Acoustic Geophysical Data

Contact person: Francois Renard    
Keywords: Acoustic data, Geohazards, Renewble energy, Environmental Monitoring
Research group: Njord Centre
Department of Geosciences    
 

Acoustic geophysical data are acquired continuously with fleets of acoustic sensors such as seismometers, hydrophones, and geophones that record the fast and slow displacements in and at the surface of the Earth’s (slow and fast earthquakes, landslide slip, glacier instabilities), the flow and sediment transport in rivers and in the subsurface, deformations leading to catastrophic rock failure in laboratory experiments, and even the migration of whales in the ocean. These data are critical for the monitoring of both natural geohazards and human activities in the fields of renewable energy (geothermal, hydropower), geological storage (carbon dioxide, hydrogen), and environmental monitoring in a context of climate change. As the quantity of acoustic data increases, novel real-time analysis methods based on computationally efficient solutions, including artificial intelligence, are timely needed. Research proposals may span methodological developments to automatically process acoustic data to unravel natural processes, to develop general purpose “hearing” artificial intelligence techniques that can identify sub-aural (e.g., small earthquakes) to super-aural (e.g., friction in a high-speed turbine) sounds, as well as to develop applications in the domains of geohazards, energy, and environmental science.

Methodological research topics:

  • Development of new unsupervised and self-supervised artificial intelligence and machine learning methods for investigating acoustic data across frequency domains
  • Development of uncertainty quantification methods for energy budgets governing acoustic events (e.g., elastic grain rotations shifting stress orientations prior to catastrophic rock failure in laboratory experiments)
  • Development of new time-series analysis methods to connect long period processes with short period events (e.g., days of fluid flow motivating seconds of rock fractures)
  • Development of new real-time analysis methods of streaming acoustic data from multiple sources

Topics from natural sciences or technology:

  • Coupling vision AI with hearing AI in the lab through analysis of acoustic data and images in laboratory earthquakes
  • Detection and analysis of bubble nucleation in sub-surface fluid flows
  • Correlative analysis of acoustic data and borehole logging data to detect fluid flow in the sub-surface
  • Automatic analyses of acoustic data, ground water data, and strain data to monitor slow and fast displacements in glaciers, landslides, or fault zones
  • Structural integrity analysis for dams and turbines used in the production of hydropower energy
  • Remote sensing of nuclear test explosions
  • Detection of wildlife

External partners:

  • Statkraft