由單船ADCP資料推算東海六月平均流

Abstract

在大多數的淺海，絕大部分的海流是伴隨潮汐產生的潮流，其中又以全日潮流和半日潮流為主要成分。若要得到低頻流速（主要是平均流及週期大於一天的海流），通常會施放流速儀，測量海流時間超過25小時，才能將全日潮和半日潮的成分濾除而得低頻流的流速。低頻流，又稱餘流，主要是平均流及週期大於一天的海流，以下以平均流代稱。使用船載ADCP (Acoustic Doppler Current Profiler) 也可以測量平均流，常見的方法是相位平均法，也就是在測線上的每個地點，測量潮流二至五等分的相位，以各相位的平均值代表平均流。但是大多數的水文調查實驗主要目標並非測量平均流，ADCP的資料只有當時當地的一筆流速資料，沒有在潮流二至五等分相位所測量的海流資料可供計算平均流，本論文利用潮流及平均流隨空間緩慢變化的特性，從散佈的ADCP資料中，估算平均流及潮流。經由處理1999年6月東海14天航次的ADCP資料，將展示如何由這類短期的航次資料中，推估平均流與半日潮流。本文的基本假設與許多的東海潮汐數值模式相類似，在此航次半天之內所經之區域，其半日潮流及平均流皆隨空間緩慢變化。分析方法是先將每筆1分鐘的ADCP 資料做11分鐘的平均以減少資料量，再做12.4小時的滑動平均得出剩餘流速（仍含一部份的全日潮流）。然後用雙線性公式擬合此剩餘流速(低頻流)的空間分布而得到大尺度的平均流。在此分析中，大陸坡上的流速資料已被刪除，因為在黑潮通過的區域，潮流及平均流的空間梯度偏高。此平均流與歷年來東海的海流(以漂浮球為主)觀測資料特性相符。在遠離無潮點及近岸的部分，潮流振幅與相位角皆與潮流模式及潮流觀測值近似。因此，在東海大陸棚中部執行探測作業的船載ADCP資料，也可以估算出該區當時的平均流場與半日潮流的振幅與相位分布。

In most shallow seas, the major components of ocean current are tides. To derive the residual low frequency current velocity, one usually deploys current meter moorings over 25 hours or more to remove the diurnal and semi-diurnal tidal components. Using ship-board ADCP (Acoustic Doppler Current Profiler) to derive the residual current, one may design ADCP measurements at different phases of tides at every location. The sum of these measurements cancels out the tidal component. But, most cruises are aimed at other goals, and ADCP measurements are only complementary data of the cruise. Here, we demonstrate a method for deriving the M2 and S2 tidal current and the mean current from a 14-day cruise in the East China Sea, June 1999. The basic assumption is that the tidal current varies smoothly in the region of survey. This assumption is valid as seen in many tidal models of East China Sea. The 1-min ADCP data were first averaged over 11 minutes to reduce data size, then averaged over 12.4 hours to derive the residual current that contains little M2 tidal component. Data over continental slope were deleted because of the influence of Kuroshio and the possibility of large gradient in tidal velocity. The residual current data were fitted with bi-linear spatial function to capture its main feature. The pattern of derived mean current matches those published earlier from mooring and drifter data. The semi-diurnal tidal current equals the total current minus the residual current which is 12.4-hour moving average of total current velocity. The co-phase and co-range line of the semi-diurnal tidal current are similar to those of model results of semi-diurnal tidal elevation and tidal current observation near cruise track. This provides a proxy proof of the reliability of this method.