Challenge #1

Marine Traffic Acoustic Footprint Assessment in the Celtic Sea

Context

The Subsea Soundscape project has been launched by the Offshore Renewable Energy (ORE) Catapult, in partnership with Celtic Sea Power and the University of Exeter, and is funded by The Crown Estate, as part of their Offshore Wind Evidence and Change programme, and pioneers a regional framework in the Celtic Sea to provide valuable insights into underwater noise conditions and marine mammal presence, informing maritime spatial planning and consenting decisions for floating wind development.

To accomplish this, 21 autonomous passive acoustic monitoring stations were deployed in October 2025 for two years, and were located to capture spatial heterogeneity in soundscape characteristics, encompassing varying depths, distances from shipping lanes, and proximity to different habitat types and proposed offshore wind lease areas. Planned analyses will characterise spatial and temporal patterns in vessel noise and biological sounds, and ambient sound levels across multiple frequency bands relevant to marine mammals, fish, and invertebrates

One of the overarching aims of the project is to assess the contribution of marine traffic noise to the overall background noise (as part of the overall subsea soundscape), as well as assessing its potential impact to marine life (as described in Figure 1).

Figure 1. Impacts of Underwater noise from ships from “Clean Arctic Alliance” 

Objectives

For this challenge, you will be tasked with developing methods for automatic detection and characterizstion of marine traffic acoustic footprints. A subsample of the 21x station broadband hydrophone data will be provided together with AIS data containing information of vessel types, locations and speeds

Stage 1: Vessel Detection

Participants are tasked with developing methods to identify the presence of marine traffic within passive acoustic data collected across the Celtic Sea monitoring network. Using the provided hydrophone recordings alongside AIS data, teams should explore approaches to detect when vessels are acoustically observable at a given station. Although solutions may use the AIS provided data for training and validation, they should not rely heavily on such examples due to its limited temporal resolution and availability for some stations. Sonar data from FPOD devices will also be provided.

Approaches may include both supervised and unsupervised clustering and classification techniques.

Stage 2: Acoustic Footprint Characterisation

Building on detection, participants should aim to characterise the acoustic signatures of different vessels and operating conditions. This may involve linking acoustic features to AIS-derived information such as vessel type, size, speed, or distance from the sensor. Teams are encouraged to explore methods for distinguishing between vessel classes, identifying key acoustic features, and examining how vessel activity influences the local soundscape.