臺灣鄰近海域環境噪音之資料分析及數值模擬

Abstract

本研究為瞭解臺灣鄰近海域的環境噪音之變動特性，對主動式聲納和被動式聲納的偵測距離之影響，利用ASIAEX、VANS和OAJEX等實驗所蒐集的環境噪音資料，以統計方法分析噪音位準的變動特性。首先由ASIAEX在大陸棚海域的觀測數據之分析結果顯示：風浪影響環境噪音的最低頻段約為500 Hz；風浪噪音在頻譜位準的衰減斜率將隨海況而變動；以1k Hz的風浪噪音觀察高低海況的噪音位準之標準偏差值，當風速在5 m/s以上時噪音的變動明顯降低；而隨著統計時間越長則噪音位準的變動可能越大，因此提高了噪音位準的不確定性；在台灣鄰近幾處海域的環境噪音分佈的特性都略有差異。其次，建構一噪音數值模式以模擬海表面噪音的聲場特性，由1k Hz的風浪噪音模擬結果顯示，海面環境改變噪音聲源強度是首要影響噪音位準變動的因素，海洋內部環境如非線性內波改變海水聲速對噪音位準的影響則相對有限；從模擬結果與實驗資料比對已經驗證模式的可信度，因此該模式可實際應用到噪音位準的預估中。最後利用ASORPS的聲納效能預估算例，討論噪音位準預估誤差和噪音不確定性對聲納偵測距離的影響，結果顯示ASORPS目前使用的經驗公式對臺灣鄰近海域的噪音位準預估之誤差略大，因此本文提出一改善方法，即建構一海洋環境噪音資料庫，結合實驗資料的統計結果與數值模式的模擬，提供噪音位準的平均值和標準偏差值給予ASORPS，以提高聲納系統的預估效能。

This study is to survey the variant characteristics of ocean ambient noise in the vicinities of Taiwan and to understand their effects on detection ranges of active and passive sonar system. The data from the experiments of ASIAEX, VANS, and OAJEX are analyzed by statistical methods to obtain varying characteristics of noise levels. The first study is to analyze the ASIAEX data that were measured on the continental shelf, which indicates the minimum frequency-dependence of wind-wave noise is about 500 Hz. The noise level slopes in noise spectrum are changed with sea states. The standard deviations decrease significantly while wind speeds are over 5 m/s. The uncertainty from noise variance may increase due to the statistical duration. There are some distribution variations of ocean ambient noises in the vicinities of Taiwan. The second study is to develop a noise numerical modeling to simulate the sea surface noise fields. The results show the changes of noise source strength from sea surface are the main factor, and the changes of sound speeds due to nonlinear internal waves in ocean are the second factor. The comparisons of resultes on simulation and data have provided good confidence in this noise modeling, which can be applied to predict noise levels. Moreover, the ASORPS study cases explain the prediction errors and the uncertainty of noise levels on detection ranges of sonar. The results show the errors in the noise levels prediction in the vicinities of Taiwan by using the present empirical formula are a little high. So this study provides a development method that is to construct a database of ocean ambient noise, which integrates with statistical results of measurement data and simulations of numerical model. The mean and standard deviation of noise levels would be provided for ASORPS to improve the prediction performance of sonar.