利用無人船進行海洋聲速場之層析重建:性能指標與調查設計

Abstract

本研究旨在發展一調查分析工具，在給定海洋現象之相關時/空特徵尺度以及考慮無人船之有限資源下，可用來決策移動船聲層析實驗時之最佳軌跡。由於大部分海洋現象為一動態過程，透過理論分析方式，本研究首先推導出調查誤差指標(survey error metric) 之解析形式，用以量化移動船進行聲層析調查時因空間取樣不足及海洋時變性之調查誤差。且該理論指標亦在本研究中以數次數值模擬實驗來進行驗證。此外，考量無人船有限承載能量及最大/最小速度之限制，本研究推導出調查包絡 (survey envelope) 之解析形式，此包絡劃定了無人船在能量與速度之限制下，可成功完成實驗之調查參數空間。整合前述所提之調查誤差指標調查包絡，此調查分析工具可應用於移動船聲層析之實驗設計問題，並協助使用者量化不同調查策略中之優劣以從中取捨。最後，則藉由實例應用該工具於給定時/空特徵尺度之隨機場調查與綠島渦流聲速場調查之實驗設計問題，並對其進行相關討論。

This study aims at developing a survey analysis tool to determine the optimal trajectories of the autonomous surface vehicles (ASVs) to conduct a moving-vehicle tomography (MVT) experiment given an oceanographic process of interest with its associated spatial and temporal scales and limited resources of survey vehicles. As most of the oceanographic process is not stationary but evolving with time, a “survey error metric” is derived analytically for the MVT survey which accounts for errors due to both spatial under-sampling and temporal evolution of the sample field. The analytic metric has been validated numerically using several synthetic MVT experiments. In addition, due to the finite energy capacity and minimum/maximum speeds of the ASVs, a “survey envelope” delineating a operational region of survey parameter space is defined within which an ASV can complete a survey successfully. The developed survey analysis tool combines the survey error metric with the survey envelope to gain insight into the MVT experiment design problem and to quantify the trade-off between different survey strategies. Finally, the usefulness of the survey analysis tool was demonstrated using several examples including both the stochastic fields with given spatial-temporal scales and the synthetic sound-speed variations induced by flow passing an island.