2-2. Underwater soundscape analysis
All data processing, analysis, and visualization were conducted using MATLAB 2021a (Mathworks, Inc.). The underwater soundscape was analyzed in the time-averaged spectrum levels and spectrograms. The spectrum levels were estimated by the power spectral density (PSD) in units of dB re 1 \(\mu\)Pa2/Hz (Han et al., 2021), which was computed from 1 minute of acoustic data with a 1-sec Hanning window, 50% overlap and a fast Fourier transform length of 96,000 points. The spectral probability density, which represents percentile spectra (Merchant et al., 2013), was then calculated to evaluate temporal variations in the sound pressure levels and the impact of leopard seal vocalizations on the underwater soundscape in the measurement region. The spectrogram was represented by stacking the intensity-averaged PSD in 10-minute intervals, as that depicted by higher temporal resolution was ineffective to show acoustic events over time.