EU Technical Group on Underwater Noise (EU TG-NOISE ... · TG Noise Thematic workshop...
Transcript of EU Technical Group on Underwater Noise (EU TG-NOISE ... · TG Noise Thematic workshop...
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
1 | P a g e
EU Technical Group on Underwater Noise (EU TG-NOISE)
Thematic workshop: Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
Torrelodones, Spain, 9-10 November 2017
WORKSHOP REPORT
Context
The Marine Strategy Framework Directive (MSFD) requires European Member States (MS) to develop
strategies for their marine waters that should lead to programmes of measures to achieve or maintain
Good Environmental Status (GES) in European Seas. As a first essential step in reaching good
environmental status, MS should establish monitoring programmes enabling the state of the marine
waters concerned to be assessed on a regular basis. Over recent years, significant progress has been
made with establishing joint monitoring programmes. Impulsive noise monitoring (through registers of
relevant activities) has been started for the Mediterranean Sea, Baltic Sea and North-East Atlantic
regions.
Joint ambient noise monitoring has been undertaken in the Baltic Sea (through the now-completed BIAS
project) and similar joint monitoring programmes for ambient noise are now in various stages of
development or implementation in other (sub)regions. Through these monitoring programmes more
information on the pressure will become available. Pressure is only one side of the equation; to assess
the extent to which good environmental status is achieved, it is necessary to understand how the
specific pressure is related to impacts on populations of marine animals.
The 2017 Commission Decision: introduction of thresholds
The Commission Decision of May 2017 requires EU Member States to establish threshold values to
ensure that levels of anthropogenic noise do not exceed levels that adversely affect populations of
marine animals. Member States should establish threshold values through cooperation at Union level
(but taking into account regional or subregional specificities). In the CIS Work Programme (2016-2019)
TG Noise was tasked to provide further advice to EU Member States on the development of thresholds.
In order to contribute to this process, TG Noise organised a thematic workshop titled Towards
thresholds for underwater noise. Common approaches for interpretation of data obtained in (Joint)
Monitoring Programmes. This meeting was hosted by the Spanish Ministry of Agriculture and Fisheries,
Food and Environment and took place on 9-10th November in Torrelodones, Spain.
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
2 | P a g e
The workshop was co-chaired by René Dekeling and Mark Tasker (TG Noise Chairs) and was attended by
32 participants (Annex 1) including members of the research community, government experts, and
representatives from NGOs (both environmental and commercial).
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
3 | P a g e
Objectives
This workshop focused on the evaluation of possible uses of the data collected by
monitoring/registration of underwater noise and to provide advice on how these data can be used to
obtain a better understanding of the environmental impacts of underwater noise. There was no attempt
to define thresholds. There are still fundamental knowledge gaps that hinder establishing concrete
thresholds, and it may take several years to establish relevant thresholds in specific (sub)regions. At this
stage, methodologies that can be used in future, common assessments need to be assessed and further
developed, and ensuring that common methods are being used at Union level is a priority.
The present workshop builds upon the results of the 2016 Hamburg thematic workshop, where
consensus was found on the main directions to better understand the effects of underwater noise and
how to develop impact indicators.
Habitat degradation
For both impulsive and increased (anthropogenic) ambient noise, the participants at the Hamburg
workshop considered that the concept of habitat degradation was worth exploring to develop indicators
of impact. Possible options that were proposed and that should be further investigated include:
• Loss of communication space (masking)
• % loss of habitat
• Habitat degradation index
Some of these approaches would be possible already, at least to some extent, if there was full input of
relevant data by Member States to noise registers and full implementation of ambient noise monitoring
programmes. The challenge is to link these indicators of effect in a meaningful way to indicators of
population impact – noting in particular that temporal scales would also need to be considered. For
example, effects such as masking would be instantaneous, but the impact would (probably) stop after
the sound source has ceased; where habitat loss (displacement) would have continue to have effects for
a period of time after the ceasing of sound outputs, and the “index of habitat degradation” would also
need a temporal “recovery” component. The choice of marine species that would be used in such
indicators also needs to be agreed – especially as there has only been research on a very narrow range
of species (despite multiple earlier recommendations from TG Noise).
The workshop was organized around 2 thematic sessions (impulsive and ambient noise) and break-out
sessions. The discussions at the workshop clearly showed that there is a need to converge to
methodologies that may be applied at European scale. TG Noise should refine the methodology with
clear proposals on methods and advice on definition of thresholds values that can be implemented by
MSs and RSCs. This task needs to be further developed in the short-term by TG Noise and is included in
its Terms of Reference.
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
4 | P a g e
Thematic Sessions and Break-out sessions After a general introduction regarding the concept of thresholds, the two thematic sessions were
introduced. The workshop attendees were then divided into three break-out groups, following each of
the invited speakers’ presentations (think pieces). Each of the break-out groups were requested to focus
on main questions relating to each of the think pieces in turn. The results of the plenary discussions and
the outcomes of the break-out groups are reported below per session theme.
SESSION 1: IMPULSIVE NOISE Chair: Mark Tasker
Think piece # 1: Putting the Commission Decision into practice
Nathan Merchant, CEFAS
This presentation first gave an overview of starting points for indicators based on a recently published
paper. Indicators need to be designed such that targets/thresholds agreed at appropriate management
levels can be implemented by regulators in practice. This will require that indicators are straightforward
to communicate to regulators and stakeholders, and that the methodology does not require unrealistic
amounts of time or funding to implement. Similarly, the indicators should be aligned with existing and
emerging marine management practices, particularly approaches for marine spatial planning and
cumulative effects assessment.
A framework for the development of such indicators was presented, that combines population or
habitat data with data on pressures, within a defined management area, to produce quantified risk
maps and exposure curves which can be used as a basis for defining indicators and setting thresholds.
This methodology was demonstrated for two case studies in the North Sea (harbour porpoise, spawning
herring), based on data from the OSPAR Intermediate Assessment 2017.
Think piece # 2: Impact assessment using the Impulsive Noise Registry
Michael Ainslie, TNO
This presentation introduced a methodology to predict the spatial and temporal distribution of the
potential for behavioural disturbance. It makes use of the data used in the OSPAR Intermediate
Assessment 2017 – the distribution of (loud) sound sources contained in the impulsive noise register
(INR). This method has been used by the Dutch government in an assessment of the cumulative effects
of the construction of offshore wind farms. The main conclusion of this presentation was that tools are
available to produce sound maps using information contained in INR through acoustic modelling. This
method naturally combines different source categories (also noise mitigation) into one meaningful map.
If an assumption is made on the received level at which a specified effect may occur, e.g. disturbance, it
is possible to map the area of potential disturbance. This area of disturbance can then be combined
with distribution data of sensitive species. Modelling in a variety of contexts shows that the size of
predicted disturbed areas varies greatly, and may be larger or much smaller than the ICES sub-grid cells.
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
5 | P a g e
Think piece # 3: Impact indicator for impulsive noise – German proposal
Alexander Liebschner
This presentation introduced a way to define an impact for impulsive noise by quantifying the overlap of
a species (e.g. harbour porpoise) and anthropogenic impulsive sounds exceeding specifically defined
levels (e.g. Sound Exposure Level of 140 dB) that disturb/displace individuals of that species, over
determined areas within selected periods (‘individual-disturbance-days’).
An example using grid cells was presented to demonstrate the methodology of the proposed indicator.
The logical steps are delimitation of an area (convention area, sub-region, EEZ, national waters,
ecological important areas, etc.), selection of a period (e.g. year), determination of number of
individuals in the area in the period concerned (e.g. based on existing monitoring data). Next the
number of ‘individual-days’ is calculated; then number of days different parts of the area are impacted
by disturbing anthropogenic impulsive noise is calculated, using defined thresholds; in the example
these parts of the of the area were grid cells, but they may be other forms/polygons. These figures were
combined to first calculate total individual-disturbance-days and then the intensity of impact, i.e. the
proportion/percentage of total ‘individual-days’ that are ‘individual-disturbance-days’.
Questions discussed in break-out groups:
1. Explore and list ideas for (methodologies for) an impact indicator based on the noise
registries, all indicators should have the potential to set a justifiable threshold – though the
actual threshold should not be considered at this stage
2. Consider the advantages and disadvantages of each (main) idea.
3. Specify, as far as possible, a work plan to develop each idea (or if insufficient time, the idea
most likely to be initially developed)
Key discussion outcomes:
A summary of issues identified in the breakout groups discussions is presented. There were only limited
discussions in the plenary meeting, so these results do not reflect an overall consensus of the workshop
participants.
It was observed that in general, all approaches presented in the plenary and some other methods
discussed in the break-out sessions are similar.
The ideas/methods discussed were:
1. Register sources and ‘disturbance maps’ based on exposure criteria and propagation modeling
(e.g. SEL harbour porpoises) (Michael Ainslie)
2. Threshold/ percentage of area /population / animals affected (Alexander Liebschner)
3. Combining population/habitat and pressure data in a defined management area and produce
risk maps/exposure curves (Nathan Merchant)
4. % of unit area affected (can be SAC or other unit; based on biological assessment, can involve
analysis of persistence) (Mark Tasker).
5. French approach: Consideration of the categories of the register (risk is higher for high sources;
Temporal component = % time; spatial component = % area; applied for sub-marine region)
(Florent le Courtois).
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
6 | P a g e
The first four ideas (Ainslie, Merchant, Liebschner, Tasker), are all based on the principle of overlaying
the noise registry information in some format with species distribution data, in effect an exposure
assessment by comparing the distribution of sounds with that of marine life. The species data could be
absolute or relative density, or distribution range with an added measure of habitat importance/quality.
There are varying degrees of complexity and also of uncertainty associated with the approaches. An
indicator with too much uncertainty might be a weak indicator, however there are ways in which one
could incorporate uncertainty explicitly.
In relation to the pressure layer, Ainslie’s proposal is the most quantitative using noise propagation
models together with received sound levels to estimate the disturbance area around each sound
source, noting that the German method would also use sound modelling at some stage. In theory this is
more precise than the other approaches and would take into consideration environmental effects on
noise propagation but adds complexity to estimates which could result in a potentially time consuming
and expensive process. Further it should be noted that actual accuracy relies on several assumptions
and there is disagreement between the methods as to whether sound received levels should be used as
a proxy for species responses.
Some approaches (Merchant /Tasker) rely on a fixed deterrence radius that can be determined based
on empirical evidence of disturbance ranges and does not rely on noise propagation, simplifying this
element of the indicator. Another option is to use the pulse block days more directly, at least for seismic
surveys, where the survey route coordinates are not recorded in the registries. All approaches have
limitations in representing the (poorly known) disturbance effect of seismic surveys.
Most approaches can incorporate a direct measure of species abundance (e.g. for harbour porpoise),
and have the potential to link to MSFD Descriptor 1 in relation to species abundance estimates and
associated monitoring. There is however great variability and uncertainty associated with species
abundances and there could be large areas where survey effort is low. Most approaches have not
defined how to deal with animal movement and when numbers of animals against time affected is
plotted in a specific area it is not possible to distinguish between whether it is the same or different
animals disturbed for X number of days. One way around this could be to use an index of habitat quality
that would be derived from the species density surfaces. Another potential way was presented by
Germany where the unit of the indicator is porpoise/days of disturbance.
An alternative to using animal densities to weight the potential importance of areas was presented in
both UK methods where certain areas such as marine protected areas or fish sensitive areas could have
a weighting factor. These areas may have been identified based on several criteria in addition to that of
density numbers (e.g. persistency, confidence in data) and therefore it may be more important to
reduce risk to individuals in these areas.
It is important to realize that ANY approach needs to be suitable for use in management (i.e.
forecasting) and to be understood/implemented by non-specialists with limited resources.
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
7 | P a g e
As described above there are pros and cons to the different approaches which need further analysis and
discussion, but it was clear that there are fundamental similarities and therefore there is potential to
merge elements of the different approaches. More discussion is needed to decide on the desirable level
of confidence for the indicator versus level of detail and complexity. The two more detailed approaches,
by Merchant and Ainslie, are more data hungry but can be modified as the scientific understanding
improves (e.g. distances of effect with/without mitigation). They also explicitly describe uncertainty but
there is scope to capture and communicate those uncertainties as part of the indicator.
Further discussion will also be needed on how the indicator could be used in management. The more
readily that one can translate the findings of the indicator into management measures the stronger the
indicator.
Finally, one of the groups discussed potential indicators in addition to those using the noise registry
information. This could include direct measurements of fish/invertebrates’ stress markers in certain
areas.
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
8 | P a g e
SESSION 2: AMBIENT NOISE Chair: René Dekeling
Think piece # 1: The use of the BIAS GES tool to determine habitat degradation
Thomas Folegot, Quiet Oceans
This presentation introduced the use of BIAS soundscape tool developed by the EU LIFE project Baltic
Sea Information on the Acoustic Soundscape (BIAS) for quantifying the pressure from ship noise in the
Baltic Sea. The planning tool can be the basis for study of the impact on marine animals once thresholds
for impact are established (Nikolopoulos et al., 2016).
The soundscape tool is based on measured and modelled soundscape data and provides a number of
functions to evaluate the spatial and temporal sound characteristics within a user-defined geographical
area. The tool is based on several hundred soundscape maps covering different frequencies, depth
intervals and exceedance levels, where e.g. 5% exceedance level describes the highest SPL in an area
that occurred 5% of the time (this will give the strongest sources in the area) and 90% exceedance level
will represent SPL 90% of the time (this level will often represent the background noise level).
Two main assessment methods are implemented in the tool. Both are designed to assess the pressure
during a certain time and in a specific area. The data from these assessments can later be used to study
the impact on marine life or to establish if a threshold for impact was exceeded or not.
Think piece # 2:
Bioacoustics-Based Impact Indicator (BBII)
Michel André
This presentation introduced a relatively new bioacoustics-based approach – the Lido approach. This
method is not species dependent and does not rely on gathering new data on species sensitivity to
noise (thresholds) to be implemented. It is global (pan-European) and transversal by combining MSFD
D11.1 and MSFD D11.2 objectives to address both impulsive and tonal (continuous) sources. It does not
rely on static thresholds but on monitoring trends (changes in the soundscape) that can be assessed on
a variety of timescales (hourly/daily/weekly/seasonally/yearly). It provides the MSFD a regional noise
budget approach for GES; biological and anthropogenic activities are comparatively assessed, and it can
be based on MSFD D11.2 existing/planned monitoring stations.
Questions discussed in break-out groups:
1. Explore and list ideas for (methodologies for) an impact indicator based on the results of
ambient noise monitoring, all indicators should have the potential to set a justifiable
threshold – though the actual threshold should not be considered at this stage
2. Consider the advantages and disadvantages of each (main) idea.
3. Specify, as far as possible, a work plan to develop each idea (or if insufficient time, the idea
most likely to be initially developed)
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
9 | P a g e
Key discussion outcomes:
One breakout group approached the concepts for methodology from the viewpoint that there could be
target expressed in terms of a reduction in source levels of a proportion of the global commercial fleet,
and this would lead to a relative change in risk to sensitive populations. The impact could be expressed
in terms of a percentage reduction in risk to sensitive populations selected for a management area
(which could be MS waters, an area within that, or an area including waters of several MS).
Using the BIAS methodology, the relative risk could be described in terms of x% of time exposed to
levels above level y for z% of the area, for both the current, or reference, situation and if measures that
achieved the target were implemented. The reported impact indicator R would be the expected risk
reduction achieved if the target were met e.g. R=xtarget/xno action. (R could also be expressed in terms of z
and consideration would need to be given to which was most appropriate). Although the group
considered this in the context of the BIAS methods, it was noted that relative risk estimates could also
be made using other approaches.
It is likely that for different management areas there will be different populations that are of concern,
so there could be very different values of R. Nevertheless, if a MS were to evaluate that for a population
of particular concern to that MS a certain relative reduction in risk was needed then the methodology
would allow the calculation a source level reduction target that could achieve this.
The process aims to give stakeholders the information that will be of most concern to them. The
shipping industry is given a target noise reduction for selected vessel types (and can comment on the
technical and economic implications) and environmental management agencies are given a set of
relative risk reduction values for species they have identified as of concern.
An overview of pros and cons of the two proposals were discussed, but since no detailed description of
the LIDO approach was available at the workshop this will need further analysis. Suggested issues that
were identified include:
• BIAS
Pros: Can be understood intuitively, it seems to address the right question (habitat degradation),
adaptable (e.g. level of masking and addition of more frequencies); to be used by for managers
taking into account biological effects
Cons: Where it may be easy to use at a local scale, could be difficult to apply at large spatial scales
(relevant for example for baleen whales) as it involves three variables that are difficult to investigate;
there are still fundamental uncertainties on masking, e.g. it not clear whether animals use all active
space;
• Bioacoustics-Based Impact Indicator (BBII)
Pros: According to developers it is not-species dependent (=quick implementation), large scale
possible (global, EU), existing monitoring can be used
Cons: Because the method could not be fully explained during the meeting there was discussion on
whether the concept was suitable (e.g. bioacoustics as proxy for biological activity), threshold for
S/N, spatial issues; i.e. sample area); it was not clear how complicated it is for use by MS
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
10 | P a g e
Summary of comparison / discussion between the two methods:
• BIAS approach: coherent method (same logic) with the approach for impulsive noise setting
habitat degradation as central concept, and easier to understand by regulators; methodology is
reasonably developed, but the method is species dependent and there may be challenges in
selecting species and defining factors for those species
• LIDO approach: a promising approach but addresses many variables that may be more difficult to
understand by regulators;
Both approaches need to be further developed.
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
11 | P a g e
SESSION 3: Conclusions and way forward
The specific aim of this workshop was to evaluate possible uses of the data collected by registration and monitoring of underwater noise and to provide advice on how these data can be used to obtain a better understanding of the environmental impacts of underwater noise. This was an early step in the process of considering if thresholds could be defined and advised upon.
During the workshop, several concepts for assessing and evaluating impulsive and ambient noise were presented. The pros and cons of these concepts were discussed, both in break out groups and later in plenary. The objective was not to recommend a single concept or method, but to review all ideas.
The overall conclusions of this workshop are as follows and need further work at the next TG meeting and/or thematic workshop in 2018:
1) To converge methodologies: TG Noise should provide WG GES proposals for methodology that can be used at European scale. If WG GES (and MSCG) agree to these proposals, TG Noise could refine the methodology with clear proposals on methods and perhaps advice on options for setting thresholds (or carrying out research aimed to provide a scientific basis for thresholds). It should be noted that MSs (potentially working through RSCs or other international organisations) will be responsible for setting threshold values;
2) When developing an EU-wide concept, TG Noise should ensure it can be readily understood and implemented by multiple regulators (incl. non-specialists);
3) Any methodology should be usable for multiple (types of) species, not only for marine mammals; The concept identified in the 2016 Hamburg workshop to address habitat has been used as basis for the discussions on methodology; this approach looks promising.
A short description of each of the methods suggested during the workshop is attached at Annex 2.
Each method is described:
• Aim/purpose of the method
• Brief description of the method and concepts (non-technical)
• Description of any pros and cons
• References and contact person
TG Noise Chairs and support contract will collate the information into one document, which will include
an overview of pros and cons, and usability of each the methods.
The workshop participants agreed that progress needs to be made during spring 2018. The first version
of a consolidated report based on round-of comments will be available early 2018. This draft document
will be sent to workshop participants and TG Noise members for comments (early 2018). TG Noise
members provide comments and additional contributions (early 2018), then an updated version will be
made based on received comments; if this document is sufficiently developed, it may be discussed with
WG GES, potentially MSCG.
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
12 | P a g e
ANNEX 1 – List of Participants
Name Surname Organization
Alain Norro Royal Belgian Institute for Natural Sciences, Belgium
Albert Willemsen International Council of Marine Industry Associations - ICOMIA
Aleksander Klauson Faculty of Civil Engineering, Tallinn University of Technology, Estonia
Alexander Liebschner German Federal Agency for Nature Conservation, Germany
Andreas Müller Müller-BBM GmbH, Germany
Annemie Volckaert ARCADIS, Belgium
Antonio Novellino EMODnet Physics
Fabrizio Borsani ISPRA, Italy
Florent Le Courtois SHOM, France
Frank Thomsen Central Dredging Association, CEDA
Jorge Ureta Ministry of Agriculture, Food and Environment, Spain
Katja Klančnik Institute for Water of the Republic of Slovenia
Lindy Weilgart Oceancare
Maria Ferreira Coastal & Marine Union (EUCC), The Netherlands
Maria-Emanuela Mihailov NIMRD, Romania
Mark Tasker Joint Nature Conservation Committee, United Kingdom
Marta Martínez-Gil
Pardo de Vera Ministry of Agriculture, Food and Environment, Spain
Mathias Andersson Swedish Defence Research Agency, Sweden
Maud Casier EC DG Environment
Michael Ainslie TNO, The Netherlands
Michel André Universitat Politècnica de Catalunya, Barcelona, Spain
Nathan Merchant CEFAS, United Kingdom
Noelia Ortega Technological Naval Centre, Spain
Pablo Cervantes Technological Naval Centre, Spain
Patrick Gorringe EuroGOOS
Predrag Vukadin Institute for Oceanography & Fishery, Croatia
Ramón Miralles University Politècnica de València, Spain
René Dekeling Ministry of Infrastructure and the Environment, The Netherlands
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
13 | P a g e
Name Surname Organization
Russel Leaper IFAW / Seas at Risk
Signe Jung-Madsen Ministry of Environment and Food of Denmark
Sonia Mendes Joint Nature Conservation Committee, United Kingdom
Thomas Folegot Quiet Oceans, France
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
14 | P a g e
ANNEX 2 - Descriptions of Methods
For impulsive noise the following methodologies are described:
1. Marine noise budgets in practice. Cefas, UK - Nathan Merchant
2. Dutch noise propagation method- Michael Ainslie
3. German noise protection concept- Alexander Liebschner
4. Habitat loss method – Mark Tasker
5. French impulsive noise data gathering - Florent LeCourtois
For continuous (ambient) noise two approaches discussed are presented (additionally the French
approach provided after the workshop is also included as information):
6. BIAS GES tool – Mathias Anderson
7. Bioacoustics-Based Impact Indicator (BBII) – Michel André
8. D11C2: French approach on continuous noise – Florent LeCourtois
These descriptions (and the suggested pros and cons) were made by the presenters/contact persons for
these methods. Further evaluation of the methods will be published in a TG Noise report.
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
15 | P a g e
1. Marine noise budgets in practice. Cefas, UK - Nathan Merchant
Aim/purpose of the methodology
This methodology is a risk-based framework which quantifies the exposure of animal populations or
habitats to noise pollution. It can be applied to derive indicators and thresholds for noise exposure,
consistent with MSFD requirements, and is suitable for both place-based approaches (e.g. marine
protected areas, designated habitats) and ecosystem-based approaches (e.g. protected/managed
populations). The methodology has been published in a high-ranking, peer-reviewed conservation
journal (Merchant et al., 2017), using data from the OSPAR Intermediate Assessment 2017.
Brief description of the methodology and concepts (non-technical)
Figure 1. Workflow for proposed assessment framework (Merchant et al., 2017).
The framework combines population or habitat data (Fig 1b) with data on pressures (Fig. 1c) within a
defined management area (Fig. 1a), to produce quantified risk maps (Fig. 1d) and exposure curves (Fig.
1e). The exposure curves introduced by this method (Merchant et al., 2017) plot the % population or
area exposed vs. the % time exposed during the assessment period. The exposure curve provides a
quantitative basis to define indicators of overall exposure to the population or habitat (Fig. 1f; see
Merchant et al., 2017 for details).
The method has been demonstrated in case studies for both harbour porpoise (Figure 2) and herring
spawning in the North Sea (Merchant et al., 2017). The framework is flexible, and can be applied to non-
gridded data, and incorporate the most appropriate animal density and noise pressure information.
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
16 | P a g e
Figure 2. Case study for North Sea harbour porpoise showing seasonal risk maps (a-c) and exposure
curves (d) during 2015 (Merchant et al., 2017). (a) Spring (Mar.-May) (b) Summer (Jun.-Aug.) (c) Fall
(Sep.-Nov.) (d) Exposure curves and exposure indices (EI). Risk maps incorporate density data from Gilles
et al. (2016).
Description of their pros and cons
Pros: meets MSFD requirements; flexible and future-proof; can be applied to populations or habitats;
risk- and evidence-based. Cons: does not incorporate (though is compatible with) modelling of
population consequences.
References and Contact person
Dr Nathan Merchant, Principal Scientist, Cefas, UK: [email protected]
Merchant, N.D., Faulkner, R.C., Martinez, R., 2017. Marine Noise Budgets in Practice. Conservation
Letters http://onlinelibrary.wiley.com/doi/10.1111/conl.12420/pdf
Gilles, A. et al. (2016). Seasonal habitat-based density models for a marine top predator, the harbor
porpoise, in a dynamic environment. Ecosphere, 7, e01367. http://dx.doi.org/10.1002/ecs2.1367
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
17 | P a g e
2. Dutch noise propagation method. TNO, The Netherlands - Michael Ainslie
Aim/purpose of the methodology
The aim of the proposed methodology is to predict the spatial and temporal distribution of the potential
for behavioural disturbance. It makes use of the data in the OSPAR Intermediate assessment 2017 –
distribution of (loud) sound sources. The approach described here is to base the method as far as
feasible on information available from the impulsive noise register (INR). It has been used by the Dutch
government in licensing the construction of offshore wind farms. At the time of writing, OSPAR had not
agreed a methodology for assessing the effects of impulsive underwater sound, nevertheless the
approach taken can be used to further explore possibilities.
Brief description of the methodology and concepts (non-technical)
Basic principles:
While there are different potential effects of underwater sound, the assumption is that disturbance
leading to temporary habitat loss is the most relevant effect and that this needs to be quantified. This
approach is also based on results of the 2016 Thematic Workshop (ref: minutes Hamburg meeting) that
quantifying habitat loss would be a likely way forward.
This methodology employs a stepwise approach to quantify the potential effects of impulsive sound on
marine animals. There are different options with this method: In its simplest form it comprises
steps 1 and 2 only; subsequent steps can be executed at a later stage where consensus exists. The
methodology links two OSPAR common indicators (impulsive underwater sound and the distribution and
abundance of cetaceans).
Step by step summary:
Step 1: Quantify underwater sound propagation
Step 2: Determine the area affected by underwater sound: effect parameters and threshold values
After this step, there is an insight in the amount of potential habitat loss (in space and time).
Options for further assessment:
Step 3: Quantify the number of affected animals (or relative part of population)
If there is no detailed information of densities, but there is information on relative population
distribution, it would be possible to make calculations on the proportion of the population that is
exposed.
Step 4: Calculate the number of ‘animal disturbance days’
Step 5: Assess the possible impact on the population
Step 6: Assess cumulative effects of multiple projects / impulsive underwater sound sources
The approach was tailored to make it compatible with the data collected in the INR. An accurate
estimate of the effects of a pressure on a population (whether cumulative or not) requires either a
qualitative or quantitative understanding of the relationship between pressures and their effects on
individual animals. Such information is available (with certain margins of error) for most of the steps
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
18 | P a g e
listed. For step 5, however, translating direct effects (disturbance) into population effects is hampered by
a fundamental lack of knowledge. A temporary solution to this problem is to use a population model
(Figure).
Transparency in terms of how outcomes are determined and how they are interpreted and used to
inform management decisions, are critical to gaining international acceptance. In this approach,
transparency is ensured by showing the uncertainties at each step, and by indicating the relative
influence of these uncertainties for the outcomes. This helps to focus future research effort. Verification
of population size (whether measured or modelled) is a clear priority to make this methodology
acceptable at a regional scale.
Steps 1 and 2 are described in more detail below:
Step 1 (Quantify underwater sound propagation): The first step is to characterize underwater sound
sources according to their energy source level (or proxy). This source level is used in combination with
known local propagation conditions to determine a map of sound exposure level, SEL (or sound pressure
level, SPL).
Step 2 (Determine the area affected by underwater sound: effect parameters and threshold values): The
value of SEL (or SPL, where appropriate) is compared with a disturbance threshold. At this stage the
number of disturbance days can be calculated at each position on the map as the number of days for
which the threshold is exceeded at that position.
The information used from the INR is:
• Start and end date of activity
• Source type (seismic/ sonar/ explosion/ pile driving)
• Source characteristics:
• source level where provided (alternatively, the TG Noise source category from ‘Very low’ to ‘Very high’ may be used).
Assumptions
The two main assumptions made are that disturbance is correlated with exceedance of a SEL (or SPL)
threshold and that the duration of each disturbance is 1 day.
Description of their pros and cons
The advantages of the proposed approach (steps one and two) are:
• It uses information from the INR
• Use of acoustic modelling facilitates the calculation of effect distances
• Addresses impact
• Naturally combines different source categories into one meaningful map
• Effects of source mitigation are included
• Allows for potentially large differences in size of disturbed areas which may strongly vary, and can be larger or much smaller than the ICES sub-grid cells
• Maps provide managers with insight into spatial/temporal distribution of disturbance
• Can be used in combination with population density to estimate effect on population (eg IPCOD)
• Transparent method.
The main disadvantages are:
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
19 | P a g e
• magnitude of predicted effects depends on information not available in INR (requires assumption of geographical distribution of sources within INR resolution cell)
• requires regional harmonization of propagation modelling
Contact person and references
Niels Kinneging, Alexander von Benda-Beckmann, Michael Ainslie, Floor Heinis, Christ de Jong.
Heinis, F., de Jong, C., & Rijkswaterstaat Underwater Sound Working Group (2015). Framework for
assessing ecological and cumulative effects of offshore wind farms: cumulative effects of impulsive
underwater sound on marine mammals. TNO 2015 R10335-A. Available from
https://tethys.pnnl.gov/publications/framework-
Heinis, F. (2017). Assessment methodology for impact of impulsive sound: Evaluation of available
methods and action plan for the development of a methodology for application in the MSFD. Version 1.1
- Final report 1.1. HWE.
Sander von Benda-Beckmann, Christ de Jong, Mark Prior, Bas Binnerts, Frans-Peter Lam, Michael Ainslie
(2017). Modelling sound and disturbance maps using the impulsive noise register for assessing
cumulative impact of impulsive sound. TNO 2017 R11282 | Final report
MSFD Common Implementation Strategy, Technical Group on Underwater Noise (TG-NOISE), Thematic
Workshop: Way forward to define further Indicators for Underwater Noise. 7- 8 June 2016, Bundesamt
für Seeschifffahrt und Hydrographie (BSH), Bernhard-Nocht Straße 78, 20359 Hamburg, Germany
Figure:
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
20 | P a g e
3. German noise protection concept. German Federal Agency for Nature
Conservation, Germany - Alexander Liebschner
Aim/purpose of the methodology
Distribution of co-occurrence of species (indicator species e.g., harbour porpoise) and anthropogenic
impulsive sounds exceeding levels that disturb/displace individuals of that species (e.g., Sound Exposure
Level of 140 dB), over determined areas within selected periods (‘individual-disturbance-days’).
The indicator reflects the quantity/intensity of the species’ spatio-temporal use of an area (‘individual-
days’) that is impaired by disturbing/displacing impulsive noise by calculating the proportion of
‘individual-days’ that are ‘individual-disturbance-days’ as well. This proportion provides a numerical
measure of the impact of impulsive noise and thus indicates the environmental status of a determined
area within a selected time period.
The indicator should help to assess the Environmental Status. The indicator so far quantifies the impact
and can be applied both for the whole area and the sub-units of it. In order to use this impact indicator
for assessing whether the environment is in a good status (GES) threshold values have to be set.
Because the indicator is rather independent from abundance/densities or sensitivity of the species there
is the need for different thresholds: The acceptable level of impact may be much higher for areas or
times with low abundance/densities of the receiver species and much lower in e.g. spawning grounds or
during periods of especially high noise sensitivity such as the reproduction phase.
Brief description of the methodology and concepts (non-technical)
Step I delimitation of an area:convention area, sub-region, EEZ, national waters, ecological important
areas, etc.
NOTE: To be able to include inhomogeneous distribution of species and/or to calculate the
extent of the surface impacted the area may be divided in reasonable sub-units such as e.g. grid cells.
Step II selection of a period: reporting period, year, season, sensitive time of the year, etc.
Step III calculation of (mean) total number of individuals in the area in the period concerned (e.g. based
on existing monitoring data)
Step IV calculation of ‘individual-days’ by multiplying the number of individuals by the length of the
period [days]
Step V identification of the number of days different parts of the area are impacted by disturbing
anthropogenic impulsive noise during the period
the threshold defining ‘disturbing’ depends on species specific sensitivity, source characteristics
(e.g., frequencies), noise level etc.
parts of the area may be grid cells, surface, etc.
Step VI a) calculation of ‘individual-disturbance-days’ by multiplying the number of individuals within
the area impacted by the duration of the impact [days], or
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
21 | P a g e
b) calculation of ‘individual-disturbance-days’ by multiplying the number of individuals within
the area impacted by the length of the period [days]
Step VII calculation of the intensity of impact, i.e. the proportion/percentage of total ‘individual-days’
that are ‘individual-disturbance-days’ as well.
regarding the whole area the total number of ‘individual-disturbance-days’;
regarding the different sub-units only the ‘individual-disturbance-days’ within the respective
sub-unit.
Description of pros and cons
Pros:
• it works with indicator species (which can be different in different area), Indicator species representing usually a group of noise sensitive animals;
• it can be adjusted to any time period
• can be applied to any area or any size of area
• therefore it is very flexible Cons:
• you need monitoring data about the abundance of the indicator species (in the area)
• you need data about the noise events in the area
References and Contact person
For further information please contact: Thomas Merck or Alexander Liebschner/ BfN/ Germany;
Email: [email protected]; [email protected];
Tel. +49 38301-122 +49 38301-163
Example how to use the Indicator:
In the following an example is presented only to demonstrate the methodology of the proposed
indicator. It does not represent an existing area or is referring to a specific species. The use of grid cells
was chosen to make the application comprehensible, but the method can also be applied for polygons.
The grid does not reflect any existing partitioning such as e.g. ICES sub-blocks.
Step I - III
Total number of individuals [N]: 1.928
78,3 71,8 69,3 66,2 79,0 66,9 72,7
17,4 69,7 77,2 73,8 53,0 50,8 49,0
15,8 10,4 10,8 73,4 53,8 63,1 85,0
6,3 6,5 18,5 18,5 71,9 60,7 68,1
4,8 7,1 7,5 10,8 36,0 62,4 68,2
6,0 7,8 8,7 4,4 22,0 59,1 66,6
0,2 0,7 1,7 4,5 6,3 15,7 69,4
number of individuals
[n]
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
22 | P a g e
9.631 8.829 8.528 8.142 9.717 8.231 8.944
2.144 8.573 9.500 9.083 6.517 6.249 6.027
1.943 1.283 1.334 9.024 6.616 7.757 10.455
776 803 2.282 2.281 8.848 7.470 8.374
596 872 919 1.331 4.428 7.679 8.387
741 959 1.075 536 2.706 7.270 8.193
31 90 211 554 777 1.932 8.530
number of 'individual-days'
[n * d]
0 0 0 0 30 30 30
0 0 0 0 22 30 30
0 0 0 0 0 0 30
0 50 50 50 0 0 0
0 50 80 50 0 0 0
0 50 50 50 0 0 0
0 0 0 0 0 0 0
number of days with disturbing impulsive noise
[dd]
0 0 0 0 2.370 2.007 2.181
0 0 0 0 1166 1.524 1.470
0 0 0 0 0 0 2.550
0 327 927 927 0 0 0
0 355 598 541 0 0 0
0 390 437 218 0 0 0
0 0 0 0 0 0 0
number of 'individual-disturbance-days'
[idd = n * dd]
Selected period: 123 days [d]
Step IV
Total number of ‘individual-days’ [ID = N * d]: 237.144
Step V
Step VI
Total number of ‘individual-disturbance-days’ [IDD]: 17.988
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
23 | P a g e
Step VII
Indicator value regarding the different sub-units
Indicator value regarding the whole area:7,58 [IDD / ID *100]
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
24 | P a g e
4. Habitat loss assessment method (using harbour porpoise as an example) – Mark
Tasker Aim/purpose of the methodology
One of the effects of underwater anthropogenic sound (and human presence) is to deter marine animals
from using an area of sea. The method summarised here is based upon the relatively simple information
that is (should become) available within the noise registers being established as part of the
implementation of the Marine Strategy Framework Directive. The method, as outlined in Tasker (2016),
requires some limited quantification of the effects of each noise source on a receptor animal but allows
for suitable limits to be expressed in terms of time and space for the area of sea that may become
unavailable to the animal.
Brief description of the methodology and concepts (non-technical)
Available information.
The noise registers are providing a record of where underwater anthropogenic impulsive sound above
agreed thresholds of intensity and below an agreed frequency level has occurred. The spatial definition
of such records is at a minimum within blocks of 100-200 km2 (e.g. UK oil & gas licensing blocks or ICES
sub-blocks) and the temporal definition is one day. A simple activity classification (pile driving, seismic
source, explosion etc) is also available.
In order to estimate the temporal and spatial extent of “habitat loss”, an understanding of the effects of
impulsive sound sources on marine animal(s) is needed. The evidence base for most species remains
scarce and TG Noise (and its predecessors) have highlighted the importance of such research. Limited
information is known for harbour porpoises, and this is used here to illustrate the approach.
Assessment process (illustrated for harbour porpoise)
Determine Effective Deterrence Radius (EDR) for the species concerned for each impulsive sound
generating activity.
Field measurements of the distance over which harbour porpoise respond to pile driving may be
expected to vary with pile diameter, hammer energy, etc. However, piles used at Alpha Ventus were
2.5m (500kj hammer energy) compared with the larger 4m piles used at the Horns Reef I and II (900kj
hammer energy) and reaction distances were broadly similar: 15-25km (Diederichs et al. 2009; Dahne et
al. 2013) and 18-21km (Brandt et al., 2011; Tougaard et al. 2009) respectively. It is proposed that for
unmitigated pile driving, the area affected should be based on an EDR of 26 km (Tougaard et al. 2013).
This results in an area of harbour porpoise ‘habitat loss’ of approximately 2,100 km2 for the day that a
single pile driving event occurs upon. If the precise location of the event is known, then it could be used
in calculation, if not the centre of the block should be used.
For seismic surveys, the EDR based upon the work of Thompson et al. (2013) would be 10 km. In this
case, the centre of the registration rectangle(s) that the seismic survey occurred within is proposed –
again on a daily basis.
For explosions, the EDR is not researched, but a precautionary approach might be to assume the same
as for seismic surveys.
An alternative, and simpler option, would be to use pulse block days, i.e. if any pulse occurs in a block in
one day, then it is assumed that those blocks are not available for the harbour porpoise on that day. This
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
25 | P a g e
approach would also lend itself more readily to management as it could use the same unit both for
recording and managing the pressure.
Calculate the effective loss of habitat per day/season/year for the assessment unit of the species
concerned.
The assessment unit might be the range of an animal, or an accepted sub-unit of that range. The unit
might also be a protected area or a specific sensitive habitat. Calculations of the area ‘lost’ within the
assessment unit can be done on a daily/seasonal or yearly basis, and are probably best expressed as a
percentage of that area, both in total extent and averages.
Consider thresholds
These could be set in a number of ways. Ideally the loss of habitat would then be translated and
expressed in terms of effects on population viability (Tougaard et al. 2013). Population level effects may
result from a reduction in carrying capacity of the habitat in the assessment area. Long-term, permanent
reduction in carrying capacity may manifest in population declines (Tougaard et al. 2013). Thresholds
could be expressed in terms of a maximum allowable ‘habitat loss’ per agreed period of time or an
average (or both).
Such indicator and potential thresholds would allow regulators to manage noisy activities by using
spatio-temporal restrictions if the proposed activities would be likely to exceed the set thresholds. This
would need to be based on information obtained from strategic plans and project specific applications.
The information obtained by the noise registries once the activities have taken place will also allow a
further check of whether those management procedures are successful.
This approach could also take account of other activities (beyond those generating impulsive noise) that
are known to displace animals. For instance, in the case of harbour porpoise, it is known that shipping
lanes that have about 10,000 or more movements per year displace porpoises from the lane area. This
could be readily incorporated into the indicator proposed here.
Description of pros and cons
Pros:
• Any species can be used where there is knowledge of disturbance area or it is assumed to be approximately the area of a noise reporting grid (it seems unlikely that a widely-distributed animal with a substantially smaller disturbance area would be affected at the population scale);
• it can be adjusted to any time period;
• can be applied to any assessment area or any size of area;
• avoids need for monitoring data for species, but can use detailed relative abundance information;
• does not make assumptions about noise propagation nor require complex noise modelling. Cons:
• relatively coarse scale, may not be fully compatible with detailed assessments of environmental impact of projects
Contact: Mark Tasker, [email protected]
References
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
26 | P a g e
Brandt, M.J., Diederichs, A., Betke, K., and Nehls, G. 2011. Responses of harbour porpoises to pile driving
at the Horns Rev II offshore wind farm in the Danish North Sea. Marine Ecology Progress Series, 421:
205–216.
Dähne, M., Gilles, A., Lucke, K., Peschko, V., Adler, S., Krügel, K., Sundermeyer, J., and Siebert, U. 2013.
Effects of pile-driving on harbour porpoises (Phocoena phocoena) at the first offshore wind farm in
Germany. Environmental Research Letters 8, 025002.
Diederichs A, Brandt MJ, and Nehls G 2009. Effects of construction of the transformer platform on
harbor porpoises at the offshore test field “alpha ventus.” Report to Stiftung Offshore-Windenergie,
BioConsult SH, Husum, Germany.
Tasker M.L. 2016. How might we assess and manage the effects of underwater noise on populations of
marine animals? Pp. 1139–1144 in Popper, A.N. and Hawkins, A. (eds.), The Effects of Noise on Aquatic
Life II. Advances in Experimental Medicine and Biology. Springer Science+Business Media, New York.
Thompson P.M., Brookes K.L., Graham I.M., Barton T. R., Needham K., Bradbury G. and Merchant N.D.
2013. Short-term disturbance by a commercial two-dimensional seismic survey does not lead to long-
term displacement of harbour porpoises. Proceedings of the Royal Society B: Biological Sciences. DOI:
10.1098/rspb.2013.2001
Tougaard, J., Carstensen, J., Teilmann, J., Skov, H., and Rasmussen, P. 2009. Pile driving zone of
responsiveness extends beyond 20 km for harbour porpoises (Phocoena phocoena, (L.)). Journal of the
Acoustical Society of America. 126, 11–14.
Tougaard, J., Buckland, S., Robinson, S. and Southall, B. 2013. An analysis of potential broad-scale
impacts on harbour porpoise from proposed pile driving activities in the North Sea. Report of an expert
group convened under the Habitats and Wild Birds Directive – Marine Evidence Group MB0138. 38pp.
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
27 | P a g e
5. French approach impulsive noise data gathering. SHOM, France - Florent
LeCourtois
Aim/purpose of the methodology
Impulsive noise sources have been identified to be responsible for population disturbance and lethal
injuries. The Good Ecological Status (GES) is defined when functional ecological zones are preserved and
noise does not induce excess mortality [1]. The methodology proposed for to evaluate descriptor 11
criterion 1 (D11C1) follows this definition.
Regarding these two risks, the aim of our evaluation is two-fold: the first one is to provide relevant
pressure indicators of D11C1 to enable a SMART (Specific, Measurable, Assignable, Realistic and Time-
related) management of the underwater noise; the second one is to use appropriate indicators as input
of the descriptor 1 to assess the impacts.
Brief description of the methodology and concepts
The impulsive noise register contains the data from impulsive source emission which have been certified
by operators. The data contain the duration (number of days), the location (in a 15’x15’ block), the
category of impulsive source and the level. The census follows the TG Noise guidelines [2].
We define three indicators: D11C1.1 concerns the temporal extent (the total number of days during a
predefined period), D11C1.2 concerns the spatial extent (the total number of blocks over an area) and
D11C1.3 concerns the distribution of emission levels [2].
Regarding these indicators, the GES is then defined depending on the risk to assess; it follows a decision
tree approach at the subregion level and over a trimester period. The GES is evaluated in regards to
temporal and spatial thresholds. The spatial extent is then assessed as a proportion of blocks where the
temporal extent exceeds a temporal threshold; the GES is achieved if this spatial extent is below the
spatial threshold. To assess the disturbance risk, the data from all certified emissions are used to
compute the extents. To assess the lethal risk, a similar decision tree is performed but only considering
high and very high level of emissions [2]. Depending on the risk, the temporal and spatial thresholds
have to be more or less restrictive: e.g. one can consider a risk of lethal injuries with a one day temporal
threshold (i.e. there is a lethal risk at the first emission). Both approaches introduce moderate risk and
high risk thresholds to adapt the of noise emission management.
Description of pros and cons
Pros:
• Easy to build: already applicable using noise register data;
• Allows noise budget in time and space for marine planning;
• Thresholds are explicitly expressed as recommended by EU;
• Easy to adapt: thresholds and decision orders can be changed depending on marine unit specificities;
• Several thresholds can be applied in order to identify low, moderate and high risks and to provide relevant mitigation solutions;
• Non-specific: it does not require dedicated and accurate knowledge on the marine life. Cons:
• Thresholds are defined empirically: work in progress to report the indicators to descriptor 1.
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
28 | P a g e
• Propagation effects are not taken into account
Contact persons and References
Florent Le Courtois, G. Bazile Kinda, Yann Stéphan
[1] MEDDE. (2012). Arrêté du 17 décembre 2012 relatif à la définition du Bon Etat Ecologique.
Journal Officel de la République Française, 27.
[2] Dekeling, R., Tasker, M., Van der Graaf, A., Ainlie, M., Anderson, M. A., Brensing, K., … Young, J.
(2014). Monitoring Guidance for Underwater Noise in European Seas. Luxembourg: Publications Office
of the European Union.
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
29 | P a g e
6. BIAS GES tool. FOI - Swedish Defence Research Agency – Mathias Anderson
Aim/purpose of the methodology
Continuous noise covers large areas of the seas and has been observed to vary both temporally and
spatially. It is known that marine species often moves over large areas and utilizes specific areas during
certain times of the year. Thus, in order to estimate the Sound Pressure Level (SPL), both time (year or
month) and space (area and water depth) need to be considered, as well what frequency (63, 125, 2 000
Hz) to study. The BIAS Soundscape planning tool was developed by the EU LIFE project Baltic Sea
Information on the Acoustic Soundscape (BIAS) for the purpose of quantifying the pressure from ship
noise in the Baltic Sea. The planning tool can be the basis for study of the impact on marina animals
once, thresholds for impact are established (Nikolopoulos et al., 2016).
Brief description of the methodology and concept (non-technical)
The soundscape tool are based on measured and modelled soundscape data and provides a number of
functions to evaluate the spatial and temporal sound characteristics within a user-defined geographical
area. The tool is based on several hundred soundscape maps covering different frequencies, depth
intervals and exceedance levels, where e.g. 5% exceedance level describes the highest SPL in an area
that occurred 5% of the time (this will give the strongest sources in the area) and 90% exceedance level
will represent SPL 90% of the time (this level will often represent the background noise level).
Two main assessment methods are implemented in the tool. Both are designed to assess the pressure
during a certain time and in a specific area. The data from these assessments can later be used to study
the impact on marine life or to establish if threshold for impact was exceeded or not.
Temporal Exposure Assessment - TEA
This method gives the proportion of area for which a SPL is exceeded a certain percentage as a function
of time for the specified month of the year, centre frequency and depth interval. For example, the
percentage of an interest area where a species is residing during mating period is exposed to SPL higher
than a certain level 5% of the time (the 5% exceedance level).
Spatial Exposure Assessment - SEA
The other method will give the proportion of area for which a SPL surpass a user defined SPL threshold,
a certain percentage of time for a specified time period, centre frequency, and depth interval. This
method is used when there is a specific threshold level of interest, for example sound levels related to
masking.
Description of pros and cons
The user of the tool needs to have a basic understanding of the topic underwater noise and impact in
order to define the parameters that the tool will use to calculate the desired statistics for the area and
time of interest. The highest temporal resolution is monthly averages the spatial resolution can be
decided by the user.
References and Contact person
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
30 | P a g e
Nikolopoulos A., Sigray P., Andersson M., Carlström J., Lalander E., 2016: BIAS Implementation Plan -
Monitoring and assessment guidance for continuous low frequency sound in the Baltic Sea, BIAS LIFE11
ENV/SE/841. Available from www.bias-project.eu.
Contact person: Mathias Andersson
FOI - Swedish Defence Research Agency, SE- 164 90 Stockholm, Sweden
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
31 | P a g e
7. Bioacoustics-Based Impact Indicator (BBII). Laboratory of Applied
Bioacoustics (LAB) of the Technical University of Catalonia, Barcelona
(UPC, Spain) – Michel André
Aim of the methodology
The Bioacoustics-Based Impact Indicator (BBII) aims at developing a decision support tool for the
continuous monitoring of ecosystem changes in the marine environment and assessment of whether
these changes were caused by the introduction of anthropogenic sound sources. It works by
automatically and continuously tracking biological and anthropogenic acoustic activities with dedicated
instruments and analysis modules. The resulting indicators allow the informed analysis of relationships
between changes in anthropogenic sound levels and changes in ecosystem health.
Brief description of the methodology and concept (non-technical)
Assessing the impact of anthropogenic noise on the marine environment requires objective data on the
sensitivity to noise of representative species. Currently, such data is scarce on most studied species
(marine mammals and fishes) and recent findings show that non-hearing specialists like invertebrates
suffer acoustic trauma when exposed to certain sounds. It will probably take decades before we can
scientifically assess species-specific thresholds for a broad spectrum of species, which would allow a
global approach to predict impact on marine fauna.
Here, we propose to use the most recent developments in bioacoustics analysis to automatically
separate and estimate the contributions to the overall noise budget in a given area, in particular (1) the
anthropogenic sources and (2) the biological sources which serves as a proxy to biological activity. This
approach assumes that a good environmental status of a given area is associated with a diversity of
biological activities, and in turn associated to the production of biological sounds. The methodology
requires the automated real-time processing of audio streams.
The automated real-time processing of audio streams is based on machine learning methods and the
post-processing to obtain the final indicators of anthropogenic and biological activity is based on
statistical methods. These indicators will inform on the composition and relative density of acoustically
active species and received levels from anthropogenic sources in the area at a fine temporal scale. Joint
analysis of these two indicators can reveal potential impact of anthropogenic sounds on ecosystem
health. Due to its ability to work continuously and with low latency, BBII is expected to be an important
decision support tool for governmental agencies, conservation NGOs, and other stakeholders.
Description of pros and cons
Pros:
• Automated processing allows an accurate quantification of relative changes. Indicators can be calibrated initially by punctual direct assessment of ecosystem health with more invasive methods which could even allow quantification of absolute changes.
• Automated method can run 24/7 over the long term, therefore offering a complete picture of an ecosystem’s temporal patterns (e.g. seasonal, diurnal, tide-related)
• Monitoring acoustic stations are required to define indicators: Indicator 11.2 monitoring stations can be used
• Indicators can be defined without awaiting further data on species-specific thresholds
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
32 | P a g e
Cons:
• As any proposed methodology and measurement method, the indicators must assume a certain level of uncertainty
• A given area may not be bio-acoustically active to be able to define a reference value
• Distant sources (biological and anthropogenic) might be taken into account locally
References and contact person
The BBII approach is jointly developed by the Laboratory of Applied Bioacoustics (LAB) of the Technical
University of Catalonia, Barcelona (UPC, Spain) and Quiet-Oceans (France)
Contact person: Michel André ([email protected])
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
33 | P a g e
8. French approach on continuous noise. SHOM, France – Florent LeCourtois
Aim/purpose of the methodology
Continuous noise sources in the oceans can cause masking of the animal communication. The Good
Ecological (GES) is defined when communication capabilities are not downgraded [1]. This definition can
be interpreted in terms of negative or null trend in the shipping noise levels. The methodology proposed
to evaluate of descriptor 11 criterion 2 (D11C2) follows this definition.
Regarding this risk, the aim of our evaluation is two-fold: the first one is to provide relevant pressure
indicators of D11C2 to enable a SMART (Specific, Measurable, Assignable, Realistic and Time-related)
management of the underwater noise; the second one is to use appropriate indicators as input of the
descriptor 1 to assess the impacts.
Brief description of the methodology and concept
Shipping is one of the dominant anthropogenic sources of underwater noise at low frequency. Our
approach is based on the modelling of the shipping noise. The modelling of the emitted and propagated
levels is explained in [2]; in addition, the model estimates the propagation uncertainties using basin-
scale scenarios of environmental mismatches.
The shipping noise is then computed in discrete meshes for 4 months (one for each season: January,
May, August and November), at several depths between 5 and 300 m, and for the one-third octave
bands centred on 63 Hz and 125 Hz. The yearly maximal levels are computed by selecting only the
maximal levels per depth and per month for each mesh. The spatial distribution of the maximal levels in
the subregion units are the indicators D11C2.1 and D11C2.2 for the one-third octave bands respectively
centred on 63 and 125 Hz.
The maps of annual maximal levels are computed for 2 years: 2012 (year of reference according to TG
Noise) [3] and 2016 (during the evaluation cycle). The difference between the 2016 and 2012 maps
leads to the spatial distribution of the trend in the maximal shipping noise levels. For each subregion
unit, the percentile values of the maximal level differences are investigated.
The spatial proportion of the positive and negative trends and the magnitude of the percentiles are
compared to the uncertainties to provide a qualitative assessment of the GES. Additional information
such as reports from the MRCC and underwater noise opportunistic measurements [4] are used to
strengthen our analysis and to reduce the uncertainties.
Description of pros and cons
Pros:
• Already applicable: scientifically validated;
• Global approach: basin scale analysis;
• No biological information is required;
• Estimate the uncertainties;
• Allows marine space planning by considering ship categories [2];
• Can be easily adapted: i.e. monthly analysis, focus on a zone and at given depth, monitoring of higher frequencies…
TG Noise Thematic workshop - Torrelodones, Spain 9-10 November 2017 Towards thresholds for underwater noise
Common approaches for interpretation of data obtained in (Joint) Monitoring Programmes
34 | P a g e
Cons:
• Trends are difficult to estimate in a robust way: long term monitoring of the trend is required to provide statistic robustness;
• AIS and environmental data are required;
• Measurements are required to validate and to correct the modelling (i.e. in shallower zone, at higher frequencies or when the AIS data are not comprehensive).
References and Contact person
Florent Le Courtois, G. Bazile Kinda, Yann Stéphan
[1] MEDDE. (2012). Arrêté du 17 décembre 2012 relatif à la définition du Bon Etat Ecologique.
Journal Officiel de la République Française, 27.
[2] Le Courtois, F., Kinda, G.B., Stéphan, Y., Boutonnier, J.-M., and Sarzeaud, O. (2016). Statistical
ambient noise maps from traffic at world and basin scales. Proceedings of the Institute of Acoustics,
Cambridge (UK).
[3] Tasker, M. L., Amundin, M., Andre, M., Hawkins, A., Lang, W., Merck, T., ... & Zakharia, M. (2010).
Marine Strategy Framework Directive Task Group 11 Report Underwater noise and other forms of
energy. Report No. EUR, 24341.
[4] Kinda, G. B., Le Courtois, F., & Stéphan, Y. (2017). Ambient noise dynamics in a heavy shipping
area. Marine Pollution Bulletin, 124(1), 535-546.