F. Basanı et al. 
Marine Pollution Bulletin 198 (2024) 115891 
O0 50100 200 300 400 
+ "— a Kilometer 
12-5C-MOR 
11-SC-HEL 
14-NO-NTR 
— 
jescenS\ 
‚ &6 
10-SC-ARB 03-DK-HRF 
AT 
05-DE-ESAY 
VD 
18-DK-EDAa 
04-DE-FN3 
06-DEPNLAN 
07-NL-TEX . 
gr 
#5 
9 sek 
)1b-SE-HOF- 
JOMOPANS 
Distribution of 
ship classes per 
station 
{ ü ) 
#7 Fishing 
[—] Passenger 
7] Cargo 
IM Tanker 
I Other 
— EEZ 
\0 dB Broadband 
Dominance 
L10-2.5% 
112.5-7% 
E17-15% 
MN 15-25% 
25 -35% 
35 - 50% 
"50 - 60% 
60 - 75% 
75 - 90% 
E90 - 100% 
16/17-DK-TN1/TN- 
Interrev 
North Sea Region 
Jomopans 
“Anegan Regional Development Fun- 
= DRAADCARI I AM 
Es 
1 
02-DK-AN 
A 
x 
Vu 
Fig. 1. Positions of 18 measurement stations indicated by pie charts depicting the share of ship classes on total traffic within 35 km of each station; contour map in 
‘he background exhibits over what percentage of the year 2019 the predicted anthropogenic noise dominates the natural ambient noise by 20 dB or more (broadband 
LO Hz - 20 kHz), underlying sound map adopted from (de Jong et al., 2022); note that the 13-NO-LOV station is not depicted in the figure as it is outside the modelled 
region; its coordinates are given in Table 2. 
pressure fluctuations that result in low frequency signals in the mea- 
surements, called flow noise. Flow noise is not part of the ambient noise 
aut is a contaminating low frequency signal in sound recordings. Cur- 
‚ents can also induce noise in the measurements by for example causing 
vibration of the setup (e.g. cable strum), which is referred to as platform 
self-noise. The range of the turbulent pressure fluctuation covers the 
frequency range between 1 - —100 Hz (cf. Robinson et al., 2014). 
Currently, there is no standard method to remove the effect of this noise 
from the measurements, but there are practicable approaches to 
describe it and exclude the contaminated data from the analysis (van 
Seel et al., 2020). This is particularly important when it comes to 
assessment (Borsani et al., 2023) and when validating numerical model 
-esults (Putland et al., 2022), 
ıt is expected that at measuring stations with a strong tidal current 
regime, the flow-noise component would be clearly visible in the low 
frequency acoustic data (see Fig. 3). The current pattern of the North Sea 
is mainly dominated by tides (Sündermann and Pohlmann, 2011). 
However, the orbital motion of sea surface waves also creates currents 
‘hat decrease with depth (e.g. Soulsby and Smallman, 1986). The depth 
:o which orbital motion is still present equals roughly half the wave- 
length. With wave periods, ranging approximately from 6 to 8 s 
(Bonaduce et al., 2019), the average wavelength in the North Sea is 
expected to be between 56 m and 99 m (Holthuijsen, 2007). It is thus 
expected that orbital motion of wind driven waves can contribute to 
low noise at stations in waters shallower than 50 m. Currents can 
further be forced geostrophically or can be caused by other phenomena 
such as Langmuir circulation or Ekman spirals (e.g. Li et al., 2013; 
Sündermann and Pohlmann, 2011). 
This study is structured as follows. First, the acoustic and environ- 
mental data are introduced. Second, similarities and differences be- 
tween geographical positions are identified, and last correlations of 
sound levels with data on wind, shipping and currents are presented. 
2. Data 
For the presented analyses, acoustic and shipping data (AIS (Auto- 
matic Identification System) data purchased from the French company 
Quiet Oceans complemented by VMS (Vessel Monitoring System) data 
from Denmark, Germany, Norway and Sweden) from the JOMOPANS 
project, wind data from the E.U. Copernicus Marine Service, and ocean 
zurrent data from the BSH operational hydrodynamic model, were used. 
These sources are elaborated in the Sections 2.1 to 2.4 below. 
2.1. Acoustic data 
Ambient sound was measured at 18 stations across the North Sea and 
at one reference station outside the North Sea at the Lofoten Islands 
(Fig. 1 and Table 2). The scope of the JOMOPANS project was limited to 
waters deeper than 10 m, thus all stations were located relatively far 
from the coast. Measurements were conducted in 8 different countries 
by representative partner organisations. 
Depending on equipment availability, these partner organisations