Volcanoes, earthquakes and seismic anisotropy EU-funded scientists combined multiple geophysical techniques to minimise any ambiguity in the interpretation of the observations of seismic activity, changes in tremor and ground deformations. Industrial Technologies © Shutterstock Αny magma storage system under excessive pressure, whether a dike, a conduit, a chamber or a combination of these, exerts a stress on the surrounding country rock. Determining and understanding the level of stress is key to predicting if and when a volcano will erupt. Shear wave splitting analysis around volcanoes provides useful indications of the horizontal stress direction and hence, of the pressure induced by magma movement. However, spatial variations may be misinterpreted as temporal changes. For this reason, scientists introduced a new method of analysis for seismic anisotropy. Within the EU-funded SAMS (Seismic anisotropy and magma systems) project, the scientists used finite element models to examine variations in seismic properties before and during magmatic activity. The Kilauea volcano on the Big Island of Hawaii, the Tungurahua volcano in Ecuador and the Upptyppingar volcano in Iceland were selected as case studies. The aim was to elucidate the relationship between measured anisotropic seismic velocities based on shear wave splitting and pressures in magma reservoirs. For this purpose, the SAMS team inverted the available data to obtain information about the sub-surface geometry, including crack structure and pore content. Notably, in the case of the Kilauea volcano, the inversion technique proved to be a powerful tool for the identification of a known region of magma storage. The results from the inversion of shear wave splitting data were in good agreement with ground deformation and gas emission data collected with petrological methods. On the other hand, finite element modelling revealed the shortcomings of traditional modelling techniques, attributed to the magma reservoir being relatively shallow. SAMS' findings are expected to help improve shear wave splitting as a volcano eruption-forecasting tool. Besides interpreting temporal changes in the measured seismic anisotropy to forecast volcanic activity with sparse data, it should also support the design of better monitoring networks. Keywords Volcanoes, earthquakes, magma storage, shear wave splitting, finite element model
