Institut de Recherche de l'Ecole Navale (IRENav), BCRM Brest, CC 600, F-29240 Brest Cedex 9, France.
J Acoust Soc Am. 2010 Oct;128(4):1685-93. doi: 10.1121/1.3483720.
This paper presents the first results of an imaging technique that measures the geoacoustic structure of a seafloor in shallow water areas. The devices used were a broadband (100 Hz-6 kHz) acoustic source towed by a ship and a vertical array. Among all the acoustic paths existing in the water column, two are used: the direct one and the seabed-reflected one, the latter being composed of the reflections from the seafloor's surface as well as that from each buried layer. Due to the good time resolution of the signal and to the short range configuration, the reflected signal can be modeled as a sum of contributions coming from image sources relative to the seabed layers. The seabed geometry and the sound speed profile can then be recovered with the detection and localization of these image sources. The map of the image sources is obtained by a function that combines back-propagation of signals and knowledge of the emitted pulse. The thickness and sound-speed of each layer is finally obtained by a position analysis of the image sources. The results obtained by this data-driven algorithm on both at-sea and synthetic data are satisfactory.
本文介绍了一种用于测量浅海海底地质声学结构的成像技术的初步成果。该技术使用的设备包括由船只拖曳的宽带（100Hz-6kHz）声源和垂直阵列。在水柱中存在的所有声路径中，使用了两种：直达路径和海底反射路径，后者由海底表面以及每个埋层的反射组成。由于信号的良好时间分辨率和短距离配置，反射信号可以建模为来自相对于海底层的图像源的贡献之和。然后可以通过检测和定位这些图像源来恢复海底几何形状和声速分布。通过结合信号反向传播和发射脉冲知识的函数来获得图像源的图谱。最后通过对图像源的位置分析来获得每个层的厚度和声速。该数据驱动算法在海上和合成数据上的结果都令人满意。
