UC San Diego

The Arctic Ocean has experienced wide-spread decreases in sea ice concentrations that may impact various marine ecosystems. This study analyzes yearlong ocean acoustic recordings from north of Barrow, Alaska, to provide baseline measurements prior to possible increases in anthropogenic activities. In September 2006, two autonomous High-frequency Acoustic Recording Packages (HARPs) were deployed to the seafloor (250m), where sound was continuously recorded by hydrophones for nine months. Ice conditions during the recordings included open water, pack ice formation, shore-fast canopies, and thermal breakup, providing a wide range of Arctic Ocean acoustic measurements. Spectral-averaging was used to determine received sound-pressure levels. Across the low-frequency band, fall was the noisiest season, reaching 87dB re [mu]Pa between 20-60Hz, while 10% of October was exposed to noise above 130dB re [mu]Pa at 10Hz and 112dB re [mu]Pa between 20-30Hz; seismic airguns were present from September to November. Acoustic data was compared with sea ice concentration and wind speed; during summer and fall, sound-pressure spectrum levels correlate directly with high wind speeds, typically indicative of low-pressure atmospheric events. Throughout winter and spring, strong winds and thermal fracturing in sea ice opens leads, resulting in correlations with spectral energy-peaks. Bioacoustic recordings of cetaceans and pinnipeds were analyzed using long-term spectral-averages to determine presence or absence on an hourly basis. Combined with ancillary measurements, long-term acoustic monitoring is an effective tool for observing changing levels of ambient sound related to sea ice dynamics, environmental noise- generating mechanisms, and anthropogenic noise, while simultaneously detecting marine mammals