高頻雷達是近代廣用的即時監測海流遙測工具，但雷達系統會因為儀器故障 或訊號干擾等因素影響而不時造成觀測區內海流資料發生缺漏不完整的問題。蘇 澳、漢本CODAR 高頻雷達測站自 2011 年 4 月開始運作以來，迄 2012 年 12 月止 在臺灣東北海域觀測並記錄15048 組表面海流資料(每小時一組)，在這些資料中，
各資料點發生缺漏資料之百分比出現率幾乎不超過5%；整體而言，在此期間內如 果扣除完全沒有任何缺漏以及完全沒有觀測資料兩種極端情況，則以觀測區內缺 漏一個資料點的情形為最多，共有126 組，佔所有有缺漏點情況的 29.37%；而缺 漏1~3 個資料點的比例則有 50.58%，即不完整資料的大部份情況缺漏個數都在 3 個以下；觀察每日出現資料缺漏的逐時變化，可以發現在本地時間晚上6 點到凌 晨4 點間比較容易出現資料有缺漏的情形，此段時間內平均每小時資料缺漏之出 現機率大約佔所有缺漏情形的4~6%。
這些觀測資料之平均流向為北至東北，與我們所認知的黑潮流況相符；統計 分析東西向以及南北向之海流流速分量資料，皆呈常態分佈；經由能譜分析，則 可發現當地海流主要是受低頻運動的影響，其次則為潮流；由於潮汐運動不屬於 隨機過程，因此使用調和分析法進一步將觀測資料中的潮流濾除，並以此濾潮後 之資料進行平穩性檢定，結果呈現為具有弱平穩性，顯示這些時序資料之二階統 計結果應當可以向外延伸用於觀測時段之外。
我們使用實向量EOF法以及KLE法分別分析濾潮後完整無缺值的CODAR觀測 資料(由2011年4月14日至2012年11月30日，共13280組)，得出觀測場的特徵值與對 應之特徵向量，並以此特徵向量做為基底，再求出流速資料在各基底的投影即可 得出該模組之振幅，如此可將各基底分量重新合成並重建流場。經二法處理後，
由實向量EOF法得到的前20個模組可解釋流速資料總變異量的97.02%，而KLE法則 可解釋97.14%。此外，另計算由前20個模組重建流場與原始觀測流場間的流速偏
差均方根值，在70個資料點中，實向量EOF法之偏差均方根整體平均值為5.92 cm/s，
而偏差均方根之最大與最小值分別為8.40 cm/s、3.68 cm/s，靠近二CODAR雷達站 處均方根較小，但觀測區外圍及北部則較大。至於KLE法偏差均方根之整體平均 值為6.39 cm/s，最大、最小值為11.38 cm/s與3.79 cm/s，同樣也是離雷達站較近處 均方根較小、較遠處均方根較大。整體而言，此結果顯示離雷達站較遠處的資料 點受到高階模組的影響較大，也就是說受到小尺度運動的影響較大。
此外，我們選用2012 年 12 月(共 734 組)完整無缺值之 CODAR 觀測資料作為 獨立樣本用以檢測不同方法填補不完整觀測資料之效果。我們分別使用了實向量 EOF 法以及 KLE 法(均使用前 20 個模組)並各自搭配最小平方法以及迭代法等共組 成四種填補方法，依照人為製造的不同資料點缺漏個數進行填補實驗，並計算填 補結果與原始觀測值之偏差均方根作為誤差用於比較填補效果。計算結果顯示，
不管資料缺漏個數多寡與填補方法為何，誤差均方根皆是在靠近雷達站處較小，
而在觀測區海域外圍少數的資料點處較大。若將四種資料缺漏填補方法之誤差均 方根互相比較(圖 4-6)，可以發現以最小平方法進行填補實驗時，當資料點缺漏數 小於40 時其誤差均方根數值隨缺漏點個數的變化並不大，但當缺漏個數大於 40 後，誤差即開始明顯增大，當缺漏個數達50 以上時，誤差則呈指數成長；若以迭 代法填補時，其誤差均方根數值會隨缺漏點個數的增多而逐漸變大，並呈現線性 增長的趨勢，其斜率皆約0.37~0.38 左右。
綜合而言，當缺漏個數在40 (57%)以下時，四種方法中以最小平方法搭配實 向量EOF 法進行填補之誤差均方根為最小，填補效果較佳；但在缺漏個數大於 50 (71%)時由於使用最小平方法填補後得到之誤差亦逐漸變大，故在此情況下不適合 使用這些方法來填補缺漏值。
參考文獻
Alvera-Azcárate, A., A. Barth, M. Rixen, and J. M. Beckers, 2005：”Reconstruction of incomplete oceanographic data sets using empirical orthogonal functions: application to the Adriatic Sea surface temperature”, Ocean Modelling, 9, 325-346.
Barrick, D. E., M. W. Evans, B. L. Weber, 1977：”Ocean Surface Currents Mapped by Radar”, Science, 198, 138–144.
Beckers, J. M., and M. Rixen, 2003：”EOF Calculations and Data Filling from Incomplete Oceanographic Datasets*”, Journal of Atmospheric and Oceanic Technology, 20, 1839-1856.
Bendat, Julius S., and Allan G. Piersol, 1971：“Random data: analysis and measurement procedures”, John Wiley & Sons, New York.
Boyd, Janice D., Eileen P. Kennelly, and Pavel Pistek, 1994：”Estimation of EOF expansion coefficients from incomplete data”, Deep-Sea Research Part I:
Oceanographic Research Papers, 41.10, 1479-1488.
Bretherton, Francis P., Russ E. Davis, and C. B. Fandry, 1976：”A technique for objective analysis and design of oceanographic experiments applied to MODE-73*”, Deep Sea Reasearch and Oceanographic Abstracts, 23, 559-582.
Bui-Thanh, T., M. Damodaran, and K. Willcox, 2004：”Aerodynamic Data
Reconstruction and Inverse Design Using Proper Orthogonal Decomposition”, AIAA Journal, 42.8, 1505-1516.
Dong, D., P. Fang, Y. Bock, F. Webb, L. Prawirodirdjo, S. Kedar, and P. Jamason,
2006：”Spatiotemporal filtering using principal component analysis and
Karhunen-Loeve expansion approaches for regional GPS network analysis”, Journal of Geophysical Research, 111, B03405, doi: 10.1029/2005JB003806.
Defant, Albert, 1961：”Physical oceanography; volume 2”, Pergamon Press.
Emery, Brian M., Libe Washburn, and Jack A. Harlan, 2004：”Evaluating Radial Current Measurements from CODAR High-Frequency Radars with Moored Current Meters*”, Journal of Atmospheric and Oceanic Technology, 21, 1259-1271.
Emery, Brian M., Libe Washburn, Milton S. Love, Mary M. Nishimoto, and J. Carter Ohlmann, 2006：”Do oil and gas platforms off California reduce recruitment of bocaccio (Sebastes paucispinis) to natural habitat? An analysis based on trajectories derived from high-frequency radar”, Fishery Bulletin, 104.3, 391-400.
Fang, Y.-C, J. Wang, Y.-J. Yang, J.-C. Mau, and M.-J. Huang, 2011：”Preliminary Results of CODAR Surface Current Observations Northeast of Taiwan”, 2011
International Workshop on Operating System for Marine Environment Monitoring and Forecasting, National Kaohsiung Marine University, College of Ocean Engineering, Kaohsiung, 59-77.
Fukunaga, Keinosuke, 1990：”Introduction to Statistical Pattern Recognition, Second Edition”, Academic Press.
Gurgel, K.-W., G. Antonischki, H.-H. Essen, and T. Schlick, 1999：”Wellen Radar (WERA): a new ground-wave HF radar for ocean remote sensing”, Coastal Engineering, 37, 219-234.
Han, Jiawei, and Micheline Kamber, 2006：”Data Mining: Concepts and Techniques,
Second Edition”, Morgan Kaufmann Publishers, 61-62.
Hardy, Donald M., 1977：”Empirical eigenvector analysis of vector observations”, Geophysical Research Letters, 4.8, 319-320.
Holmes, Philip, John L. Lumley, and Gal Berkooz, 1996：”Turbulence, Coherent Structures, Dynamical Systems and Symmetry”, Cambridge University Press.
Jan, Sen., Joe Wang, Ching-Sheng Chern, and Shenn-Yu Chao, 2002：”Seasonal variation of the circulation in the Taiwan Strait”, Journal of Marine Systems, 35, 249-268.
Jan, Sen, Ching-Sheng Chern, Joe Wang, and Ming-Da Chiou, 2012：”Generation and propagation of baroclinic tides modified by the Kuroshio in the Luzon Strait”, Journal of Geophysical Research, 117, C02019, doi: 10.1029/2011JC007229.
Kaihatu, James M., Robert A. Handler, George O. Marmorino, and Lynn K. Shay, 1998：”Empirical Orthogonal Function Analysis of Ocean Surface Currents Using Complex and Real-Vector Methods*”, Journal of Atmospheric and Oceanic Technology, 15, 927-941.
Kaplan, Alexey, Yochanan Kushnir, Mark A. Cane, and M. Benno Blumenthal,
1997：”Reduced space optimal analysis for historical data sets: 136 years of Atlantic sea surface temperatures”, Journal of Geophysical Research, 102.C13, 27835-27860.
Kerschen, Gaetan, Jean-Claude Golinval, Alexander F. Vakakis, and Lawrence A.
Bergman, 2005：”The Method of Proper Orthogonal Decomposition for Dynamical Characterization and Order Reduction of Mechanical Systems: An Overview”, Nonlinear Dynamics, 41, 147-169.
Kohut, Josh, Scott Glenn, and Don Barrick, 1999：”SeaSonde Is Integral to Coastal Flow Model Development”, Hydro International, 3.3, 32-35.
Kohut, Josh T., and Scott M. Glenn, 2003：”Improving HF Radar Surface Current Measurements with Measured Antenna Beam Patterns”, Journal of Atmospheric and Oceanic Technology, 20, 1303-1316.
Kondrashov, D., and M. Ghil, 2006：”Spatio-temporal filling of missing points in geophysical data sets”, Nonlinear Processes in Geophysics, 13, 151-159.
Kundu, Pijush K., J. S. Allen, and Robert L. Smith, 1975：”Model Decomposition of the Velocity Field near the Oregon Coast”, Journal of Physical Oceanography, 5, 683-704.
Kundu, Pijush K., and J. S. Allen, 1976：”Some Three-Dimensional Characteristics of Low-Frequency Current Fluctuations near the Oregon Coast”, Journal of Physical Oceanography, 6, 181-199.
Kutzbach, John E., 1967：”Empirical Eigenvectors of Sea-Level Pressure, Surface Temperature and Precipitation Complexes over North America”, Journal of Applied Meteorology, 6, 791-802.
Metropolis, Nicholas, and S. Ulam, 1949：”The Monte Carlo Method”, Journal of the American Statistical Association, 44.247, 335-341.
Oliveira, P., and L. Gomes, 2010：”Interpolation of signals with missing data using Principal Component Analysis”, Multidimensional Systems and Signal Processing, 21, 25-43, doi: 10.1007/s11045-009-0086-3.
Paduan, Jeffrey D., and Leslie K. Rosenfeld, 1996：”Remotely sensed surface currents in Monterey Bay from shore-based HF radar (Coastal Ocean Dynamics Application
Radar)”, Journal of Geophysical Research, 101.C9, 20669-20686.
Pawlowicz, Rich, Bob Beardsley, and Steve Lentz, 2002：”Classical tidal harmonic analysis including error estimates in MATLAB using T_TIDE”, Computers &
Geosciences, 28, 929-937.
Preisendorfer, Rudolph W., 1998：”Principal Component Analysis in Meteorology and Oceanography”, Elsevier.
Ramp, Steven R., Donald E. Barrick, Tomotaka Ito, and Michael S. Cook,
2008：”Variability of the Kuroshio Current south of Sagami Bay as observed using long-range coastal HF radars”, Journal of Geophysical Research, 113, C06024, doi:
10.1029/2007JC004132.
Reynolds, Richard W., and Thomas M. Smith, 1994：”Improved Global Sea Surface Temperature Analyses Using Optimum Interpolation”, Journal of Climate, 7, 929-948.
Smith, Thomas M., Richard W. Reynolds, Robert E. Livezey, and Diane C. Stokes, 1996：”Reconstruction of Historical Sea Surface Temperatures Using Empirical Orthogonal Functions”, Journal of Climate, 9, 1403-1420.
Snihir, Iryna, William Rey, Evgeny Verbitskiy, Afifa Belfadhel-Ayeb, Peter H. L. Notten, 2006：”Battery open-circuit voltage estimation by a method of ststistical analysis”, Journal of Power Sources, 159, 1484-1487.
Teague, Calvin C., John. F. Vesecky, and Daniel M. Fernandez, 1997：”HF Radar Instruments, Past to Present”, Oceanography, 10.2, 40-44.
Tiampo, K. F., J. B. Rundle, W. Klein, Y. Ben-Zion, and S. McGinnis, 2004：”Using Eigenpattern Analysis to Constrain Seasonal Signals in Southern California”, Pure and
applied geophysics, 161, 1991-2003.
劉文俊，1999：”臺灣的潮汐”，自費出版。