Matrix imaging from seismic noise maps deep volcanic plumbing at 100m resolution

A team led by Alexandre Aubry (Institut Langevin, ESPCI Paris) and Jean-Christophe Komorowski (IPGP) has demonstrated that matrix imaging of seismic noise can map deep volcanic plumbing systems with unprecedented resolution. Published in Communications Earth & Environment, the study used data from a sparse geophone array on La Soufrière volcano in Guadeloupe. Traditional seismic migration techniques fail because volcanic rock is highly heterogeneous, scattering waves and distorting images. The researchers instead applied a matrix formalism to seismic noise interferometry: they built a reflection matrix from ambient noise correlations, then used focusing algorithms that are robust to multiple scattering to recover echoes from deep structures. The resulting images reveal magma and hydrothermal systems down to 10 kilometers depth with a lateral resolution of 100 meters—far better than the array's aperture would normally allow.

This breakthrough is significant because deep magma storage is notoriously hard to image, yet critical for assessing eruption hazards. The method works with sparse, low-cost geophone networks, meaning it can be deployed at many volcanoes worldwide. The team sees matrix imaging as a “revolutionary tool” for volcano observatories, enabling real-time monitoring of pressure changes and magma movement. The paper (arXiv:2311.01296v2) includes 55 pages and 18 figures detailing the technique and its validation. By overcoming the diffraction limit imposed by sensor density, this approach could also be adapted for other geophysical imaging challenges, such as fault zones or geothermal reservoirs.