F.Félix Zarzuela ffelixzarzu(at)hotmail.com
Institute of Marine Sciences – OKEANOS, University of the Azores & Institute of
Marine Research – IMAR, Horta, Portugal,
S.M. Jesussjesus(at)ualg.pt
LarSys, Universidade do Algarve, Campus de Gambelas, PT-8005-139 Faro, Portugal,
P.J. Wensveen pjw(at)hi.is
Westman Islands Research Centre, Institute of Research Centres, University of Iceland,
Vestmannaeyjar, Iceland
M. Romagosa and M.A. Silva
Institute of Marine Sciences – OKEANOS, University of the Azores & Institute of
Marine Research – IMAR, Horta, Portugal.
Comments: download presentation pdf.
Ref.: in UACE'2025, June 2025.
Abstract:
Over recent decades, human activities have significantly increased ocean noise levels,
affecting marine ecosystems. While wind-generated sound is the dominant natural noise
source for most ocean locations, commercial shipping is the primary anthropogenic contributor,
potentially interfering with marine animals' communication, causing stress and disrupting
vital behaviours. Here, we present a novel modelling framework to estimate underwater
sound levels across a large North Atlantic area (~1.85 million km²). This approach
combines estimations of shipping noise distribution—derived from sound propagation
modelling—with an empirical wind model, as proposed by Hildebrand. Open-source shipping
data from the Global Maritime Traffic Density Service, along with environmental variables
(bathymetry from GEBCO, generic bottom composition, and salinity and temperature from
Copernicus), serve as key inputs. The BELLHOP beam-tracing model was used as the acoustic
propagation engine, and monthly shipping densities (ships per hour per 1 km2) were
combined with source levels from ships (cargo, tanker, fishing and passenger from
JOMOPANS-ECHO model). The framework was applied with cetaceans as an initial target
species but is broadly applicable to marine life. We generated 108 high-resolution
(2.5 km²) underwater noise maps for the Azores’ EEZ during 2023, covering combinations
of months, central ⅓ octave frequency bands (64 Hz, 126 Hz and 15849 Hz) and depth
(100 m, 500 m and 1000 m). Model outputs were validated against acoustic recordings
from a calibrated hydrophone deployed at 1385 m depth, showing model underestimations
ranging from 16.18 dB to 7.65 dB. Further validation is needed to assess model
performance across the water column. However, this approach provides a scalable,
open-source data method for ocean noise estimation which can contribute to risk
assessment across deep offshore areas, supporting conservation efforts and
environmental management.