻²，平均直徑與振幅分別約 200 km與 0.15 m。生存期、振幅、直徑均不及西北太平洋渦旋，但渦旋動能僅略小於大洋渦旋。

Previous studies have indicated that the Kuroshio Current, influenced by the northeast monsoon and mesoscale eddies from the North Pacific interior, can intrude into the northern South China Sea (SCS) through the Luzon Strait and form a loop current southwest of Taiwan. This loop current may eventually shed into an anticyclonic eddy (AE). In this study, we revisit this phenomenon using 24 years (2000–2023) of satellite altimetry data. The summer climatology reveals a meandering Kuroshio entering the Luzon Strait and recirculating along the SCS western boundary before returning to the east of Taiwan. In contrast, the winter climatology shows a more developed loop current and AE structure extending into the northern SCS, typically centered in a 200 × 200 km area off southwestern Taiwan. The time series of area-averaged relative vorticity (ζ ) within this region indicates that loop current/eddy events (ζ = 0.3f–0.4f) predominantly occur between November and March, with a notable increase in frequency over the past decade. Based on time series analysis, 26 eddy events were identified: 12 driven by monsoonal forcing, 3 triggered by impinging Pacific cyclonic eddies, and 11 resulting from a combined mechanism. Complex Empirical Orthogonal Function (CEOF) analysis of Absolute Dynamic Topography (ADT) reveals three dominant modes: Mode-1 (30.5%) represents seasonal variation in the study area; Mode-2 (25.7%) corresponds to Kuroshio meandering and loop current formation; and Mode-3 (11.5%) describes the detachment and southwestward propagation of AEs. The development of these AEs is closely associated with surface wind forcing. Statistical analyses were conducted to evaluate the relationship between wind stress (τ), wind speed (U₁₀), and eddy properties including normalized relative vorticity (|ζ/f|), eddy area (A), and circulation strength (|ζ/f| A). Correlation coefficients between τ and the three eddy metrics were 0.52, 0.48, and 0.56, respectively—suggesting weak positive correlations. However, integrating τ and ζ over time improved the correlation (R = 0.64), implying that cumulative wind forcing better reflects the eddy growth process. Wind speed showed significantly stronger correlations, particularly with eddy area (R = 0.93) and circulation strength (R = 0.90), suggesting that wind speed may play a primary role in controlling horizontal eddy size.

To characterize general eddy features, 10 well-formed and symmetrical AEs were selected based on the Okubo–Weiss parameter. The results show an average lifespan of 200–350 days, with growth, maturity, and propagation phases accounting for approximately 20%, 30%, and 50% of their lifecycles, respectively. The maximum eddy kinetic energy reached 0.3 m² s⁻², with average diameter and amplitude of approximately 200 km and 0.15 m. While their lifespan, amplitude, and size are smaller than those of typical open-ocean eddies in the northwest Pacific, the kinetic energy of these coastal AEs is only slightly lower, indicating robust eddy activity under the influence of regional forcing.