Toward seismic source imaging using seismo-ionospheric data

The worldwide coverage offered by global navigation space systems (GNSS) such as GPS, GLONASS or Galileo allows seismological measurements of a new kind. GNSS-derived total electron content (TEC) measurements can be especially useful to image seismically active zones that are not covered by conventional instruments. For instance, it has been shown that the Japanese dense GPS network GEONET was able to record images of the ionosphere response to the initial coseismic sea-surface motion induced by the great Mw 9.0 2011 Tohoku-Oki earthquake less than 10 minutes after the rupture initiation (Astafyeva et al., 2013). But earthquakes of lower magnitude, down to about 6.5 would also induce measurable ionospheric perturbations, when GNSS stations are located less than 250 km away from the epicenter. In order to make use of these new data, ionospheric seismology needs to develop accurate forward models so that we can invert for quantitative seismic sources parameters. We will present our current understanding of the coupling mechanisms between the solid Earth, the ocean, the atmosphere and the ionosphere. We will also present the state-of-the-art in the modeling of coseismic ionospheric disturbances using acoustic ray theory and a new 3D modeling method based on the Spectral Element Method (SEM). This latter numerical tool will allow us to incorporate lateral variations in the solid Earth properties, the bathymetry and the atmosphere as well as realistic seismic source parameters. Furthermore, seismo-acoustic waves propagate in the atmosphere at a much slower speed (from 0.3 to ~1 km/s) than seismic waves propagate in the solid Earth. We are exploring the application of back-projection and time-reversal methods to TEC observations in order to retrieve the time and space characteristics of the acoustic emission in the seismic source area. We will first show modeling and inversion results with synthetic data. Finally, we will illustrate the imaging capability of our approach with, among other possible examples, the 2011 Mw 9.0 Tohoku-Oki earthquake, Japan, the 2012 Mw 7.8 Haida Gwaii earthquake, Canada and the 2011 Mw 7.1 Van earthquake, Eastern Turkey.