Dynamic Marine Mammal Distribution Estimation Using Coupled Acoustic Propagation, Habitat Suitability and Soundscape Models
Abstract
The use of loud hydroacoustic sources, such as naval sonars or airguns as used in marine geophysical prospecting, has been increasingly criticized for its contingent negative effects on marine mammals and was even implied in some stranding events. To minimize possible risks, competent authorities now regularly request the implementation of mitigation measures, including the assessment of potential shifts in marine mammal distributions or density after naval exercises. This fundamentally depends on a good prior knowledge of the marine mammal distribution in US waters as well as in other open oceans. Most of the up-to-date marine mammal distribution estimations are based on aerial or ship-board surveys. While this standard approach delivers reliable numbers, it requires high effort and costs, and it is mainly limited to coastal waters, as dedicated cruises for marine mammal density estimation in the open ocean require even more resources. Developments of density estimation using passive acoustic monitoring (PAM) have attracted lots of attention in the last decade and the PAM approach can also be extended for spatial distribution estimation given a good understanding of the mammal s vocalization behavior. However, this type of approach is currently limited to areas close to acoustic recorders or ship tracks. Furthermore, except for a few well studied populations, most of the current marine mammal distribution estimations are static in time, density estimations using PAM are still based on the idea of counting single calls and translating an average call rate to an average number of animals. While this is a valid approach, it is limited to short ranges. To overcome these shortcomings limiting PAM methods, we propose to develop a Bayesian inference framework incorporating all available acoustic and visual marine mammal observations, and considering environmental and behavioral parameters to estimate marine mammal distribution with a large spatial coverage, up to the ocean basin scale. The proposed framework will consist of a maximum likelihood distribution estimator that fits a modelled acoustic environment (soundscape) to a measured one to obtain an optimized marine mammal distribution in the study area. With this approach, we will be able to account for uncertainties such as detection probability, acoustic propagation, behavioral parameters and anthropogenic noise. By utilizing the whole soundscape information and not only extracting calls, we will be able to extend the range of current and future PAM deployments to areas form which no single calls can be classified.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Sep 04, 2018
- Source ID
- N000141812811
Entities
People
- Daniel Paranhos Zitterbart
Organizations
- Office of Naval Research
- United States Navy
- Woods Hole Oceanographic Institution