東沙島附近之潮流-當地模式模擬

Abstract

本論文之研究目的在於：(1)以資料分析及當地模式探討東沙島附近水位與正壓潮流(barotropic tidal current)之間的關係，(2)藉由水位資料來預報正壓潮流。水位與流速的觀測資料分別取自置放於東沙島(DongSha Island)西岸的潮位計(tide gauge)與東沙島東北方約100公里(當地水深350公尺)的ADCP (Acoustic Doppler Current Profiler)流速資料。水位資料涵蓋期間為2004年11月初至2006年3月底，流速資料為2005年7月28日至11月1日，兩者約有3個月的同時(simultaneous)資料。東沙島的水位資料經由調和分析，結果顯示主要的分潮(tidal constituent)為全日潮K1、O1以及半日潮M2、S2，其振幅分別為27、24、14、以及5公分。由深度平均法計算流速的正壓潮分量，經由調和分析結果顯示，K1、O1、M2的流速大小約4 ~ 5公分/秒，而S2的流速約2公分/秒。 利用轉換函數(transfer functions)與調和分析兩種方法，驗證水位預報正壓潮流的可行性，將觀測資料分為等長的兩段時期，藉由第一段時期約45天的正壓潮流與水位的關係，預報下一段時期約45天的正壓潮流，結果顯示半日潮流的觀測值與預報值大致符合，其均方根(root mean square)約在3公分/秒以下；但是全日潮流的觀測值與預報值的差異大，其均方根最大可達到約19公分/秒，推測其原因為斜壓潮分量(internal tide)大而影響潮流，但是不會對水位資料造成影響。 根據一維正壓潮流模式(Clarke and Battisti 1981; Battisti and Clarke 1982 a,b; Clarke 1991)，以及M2的能量傳輸(transport)往西北方傳播幾乎不變(nondivergent)(Beardsley, 2004)，應用到東沙島的M2潮位資料以及當地模式(local model)推算附近海域的正壓潮流大小，模式結果顯示當底部摩擦係數r = 3×10-2 公尺/秒，半日潮M2的觀測值與模式值大致吻合，而底部摩擦係數的數量級接近於澳洲大堡礁，可能由於兩地皆受到珊瑚礁地形的影響，所以造成底部摩擦係數的數量級大於多數的沿岸地區。全日潮O1、K1的觀測值與模式值的比對結果雖然在可接受的範圍，但是無法比擬半日潮M2觀測值與模式值的吻合結果。 綜合資料分析與模式的結果，皆證實半日潮M2為主要的正壓潮流，可藉由水位資料預報；但是全日潮O1、K1的斜壓分量明顯且非規律，故無法藉由水位資料預報其正壓潮流大小。

關鍵詞: 東沙島，潮流，內潮，正壓，斜壓。

The purpose of this thesis is using data analysis to study the relationship between sea level and barotropic tidal current near DongSha Island, and developing a local model to examine the tidal dynamics in a shallow water environment. Right at the coast of the DongSha Island, a tide gauge was deployed from November, 2004 to March, 2006, 100 km offshore in a water depth of 350 m, an ADCP (Acoustic Doppler Current Profiler) was moored to record the current velocity from July 28 to November 1, 2005. There have been simultaneous measurements of both sea level and current for 3 months. Harmonic analysis of the DongSha Island sea level data shows that the most significant tidal constituents in this region are the K1 (amplitude: 27 cm) and O1 (24 cm) diurnal tides and the M2 (14 cm) and S2 (5 cm) semidiurnal tides. The barotropic tidal current was computed using the depth-averaged velocity. Harmonic analysis reveals that the tidal current velocity of K1, O1 and M2 are about 4 ~ 5 cm s-1, and S2 is about 2 cm s-1. The relationship between barotropic tidal currents and sea level were established for the first 45-day period using the transfer functions and harmonic analysis, and then carry the prediction for the second 45-day period. There is very good agreement between the observed and predicted semidiurnal tidal current, the rms (root mean square) error is less than 3 cm s-1. However, prediction on diurnal tidal current is rather disappointing, the rms error is about 19 cm s-1, suggesting strong baroclinic tidal component (internal tides) may have a significant contribution to the tidal current, which is not registered in the sea level records. Knowing the tides propagate in the northwestward direction (Beardsley, 2004) toward the coast of mainland China, the barotropic tidal current can be computed through a simple one-dimensional, barotropic tidal model (Clarke and Battisti 1981; Battisti and Clarke 1982 a,b; Clarke 1991). Giving the M2 sea level at the DongSha Island and topography, the local model successfully computes the tidal current near the observational site with a bottom friction coefficient r = 3×10-2 m s-1. This friction coefficient is rather large compared with that of most known coastal waters, but closed to what have been found near Great Barrier Reef, Australia, where both have similar coral reef environment. Computed K1 and O1 tidal currents are within acceptable range, but not as good as the M2 component. In essence, both the data analysis and model results confirmed that the M2 semidiurnal tidal current is largely barotropic, and can be predicted through the coastal sea level. However, the baroclinic component of O1 and K1 diurnal tidal current are rather obvious and prohibited us from establishing a tight relationship, and hence, a prediction scheme from the coastal tidal data.

Key Words: DongSha, tidal current, internal tides, barotropic, baroclinic.