IVL Swedish Environmental Research Institute

ivl.se
Change search
Refine search result
1 - 7 of 7
CiteExportLink to result list
Permanent link
Cite
Citation style
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Rows per page
  • 5
  • 10
  • 20
  • 50
  • 100
  • 250
Sort
  • Standard (Relevance)
  • Author A-Ö
  • Author Ö-A
  • Title A-Ö
  • Title Ö-A
  • Publication type A-Ö
  • Publication type Ö-A
  • Issued (Oldest first)
  • Issued (Newest first)
  • Created (Oldest first)
  • Created (Newest first)
  • Last updated (Oldest first)
  • Last updated (Newest first)
  • Disputation date (earliest first)
  • Disputation date (latest first)
  • Standard (Relevance)
  • Author A-Ö
  • Author Ö-A
  • Title A-Ö
  • Title Ö-A
  • Publication type A-Ö
  • Publication type Ö-A
  • Issued (Oldest first)
  • Issued (Newest first)
  • Created (Oldest first)
  • Created (Newest first)
  • Last updated (Oldest first)
  • Last updated (Newest first)
  • Disputation date (earliest first)
  • Disputation date (latest first)
Select
The maximal number of hits you can export is 250. When you want to export more records please use the Create feeds function.
  • 1. Genell, Anders
    et al.
    Johansson, Torbjörn
    IVL Swedish Environmental Research Institute.
    Andersson, Carl
    IVL Swedish Environmental Research Institute.
    Glebe, Dag
    IVL Swedish Environmental Research Institute.
    Krång, Anna-Sara
    IVL Swedish Environmental Research Institute.
    Ship noise in coastal regions - final report of the SHIPNOISE project2024Report (Other academic)
    Abstract [en]

    Ships carry cargo and passengers around our planet. In coastal regions, complaintsabout ship noise have been raised from nearby residents. Ships are also thedominant source of continuous underwater noise in the seas.

    In the SHIPNOISEproject, we investigate levels of ship noise in coastal regions using a custommeasurement station for both airborne and underwater noise from shipsunderway. The measurement station is developed using embeddedmicroprocessors for low power consumption. The station is deployed at Böttö atthe inlet to Gothenburg harbor, and then at Lurö in Lake Vänern.

    These sites areclose to shipping lanes of different traffic intensity. The measured airborne noiselevels indicate that there is a risk to exceed recommended indoor low-frequencynoise limits for dwellings positioned up to several hundred meters from thepassing ships, although the effect on public health is uncertain.

    The underwaternoise recorded at 200-300 m range at the SHIPNOISE measurement locationsduring ship passages is strong enough to exceed levels for environmental impactpreviously demonstrated on local marine mammals, fish and possiblyinvertebrates. For example, harbor porpoises, herring and salmon are likely toavoid or escape the area when a loud ship passes.

    Download full text (pdf)
    SHIPNOISE
  • 2.
    Glebe, Dag
    et al.
    IVL Swedish Environmental Research Institute.
    Johansson, Torbjörn
    IVL Swedish Environmental Research Institute.
    Martinsson, John
    RISE.
    Genell, Anders
    VTI.
    Bullerdatainsamling och autonom artidentifiering för att underlätta miljöövervakning2022Report (Refereed)
    Abstract [en]

    The trend for Sweden’s environmental goals is in several cases negative, and one ofthe areas that shows an undesirable trend is the goal “A rich diversity of plant andanimal life”. Sweden’s follow-up to the habitat and bird directive and the shows acontinued vulnerable situation for biological diversity. But today it is difficult andcostly to monitor the Swedish environmental quality targets are met.The collection of sound and image data already takes place on a large scale today,but there is a large and untapped potential to simplify environmental monitoringthrough new cheap data collection devices and above all through new automaticAI-based analysis systems. Manual sampling and data collection is time-consumingand costly, which makes autonomous audio and video data collection attractive,especially in inaccessible locations such as underwater. In several cases, the publiccould be enlisted to help detect invasive species using a cell phone application thatuses artificial intelligence for species identification.This report reviews the state of the art in the use of sound and image data fornoise monitoring and noise mapping, for species identification of animals and plants,and for invasive species monitoring. The report reviews current and emergingtechnologies and methods and assesses their maturity, availability and reliability.This report aims to report opportunities and challenges linked to:• Systems to collect noise data to model noise impacts in the terrestrial andaquaticdomains and enable noise mapping• System for autonomous sound and image-based species identification,populationestimation and biodiversity monitoring• System for the detection of invasive species, e.g. mobile phone appsTechnical solutions and methods must be based on open data and open design.Echoacoustic noise data acquisitionBiological applications in acoustics are called bioacoustics or echoacoustics. Themost explored areas of passive acoustic monitoring, or PAM, are animal sounds inthe ultrasonic range (e.g.bats and whales), as traditional analysis methods can beused in those cases. For example, PAM can be realized with sound boxes, single orin arrays, or with collars or implanted devices on individual animal individuals.Acoustically active animals align themselves with each other and other sounds,which is why soundscapes are considered to provide information about the healthof the biosystem.It is important to address the entire measurement chain in PAM technology.The report’s overview of hardware is focused on audio, because camera technologyis established in Swedish wildlife management. Most hardware components in themeasurement chain need to be chosen from a practical point of view to work wellin measurement systems, such as battery life, memory management or connectivityoptions, and low cost if many acquisition devices are needed.

    The report also contains an overview of autonomous systems or integrated devicesthat cover the entire measurement chain when collecting sound data. The possibilitiesof combining audio and visual data for analysis are rarely used today, andthere can be great gains to be made in this field.Successful AI-based analysis methods have not made it into commercial applications.There is great pressure within the research community to make analysisresults, analysis tools and collected data available, to be able to reuse data andresults for resource reasons, and to offer greater data coverage. Standardized formatsfor metadata are also requested, aiming at international research practice.The success of citizen research can be partially attributed to new tools implementedin mobile apps, but there is great development potential to tailor toolsand methods to businesses and actual needs.AI in bioacoustics – State of the artThe AI-based methods most used in bioacoustics are deep learning methods, aboveall different forms of neural networks that are suitable for resource-demandinglistening and image review. The largest area is bird classification.The most common type of neural network in bioacoustics is CNN (convolutionalneural network), which is important in image and sound analysis, but new variantsare constantly being developed. Spectrograms (image representations of sound)are often used as input to deep learning models, but many variants are in use. Melspectrogramis the one that has worked best in bioacoustic contexts, because thefrequency scaling depicts the sound the same regardless of pitch, which is suitablefor CNN-type networks. Note that the same raw data can form the basis of differenttypes of training input, if they are of sufficient quality and resolution. The requirementsfor high-resolution data are expected to increase, which is important whencollected raw data is to be future-proofed, as are the requirements to link metadatato registered observations.In the machine listening area, there has been a fifteen-fold increase in thenumber of publications between 1998 and 2018. Estimation of population densityof birds using machine listening has been shown to give as good results as manualpoint counting, with respect to key parameters such as number of recorded birdspecies.Methods that do not require manual classification of training data are apromisingway forward for noise reduction and source separation. Annotationscarcity of training data can be addressed with embedding functions. In the areaof urban noise, there are methods for real-time data streaming using distributednetworks. Active learning methods, i.e. methods where experts actively participatein the learning process, quickly produce powerful results. An interesting variantis to train animals to make choices that become annotations of input data. Thisprovides a model representing the animals’ own perception, which must be usedwith caution.In summary, an extensive investment in noise data collection outside Sweden’sbuilt-up areas or around Sweden’s coasts does not seem realistic in today’s situation,but there are great opportunities to fruitfully integrate autonomous species identificationinto the ongoing monitoring activities in Swedish nature.

    Download full text (pdf)
    fulltext
  • 3.
    Jacobson, Anton
    et al.
    IVL Swedish Environmental Research Institute.
    Johansson, Torbjörn
    IVL Swedish Environmental Research Institute.
    Fridén, Håkan
    IVL Swedish Environmental Research Institute.
    Björk, Anders
    IVL Swedish Environmental Research Institute.
    Detektion av Givar- och Kommunikationsfel vid Dammövervakning - Projektet Optimerade och Anpassade Datadrivna Metoder för Dammsäkerhet2021Report (Other academic)
    Abstract [sv]

    Dammövervakning är ett stort och viktigt område för svenska dammägare. Det är en utmaning att kontinuerligt övervaka en dammkonstruktion och snabbt kunna reagera på förändringar som visar på en avvikelse i dammen. Samtidigt kan även avvikelser i själva övervakningsutrustningen förekomma, som skulle kunna uppfattas som en avvikelse i dammen.  Sådana fel är vanligt förekommande och beror typiskt på fel i givare eller i mätvärdets överföring från givare till datacentral. För att felen inte ska störa dammövervakningen behöver dessa fel detekteras, identifieras och skiljas från avvikelser som kan tyda på en förändring i dammen. Denna rapport behandlar just denna detektion och identifiering av givar- och kommunikationsfel vid dammövervakning. 

    Rapporten är resultatet av ett forskningsprojekt som drivits av IVL Svenska Miljöinstitutet AB under 2020–2021. Projektet finansierades av Energiforsk och Stiftelsen Institutet för Vatten- och Luftvårdsforskning (SIVL) och hade som slutmål att implementera metoder för detektion av givar- och kommunikationsfel i ett dammövervakningssystem. Detta mål uppnåddes, och för att nå dit utfördes flera steg.

    Datadrivna analysmetoder utvecklas till att börja med typiskt off-line. Detta kräver en stor mängd historiska data. Att samla in, bygga upp en gemensam förståelse av och förbehandla dessa data kräver mycket tid och ett gott samarbete mellan dataspecialister och dammägare. Förbehandling är ett nödvändigt steg för att kunna utveckla en övervakning av dammens tillstånd. Tillståndsövervakning var dock inte del i detta projekt. Genom ett gott samarbete med en anläggningsägare fick projektet tillgång till 5 års data från 81 givare placerade på en damm i Sverige. Data innehöll ett nytt mätvärde var 15:e minut. Totalt utgör denna datamängd 14,2 miljoner mätvärden. Metoder för detektion av givar- och kommunikationsfel byggdes upp efter noggrann genomgång av data och identifiering av olika typer av förekommande fel. Här har samarbete med vår referensgrupp med medlemmar från olika kraftbolag varit ovärderlig. Metoderna byggdes upp från grundläggande signalanalysteori, men fanns trots sin relativa enkelhet fungera väl. Sju olika metoder har tagits fram och utvärderats i denna rapport. Tillsammans med dammägaren, Vattenfall, har en pilotimplementation genomförts i deras system. På grund av de krav som säkerhetslagstiftningen ställer kunde IVL endast få ett begränsat interface till realtidsdata. Trots detta har en implementation sjösatts och varit i drift i 5 veckor när denna rapport skrivs. För att effektivisera utvecklingen byggdes först en lokal utvecklingsversion som sedan installerades på en server hos Vattenfall. Den lokala versionen kommer också att användas i det fortsatta metodutvecklingsarbetet.

    Detta projekt har tagit fram feldetektion och förbehandlingsmetoder som kan ligga till grund för en kostnadseffektiv utveckling av multivariata och andra maskininlärningsmetoder för övervakning av en eller flera dammars tillstånd. 

    Download full text (pdf)
    fulltext
  • 4.
    Johansson, Torbjörn
    et al.
    IVL Swedish Environmental Research Institute.
    Andersson, Carl
    IVL Swedish Environmental Research Institute.
    Genell, Anders
    Winroth, Julia
    IVL Swedish Environmental Research Institute.
    Von Elern, Fredrik
    Noise from ships powered by LNG or electricity and its effects: a cross-domain investigation: Final report of the Silent@Sea project2023Report (Other academic)
    Abstract [en]

    Electrification of ships offers zero-emission travel and is spreading rapidly, and more and more ships are operating on liquid natural gas, LNG, or other alternative fuels. However, the relation between these modern forms of ship propulsion and noise pollution is not generally understood.

    The Silent@Sea project has investigated this through four case studies, where modern vessels have been measured in different propulsion modes and compared to sister vessels.

    This has mainly been done in route, which permitted us to gather unique data on the noise radiation of large ships in commercial operation. The project has investigated radiated airborne and underwater noise as well as onboard noise and its impact on work environment and passenger comfort.

    The results show that the modern forms of propulsion lead to lower noise levels onboard, which are coupled to a better work environment and greater passenger comfort.

    The radiated airborne noise of electrical hybrid vessels is reduced in battery powered operation at certain low frequencies associated with the diesel engine. The same holds for the radiated underwater noise, but the differences are smaller there, indeed smaller than differences between sister vessels.

    Finally, a new generation of LNG-powered vessels are found to be quieter than an older generation with similar specifications.

    Download full text (pdf)
    Silent@Sea slutrapport
  • 5.
    Johansson, Torbjörn
    et al.
    IVL Swedish Environmental Research Institute.
    Andersson, Carl
    IVL Swedish Environmental Research Institute.
    Krång, Anna-Sara
    IVL Swedish Environmental Research Institute.
    Andersson, Cecilia
    IVL Swedish Environmental Research Institute.
    Underwater noise from fairways – policies, incentives and measures to reduce the environmental impact2023Report (Other academic)
    Abstract [en]

    Underwater noise and its negative impact on marine life is a growing environmental concern where scientific knowledge is increasing but mitigation is scarce. This report is the outcome of a joint effort of the IVL Swedish Environmental Research Institute and the Swedish Maritime Administration that addresses this challenge.

    Motivated by environmental concerns and coming EU legislation, our vision is that Sweden should become the first country to implement national incentives for underwater noise mitigation. The technical aspects of ship underwater noise are relatively well known.

    At cruise speed, cavitation at the propeller is typically the dominant source of underwater noise, but this is not true for all ships. Standardised measurement methods exist but are costly to implement. Prediction models are useful for noise mapping and fleet-wide estimates but not sufficiently accurate for individual ships. 

    The environmental impact of underwater noise from shipping has gained increased scientific attention in recent years. While many studies have been made, dose-response relationships and thresholds for different effects are largely unknown. Behavioural effects, including escape reactions, difficulty to avoid predators and masking of important communication calls, have been observed across a large number of species upon exposure to ship noise.

    There are no national or international binding rules on ship underwater noise emissions. The International Maritime Organisation is currently updating its voluntary guidelines on ship underwater noise. The EU is introducing legislation on permissible levels of ship underwater noise in the environment, which is expected to come into force in member states within a few years.

    Technical methods for mitigation of underwater noise are known but not independently validated. Ship speed reductions may reduce underwater noise but may incur increased operational costs at the ship owners. Stakeholders in ship underwater noise mitigation are found across ship owners, the ship design and technology industry, research bodies and authorities.

    Through interviews and workshops a network of relevant stakeholders in Sweden has been established. A stakeholder analysis showed that there is a need for more knowledge on ship underwater noise and its environmental impacts as well as its mitigation. Fairway design for reduced transmission of underwater noise to the environment was investigated by long-term measurements at different sections of the fairway leading to Västerås in lake Mälaren. Neither depth nor a turn could be demonstrated to have an effect on the radiated noise.

    A more detailed experiment would be required to clarify if fairway design is a viable alternative for noise mitigation. Six different ways of designing a financial incentive for ship underwater noise reduction were described. Rewarding speed reductions or technical measures for noise mitigation is feasible but the scientific basis is not clear. An incentive may be based on a silent ship notation from a classification society, but these are not commonly issued.

    A noise inquiry may be performed, but it may be difficult to identify the most relevant mitigations without underwater noise measurement. Bespoke measurement stations at or near port inlets may be a cost-effective way to collect measurement data, but the accuracy of such opportunistic measurements would need to be improved if the data is to be used for a financial incentive.

    Download full text (pdf)
    BUFF report_FINAL
  • 6.
    Johansson, Torbjörn
    et al.
    IVL Swedish Environmental Research Institute.
    Fridén, Håkan
    IVL Swedish Environmental Research Institute.
    Samuelsson, Oscar
    IVL Swedish Environmental Research Institute.
    Björk, Anders
    IVL Swedish Environmental Research Institute.
    Sundberg, Mikael
    Datadrivna metoder för att detektera avvikande mätvärden inom dammsäkerhet2020Report (Other academic)
    Abstract [sv]

    Vi gör en litteraturstudie av state of the art i dammsäkerhet och tar reda på vad som har gjorts i andra länder. Vi tittar även på teorier och metoder som inte ännu har använts i dammsäkerhet.

    Det är lätt att komma igång med datadriven analys med hjälp av färdiga verktygslådor som finns öppet tillgängliga. Vi diskuterar därför kring vanliga fallgropar vid dataanalys och hur man undviker dem.

    Mycket av den vetenskapliga litteraturen inom datadriven dammsäkerhet handlar om tillämpning av maskininlärningsmetoder. Den struntar ofta i förbehandling och antar att alla sensorer fungerar. Vi summerar resultat om förbehandling av sensordata och hur avvikelser kan detekteras, och ser att här finns ett gap i dammsäkerhetslitteraturen.

    Vi ser att väl förstådda och utredda metoder som PCA och PLS, tidsseriemodellering, SPC/MSPC och neurala nätverk är lämpliga för vidare studier.

    Med stöd av tiotals års erfarenhet av forskning och utveckling i datadrivna metoder ger vi rekommendationer för hur ett arbete för att implementera datadrivna metoder för att höja dammsäkerheten i Sverige skulle kunna utformas.

    Vi ser att datadrivna metoder för dammsäkerhet är redo att implementeras i Sverige, och bedömer att i kombination med väl utformad instrumentering och rätt givarplacering kan de bidra till väsentligt höjd dammsäkerhet.

    Download full text (pdf)
    FULLTEXT01
  • 7.
    Wilhelmsson, Jens
    et al.
    IVL Swedish Environmental Research Institute.
    Strand, Åsa
    IVL Swedish Environmental Research Institute.
    Wrange, Anna-Lisa
    IVL Swedish Environmental Research Institute.
    Hunter, Karl
    IVL Swedish Environmental Research Institute.
    Johansson, Torbjörn
    IVL Swedish Environmental Research Institute.
    Klassificering av ostronyngel med hjälp av artificiell intelligens2020Report (Other academic)
    Abstract [sv]

    IVL har i ett samarbete mellan projekt finansierade av Jordbruksverket, Interreg (MarGenII) och H2020 (AquaVitae) utvecklat ett bildidentifieringsprogram som kan särskilja yngel av de inhemska platta ostronen i Sverige från de invasiva stillahavsostronen. Över 98 procent av bilderna klassificeras korrekt, visar projektets slutrapport

    Download full text (pdf)
    FULLTEXT01
1 - 7 of 7
CiteExportLink to result list
Permanent link
Cite
Citation style
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf