Noise Source Direction of Arrival Estimation in the New England Mud Patch Using Acoustic Vector Sensors

The Seabed Characterization Experiment 2022 (SBCEXP 2022) was conducted in the New England Mud Patch (NEMP) and shelf break areas from May 6, 2022 to June 6, 2022. The experiment aimed to collect active and passive acoustic data supporting the development and comparison between geoacoustic inversion data sets. Various acoustic sensors were deployed to collect this data, including Vertical Line Arrays (VLA) and Ocean Bottom Recorders (OBX). Additional contextual data, such as Automatic Identification System (AIS) ship tracking data showing vessel traffic in the experiment area, was also collected. The measurements presented in this paper utilize data collected from OBXs and the logged AIS data to estimate the direction of arrival (DOA) of ship noise and active acoustic sources using only passive acoustic data. The OBX sensors deployed in the experiment are acoustic vector sensors capable of capturing sound pressure and particle velocity measurements in all three dimensions. This dimensional information can be leveraged to generate DOA estimations using beamforming techniques. Multiple DOA estimation algorithms, including the Multiple Signal Classification (MUSIC) method and the Capon method (also known as minimum-variance distortionless beamforming), are used to analyze OBX data and make DOA estimates for noise sources. These estimates are then compared to processed AIS data, showing ship position and track in the NEMP. Underwater noise pollution from shipping traffic is pervasive globally [1]. DOA data from ship noise sources on the ocean surface can provide contextual information for maritime surveillance and security, environmental monitoring, ship navigation and traffic management, scientific research, and search and rescue operations. The data processed and presented in this paper show the detection and azimuth tracking of vessels in the shipping lane and the active sources deployed for SBCEXP 2022. The results demonstrate the possibility of monitoring underwater noise pollution sources using a single acoustic vector sensor. The successful application of the DOA estimation techniques with acoustic vector sensors underscores their utility in passive acoustic monitoring by providing DOA data essential for operational and environmental noise mitigation and monitoring strategies in heavily trafficked marine areas. The advantages of using a single sensor package are that it allows noise sources to be monitored with relative ease of deployment and low cost. Additional sensors can be deployed to enhance detection capabilities and provide more accurate noise source localization. This paper compares different DOA estimation techniques and scenarios where they can be applied.