Electrical self-potential anomalies from magmatic stressing: a tool for volcano monitoring?

Abstract:

Pre-eruptive electrical signals at active volcanoes are generally interpreted in terms of electrokinetic processes in the subsurface. Spatio-temporal self-potential (SP) signals can be caused by strain-induced fluid flow in volcanic aquifers, however, previous studies lack the quantitative assessments of these phenomena and the underpinning poroelastic responses. Here, we use Finite-Element Analysis to study poroelastic responses and emerging SP signals induced by subsurface stressing from sill and dike sources by jointly solving for resultant ground displacements, aquifer pressure changes and SP signals. We evaluate the influence of pressure source orientation on the SP response in two different volcanic aquifers (pyroclastic and lava flow) and provide insights on the sensitivity of SP signals to governing model parameters. Strain-induced SP amplitudes deduced from a reference parameter set vary in both aquifer models and are of negative polarity (-0.35 mV and -22.6 mV) for a pressurized dike and of positive polarity (+4 mV and +20 mV) for a pressurized sill. Overall, we find that SP amplitudes are most sensitive to elastic and electrical properties of the subsurface and source orientation with uniquely different SP and ground displacement patterns from either sill or dike intrusions. We observe SP amplitudes of up to -947 mV in the parametric study for a dike-generated poroelastic response, indicating that predicted SP signals are broadly representative of records from volcanic areas (mV to few V). Our study demonstrates that electrokinetic processes can reflect subsurface straining and highlights the potential of joint geodetic and SP studies for volcano monitoring to gain new insights on the causes of volcanic unrest. SP surveying in particular is inexpensive, non-intrusive and efficient to execute in challenging volcanic terrain and we propose its adaptation as a valuable contribution to volcano monitoring.