臺灣東北海域之音傳損耗不確定性研究

Uncertainty Quantification of Acoustic Transmission Loss in the Sea Area Northeast of Taiwan

張元櫻(Yuan-Ying Chang)

指導教授 ： 陳琪芳

國立臺灣大學/工學院/工程科學及海洋工程學研究所/博士(2011年)

https://doi.org/10.6342/NTU.2011.10025

摘要

水下音響的應用層面非常廣泛，且大部分應用常牽涉到水下音響傳播(聲場)之預估，如何輸入足以描述真實海洋環境之參數予數值模式進行計算是在聲場預估過程中最大的挑戰。臺灣東北的北棉花峽谷附近海域在水文環境、地形及底質上都具有複雜的變化，使得北棉花峽谷海域具有高度的環境不確定性，在此海域進行聲場預估時也會存在不可忽略的誤差。由於臺灣東北海域在經濟、軍事或海洋資源開發方面都屬於重要海域，在以現有的技術尚無法完全避免海洋環境與聲學不確定性的前提下，量化此海域的不確定性與定義關鍵環境因子之相關研究工作就顯得更為重要。本論文的研究目標為定量並降低北棉花峽谷附近海域之音傳損耗不確定性，了解其在陸棚區域與陸棚邊緣區域之差異，並基於海洋模式輸出獲得此海域中水文不確定性與音傳損耗不確定性之關係。
本論文在研究過程中應用了「Quantifying, Predicting, and Exploring Uncertainty」(QPE)實驗中的中頻聲學資料、水文資料、地形資料以及底質資料，完成的研究內容包含(1)定義及定量音傳損耗不確定性，蒐集地形及底質資料，以降低音傳損耗不確定性為目標，定義最佳地形及底質參數組合；(2)結合海洋模式建構四維模擬系統，定量運用此四維模擬系統之音傳損耗不確定性，並分析陸棚與陸棚邊緣區域之差異；(3)定義海洋模式預估水文場之不確定性，分析在不同物理條件下，水文不確定性對音傳損耗不確定性之影響。
研究結果顯示，基於最佳地形與底質參數組合，以現場水文資料進行聲場預估之音傳損耗不確定性可以使用平均值為0dB、標準差為3.8dB之常態分布表達，達到沒有系統誤差之成效，且在陸棚區域與陸棚邊緣區域差異不大。若以完整的預估過程為出發點，結合Princeton Ocean Model (POM海洋模式)與Gaussian Beam Model (GBM聲學模式)建構四維模擬系統，針對臺灣東北海域進行水文不確定性分析以及音傳不確定性分析，結果顯示海洋模式預估結果之水文不確定性隨海洋物理現象改變，且可依其聲速誤差剖面之特性切分為由內潮通過所帶來之等溫面下沉期間、等溫面抬升期間以及冷水通過期間，三段期間之水文不確定性反映在音傳損耗不確定性上，對陸棚區域之音傳損耗不確定性影響較大，分別造成高估、接近以及低估真實音傳損耗之趨勢，而陸棚邊緣區域之音傳損耗不確定性則比較不受影響。
本論文對於東北海域之音傳不確定性提供定量之結果，相關研究成果可被應用在評估及修正聲場預估結果上，降低由音傳損耗預估誤差在水下音響應用層面上產生之誤差。

關鍵字

音傳損耗不確定性 ； 臺灣東北海域 ； 聲速不確定性

並列摘要

Uncertainty of transmission loss (TL), resulted from the uncertainty of geophysical and physical oceanographic parameters (features), could consequently contribute to the uncertainty of sonar performance prediction. This research adopts coupled and combined modeling and data of field observations over the continental shelf and slope close to North Men-Hua Canyon offshore northeastern Taiwan in order to study the uncertainty of transmission loss. This area contains range-varying sediment type, rapid-changing bathymetry, and complex water column activities brought by the Kuroshio intrusion, which jointly induce a highly uncertain environment for sound propagation. Sound propagation in this area is observed and modeled to interpret the propagation effect of the water column and seabed. Finally, the uncertainty of ocean model output and transmission loss in this area is quantified. And the spatial and temporal effects on the ocean and acoustic field are also quantified in this thesis.
In this thesis, field survey data of Quantifying, Predicting, and Exploring Uncertainty (QPE) experiment is applied to study the uncertainties, include mid-frequency acoustic data, bathymetry data, sub-bottom survey data, and temperature data. The works finalized in this thesis includes (1) quantifying the uncertainty of TL and defining the optimal data set of bathymetry and geoacoustic, (2) combing ocean model and acoustic model, and giving the uncertainty of TL resulted by using this combined model, and (3) quantifying the uncertainty of ocean model output and analyzing it’s acoustic effects.
According to the results, the lower bound of the uncertainty of TL could be described by a Normal distribution with 0 dB mean and 3.8 dB standard deviation based on applying optimal bathymetry and geoacoustic data set and the CTD data. The uncertainties are similar on shelf region and shelf break region. On the other hand, while combining the Princeton Ocean Model and acoustic model to predict watercolumn and TL, the uncertainty of ocean model output varies with oceanographic features and could be categorized into three periods corresponding to the descending and the ascending of isothermal brought by internal tides and the emerging of cold water. The uncertainty of ocean model output would project on the uncertainty of TL and result in larger variation on the shelf region. The predicting results would likely to overestimate, close to, and underestimate the TL on the shelf region but have little influence on the predicting TL on the shelf break region.
The study of this thesis provides the quantification of the uncertainty of TL close to North Men-Hua Canyon offshore northeastern Taiwan. The results could be applied to verify and adjust the simulated TL and reduce the errors while applying TL prediction.

並列關鍵字

uncertainty of transmission loss ； uncertainty of sound speed ； North Men-Hua Canyon

參考文獻

5. 楊穎堅，臺灣周邊海域(東北)水下偵測整合計畫─子計畫三：水文及聲速分佈特性研究(III)，國科會專題研究計畫成果報告，2011

42. 羅建育，臺灣周邊海域(東北)水下偵測整合計畫─子計畫四：臺灣周邊海域(東北)海床沉積物之聲學特性研究(III)，國科會專題研究計畫成果報告，2011

2. E. S. Livingston, J. A. Goff, S. Finette, P. Abbot, J. F. Lynch, and W. S. Hodgkiss, “Guest editorial capturing uncertainty in the tactical ocean environment,” IEEE J. Ocean. Eng. 31, 245–248, 2006.

3. S. Finette, “A stochastic response surface formulation of acoustic propagation through an uncertain ocean waveguide environment,” J. Acoust. Soc. Amer., 126, 2242–2247, 2009.

7. R. H. Headrick, J. F. Lynch, J. N. Kemp, A. E. Newhall, K. von der Heydt, J. Apel, M. Badiey, C. Chiu, S. Finette, M. Orr, B. Pasewark, A. Turgot, S. Wolf, and D. Tielbuerger, “Acoustic normal mode fluctuation statistics in the 1995 Swarm Internal Wave Scattering experiment,” J. Acoust. Soc. Amer., 107, 201–220, 2000.