Is it too loud? Approximations for underwater noise impact assessments and reporting for highresolution seismic surveys

Introduction

As offshore exploration intensifies in ecologically sensitive regions like the North and Baltic Seas, the lack of standardised, accessible tools for underwater noise impact assessment threatens both regulatory compliance and marine conservation. Anthropogenic underwater noise emissions pose a well-documented threat to marine ecosystems (Green et al., 1994; Schack et al., 2019; Southall et al., 2007, 2019). As a result, environmental impact assessments are now a regulatory requirement for seismic survey planning, aimed at preventing or mitigating adverse effects on marine flora and fauna. This requirement is enshrined in EU legislation, notably Commission Decision (EU) 2017/848 of 17 May 2017 (European Commission, 2017). The decision mandates reporting to underwater noise registries for seismic surveys, such as Germany’s MarineEars system (marinears.bsh.de). These registries are becoming increasingly critical as coastal waters experience heightened noise levels, e.g., from offshore wind farm development. Best available techniques and best environmental practices for seismic surveys are established in IAU (2001) and OSPAR (2016). Also industry-led initiatives, such as the Sound & Marine Life Joint Industry Programme (SML JIP), have advanced source characterisation efforts (Prior et al., 2021). These characterisations inform impact assessment and mitigation strategies (see, e.g., Breitzke et al., 2008; Laws, 2012, 2013). Recent advances include marine vibratory seismic sources that emit non-impulsive sounds with controlled frequency spectra, minimising overlap with the sensitive hearing ranges of most marine mammal (Duncan et al., 2017; Matthews et al., 2021; Austin et al., 2026).

Noise assessments rely on quantitative metrics to evaluate potential impacts, including permanent hearing damage (PTS), temporary hearing impairment (TTS), behavioural disturbance, and audibility with severity decreasing with distance (Figure 1). For PTS and TTS, dual exposure criteria are established as defined by Southall et al. (2007, 2019): Unweighted zero-to-peak Sound Pressure Level ($SP{L}_{uw,0-peak}$; Unit: dB re 1 µPa) reflecting peak pressure as well as cumulative weighted Sound Exposure Level ($SE{L}_{W,cum}$; Unit: dB re 1 µPa2s) integrating over time and accounting for species-specific hearing sensitivity. Southall et al. (2007, 2019) proposed a 24-hour integration period for $SE{L}_{W,cum}$. However, NMFS (2024) revised the period, stating that sound exposures below the Effective Quiet threshold do not contribute to PTS or TTS, regardless of duration or cumulative exposure. For behavioural disturbance, the Root Mean Square Sound Pressure Level ($SP{L}_{RMS}$; Unit: dB re 1µPa) with an averaging period of 125 ms is recommended, as this timescale approximates the integration time of the mammalian auditory system (Tougaard, 2014; Tougaard et al., 2015; Tougaard, 2016; DEA, 2023; NMFS, 2024). $SP{L}_{RMS}$ is used both weighted and unweighted for the hearing capabilities, with thresholds varying across species and exposure scenarios (see Table 1). A sound is considered audible if its weighted zero-to-peak level ($SP{L}_{w,0-peak}$; Unit: dB re 1 µPa) exceeds the ambient sound level.

In the North Sea and Baltic Sea, harbour porpoises (Phocoena phocoena) serve as a key case study for underwater noise impact assessments. Harbour porpoises are designated as a priority noise-sensitive species by Schack et al. (2019), owing to their high hearing sensitivity, vulnerability to noise impacts, and conservation status. Their noise sensitivity has been extensively studied (Kastelein, Helder-Hoek, & Van de Voorde, 2017; Kastelein, Helder-Hoek, Van de Voorde, et al., 2017; Kastelein et al., 2018, 2020). Following Southall et al. (2007, 2019), harbour porpoises are classified as Very High Frequency (VHF) cetaceans, and PTS/TTS thresholds are derived accordingly (Table 1). Data on the onset of behavioural responses in VHF cetaceans remain limited. Based on personal communication with Jakob Tougaard (2021), a behavioural response threshold of $SP{L}_{W,RMS}$=95–110 dB re 1 µPa is adopted for VHF cetaceans. This threshold aligns with the 103 dB re 1 µPa limit for offshore wind construction in Danish waters (DEA, 2023). Alternatively, an unweighted threshold of $SP{L}_{UW,RMS}$=160 dB re 1 µPa is reported in NMFS (2024). However, recent experimental evidence presents a complex picture. Small airguns induce only low-level TTS (Kastelein, Helder-Hoek, Van de Voorde, et al., 2017). Harbour porpoises may self-mitigate noise exposure by increasing swimming speed (Kastelein, Helder-Hoek, & Van de Voorde, 2017; Kastelein et al., 2018), adjusting head orientation, and modulating hearing thresholds via physiological mechanisms (Kastelein et al., 2020). Harbour porpoises exhibit highly directional hearing, particularly at high frequencies (Schack et al., 2019).

Impact assessments prior to seismic and hydroacoustic surveys (e.g., Single or Multi Beam Sub Bottom, and Side Scan Echo Sounders) are essential for the safe execution of these projects and for the sustainable management of marine resources. Post-survey reporting to underwater noise registries enables authorities to coordinate activities temporally and spatially, and to monitor and manage noise emissions effectively. However, no freely available, standardised software package currently exists to support report generation. In many regions, detailed reporting has not been required historically. With increasing regulatory mandates in specific areas, significant additional effort is now required to establish data collection workflows and compile the necessary input data.

To address this gap, we develop a modular, open-source workflow for consistent and auditable noise impact assessments of high-resolution seismic and hydroacoustic surveys. The workflow is implemented in a freely available Python package, enabling community use, transparency, and continuous improvement. The tool supports impact assessments for single- and multi-beam echosounders based on Lurton (2016) and facilitates automated reporting to the MarineEars noise registry (BSH, 2021). Unlike proprietary approaches, our workflow is open-source, reproducible, and aligned with current regulatory standards, enabling transparent, scalable assessments across diverse survey types.

Material and methods

Calculation of the sound metrics

The basis of the different sound metrics is the pressure signal of the seismic source $s(t,d=1m)$ which is measured in Pascal (Pa), recorded as a time series $t$, and corrected to a reference distance $d=1m$. As shown in Figure 1, an animal’s reception of the signal $p$ depends on the distance $d$ to account for transmission loss $G(d)$ by geometric spreading, the absorption $A(f,d)$ of the signal in the water as a function of the instantaneous frequency $f$ and $d$, and the source specific directivity $D(f,x,y,z)$ as a function of $f$ and the spatial coordinates $x,y,z$ relative to the source. For the weighted sound metrics, an animal group specific auditory weighting function $V(f)$ is included to account for the specific hearing capabilities. The described factors are formulated as frequency filter functions denoted by capital letters in the Formulae 1 and 2. A zero phase filter is assumed. As advised by Smith (2003) filtering is performed in the time domain by convolution after inverse Fast Fourier Transform $F^{-1}$, shifting, truncating and windowing with a Tukey window.

$$p(t,d,x,y,z)=s(t,d=1m)\ast F^{-1}\{G(d)\cdot A(f,d)\cdot D(f,x,y,z)\}\text{\hspace{1em}(1)}$$

$$p_W(t,d,x,y,z)=s(t,d=1m)\ast F^{-1}\{V(f)\cdot G(d)\cdot A(f,d)\cdot D(f,x,y,z)\}\text{\hspace{1em}(2)}$$

The original weighting function $W(f)$ by Southall et al. (2019) with its parameters $C,a,b,f_1$ and $f_2$ are defined by:

$$W(f)=C+10{\log }_{10}\left\{\frac{{\left(f/f_1\right)}^{2a}}{{\left[1+{\left(f/f_1\right)}^2\right]}^a{\left[1+{\left(f/f_2\right)}^2\right]}^b}\right\}\text{\hspace{1em}(3)}$$

As $W(f)$ returns decibel (dB) amplitudes, the weighting function is transformed to the weighting function $V(f)$:

$$V(f)=10^{\left[W(f)/20\right]}\text{\hspace{1em}(4)}$$

An approximation for chemical absorption in sea water $[a(f)]=1dB/km$ is e.g. given by Ainslie and McColm (1998). This approximation needs to be converted similarly to Formula 4:

$$A(f,d)=10^{\left\{\left[a(f)\cdot d\right]/\left(1000\cdot 20\right)\right\}}\text{\hspace{1em}(5)}$$

A range of different options are available for the geometric spreading approximation. For deep water a spherical spreading approximation can be used:

$$g_{sph}(d)=20{\log }_{10}\left(\frac{1}{d}\right)\text{\hspace{1em}(6)}$$

For shallow water <200 m (Müller & Zerbs, 2013) assuming no losses at the sea to air and bottom interface, cylindrical spreading is a maximum approximation:

$$g_{cyl}(d)=10{\log }_{10}\left(\frac{1}{d}\right)\text{\hspace{1em}(7)}$$

More realistic approximations for geometric spreading in shallow water are given by Elmer et al. (2007) or Duncan and Parsons (2011). Alternatively, models with the expected range of seafloor substrates and considering frequency-dependent effects can be run e.g. with the KRAKEN normal mode model (Acoustic Toolbox from HLS Research with the AcTUP V2.2L interface of Alec Duncan, 2021). Thereby, the shallow water model by Duncan and Parsons (2011) is a lower limit estimation for KRAKEN models. In this model, the transmission loss considers the water depth $h$ and is determined by Formula 8:

$$g(d)=10{\log }_{10}(d)+10{\log }_{10}\left(\frac{h}{2}\right)-B\text{\hspace{1em}(8)}$$

Choosing a threshold of $B=6.6$ dB in this Formula 8, the approximated spreading loss is a lower limit estimation in 99% of the modelled cases (probability of exceedance is 1% according to Duncan & Parsons, 2011). Thus, the real spreading loss is with a high probability more pronounced and the received sound levels are thus smaller than approximated. Similar to Formula 4 and 5, $[g(d)]=1$ dB need to be converted:

$$G(d)=10^{\left[g(d)/20\right]}\text{\hspace{1em}(9)}$$

The directivity of a pressure source is a design and operation dependant parameter. In the case of Single or Multi Beam Echosounders, the transducer array dimensions and the signal frequency are the critical design parameters (Lurton, 2016). Thus, the orientation, distance and receiver depth $z_r$ affect the received signal level. For single seismic sources, the emitted signal characteristics and the source tow depth $z_s$ are most relevant (Parkes & Hatton, 1986). Thereby, the wavenumber $k=\frac{2\pi }{{\lambda }_{dom}}$ can be used to describe the signal characteristics based on the dominant wavelength ${\lambda }_{dom}=\frac{v}{f_{dom}}$, the dominant frequency $f_{dom}$ and the water velocity $v$. The phase reversed reflection from the sea surface interferes with the primary signal which is referred to as the ghost reflection or Lloyd’s Mirror effect, and causes the source to be directive (Carey, 2009). Following Carey (2009), and Sertlek & Ainslie (2015) the field intensity at a receiver $I_r(d,z_r,z_s)$ caused by a seismic source with intensity $I_s$ for flat sea surfaces with a mirror like reflection is given by:

$$I_r(d,z_r,z_s)=I_s\cdot \frac{1}{d^2}\cdot 4{\sin }^2\left(\frac{k{z}_r{z}_s}{d^2}\right)\text{\hspace{1em}(10)}$$

Excluding the constant factors $(4I_s)$ as well as the geometric spreading term ($\frac{1}{d^2}$), an expression for the range $d$ and receiver/ source depth $(z_r,z_s)$ dependant directivity $D(d,z_r,z_s)$ is determined by:

$$D(d,z_r,z_s)={\sin }^2\left(\frac{k{z}_r{z}_s}{d^2}\right)\text{\hspace{1em}(11)}$$

In the notation of Formula 11, the frequency dependant directivity is neglected. This simplification holds true, e.g. for chirps, airguns, sleeve-guns, or water-guns as Verbeek & McGee (1995) describe those source signals to be effectively non-directional. Additional directivity is introduced if multiple spaced apart seismic sources are fired in arrays, e.g. in the form of multi-electrode sparkers (Verbeek & McGee, 1995; Giordano et al., 2020), or air-gun arrays (Parkes & Hatton, 1986).

After the pressure signal at a given position and for a specific functional hearing group has been determined (see formulae 1-11), the sound exposure metrices are calculated to assess the sound impact (see Table 1). According to Southall et al. (2019), the metric $SP{L}_{UW,0-peak}$ for the onset of mortalilty, PTS and TTS depends only on the pressure signal $p(t,d)$ and the reference pressure for water $p_0=10^{-6}Pa$. Thus, $SP{L}_{UW,0-peak}$ is defined as:

$$SP{L}_{UW,0-peak}=20{\log }_{10}\left[\frac{\max \left|p(t,d)\right|}{p_0}\right]\text{\hspace{1em}(12)}$$

Similarly, the weighted metric $SP{L}_{W,RMS}$ to determine behavioural impact is calculated according to Tougaard (2016):

$$SP{L}_{W,RMS}=20{\log }_{10}\left[\frac{p_{RMS}(T=125ms)}{p_0}\right]=20{\log }_{10}\left[\frac{\sqrt{\frac{{\sum }_{t=0}^{t=125ms}{p}_W^2(t,d)\Delta t}{125ms}}}{p_0}\right]\text{\hspace{1em}(13)}$$

For $SP{L}_{UW,RMS}$, the unweighted received signal $p(t,d)$ is used in formula 13.

As a second measure for the onset of PTS and TTS, the $SE{L}_{W,cum}$ is calculated, which is a cumulated measure for a theoretic profile and a stationary animal as proposed in Tougaard (2016) and shown in Figure 2. The assumption of a stationary animal is a conservative estimate, as most animals are likely to be moving during the survey and thus experience less underwater noise. For the calculation of the $SE{L}_{W,cum}$, the signal spreading, absorption, and weighting with the frequency filter of the functional hearing group is applied to the signal as a first step. In a second step, the signals in a integration interval $T$ determined by the Effective Quiet criterium are squared and summed. Finally, the summed signal is cumulated for all shot points of the theoretic profile:

$$SE{L}_W=10{\log }_{10}\left[\frac{{\sum }_{t=0}^T{p}_W^2(t,d)}{p_0^{2}s}\right]\text{\hspace{1em}(14)}$$

Data basis for sound metric calculation

Accurate source characterisation is the foundation of any reliable noise impact assessment. Source signals must be recorded using calibrated hydrophones and analog-to-digital converters to ensure measurement accuracy. Recording duration must capture the full primary signal, excluding reflections or other secondary effects that may distort the source signature. The sampling frequency must satisfy the Nyquist criterion, exceeding twice the highest frequency of interest, typically 200 kHz for Very High Frequency (VHF, see Figure 3) cetaceans. Figure 3 displays the source signal of a custom Sercel Micro-GI airgun of the University of Bremen, which is compiled from measurements across varied distances, azimuths, and depths. The airgun is towed at a depth of 0.7 m and fires compressed air at 130 bar from two 0.1 L (6 in³) chambers. A 15 ms delay between the first and second chamber firing suppresses bubble oscillation. The source output at 1 m is estimated at $SP{L}_{UW,0-peak}=226\pm 2$ dB re 1 µPa. The measured $SP{L}_{UW,0-peak}$ is consistent with high-resolution source characterisations reported by Crocker & Fratantonio (2016, see Table 12 therein).

For true point sources, a single far-field measurement in the water column is often sufficient, though spatially distributed recordings help to minimise biases in source waveform estimation. However, most seismic sources are directive, necessitating measurements across a wide range of relative positions in the far field to capture directional variability. For single sources, directivity is accounted for using Equation 11. For more complex configurations, such as gun arrays, source signature modelling tools like AGORA (Sertlek & Ainslie, 2015; Ainslie et al., 2016; Blacquière & Sertlek, 2019), GUNDALF, or NUCLEUS+ are recommended. Lurton (2016) provides simplified formulas for modelling directionally radiated sound fields and approximating key sound metrics when only the unweighted zero-peak sound pressure level ($SP{L}_{UW,0-peak}$) and central frequency are known.

Reporting of underwater noise emissions

Although the noise registry for seismic surveys in the German Exclusive Economic Zone (EEZ) is still under development (source), data requirements have been made available to survey operators. Reporting to MarineEars requires a single, standardised HDF5 file, containing survey metadata, detailed source information, and acquired transect data. Currently, no established software package exists to generate this file. However, HDF5 files can be created using Python, and a suite of scripts and functions is provided in the WhaleWatchWorkhorse package. Both Danish and German authorities require noise emission reports to be spatially and temporally resolved. These reports must reference ICES statistical rectangles, and the package includes tools to generate corresponding tables from survey protocols and navigation logs.

Results and discussion

Workflow underwater noise impact assessment and mitigation

Following successful deployment during research cruises in Danish and German waters, the following workflow for high-resolution seismic noise impact assessment has been established. Given that permit processes can take up to six months, early initiation of impact assessments is critical.

1. Identification of noise sources (seismic and hydroacoustic sources): Wavelet records for seismic sources or centre frequency information for echosounders, and $SP{L}_{UW,0-peak}$ at a reference distance of 1 m are a minimum requirement. For robust assessments, multiple calibrated source records across varying incidence angles are recommended to capture directional variability and to avoid bias.
2. Characterisation of regions of interest: Environmental parameters have to be gathered including the composition of the seawater to determine the absorption, typical water depths and seafloor substrates to approximate the geometrical spreading.
3. Identification of possibly affected animal species: Marine mammals are grouped into functional hearing categories (e.g., harbour porpoises to very high frequency cetaceans). Annual periods of biological significance (e.g., breeding, feeding) are considered, and exposure thresholds (PTS, TTS, behavioural disturbance) are compiled from authoritative sources.
4. Calculation of the sound metrics: To ensure conservative, precautionary estimates, impact zones are calculated under worst-case propagation conditions using the shallowest water depths and most reflective, acoustically hardest seafloor substrates. This approach yields maximum sound exposure levels at the seafloor, which is particularly relevant for species like harbour porpoises, known to forage near the seabed.
5. Definition of safety distances: Derived impact zones are used to establish exclusion zones, generate buffers around specially protected areas, or define the extent of affected regions.

During the surveys and as e.g. documented in Weir & Dolman (2007), OSPAR (2016) and JNCC (2017) mitigation measures are implemented including pre-shoot watch keeping, watch keeping of marine mammal observers, Passive Acoustic Monitoring, shut-down for animals within exclusion zones and soft start procedures. Also, procedures for shooting breaks, line changes, night-time operations or sensitive areas can be defined. The use of mitigation guns or a new soft start after extended breaks are encouraged. After a survey, the relevant data has to be submitted to the national noise registries.

Calculation basics: Auditory weighting, absorption, geometric spreading, and drectivity

The choice of auditory weighting function and geometric spreading approximation significantly influences the outcome of underwater noise impact assessments. The hearing threshold (Figure 3C) and the corresponding frequency-domain filter function (Figure 4B) form a complementary pair, both derived from Southall et al. (2019). The filter enables accurate weighting of the source signal in the frequency domain, ensuring that exposure metrics reflect species-specific hearing sensitivity. For Very High Frequency (VHF) cetaceans, the frequency range of the most sensitive hearing is between 12-140 kHz (compare Figure 3C, Figure 4B ; Table 5 in Southall et al., 2019).

Absorption plays a critical role in high-frequency propagation. As established by Ainslie and McColm (1998), viscous absorption dominates above 100 kHz, while lower frequencies are primarily attenuated by chemical relaxation effects primarily due to boric acid. In the Baltic Sea, this results in significant attenuation only above ~10 kHz (Figure 4A), consistent with oceanographic setting of the region. This pattern is expected to hold across oceanic environments, though low-frequency absorption increases with higher pH, lower salinity, and lower temperature (Ainslie & McColm, 1998).

These findings underscore the importance of using region-specific absorption models and species-specific weighting functions. In shallow waters, the choice of geometric spreading approximation has a profound influence on noise impact assessments. While spherical spreading (Formula 6) applies in deep water, shallow environments exhibit complex interference between the direct wave and reflections from the seafloor and water surface. In the absence of energy loss, this would lead to cylindrical spreading (Formula 7). However, energy is lost both into the seafloor and into the air in reality, reducing the effective spreading rate. Shallow water approximations account for this loss. The Elmer et al. (2007) model is based on empirical data, while the model by Duncan and Parsons (2011; Formula 8) assumes a basaltic seafloor. In the North and Baltic Seas, where soft sediments (e.g., clay, sand) dominate, this assumption leads to underestimation of spreading loss. As shown in Figure 4C, the KRAKEN normal-mode model with a clay substrate predicts transmission loss ~5 dB lower than the Duncan and Parsons (2011) approximation. In contrast, the sandy substrate case closely aligns with the shallow water approximation, suggesting that the Duncan and Parsons model may serve as a conservative, practical choice for sandy environments. The difference between spherical and cylindrical spreading is substantial with approximately 30 dB at 250 m (Figure 4C) underscoring the critical importance of model selection. For realistic assessments, full-wave models (e.g., KRAKEN) that account for bathymetry, water column stratification, and seafloor composition are essential.

For the Micro-GI airgun, directivity is accounted for by using Formula 11, with the maximum water depth used to represent a maximum noise exposure scenario. Figure 5A shows a dipole-like radiation pattern, with signal intensity decreasing toward the surface and vanishing at the water surface. Consequently, maximum sound exposure occurs at the seafloor. This conservative assumption is also justified by the foraging behaviour of species like harbour porpoises (Phocoena phocoena), which forage near the seabed.

Impact of a high-resolution seismic survey

The calculation of sound exposure metrics across a range of distances (Figure 6) enables the delineation of distinct impact zones, which is a critical step in risk-informed mitigation planning (Figure 1). For the Micro-GI airgun, with a source level of $SP{L}_{UW,0-peak}$=226±2 dB re 1 µPa, the mortality zone is not reached at any distance, as this level is below the threshold for VHF cetaceans (Table 1). This finding is consistent with the BSH (2021) classification of the Micro-GI as a ‘very low’ energy source.

While the Micro-GI airgun source poses no risk of mortality, its signal remains capable of inducing hearing impairment and behavioural disturbance within tens of metres. Permanent (PTS) and temporary (TTS) hearing impairment are predicted within 15 m and 20 m, respectively. At distances beyond 700 m, the signal falls below the Effective Quiet threshold, meaning no cumulative exposure is expected. Thus, a track length of 1400 m is used for the cumulative weighted $SE{L}_{W,cum}$ calculation. The safety distances are orders of magnitude smaller than those typically used in industrial surveys as compiled by Weir and Dolman (2007) and confirmed by Laws (2012, 2013). The safety distances are further reduced when using realistic propagation models. The Duncan and Parsons (2011) shallow water approximation, while conservative, assumes a basaltic seafloor. In contrast, the KRAKEN normal-mode model with a clay substrate predicts stronger geometric spreading losses reducing the impact zone even further. Additionally, surface wave effects, which are not modelled here, would further increase transmission loss, making the actual impact zone smaller than estimated. Particularly signals of high-frequency sources and for small incidence angles are attenuated by a rough sea surface (Asgedom et al., 2017; Orji et al., 2013).

The weighted, cumulative metric $SE{L}_{W,cum}$ yields smaller safety distances than the metric $SP{L}_{UW,0-peak}$, due to the auditory weighting function (Figure 4A), which downweights frequencies outside the hearing range of the functional hearing group under consideration. The Micro-GI airgun emits energy primarily below 1 kHz, while VHF cetaceans are most sensitive above 10 kHz (Figure 4A). As a result, there is minimal overlap between the source spectrum and the hearing capabilities of these species (Figure 3B, C). Following the approach of the dual exposure criterion (Southall et al., 2007, 2019), the larger safety distances determined from $SP{L}_{UW,0-peak}$ are considered.

Behavioural disturbance is predicted to occur if the weighted metric $SP{L}_{W,RMS}$ exceeds 95-110 dB re 1µPa, which is predicted to be within a 22 m radius around the source. The unweighted threshold ($SP{L}_{UW,0-peak}$ =160 dB re 1µPa; NMFS, 2024) would allow a larger zone, but the weighted metric is more conservative and biologically relevant. Beyond 700 m, the signal is below the Effective Quiet threshold and thus too silent to contribute to the cumulative sound exposure (NMFS, 2024). Further away than 1250 m, the signal falls below the ambient noise floor in the Baltic Sea (Mustonen et al., 2019), making masking and other indirect impacts unlikely.

The workflow presented here enables a transparent, reproducible, and science-based assessment of underwater noise impacts. By quantifying impact zones at the scale of tens of metres, it supports the design of mitigation measures that are proportionate to actual risk while avoiding unnecessary survey limitations on the one hand and ensuring ecological protection on the other hand. These findings underscore that high-resolution seismic sources can have a comparatively small environmental footprint when conducted carefully and responsibly.

Conclusion

This study demonstrates a transparent, reproducible, and science-based workflow for underwater noise impact assessment for high-resolution seismic surveys. This workflow has been applied to the Sercel Micro-GI airgun and harbour porpoises in the Baltic Sea. Our results show that, despite significant sound levels, the environmental impact of this source is comparatively small, with impact zones confined to tens of metres.

Thereby, the unweighted zero-to-peak sound pressure level $SP{L}_{UW,0-peak}$ is the most relevant metric, determining safety distances of approximately 15 m for permanent, 20 m for temporary hearing impairment and 30 m for behavioural disturbance. The cumulative weighted sound exposure level $SE{L}_{W,cum}$ is less restrictive, and the signal falls below the Effective Quiet threshold beyond 700 m, meaning no cumulative exposure is expected. At distances greater than 1250 m, the signal is below the ambient noise floor, making masking and other indirect effects unlikely.

Crucially, the impact is not only due to low source levels compared to large airguns or airgun arrays in the energy industry, but also to a spectral mismatch: the Micro-GI emits energy primarily below 1 kHz, while harbour porpoises are most sensitive above 10 kHz. This mismatch, combined with conservative assumptions (e.g., acoustically hard seafloor, no surface wave effects), ensures overestimated safety distances for a precautionary assessment.

The workflow, implemented in the open-source Python package WhaleWatchWorkhorse, has already enabled impact assessments for past seismic cruises. It integrates source characterisation, propagation modelling, species-specific thresholds, and automated reporting and thus closes a critical gap in regulatory compliance and transparency. With the package release to the public domain, transparency and a broad peer review of the implementation is enabled. Regulatory changes and new scientific insights can be incorporated, ensuring long-term relevance and adaptability. Future users can use and adopt the package to their use cases. Thereby, a critical assessment of the used parameters for every specific case study is a key requirement.

While the impacts of high-resolution seismic surveys are small, opportunities remain to further reduce noise impacts. We recommend that source selection criteria include:

• Spectra outside the hearing range of local noise-sensitive species,
• Minimum source level (quietest possible source and/or setting) for survey target,
• Directive sources to restrict sound emission locally,
• Survey timing outside periods of biological significance.

Future research should focus on validation under diverse sea states, large distances, and varying incidence angles, using specialised equipment. A more holistic approach to offshore activities could further minimise cumulative noise., e.g., by optimised seismic data use by inversion to reduce geotechnical drilling.

Acknowledgements

This study was conducted within the framework of the project SynCore (German Federal Ministry for Economic Affairs and Climate Action [BMWi], grant number 03EE3020C).

We gratefully acknowledge Jakob Tougaard (Aarhus University, Denmark), Mads Kløve Hallstrøm, Laura Strøm Magner, and Christina Pommer (Danish Energy Agency) for their valuable feedback on environmental impact assessments for research cruises in Danish waters.

We also thank Dr. Laws, Dr. Goertz, an anonymous reviewer, and the Editor-in-Chief, Dr. Clement Kostov, for their expert insights, constructive comments, and the time they dedicated to reviewing this manuscript. Their contributions were instrumental in strengthening the scientific rigour and clarity of this work.

References

• Ainslie, M.A. and McColm, J.G. [1998] A simplified formula for viscous and chemical absorption in sea water. The Journal of the Acoustical Society of America, 103(3), 1671-1672. https://doi. org/10.1121/1.421258
• Ainslie, M.A., Halvorsen, M.B., Dekeling, R.P.A., Laws, R.M., Duncan, A.J., Frankel, A.S., et al. [2016] Verification of airgun sound field models for environmental impact assessment (p. 070018). Proceedings of the Fourth International Conference on the Effects of Noise on Aquatic Life. https://doi.org/10.1121/2.0000339
• Asgedom, E.G., Cecconello, E., Orji, O.C. and Söllner, W. [2017] Rough Sea Surface Reflection Coefficient Estimation and Its Implication On Hydrophone-Only Pre-Stack Deghosting (p. 5). Houston, Texas: TGS. Retrieved from https://www.tgs.com/hubfs/Technical%20 Library/Technical%20Library%20Files/2017June_Asgedom_etal_ RoughSea.pdf
• Austin, M., Yancey, D., Dushane, J. and Castellote, M. [2026] Sound characterization and marine mammal impact assessment of a marine vibroseis source in a shallow, near-shore environment. The Journal of the Acoustical Society of America, 159(4), 3119-3129. https://doi. org/10.1121/10.0043144
• Blacquière, G. and Sertlek, H.Ö. [2019] Modeling and assessing the effects of the sea surface, from being flat to being rough and dynamic. Geophysics, 84(2), T13-T27. https://doi.org/10.1190/geo2018-0294.1
• BOEM. [2014] Marine Mammal Hearing and Sensitivity to Acoustic Impacts (Expert Workshop on Underwater Noise and its Impacts on Marine and Coastal Biodiversity No. Appendix H) (p. 30). London, UK: Bureau of Ocean Energy Management. Retrieved from https://www.cbd.int/doc/meetings/mar/mcbem-2014-01/other/ mcbem-2014-01-submission-boem-02-en.pdf
• Breitzke, M., Boebel, O., El Naggar, S., Jokat, W. and Werner, B. [2008] Broad-band calibration of marine seismic sources used by R/V Polarstern for academic research in polar regions. Geophysical Journal International, 174(2), 505-524. https://doi.org/10.1111/j.1365246X.2008.03831.x
• BSH. [2021] Datenanforderungen für Seismische Untersuchungen (MarinEARS – Marine Explorer and Registry of Sound) (p. 8). Hamburg and Rostock, Germany: Bundesamt für Seeschifffhart und Hydrohgraphie.
• Carey, W.M. [2009] Lloyd’S Mirror—Image Interference Effects. Acoustics Today, 5(2), 14. https://doi.org/10.1121/1.3182842
• Crocker, S.E. and Fratantonio, F.D. [2016] Characteristics of Sounds Emitted During High-Resolution Marine Geophysical Surveys (Technical Report No. 12,203 (BOEM 2016-044)) (p. 266). Newport, Rhode Island: Naval Undersea Warface Center Division.
• Danish Energy Agency (DEA). [2023] Guideline for underwater noise: Installation of impact or vibratory driven piles. Retrieved from https://ens.dk/en
• Dooling, R.J. and Therrien, S.C. [2012] Hearing in Birds: What Changes From Air to Water. In A. N. Popper & A. Hawkins (Eds.), The Effects of Noise on Aquatic Life (pp. 77–82). New York, NY: Springer.
• Duncan, A.J. and Parsons, M.J.G. [2011] How wrong can you be? Can a simple spreading formula be used to predict worst-case underwater sound levels? Proceedings of Acoustics 2011. Gold Coast, Australia. Retrieved from https://www.semanticscholar.org/paper/ How-wrong-can-you-be-Can-a-simple-spreading-formula-Duncan-Parsons/46e141052c757c3fba7468387dee04a934bca057
• Duncan, A.J., Weilgart, L.S., Leaper, R., Jasny, M. and Livermore, S. [2017] A modelling comparison between received sound levels produced by a marine Vibroseis array and those from an airgun array for some typical seismic survey scenarios. Marine Pollution Bulletin, 119(1), 277–288. https://doi.org/10.1016/j.marpolbul.2017.04.001
• Elmer, K.-H., Betke, K. and Neumann, T. [2007] Standardverfahren zur Ermittlung und Bewertung der Belastung der Meeresumwelt durch Schallimmission von Offshore-Windenergieanlagen : Abschlussbericht zum BMU-Forschungsvorhaben 0329947 ; [Verbundprojekt]. Hannover.
• European Commission. [2017] Commission Decision (EU) 2017/848 of 17 May 2017 laying down criteria and methodological standards on good environmental status of marine waters and specifications and standardised methods for monitoring and assessment, and repealing Decision 2010/477/EU (Text with EEA relevance.), 125 OJ L §. Retrieved from http://data.europa.eu/eli/dec/2017/848/oj/eng
• Giordano, A., De Luca, L. and Giordano, P. [2020] Design and implementation of a marine seismic source by electrical discharge. Bollettino Di Geofisica Teorica Ed Applicata, 61, 589-606. https:// doi.org/10.4430/bgta0326
• Green, D.M., DeFerrari, H., McFadden, D., Pearse, J.S., Popper, A.N., Richardson, W.J., et al. [1994] Low-Frequency Sound and Marine Mammals: Current Knowledge and Research Needs. Low-Frequency Sound and Marine Mammals: Current Knowledge and Research Needs. Washington, D.C.: National Research Councili (US) Committee on Low-Frequency Sound and Marine Mammals. Retrieved from https://www.ncbi.nlm.nih.gov/books/NBK236693/
• IAU. [2001] Assessment of plans and projects significantly affecting Natura 2000 Sites - Methodological guidance on the provisions of Article 6(3) and (4) of the Habitats Directive 92/43/EEC (p. 81). Luxembourg: Impacts Assessment Unit, School of Planning Oxford Brookes University. Retrieved from https://ec.europa.eu/environment/nature/ natura2000/management/docs/art6/natura_2000_assess_en.pdf
• Joint Nature Conservation Committee (JNCC). [2017] JNCC guidelines for minimising the risk of injury to marine mammals from geophysical surveys (p. 28). Retrieved from http://data.jncc.gov.uk/data/ e2a46de5-43d4-43f0-b296-c62134397ce4/jncc-guidelines-seismicsurvey-aug2017-web.pdf
• Kastelein, R., Helder-Hoek, L. and Van de Voorde, S. [2017] Hearing thresholds of a male and a female harbor porpoise (Phocoena phocoena). The Journal of the Acoustical Society of America, 142(2), 1006. https://doi.org/10.1121/1.4997907
• Kastelein, R., Helder-Hoek, L., Van de Voorde, S., von Benda-Beckmann, A.M., Lam, F.-P.A., Jansen, E., et al. (2017). Temporary hearing threshold shift in a harbor porpoise (Phocoena phocoena) after exposure to multiple airgun sounds. The Journal of the Acoustical Society of America, 142(4), 2430-2442. https://doi.org/10.1121/1.5007720
• Kastelein, R., Voorde, S. and Jennings, N. [2018] Swimming Speed of a Harbor Porpoise (Phocoena phocoena) During Playbacks of Offshore Pile Driving Sounds. Aquatic Mammals, 44, 92–99. https://doi. org/10.1578/AM.44.1.2018.92
• Kastelein, R., Hoek, L., Cornelisse, S., von Benda-Beckmann, S., Lam, F.-P., Jong, C. and Ketten, D. [2020] Lack of reproducibility of temporary hearing threshold shifts in a harbor porpoise after exposure to repeated airgun sounds. The Journal of the Acoustical Society of America, 148, 556-565. https://doi.org/10.1121/10.0001668
• Laws, R. [2012] Cetacean Hearing-Damage Zones Around a Seismic Source. In A. N. Popper & A. Hawkins (Eds.), The Effects of Noise on Aquatic Life (pp. 473-476). New York, NY: Springer.
• Laws, R. [2013] Modifying the Seismic Source Array Spectrum to Reduce the Risk of Injury to Marine Mammals. Proceedings of 75th EAGE Conference & Exhibition incorporating SPE EUROPEC (p. 3). London, UK.
• Lurton, X. [2016] Modelling of the sound field radiated by multibeam echosounders for acoustical impact assessment. Applied Acoustics, 101, 201-221. https://doi.org/10.1016/j.apacoust.2015.07.012
• Matthews, M.-N.R., Ireland, D.S., Zeddies, D.G., Brune, R.H. and Pyć, C.D. [2021] A Modeling Comparison of the Potential Effects on Marine Mammals from Sounds Produced by Marine Vibroseis and Air Gun Seismic Sources. Journal of Marine Science and Engineering, 9(1), 12. https://doi.org/10.3390/jmse9010012
• Müller, A. and Zerbs, C. [2013] Prediction of Underwater Sound: Minimum Requirements on Documentation (Research Project Report No. M100004/29) (p. 16). Hamburg (Germany): Bundesamt für Seeschifffhart und Hydrohgraphie. Retrieved from https://marinears. bsh.de/FIS_SCHALL_PORTAL/pages/index.jsf
• Mustonen, M., Klauson, A., Andersson, M., Clorennec, D., Folegot, T., Koza, R., et al. [2019] Spatial and Temporal Variability of Ambient Underwater Sound in the Baltic Sea. Scientific Reports, 9(1), 13237. https://doi.org/10.1038/s41598-019-48891-x
• NMFS. [2024] Update to: Technical Guidance for Assessing the Effects of Anthropogenic Sound on Marine Mammal Hearing (Version 3.0): Underwater and In-Air Criteria for Onset of Auditory Injury and Temporary Threshold Shifts (NOAA Technical Memorandum No. National Marine Fisheries Service-OPR-71) (p. 182). U.S. Dept. of Commer., NOAA. Retrieved from https://www.fisheries.noaa.gov/ resources/documents
• Orji, O.C., Sollner, W. and Gelius, L.J. [2013] Sea Surface Reflection Coefficient Estimation. SEG Technical Program, Expanded Abstracts (pp. 51-55). https://doi.org/10.1190/segam2013-0944.1
• OSPAR. [2016] OSPAR inventory of measures to mitigate the emission and environmental impact of underwater noise (p. 72). Convention for the Protection of the Marine Environment of the North East Atlantic. Retrieved from https://www.ospar.org/documents?v=37745
• Parkes, G. and Hatton, L. [1986] The Marine Seismic Source. Dordrecht: Springer Netherlands. https://doi.org/10.1007/978-94-017-3385-4
• Prior, M.K., Ainslie, M.A., Halvorsen, M.B., Hartstra, I., Laws, R.M., MacGillivray, A., et al. [2021] Characterization of the acoustic output of single marine-seismic airguns and clusters: The Svein Vaage dataset. The Journal of the Acoustical Society of America, 150(5), 3675-3692. https://doi.org/10.1121/10.0006751
• Schack, H., Ruiz, M., Andersson, M. and Zweifel, U.L. [2019] Noise sensitivity of animals in the Baltic Sea (Underwater noise No. 167) (p. 46). Helsinki, Finland: Helsinki Commission - HELCOM. Retrieved from https://www.helcom.fi/wp-content/uploads/2019/08/ BSEP167.pdf
• Sertlek, H.O. and Ainslie, M. [2015] AGORA: Airgun source signature model: its application for the Dutch seismic survey. Conference Proceedings of UAC 2015.
• Smith, S.W. [2003] Digital signal processing: a practical guide for engineers and scientists. Newnes. Retrieved from https://search. ebscohost.com/login.aspx?direct=true&scope=site&db=nlebk&AN=240210
• Southall, B.L., Bowles, A.E., Ellison, W.T., Finneran, J.J., Gentry, R.L., Greene, C.R., et al. [2007]. MARINE MAMMAL NOISE-EXPOSURE CRITERIA: INITIAL SCIENTIFIC RECOMMENDATIONS. Bioacoustics, 17(1–3), 273-275. https://doi.org/10.1080/09524622. 2008.9753846
• Southall, B.L., Finneran, J.J., Reichmuth, C., Nachtigall, P.E., Ketten, D.R., Bowles, A.E., et al. [2019] Marine Mammal Noise Exposure Criteria: Updated Scientific Recommendations for Residual Hearing Effects. Aquatic Mammals, 45(2), 125-232. https://doi.org/10.1578/ AM.45.2.2019.125
• Tougaard, J. [2014] Vurdering af effekter af undervandsstøj på marine organismer (Technical Report No. Del 2-Påvirkninger) (p. 51). Aarhus Universitet, DCE – Nationalt Center for Miljø og Energi. Retrieved from http://dce2.au.dk/pub/TR45.pdf
• Tougaard, J. [2016] Input to revision of guidelines regarding underwater noise from oil and gas activities - effects on marine mammals and mitigation measures (Background document No. 202) (p. 52). Aarhus, Denmark: Aarhus University, Department of Bioscience. Retrieved from http://dce2.au.dk/pub/SR202.pdf
• Tougaard, J. [2021] Thresholds for Behavioural Responses to Noise in Marine Mammals - Background note to revision of guidelines from the Danish Energy Agency (Technical Report from DCE – Danish Centre for Environment and Energy No. 225) (p. 34). Danish Centre for Environment and Energy. Retrieved from https://dce2.au.dk/pub/ TR225.pdf
• Tougaard, J., Wright, A.J. and Madsen, P.T. [2015] Cetacean noise criteria revisited in the light of proposed exposure limits for harbour porpoises. Marine Pollution Bulletin, 90(1), 196-208. https://doi. org/10.1016/j.marpolbul.2014.10.051
• Verbeek, N.H. and McGee, T.M. [1995] Characteristics of high-resolution marine reflection profiling sources. Journal of Applied Geophysics, 33(4), 251–269. https://doi.org/10.1016/0926-9851(95)90045-4
• Weir, C.R. and Dolman, S.J. [2007] Comparative Review of the Regional Marine Mammal Mitigation Guidelines Implemented During Industrial Seismic Surveys, and Guidance Towards a Worldwide Standard. Journal of International Wildlife Law & Policy, 10(1), 1-27. https:// doi.org/10.1080/13880290701229838