| 摘要: | 本研究之目的在於定量觀測風、波浪與表面海流特徵於次網格尺度的空間變異性。本研究使用76顆漂流式微型資料浮球，於2021年9月22日起為期一個月在東海南部布放，觀測區域內表面流速、示性波高、平均週期與海表風速等海氣象參數，浮標等間距布放，形成一個空間陣列，每30分鐘並將數據透過衛星回傳，後續對其進行時序上之資料品管。
本文討論海氣象因子觀測數據之定律性(Deterministic)過程與儀器相互誤差及自然界中隨機(Random)特性的比例。首先比較海洋現場觀測實驗資料與衛星遙測資料及數值模式結果之差異，不同資料來源海氣象因子的時空變化趨勢一致，但是海上實測資料呈現更高的空間變異性，討論研究區域海氣象參數變異與空間尺度(Length Scale)的關係。
在海表風、波浪參數方面進行統計分析，發現變異特徵隨空間尺度增加而增加，其中變異係數(C.O.V.)與空間尺度的關係與前人研究進行比較發現有一致的趨勢，較前人研究的空間尺度涵蓋範圍更廣達O(1)~O(2)公里(10~400公里)。並發現數值模式資料於次網格尺度的空間變異性與實際的觀測有差異，各海氣象因子空間變異性，實際觀測與數值模式的比值最大達6至17.9，此比值關係隨空間尺度的增加而降低。進一步利用機率分布擬合，以半常態分布代表不考慮儀器觀測誤差、韋伯分布代表考慮儀器觀測誤差的空間變異機率特性參數化方法。
在海表流的方面，本研究採用兩種拉格朗日(Lagrangian)分析方法針對海表流進行分析，一為利用浮標位置計算在研究區域內的水平延散係數(Dispersion coefficient)；二為計算所對應之有限大小李雅普諾夫指數 (FSLE，Finite Size Lyapunov Exponent)特性。本文比對有限大小李雅普諾夫指數並與海洋擴散條件與空間尺度之關係，結果顯示，研究期間之東海南部海域，擴散距離(δ)於10~100公里李雅普諾夫指數依循δ-2/3斜率、延散係數依循δ4/3斜率，皆呈現理查森定律(Richardson’s Law)之紊流擴散特性，與Corrado et al. (2017)描述於北太平洋環流之擴散特徵差異不大。
海氣象參數實際觀測數值的空間變異性參數化結果，可以提供數值模式、遙測資料在次網格尺度下數據驗證及品管的依據。;The purpose of this study is to quantify the spatial variability of wind, wave, and surface current features at the subgrid scale. In this study, 76 miniature wave buoys were deployed in the southern part of the East China Sea for one month starting from September 22, 2021, to observe the surface current velocity, significant wave height, mean period, and surface wind speed and other meteorological parameters in the area.
In this study, we discuss the ratio of the Deterministic process to the mutual error of the instruments and the random characteristics of nature in the observations of meteorological factors. We first compare the differences between the experimental data and the satellite remote sensing data and numerical model results. The temporal and spatial variability of the meteorological factors from different data sources are consistent, but the spatial variability of the measured data is higher.
Further statistical analysis of the surface wind and wave parameters reveals that the variability characteristics increase with the spatial scale, and the relationship between the coefficient of variation (C.O.V.) and the spatial scale is consistent with previous studies. The spatial variability of the numerical model data at the subgrid scale is found to be different from the actual observations, and the spatial variability of each meteorological factor, the ratio between the actual observations and the numerical model is up to 6-17.9, and this ratio decreases with the increase of the spatial scale. The spatial variability of the spatial variability is parameterized by using a half Normal distribution for the spatial variability without considering the instrumentation error and a Weibull distribution for the spatial variability with considering the instrumentation error.
For the surface currents, two Lagrangian analysis methods are used to analyze the surface currents: one is to calculate the horizontal dispersion coefficient in the study area using the buoy array position; the other is to calculate the corresponding Finite Size Lyapunov Exponent (FSLE). In this paper, we compare the finite-size Lyapunov exponent with the spatial scale of the ocean dispersion conditions, and the results show that the dispersion distance (δ) in the southern part of the East China Sea during the study period follows the slope of δ-2/3 and the dispersion coefficient follows the slope of δ4/3, both of which exhibit Richardson′s Law. The diffusion characteristics of the turbulent flow according to Richardson′s Law are similar with described by Corrado et al. (2017) for the North Pacific.
The results of spatial variability parameterization of meteorological parameters can provide the basis for data validation and quality control of numerical models and remote sensing at subgrid scale. |
