測海雷達可用於長時間與廣範圍的偵測海面目標特性，其中包括海浪參數、洋流特性、海面風場與船隻定位等，然而海面狀況相當複雜，使得雷達接收到的回波變化很大，因此建立模擬頻譜用以了解回波特性為測海雷達資料分析處理的首要工作。 本研究利用散射截面積理論方程式與風浪模型模擬測海雷達回波頻譜特性，包括流速、風速、風向對功率及都卜勒速度的影響，並藉此驗證多種來向角定位演算法的可行性與優劣，於實際船隻定位及洋流徑向速度解算提供很大的助益，同時針對高頻測海雷達受電離層干擾問題進行分析討論，提供污染距離辨識方法。 在船隻定位上利用HHT與濾船演算法(Cspro)開發自動偵測系統，應用在特高頻測海雷達上偵測率高達81.25%，並且距離誤差僅0.3公里，角度誤差也1度以內，可望用於海域交通的控管。 實驗結果發現特高頻測海雷達受潮汐影響明顯，在功率與徑向速度的比對上，皆符合潮位計資料的特性，並且分析探討使用波束成形技術觀測二維海面潮位變化的可行性。 在徑向流速的比對上，利用漂流浮標比對特高頻與高頻測海雷達的量測結果，發現高頻測海雷達有明顯的偏差，此偏差可用表層流與次表層流受Stoke’s drift與Ekman drift來解釋，估算結果也大致符合實際觀測偏差量，另外，利用特高頻測海雷達觀測資料比對兩種方法在徑向流速的解算結果，發現Capon method的空間覆蓋率較高，但容易在高角度的時候因旁波瓣影響造成計算誤差，與模擬結果一致，而MUSIC解算受限於頻率解析度與訊雜比，在頻率解析度與訊雜比夠高的情況下，亦可得到良好的資料品質，並且擁有不需要大型天線陣列的空間使用效率。 在示性波高反演中發現特高頻測海雷達觀測資料與龜山島浮標量測資料存在倍數關係，需律定參數得到較為可靠的雷達反演資料，本論文利用浮標與雷達反演資料在不同律定參數下的RMSE變化，計算龜山島周回海域利用特高頻測海雷達反演示性波高的律定參數為1.5。;Sea radar can be used to detect characteristic of sea surface, including wave, current, wind, and ship. However, the radar echo has high fluctuation due to the complicate sea sate that constructing a spectrum model to know the echoing feature is the first priority of radar analysis. This study uses the theoretical radar cross section formula and wind wave model to simulate power spectrum of sea echo to know the influence of Doppler velocity and echo power causing by current and wind. It is also helpful for ship detecting and radial current estimation by comparing different algorithm of direction of arrival from the simulating result. The interference of ionosphere has been analyzed and providing an algorithm for detecting contamination area. The automatically ship detecting algorithm has been developed by using Hilbert Huang Transform(HHT) and ship filtering program(Cspro). The detecting rate of ship is 81.25% and the root mean square error of range and direction are 0.3km and less 1 degree. The influence of Chung-Li Sea Radar(CLSR) radar echo causing by tide is obvious from the experiment. The echo power and radial velocity measuring from VHF radar meet the characteristic of tide height measured from tidal gauge. This study make assessment of two-dimensional tidal gauge by using VHF radar through beamforming technique. As for the radial current comparison of radar and buoy, it has a bias from the result of CODAR and buoy. This bias can be explained by the difference of surface current and sub-surface current causing by the Stoke’s drift and Ekman drift. And the comparison of radial current calculation between MUSIC and Capon method performs by the CLSR. The Capon method has higher coverage of field of view, but also has higher estimating error when steering angle is too large. The MUSIC needs high signal to noise ratio and frequency resolution to have good quality result without large space of antenna deployment The significant wave height estimation of CLSR is comparing with buoy at Guishan Island and having a systematic bias that need a calibration factor to have more reliable data. By calculating the root mean square error between radar and buoy measurement of different calibration factor, the calibration factor of CLSR to measure significant wave height around Guishan Island is 1.5.
