This is a past event.

Wendy McCausland 
USGS

Seismic monitoring data, when combined with geologic, geochemical, geodetic and petrologic data, provide important constraints on magmatic ascent and eruption. However, not all volcanoes are so thoroughly monitored or studied prior to the onset of unrest. By combining insights from well-studied volcanoes and insights from the seismic precursors that have preceded other volcanic unrest or eruption, one can develop a process-based model that provides the conceptual basis for eruption forecasting. Precursory trends may include deep (e.g., >25 km) low-frequency (LF) seismicity that records mass movement of magma through the lower crust, and distal short-period volcano-tectonic (VT) earthquakes produced by activation of tectonic faults as magma rises through the mid-crust perturbing the regional stress regime. As magma rises to the uppermost crust, LF and hybrid earthquakes result from fluid movements associated with hydrothermal systems, from faulting along conduit boundaries, or possibly even from fracturing within the magma itself. Earthquake-free and tomographically determined regions of high attenuation beneath the volcano coincide with petrologic equilibration depths below well-studied volcanoes. These relations give confidence to a simplified conceptual model of volcanic plumbing systems below stratovolcanoes consisting of magma mush-filled reservoirs at depths ranging from ~5 to >15 km that are fed by magma from the lower crust or mantle and that are connected to the surface during eruptions through narrow conduits where most of the precursory LF and hybrid seismicity takes place. By thinking in terms of this process-based conceptual model, one can use seismicity and other monitoring data to generate reasonable estimates for eruption likelihood, size or style.

Host: Greg Waite