| 摘要: | 本研究應用航海用S頻段(S-Band)和X頻段(X-Band)低掠角(Low Grazing Angle,LGA)微波雷達為主要觀測工具，進行現場實驗從觀測結果探討雷達回波強度受到海面波浪影響的程度。 本研究首先進行雷達接收強度之率定工作。第一步以佈放於觀測海域上的自製金屬浮球做為雷達觀測之標的物，再使用Gommenginger et al.(2000)的外部率定法(External Calibration Approach)，回歸出雷達系統傳遞方程式(Radar System Transfer function)中所需要的待定參數，分別完成兩架不同頻段的雷達之率定。 本研究進一步分別進行了兩次現場實驗：颱風時期實驗和季風時期實驗，再將觀測結果以兩個層面進行討論，分別是長時間之雷達回波統計分佈分析及短時間的波浪相位變化與雷達回波之關係兩者進行討論。討論項目第一為，平均雷達回波(Sea clutter)強度於統計上的特徵及其與風速、海面平均傾度(Mean Square Slope,MMS)和粗糙參數(Roughness Parameter)間之相關性；第二，本研究依據雷達連續觀測所得之定點上回波時序列與佈放於海中之聲學式都普勒流速剖面儀(Acoustic Doppler Current Profiler,ADCP)的音鼓測得之水位和傾度，將水位以零上切法解析出個別波引致之水位，探討雷達回波特性與其對應之水位和傾度之相關性。 本研究結果發現，平均雷達回波在觀測距離較遠時，掠角低觀測海面區域所反射之區域(footprint)大於海面波長時，回波訊號(spike)可以雙峰高斯分佈描述，可視為兩個不同平均與標準差的常態分佈線性之和。其與低掠角微波雷達回波的兩種機制一致，亦即本研究獲得觀測上的證據，支持雙尺度模式(two scale model)中的布拉格背向散射(Bragg Backscattering)及重力波調變作用之效應。此外，海面背向散射影像因海面波浪起伏所呈現的亮暗帶分佈，在一水位波長內雷達的回波強度僅有一最大值與最小值，最大值非發生於波峰，最小值也非發生於波谷，其間存在一相位差，此相位差隨波浪尖銳度增加而變大，顯示低掠角雷達受波峰遮蔽影響程度甚大，其效應不可忽視。 In present study, the commercial marine S-Band and X-Band low grazing angle radar systems were utilized as the main tools in field experiments, and investigation of the influence of the surface waves was carried out. The first step in this study is the calibration of radars. Metallic balls with various sizes were launched in the observational area and used as the reflectors of radar detection. The External Calibration Approach (Gommenginger et al. 2000) was then applied to obtain the parameters of Radar System Transfer Function. In this study, the calibration of two different bands of the radar systems were achieved, respectively. Secondly, field experiments including, typhoon Fanapi and a winter monsoon events, were carried out. Synchronized observations of gravity waves, air-sea fluxes, and dual-band Radar systems were implemented. The strategy of data analysis is twofold, i.e. the statistical analysis on the probability of long-term records of Radar Backscatter intensity, and the cross-correlation analysis on the dependency of Radar backscatter intensity to the wave elevation and slope. In the first part, the discussions were focused on the temporal averaged radar backscatter intensity to the wind speeds, significant wave heights, Mean Square Slope of gravity waves and the Roughness Parameters. The second part we emphasized on,the dependency of Radar backscatter intensity to the phase resolved water elevation and slope, which were obtained using zero-up crossing to separate individual wave form from the ADCP measured time series. The result of present study shows that the probability of return signals (spike) are better described using Bimodal Gaussian distribution, especially for those whose footprints were greater than the lengths of surface wave. The Probability distribution can be regarded as the summation of two normal distributions with two individual means and standard deviations. The lower ones represent the wind induced wavelet effects, whereas, the higher peak represent the gravity waves. Meanwhile, this is the observation evidence of the two scale theory of low grazing angle microwave radar. Furthermore, the distribution of the high gray value (GN) and low gray value band on the radar image which affected by gravity waves features single maximum and minimum. The maximum value does not coincide with the wave crest; likewise minimum value does not coincide with wave trough. In this case, phase lag could be identified. Moreover, with increasing wave steepness, phase lag increases. It could be concluded that the low grazing angle radar is affected by the sheltering effect which should not be neglected in any circumstances. |
