| Summary: | More than 30 years after the institution of the moratorium on commercial whaling by the International Whaling Commission, one of the biggest challenges in whale monitoring, e.g. Antarctic Blue Whales (ABW), is assessing populations to evaluate their recovery from persecution by mankind. To do this, passive acoustic monitoring is widely used all around the globe focusing on call detection of vocalizing whales. It has been demonstrated that ultra low frequency Ocean Bottom Seismometer hydrophones (OBSh) provide high quality data for whale monitoring, especially because of their large covering areas. Due to high energy seismic noise, previously developed autonomous methods based on signal cross-correlation are not suitable for this kind of approach. As a first step to ABW monitoring, a new method is used, based on a passive application of the Stochastic Matched Filter (SMF). This is an extension of the matched filter for degraded signals in colored noise. The SMF has been adjusted to perform ABW Z-call detection in noisy environments (seismic broadband noise, boat direct tonal spectrum, other whale calls…). In addition, its properties allow degraded calls (frequency dependent attenuation, multipath distortion) to be detected, covering large ranges. This new method is applied on a representative dataset -in terms of noise environment and signal variety- recorded by an OBSh in the [0-50]Hz frequency band, deployed at 4000 m depth on a mid-ocean ridge in the South West Indian ocean (RHUM-RUM project). The SMF’s robustness against noise is compared to the classical Matched Filter: the output’s SNR is maximized and the false alarm drastically decreases. It can then be arranged to provide call time of arrival as an input to a localization algorithm, even for remote calls.
|
|---|