Infrasound, or very low frequency acoustical waves below the human ear audible frequency limit of 20 Hz, propagate all around the Earth atmosphere over several thousands of kilometers with very little attenuation. Infrasound were first observed in the aftermath of the 1883 eruption of Krakatoa volcano in Indonesia. Infrasound provide a continuous source of information on both the sources and the upper atmosphere (tropo-, strato-, and thermospheres). Natural sources are very diverse : storms, auroras, lightning, exploding volcanoes, earthquakes, meteoroids, oceanic waves... Anthropic sources include supersonic aircraft and rockets, and nuclear, mining or large chemical explosions. Infrasound monitoring is one of the four technologies selected by the Comprehensive nuclear-Test-Ban Treaty Organization (CTBTO) to verify compliance with the treaty. The construction of the global Infrasound Monitoring System (IMS) with 60 microbarometer stations around the world, has contributed to a revival of scientific interest in infrasound studies. In addition to local arrays, IMS provides a new, permanent and global observation system of Earth, with a unique 24/7 database.
For data analysis, modeling of infrasound is essential, in particular for inverse problems. Modeling is of interest for both sources whose characteristics and/or infrasound emission mechanisms may be uncertain, and for long range propagation. Source models are generally specific, several of them will be presented during the school. Propagation models are generic and vary from ray tracing to Direct Numerical Simulation of Euler or Navier-Stokes equations. The training school will review the main methods and propose for three of them (ray tracing, finite differences, one-way method) dedicated numerical training sessions. Beyond this state of the art review, interactions with atmosphere will be outlined. Indeed, infrasonic propagation is strongly influenced by atmospheric phenomena at large (vertical stratification, seasonal variation, global warming), medium (gravity waves, planetary waves) and small (turbulence in the planetary boundary layer, extreme events) scales. Sensitivity to atmospheric conditions relevant to simulations compared to atmospheric models (local meteorological predictions, reassimilated database) is yet poorly evaluated. A stochastic prediction including uncertainties on sources and on atmosphere, is a promising approach.
Modeling infrasound - OHP 2017