Coherent processing of shipping noise for ocean monitoring

Abstract

Extracting coherent wavefronts between passive receivers using cross-correlations of ambient noise may provide a means for ocean monitoring without conventional active sources. Hence applying this technique to continuous ambient noise recordings provided by existing or future ocean observing systems may contribute to the development of long-term ocean monitoring applications such as passive acoustic thermometry. To this end, we investigated the emergence rate of coherent wavefronts over 6 days using low-frequency ambient noise (f < 1.5 kHz) recorded on two vertical line arrays-separated by 500m- deployed off San-Diego CA in ~200m deep water. The recorded ambient noise was dominated by nonstationary distributed shipping activity with the frequent occurrence of loud isolated ships. Noise data were first processed to mitigate the influence of these loud shipping events in order to primarily emphasize the more homogenous and continuous background ambient noise in the frequency band. Furthermore, the coherent noise field propagating between the VLAs was beamformed using spatio-temporal filters to enhance the emergence rate of specific coherent wavefronts. This presentation will discuss various strategies for the selection of these spatio-temporal filters (either data-derived or model-based) in order to improve the continuous tracking of these coherent wavefronts over 6 days.