Application of Distributed Acoustic Sensing to Detect and Identify of Vessels and Natural Events in the Northeastern Offshore Region of Taiwan

This study aims to develop a system for the identification of vessels, seismic events, and volcanic activity through analysis of the spatiotemporal characteristics of wavefields recorded by distributed acoustic sensing (DAS) using a submarine fiber-optic cable. DAS provides unprecedented spatial coverage and resolution, making it highly suitable for monitoring dense wavefield variations and anthropogenic activities, whereas traditional seismometers remain indispensable for quantitative seismic analysis and low-frequency observations. In this study, continuous DAS records acquired from a submarine fiber-optic cable located in the northeastern offshore region of Taiwan near Guishan Island, an active volcano. This region experiences frequent seismic activity due to the northwestward subduction of the Philippine Sea Plate beneath the Eurasian Plate. In addition, the passage of the Kuroshio Current, a warm ocean current, brings abundant fish resources, resulting in frequent activities of fishing vessels and whale-watching boats. Event detection is first carried out using the recursive short-time-average/long-time-average (STA/LTA) method which uses two time windows with different durations and computes the average signal amplitude within each window. When a signal arrives, the average amplitude within a short time window changes rapidly, thereby increasing the ratio of the short-time average to the long-time average. An event is detected when this ratio exceeds a predefined threshold and manual secondary inspected. However, low signal-to-noise ratios (SNR) can significantly reduce the sensitivity of STA/LTA-based detection, leading to missed events. To overcome this problem, signal processing adjustments were applied to enhance detection performance. To validate the detection performance, the detected ship-related events were compared with records from the Automatic Identification System (AIS), while earthquake events identified from the DAS data were compared with the earthquake catalog of Taiwan Seismological and Geophysical Data Management System (GDMS). Subsequently, a regression analysis of catalog magnitudes against hypocentral distance and maximum DAS-recorded amplitude was applied to determine the minimum detectable earthquake magnitude. The proposed framework demonstrates the potential of DAS as a complementary tool for offshore geophysical and maritime monitoring, providing a basis for future studies on vessel tracking, seafloor topography, and earthquake monitoring.