台灣海峽北部潮流的主要分量

Abstract

台灣海峽輸送到東海的海水超過長江流量的五十倍，輸送到東海的沉積物、營養鹽佔其總量三分之一，對於台灣近海水域及東海的水質與生態有決定性的影響，在東海與南海物質通量交換過程中扮演著重要橋梁。 本研究即著重於北台灣海峽潮汐主要分量及流量的探討；有鑑於前人研究此區域使用的資料在時空分布上密度皆很低，缺乏長期及完整段面上的實測資料，我們利用裝設在航行於基隆與馬祖間台馬渡輪上的都普勒流速儀 (ADCP) 測量並收集沿航線上流速資料，將一年又四個月經篩選後約160~280筆每個網格位置不等的資料用調和分析方法找出沿航線半日潮 (M2, S2)、全日潮 (K1, O1) 等各分潮的振幅與相位。 由調和分析出各分潮潮流振幅判斷，台灣海峽的潮流由M2半日潮所主宰，最大振幅超過1 m/s，遠大於S2的1~30 cm/s，及不到15 cm/s的O1和K1潮流振幅，且台灣沿岸的潮流較大陸沿岸的潮流為強。由於渡輪資料的取樣頻率低，取樣時間又不固定，所以調和分析時能量較小或週期較相近的分潮間難以清楚區分。M2分潮潮流振幅調和分析結果與OSU潮流模式 (TPXO) 相關性可達0.9左右、潮流相位相關性可達0.75，顯示，大部份的潮流還是可以濾出的，但是S2能量會受到M2影響，K1和O1也因週期較長、能量太小，以目前的資料亦無法完全正確分析出來。如果分析所得的潮流振幅誤差小於20%, 則M2、S2 及K1各需資料長度約為3年、5年及3年。 用調和分析和TPXO濾潮結果皆得此處流量約 1~3 Sv、去回流量差值大多分布於 -1~1 Sv，其差異有可能是受到風的影響，但也有可能潮流的濾除不夠正確。另外，海峽東北部位處基隆海谷邊緣、海峽東側有觀音凹陷，地形起伏皆較大，由資料可明顯觀測到海峽東側有斜壓潮存在，潮流振幅上下層相差約5~10 cm/s，因此，本文的調和分析結果不會與正壓潮流模式完全符合。

The volume transport from Taiwan Strait to East China Sea is 50 times more than that from Yangtze River, and the sediments and nutrients from Taiwan Strait to East China Sea is one third of the total. It has a great influence on the water mass exchanging and ecology of Taiwan coast and East China Sea. The research focuses on the major constituents of tidal currents and the volume transport (VT) of the northern Taiwan Strait. Because the of data density previous studies is low in time and space in this region, lacking of long-term and a cross section of the strait actual measurement data, we installed a shipboard ADCP on the TaiMa Ferry that navigates between Keelung and Matsu to measure and to gather along track current velocity data. We obtain 160~280 data sets in different grids after screening sixteen months, and then use Harmonic Analysis to find out along-track tidal amplitude and phase of M2, S2, K1, O1 constituents. The result from Harmonic Analysis showed that M2 is the major constituent and its energy is more than S2, K1 and O1 in our study area. Owing to the low and discontinuous sampling frequency on the fixed station, we can’t distinguish constituents that have similar period and low energy. For example, the relation between Harmonic Analysis result and OSU tidal model (TPXO) result of M2 reach 0.9 shows that most tidal current can be filtered. On the other hand, S2 is affected by M2, K1 and O1 still can’t be filtered correctly due to longer period and lower energy. To have 20% or less error on tidal current constituents, it requires data length of 3, 5 and 3 years for M2, S2 and K1 tides. Use two different methods that TPXO and Harmonic Analysis to remove tidal contribution show 1~3 Sv VT through TWS and VT difference value of to- and return-cruises may be -1~1 Sv. The cause of the difference of VT between to- and return-cruises maybe the influence of wind, but it also implies the inaccuracy of remove tidal contribution. In addition, the northeast of the TS is located at the edge of the Keelung Valley and the east side of TS has a hollow and large variation on topography. This research also suggests the existence of baroclinic tide in these two regions, the associated tidal current amplitudes differ about 5~10 cm/s between upper and lower water level. It means that the result of Harmonic Analysis can never match the result of barotropic tidal model of TPXO.