Novel low-cost methods for marine mammal and environmental monitoring

Lead Research Organisation: Newcastle University
Department Name: Sch of Natural Sciences & Env Sciences

Abstract

The UK is rapidly entering a phase where the majority of its energy comes from renewable resources. However, to
maintain this push towards green energy providers will need to find more sites to generate power. In recent years there
have been moves towards using offshore wind generation instead of onshore wind farms. Whilst offshore wind has many
advantages there is potential impact on marine wildlife that needs to be considered. New applications for offshore windfarm
sites require a thorough Environmental Impact Assessment and the offshore environment makes it hard to accurately
assess current marine populations, making this a time-consuming and costly operation.

In this research we propose to develop a real-time acoustic monitoring platform that will enable operators to assess a site
for marine mammal activity at a fraction of the current cost and effort and moreover, produce better and more accurate
results. This will be done by deploying large numbers of novel low cost, low power underwater wireless sensor devices
(NanoPAMs) which detect the sounds that marine mammals make to navigate, hunt and communicate. The data from each
NanoPAM is transmitted using acoustic signals over distances up to 10 km to a surface buoy. Since the animals may be
able to hear these acoustic signals, the NanoPAMs transmit very low energy acoustic signals which blend into the
background noise and are also smart enough to wait until animals have left the area before transmitting so as not to
influence behaviour. The surface buoy then uses radio to send the received data back to shore where it can be visualised
and analysed.

On shore, the data collected can be turned into an understanding of the current activity of marine mammals and also to
monitor the ongoing impact of offshore windfarm developments. In future, the system could be rapidly deployed in multiple
locations to help select sites that would minimise the impact on our marine wildlife.

Planned Impact

The research conducted and the resulting technologies and methods will benefit end-users that plan to conduct marine
anthropogenic activities e.g. offshore construction. It will provide the complete hardware and software system, and
associated analysis tools necessary to conduct assessment of marine mammal occurrence and noise in the area of
interest. The results from using the system will provide the necessary data and information for end-user and government
agencies (e.g. MMO, Cefas and IFCAs) to inform decision making and inform policy. It will further allow end-users to be proactive
and conduct Environmental Impact Assessments in multiple areas simultaneously (given the low cost of the methodology) to
select candidate sites for their intended development that would represent minimum environmental impact and therefore
increase the likelihood that applications are given consent.

In contrast to currently available autonomous passive acoustic recorders, the proposed methodology will be underwater
acoustically networked passive acoustic monitor devices (NanoPAM) and a surface gateway buoy with radio
communication back to operators on shore. The proposed system will transmit data back to shore in near real time,
practically eliminating data loss (recording devices being commonly lost) and facilitating much more timely analysis and
decision making. The in-built acoustic communication and positioning capability also facilitates recovery and minimises the
risk of hardware loss. The low cost of the hardware and combined with the reduction in vessel time for
deployment/recovery and data upload will bring about a step change in the cost effectiveness of marine mammal surveys.

This development also represents an opportunity for significant commercial impact via enhanced product offerings from
marine instrumentation companies with whom the investigators have close links. Marine mammal monitoring technology is
currently relatively high cost and with a high risk of deployment, whereas this technology opens up the market for much
lower cost, higher volume products with worldwide market potential. The research team have a strong track record in
transferring underwater acoustic technologies to industrial partners and successful commercialisation via licensing.

Publications

10 25 50
 
Description This award has developed NanoPAM a novel low-cost networked system for monitoring dolphin, porpoise and vessel activity in the marine environment that deliver the data to the user on land. NanoPAM consists of miniature broadband hydrophones and ultra-low energy digital signal processing that detect and classify high frequency (20-160 kHz) click sounds.

The high-level data of animal detections are transferred back to the user on shore in near real-time via a multi-hop, underwater acoustic communication network. A single surface gateway buoy, with Wi-Fi link to shore, may support networks of up to 255 NanoPAMs spaced up to 2km from one another, allowing near real-time monitoring of up to ~450km2 , with battery capacity for deployments of =1 year. Using underwater communication eliminates the need for surface buoys or markers at each node, making the devices more secure by greatly reducing the risk of entanglement with vessels, fishing gears and animals.

Furthermore, the acoustic communication network is based on very low energy signals whose characteristics blend into the background noise. NanoPAM sound emissions have been demonstrated to be a tiny fraction of that from vessels and hence the technology achieves a net reduction in anthropogenic noise compared to conducting vessel surveys. NanoPAM is a novel alternative to currently available archival PAM recorders that greatly reduces the costs of periodic PAM servicing, whilst minimising the risk of data loss resulting from lost or stolen recorders. The in-built acoustic communication and positioning capability also facilitates recovery and further minimises risk of hardware loss. Prototype hardware and software for NanoPAM have been completed, validated via lab and shore tests and full system trials conducted off Blyth, Northumberland.

As part of the award and system development we have developed algorithms for identification of porpoise, dolphin and vessel echolocation signals which have been incorporated into the devices' hardware. Analysis of recorded echolocation signals has led to the development of classification techniques which can discriminate porpoise and dolphins near 100% reliability. The bio-acousticians have worked together with the electronic engineers to translate these findings into algorithms and software for implementation on low energy battery powered instruments.
Exploitation Route Detection devices for marine mammals may be used to provide near real time reporting of animal distribution and occurrence for scientific studies on marine biodiversity and impact assessment (from offshore construction activities, fisheries etc).

The project Team has submitted a new NERC grant to the call for Demonstration Projects for Constructing a Digital Environment. This proposal will develop and demonstrate a second generation of the NanoPAM system to provide a sensor network (NanoPAM2) that expands the capability to include lower frequency signal monitoring (i.e. whistles and calls, and vessel and other noise) in addition to the echolocation signals. The NanoPAM2 will also have long-range Wi-Fi and/or satellite communication allowing for the network to be deployed and transmit data far offshore expanding the potential areas of applications.
Sectors Communities and Social Services/Policy,Construction,Electronics,Energy,Environment

 
Description Citizen scientists are engaged in the NanoPAM project - where Newbiggin Dolphin Watch Association (https://www.newbiggindolphinwatchassociation.com) can access the deployed data buoy for real-time acoustic detection of cetaceans off Blyth, Northumberland (https://www.youtube.com/angy-sealab).
First Year Of Impact 2019
Sector Communities and Social Services/Policy,Education,Leisure Activities, including Sports, Recreation and Tourism,Culture, Heritage, Museums and Collections
Impact Types Societal

 
Description Global listening Network for Turbidity Currents and Submarine Geohazards
Amount £846,911 (GBP)
Funding ID NE/S010068/1 
Organisation Natural Environment Research Council 
Sector Public
Country United Kingdom
Start 04/2019 
End 03/2024
 
Title NanoPAM cetacean click train detectors 
Description Prototypes for "NanoPAM" cetacean click train detectors have been designed, constructed and successfully tested in a realistic environment. The devices consume approximately 2mW of power when listening and so are able to operate for periods of 6-12 months from a small battery pack. The devices detect and lock onto potential click trains, by low energy statistical analysis of inter-click-interval (ICI) patterns, and then sample a series of clicks to analyse signal parameters such as centre frequency, bandwidth and peak to average ratio. These parameters are then used to do a multivariate classification of the train to decide whether it is a dolphin, porpoise or unclassified (e.g. man made sonar, or other random events). Distinction is also made between low rate click trains (typically navigation) and high rate or "buzz" clicks typically used during foraging/hunting. Each unit collates its detection data over each hour (with a resolution of 1 minute) and transmits this back to a surface gateway buoy via a low energy underwater acoustic modem and then back to shore via a long range radio link. Hence animal detections are reported to the end users with a maximum latency of 1 hour (compared to weeks or months for a typical logging instrument). The devices have been deployed in the North Sea for a period of 6 weeks to date and have demonstrated the detection of both dolphins and porpoises in the area throughout this period. Comparison with (time consuming) human analysis of continuous logging hydrophone recordings in the same area suggest that all ~70 of the detections reported are genuine and so the false positive rate of the detectors is very low. However a significant number of weaker (more distant) signals were not detected so work is on-going to improve the sensitivity of the hydrophone element and analog front end of the NanoPAM instrument to increase the detection range. Improved NanoPAM units will be tested before the end of the project. 
Type Of Material Improvements to research infrastructure 
Year Produced 2020 
Provided To Others? No  
Impact Too early for impact from this development at the time of writing. 
 
Title Algorithms for species classification from echolocation click sounds 
Description Species classification algorithms in development using field recordings of echolocation click signals from harbour porpoise, white-beaked dolphin and common bottlenose dolphin (the latter supplied by Aberdeen University), the three most common echolocating species off the UK northeast coast. The sound data have been digitally filtered to minimise the influence of low frequency sounds, including whistles and ambient noise. The data have been examined using spectrograms to select and extract loud and clear echolocation click signals for each species for species classification algorithm development. In addition, signals originating from vessel echo-sounders were also identified and extracted. Eleven spectral parameters were measured and computed for all click signals using custom written scripts in MATLAB. The results have been used to develop algorithms that separate vessel sonar from porpoises and dolphin click signals by near 100% probability, porpoises and the two dolphin species by near 100% probability, and the two dolphins species by 85% probability. The next steps in the algorithm development include: - To incorporate the sonar signals, dolphin and porpoise clicks separation algorithms to the hardware and to conduct tank test with test click signal data. This will be followed by field performance testing at sea. - Test how sample size, frequency resolution, selection and classification method affect the probability of correct classification to design a robust algorithm for the real-time system. - Explore how to improve the correct rate of the click signal separation algorithm between the two dolphin species. 
Type Of Material Computer model/algorithm 
Year Produced 2019 
Provided To Others? No  
Impact The species classification algorithms will be published in the scientific literature and made open source once finished which will benefit other scientists and developers in the field. 
 
Title Open Source Geospatial Database for NanoPAM data 
Description A Docker container, which is a lightweight, standalone, executable package of software , was configured. It has a PostGIS enabled PostgreSQL database instance. Which is an open source database with a spatial extension allowing for spatial query to be performed on the data. The schema of the database itself was generated via Django which is a high-level Python Web framework. The data is sharded/subdivided using the Django Sharding a python library/ framework for subdividing Django database tables. This means not all reading are stored in one massive table but instead are divided into sub tables given an attribute, in this case their timestamp. And all interactions, insertions and queries, are performed via HTTP requests using the Django REST framework , a toolkit designed for building APIs. 
Type Of Material Database/Collection of data 
Year Produced 2020 
Provided To Others? No  
Impact Using a sharded database built under a restful API for the NanoPAM data will mean that data can be continuously ingested into the database with minimal impact of data query time as the amount of the data stored continues to grow. 
 
Title Containised Platform for NanoPAM Data 
Description A number of Docker containers,a lightweight, standalone, executable packages of software, have been setup for visualisation and delivery of NanoPAM data. The first is a database (as described in Open Source Geospatial Database for NanoPAM data). Secondly there is a Django API container. This container uses Django, a high-level Python Web framework and the Django REST api framework. to manage all interactions with the database. This not only ensures, data is easy accessible but manages all the authentication with creating new records or deleting records, without having to expose the database itself. There is also a microservices container which is a collection of lightweight python scripts which, via HTTP request made through the API , ingests all the raw data files into the database. Finally there is a second Django container that acts as a portal, which queries data from the API then displays the data, both in terms of maps and graphs. 
Type Of Technology Webtool/Application 
Year Produced 2020 
Impact The platform allows easy ingestion of marine data, an API allowing access to the data and a front end allowing fast on-the-fly visualisations of the marine data in near real time. 
 
Title Development of algorithms for acoustic detection of marine mammals on very low energy hardware platforms. 
Description Algorithms and software implementation for the detection and classification of Dolphin and Porpoise click trains using very low power embedded microprocessors. 
Type Of Technology Software 
Year Produced 2019 
Impact This enables the construction of Passive Acoustic Monitoring devices with real time reporting of animal detections which may be deployed for long periods on battery power sources. 
 
Description ANGY (Acoustic Network Gateway) Youtube channel 
Form Of Engagement Activity Engagement focused website, blog or social media channel
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Public/other audiences
Results and Impact The acoustic gateway buoy constructed for the USMART project was deployed in August 2019 and the hydrophone signals are streamed live to a Youtube channel for public engagement and citizen science activities. This has generated major interest from local groups studying dolphin distribution who have been able to listen to whistles and echolocation signals from Dolphins off the Northumberland coast. Plans are being made for how this facility can be used more widely for similar research and public engagement and it has been written into a number of grant proposals.
Year(s) Of Engagement Activity 2019,2020
URL https://www.youtube.com/ANGY-SEALAB
 
Description Challenger Society Conference 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Industry/Business
Results and Impact Presented the project to a wide audience of research, industry and government representatives at the Challenger Society annual conference in Newcastle.
Year(s) Of Engagement Activity 2018
 
Description Low cost dolphin & porpoise bycatch mitigation & assessment 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Policymakers/politicians
Results and Impact Presentation on Friday 22nd November 2019 to the UK Bycatch Focus Group with participants from government agencies: MMO, DEFRA, JNCC, CEFAS, IFCA, Environment Agency and NGO: WWF, WDC, SMASS, CSIP and Seafish.
Year(s) Of Engagement Activity 2019
 
Description NanoPAM: A novel low-cost acoustically networked system for marine mammal monitoring 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Professional Practitioners
Results and Impact Invited presenter and participation in a Workshop on acoustic monitoring of population trends at the World Marine Mammal Conference in Barcelona Spain 8 December 2019.
Year(s) Of Engagement Activity 2019
URL https://www.wmmconference.org/workshops/
 
Description NanoPAM: A novel low-cost acoustically networked system for marine mammal monitoring 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Professional Practitioners
Results and Impact Poster presentation at the World Marine Mammal Conference in Barcelona, Spain, 9-12 December 2019.
Year(s) Of Engagement Activity 2019
URL https://www.wmmconference.org/detailed-program/
 
Description Presentation to Northumberland Inshore Fisheries Conservation Authority and stakeholders 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Policymakers/politicians
Results and Impact Presented details of this project and planned trial activities to NIFCA and a variety of stakeholders including local fishermen and Natural England. This raised awareness of the technical possibilities of the system and led to several discussions on collaboration, participation trials or long term applications.
Year(s) Of Engagement Activity 2018
 
Description Presentation to USMART project end user workshop 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Industry/Business
Results and Impact Per Berggren presented on the NERC project and acoustic monitoring of cetaceans to an audience of industry end users, including several parties interested in impact assessment/mitigation of offshore construction activities.
Year(s) Of Engagement Activity 2018
 
Description UK Bycatch Focus Group 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Professional Practitioners
Results and Impact Per Berggren gave an invited presentation to the UK Marine mammal Bycatch Focus Group meeting held at the Welsh Government Offices in Cardiff on 14 December 2018. The talk included a presentation of the NERC Novel low-cost methods for marine mammal and environmental monitoring development. There was considerable interest and follow-up questions from the meeting attendees that highlighted the usefulness of the system in development. The Bycatch Focus Group meeting was attended by the following stakeholders: JNCC, DEFRA, Welsh Government, Scottish Government, Scottish Natural Heritage (SNH), Ministry of Defence (MOD), Natural Resources Wales (NRW), Centre for Environment, Fisheries and Aquaculture Science (Cefas), Natural England (NE), and the Cetacean Strandings Investigation Programme (CSIP).
Year(s) Of Engagement Activity 2018