Analysis of under-ice ambient noise characteristics of Gakkel Ridge in the Arctic

This paper presents an analysis of the under-ice acoustic data and environmental parameters measured over a three-month period from August 31 to November 28, 2021, within the area of the Gakkel Ridge in the Arctic. After "spikes" caused by micro-level events are removed, the distribution of the retained under-ice noise related to macro-level events can be described satisfactorily by a Gaussian distribution, as verified by Q-Q plots and kurtosis/skewness analysis. We use sliding window analysis to deal with the features of under-ice ambient noise and model the data by Gaussian interpolation. This shows that the ambient noise level over the low-frequency range (10-100 Hz) is comparatively flat at about 60 dB; with the frequency increases from 100 to 2560 Hz, the ANL decreased to about 40 dB. We then introduce canonical correlation analysis (CCA) to analyze the potential relation between environmental forcing and the under-ice noise level. The results of CCA indicate that the seawater parameters (including temperature, salinity, and sound velocity) close to the ice-water interface have the greatest influence on the under-ice noise level among all environmental parameters recorded in the air, sea-ice, and seawater. Additionally, the under-ice noise level forced by the environment does not exhibit any particular frequency dependence.