:::
| 論文分類 | 碩士 |
|---|---|
| 學號 | 108626004 |
| 姓名 | 陳銘誼 Ming-Yi Chen |
| 標題 | 應用微波雷達監測海流之演算法流程改善 Process Improvement of Algorithm for Monitoring Sea Surface Currents with X-band Microwave Radar |
| 指導教授 | 錢樺 Hwa Chien |
| 畢業日期 | 2021-06 |
| 附件檔案 | |
| 參考連結 | https://etd.lib.nycu.edu.tw/cgi-bin/gs32/ncugsweb.cgi?o=dncucdr&s=id=%22GC108626004%22.&searchmode=basic |
| 摘要 | 本研究以岸基X-band微波雷達作為主要觀測工具。本研究之目的在於建立一套可供作業化應用之近岸流場反演流程，與評估水深資料反演之可行性。本文研究方法為現場實驗資料之分析，包括為降低微波雷達量測波浪的回波訊號對距離和方向之依賴性，增進圖像中波紋的亮暗對比，改善雷達影像的品質，並開發計算流程與各階段產品之檢視與比對，根據分析視窗之子影像的大小、解析度以及子影像時序列張數在分析海表面流場時，其對方法所造成的影響決定所採用之參數。 本研究以位於新北市的中央氣象局白沙灣近岸海象微波雷達於2020年1月27至28日觀測資料，以及位於桃園市觀音海岸的大潭微波雷達於2021年1月5日至11日觀測資料作為案例，藉由對比度受限自適應直方圖均衡化對雷達影像進行預處理，強化回波圖像的波紋特徵後，利用二維及三維傅立葉轉換分別求取一維雷達徑向及二維海面空間的角頻率及波數關係，並由波向和轉速對頻率和波數設定值進行修正，後以分散關係式擬合此頻域訊號求得雷達徑向流速，並以都卜勒流速剖面儀之現地實測海流資料進行比對，最後試圖以速度-方位角處理(velocity-azimuth processing, VAP)技巧，將徑向流速轉化為空間流場。 應用對比度受限的自適應直方圖均衡化結果，可增進分散關係式在受海流影響的範圍內訊號與其他雜訊之訊噪比約10%，有關於海流的演算改善約6.7%。 In this study, coastal-based X-band microwave radar is used as the subject observation tool. The purpose of this paper is to reduce the dependence of the echo signal on the range and direction of the microwave radar when doing wave measurement, to improve the contrast between the brightness and darkness of the wave pattern, to reconstruct the bathymetry from the wave signal, to retrieve the radial velocity of the radar, and to estimate the current fields. In addition, the effect of the size and resolution of the sub-images (analyzed window) and the number of pictures of the sub-images time-series on the analysis of the sea surface flow field and their applicability to the method are investigated. The data of this study, the microwave radar of the Central Weather Bureau's Baisha Bay near-shore in New Taipei City was observed from January 27 to 28, 2020, and the microwave radar of Daitan in Guanyin Coast of Taoyuan City was observed from January 5 to 11, 2021, as case studies. The radar image is pre-processed by contrast-limited adaptive histogram equalization (CLAHE) to enhance the stripe characteristics of the radar echo image, and then the relationship between angular frequency and wavenumber are obtained using two-dimensional and three-dimensional fast fourier transformations for the one-dimensional radar radial and two-dimensional sea surface space, respectively. The radar radial velocities are then compared with the in situ current data from the acoustic doppler current profiler, and finally, the velocity-azimuth processing (VAP) technique is used to convert the radial velocities into spatial flow fields. The result of applying contrast-limited adaptive histogram equalization, it can improve the signal-to-noise ratio of the signal affected by ocean currents inside the dispersion relation shell and other noise by about 10%, and the accuracy of retrieving ocean currents improved by about 6.7%. |
| 參考文獻 |
