Marine mammal choruses in the Southern Ocean – An inverse theory approach to estimate animal distribution from acoustic power spectra

Mary Woodcock Kroble
Tuesday 10 November 2015
Date: 10 February 2016
Time: 2:00 pm - 3:00 pm

Speaker: Sebastian Menze (CREEM)


This talk will at first present recordings of the Southern Ocean acoustic environment and then discuss an inverse theory approach to estimate the distribution of vocalizing animals.

Over a 3 year period, two autonomous recorders were moored along the Greenwich meridian. Sound levels were strongly affected by the annual variation of the sea ice cover, which decouples local wind speed and sound levels during austral winter. With increasing sea ice concentration, area and thickness sound levels decreased and the contribution of distant sound sources increased. Marine mammals contributed considerably to the ambient sound, distinct choruses were observed as peaks in the acoustic power spectra and associated to blue whale, fin whale, minke whale and leopard seal vocalizations. The temporal and latitudinal variation in chorus levels suggests annual patterns in migration or behaviour.

Inspired by these observations, a method to utilize the information present in the spatial variation of chorus levels is proposed. As case study, a set of simulated fin whale chorus observations were used to estimate the distribution of vocalizing animals on a basin wide scale. This is an extremely under-determined inverse problem. The estimation method is based on geophysical inverse theory and uses simulated annealing to determine the maximum likelihood distribution of sound sources (vocalizing whales) on a geodesic grid. This includes calculating a transmission loss matrix connecting all grid nodes and recorders, using an arbitrary sound propagation model. Two models were successfully implemented: geometrical spreading and the ray trace model BELLHOP. The estimation method was tested under different scenarios. The results indicated that an imprecise transmission loss matrix is tolerated well and that the accuracy of the method depends mainly on the number and distribution of recorders. Further development of the proposed method could admit monitoring of marine mammal distribution on an unprecedented temporal and spatial scale and in inaccessible regions.

Share this story