This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5).
Non-technical Explanation: It is generally believed that with adequate monitoring it is possible to detect premonitory signals prior to volcanic eruptions. Eruptions are often preceded by swarms of earthquakes and bulging of the volcano as magma rises from the earth?s mantle into the crust. However, the length of the period of` ?unrest? can vary greatly, and eruptions can be either explosive or passive (effusive). Major goals of volcano science are to provide more accurate forecasts of future behavior based on physical and chemical models of the eruptive process combined with seismic and geodetic monitoring. The 2004-08 eruption of Mount St. Helens provides a unique opportunity to develop such methods. This episode erupted magma with essentially the same chemical composition as the devastating 1980 eruption; however this eruption was very gas poor and thus non-explosive. The 2004 onset was preceded by only a few days of seismic activity and no detectable ground deformation. We will develop rigorous physical and chemical based models of the eruption and test these against observed seismic and Global Positioning System (GPS) measurements of ground deformation.
We will develop physically-based models of magma chambers and conduits that are coupled to the elastic surroundings. As magma ascends the decrease in pressure results in volatile exsolution. This decreases magma density, but increases viscosity and compressibility. Exsolution also promotes microlite crystallization; at roughly 1 km depth at MSH the magma becomes essentially a solid plug, the upward motion of which is resisted by frictional sliding on its margins. Changes in chamber pressure as well as shear and normal tractions on the conduit walls are used to predict surface deformation, which can be compared to GPS data. Markov Chain Monte Carlo (MCMC) inversions will be used to determine posterior probability distributions for magma chamber depth, shape, volume, initial overpressure, and recharge, given the GPS data and estimates of extrusion volume. We will model the cessation of the eruption in January of 2008, and surface deformation data since that time, to better constrain the rate of recharge into the crustal magma chamber. We will determine whether rate and state dependent friction effects on the boundary of the shallow magma plug can explain both the onset of the eruption ? increasing pressure overcomes frictional resistance which then weakens with sliding ? and also the rapid early deflation observed at the one continuous GPS site operating at the eruption onset.
We will also test possible explanations of cyclic ground tilt observed in the crater of Mount St. Helens. Preliminary analysis suggests that the tilts may be due to shear on the margin of the plug near the bend in the conduit, where ascending magma is redirected to the south beneath the 1980?s lava dome before extruding onto the surface. We suggest that careful analysis of accurately located earthquakes associated with the tilts will provide important clues to the processes controlling both seismogenesis and extrusion. For example, variations in the depths of shallow earthquakes during tilt events might point toward migrating slip on the margin of the plug.
