No abstract

No abstract

No abstract

An acoustic recorder was deployed near the Port of Sept-Îles, Quebec in the fall of 2020 and collected six months of data on a four-channel orthogonal array. The system, a JASCO C-lander, operated on a duty cycle consisting of 340s of data recorded at 32 kHz sampling rate, 1 min of data recorded at 256 kHz sampling rate followed by 500s of sleep. Data were stored on SD memory cards for post-retrieval analysis. Vessels were detected using narrowband tonals produced by their propulsion system and other rotating machinery and the sound pressure level (SPL) for each minute of data in the 40–315 Hz shipping frequency band was then computed. A 10min shoulder period before and after the detection was then searched for the highest 1 min SPL which was identified as closest point of approach (CPA) time for each acoustic contact. Vessel track data from the automatic identification system (AIS) were used to compute CPAs for vessels carrying an AIS transponder during the deployment period. A comparison of the two results was used to identify missed and false detections, and to assess the algorithm’s performance. Recommendations for implementing an improved detection approach will be discussed.