Southeastern Section - 58th Annual Meeting (12-13 March 2009)

Paper No. 3
Presentation Time: 9:20 AM

RHEOLOGIC AND STRUCTURAL CONTROLS ON THE DEFORMATION OF OKMOK VOLCANO, ALASKA, FROM InSAR AND AMBIENT SEISMIC NOISE


MASTERLARK, Tim1, HANEY, Matthew2, DICKINSON, Haylee1, FOURNIER, Tom3 and SEARCY, Cheryl2, (1)Geological Sciences, The University of Alabama, Tuscaloosa, AL 35487, (2)USGS Alaska Volcano Observatory, Anchorage, AK 99508, (3)Earth Science, Rice University, Houston, TX 77005, masterlark@geo.ua.edu

InSAR data reveal that the caldera of Okmok volcano, Alaska, subsided more than a meter during its eruption in 1997. Due to the strong radial symmetry and high signal-to-noise ratio of the InSAR data, inverse analyses of InSAR data precisely define the source of deformation as a spherical pressure source (–5.5•108 Pa) at a depth of 3.1 km, assuming a problem domain that simulates a homogeneous elastic half-space. Although a forward model driven by these estimated source parameters predicts about 95% of the observed deformation, the magnitude of depressurization severely exceeds lithostatic pressure constraints. Seismic tomography using ambient ocean noise reveals a significantly deeper magma chamber (>4 km) and a low velocity zone corresponding to a zone of weak, highly saturated materials within the caldera, extending from the caldera surface to a depth of 2 km. The deep low velocity zone associated with the magma chamber suggests magma remains in a molten state between eruptions. We construct finite element models (FEMs) to simulate deformation caused by mass extraction from a magma chamber that is surrounded by a viscoelastic rind and embedded in an elastic problem domain partitioned to account for the weak caldera materials observed with tomography. Thermal models define the brittle-ductile transition near the magma chamber. This configuration allows us to reduce the estimated magma chamber depressurization, while simultaneously maintaining the magnitude of deformation required to predict the InSAR data, because the total surface deformation is the combination of the elastic and viscous response. The viscosity of the viscoelastic rind must be less than 2•1017 Pa•s to adequately predict the InSAR data and simultaneously satisfy lithostatic constraints on magma chamber pressure. More precisely, the InSAR data are best predicted by an FEM simulating a rind viscosity of 7.5•1016 Pa•s and a mass flux of –4.2•109 kg/d from the magma chamber. The shallow weak layer within the caldera also provides a stress regime that supports dike arrest and lateral magma propagation to the rim of the caldera, which explains the lateral offset of the magma extrusion compared to the location of the magma chamber beneath the center of the caldera.