Transmission Loss of Fin Whale Vocalizations in the North Atlantic: Ground Truthing a Mathematically Derived Transmission Loss Model

Abstract

The United States’ demand for electricity is forecast to grow at a rate of 1.7%/year, representing a significant increase relative to the 0.1%/year demand increases from 2005 – 2020. Offshore wind represents a carbon-neutral, renewable energy source to bolster the nation’s energy profile, but the environmental impacts merit further study. Fin whales (Balaenoptera physalus) are an ecologically and culturally significant whale species whose habitat overlaps with offshore wind farms and leasing areas in the North Atlantic. Fin whales primarily sense their environment through sound, communicating to conspecifics through low-frequency calls (20 – 40 Hz). Interaction between fin whale communication and offshore wind construction/installation activities are poorly understood, especially in light of construction and installation of monopile foundations. These activities include hydraulic hammering, creates impulsive noise (short-duration, high intensity sound) that permeates the marine soundscape. Impulsive noise is broadband, covering a wide range of frequencies at high amplitudes, and may result in acoustic masking, stress, behavioral changes or temporary/permanent threshold shifts. This project, in collaboration with Project Wildlife and Offshore Wind, aims to inform development of a mathematical fin whale vocalization transmission loss (TL) model run by project partners at Pacific Northwest National Laboratory (PNNL). Results will be shared with PNNL to better the TL model and generate data contributing to the body of knowledge surrounding fin whale communication in the context of offshore wind development. For this analysis, satellite location data and call times derived from tagged whales were used to find points nearest to when an individual called. A geospatial analysis was used to find three PAM sensors nearest to these surfacing locations. Call frequency and call time from tags were compared to received signals in PAM recordings to identify if calls were detectable. Detected calls were further analyzed to find inband power and a signal-to-noise ratio. A total of 66 candidates calls were selected across 3 PAM sensors and assigned a score from 0 (best) to 3 (worst) based on dissimilarity to tagged calls. Calls receiving a score of 0 were considered positive identifications. Of these 66 candidate calls, 5 positive identifications were detected. Background noise in PAM data and misalignments in data acquisition reduced analytical confidence, so this report concludes with suggestions for future transmission loss work such as the integration of more tag and PAM data, considerations for multiple methods of detection, and the use of environmental covariates that may assist in understanding how fin whale calls propagate through the water column.