Volcano deformation source parameters estimated from InSAR: Sensitivities to uncertainties in seismic tomography

详细信息    查看全文

• 作者：Timothy Masterlark ; Theodore Donovan ; Kurt L. Feigl ; Matthew Haney ; Clifford H. Thurber ; Sui Tung
• 刊名：Journal of Geophysical Research: Solid Earth
• 出版年：2016
• 出版时间：April 2016
• 年：2016
• 卷：121
• 期：4
• 页码：3002-3016
• 全文大小：6.0MB
• ISSN：2169-9356

文摘

The eruption cycle of a volcano is controlled in part by the upward migration of magma. The characteristics of the magma flux produce a deformation signature at the Earth's surface. Inverse analyses use geodetic data to estimate strategic controlling parameters that describe the position and pressurization of a magma chamber at depth. The specific distribution of material properties controls how observed surface deformation translates to source parameter estimates. Seismic tomography models describe the spatial distributions of material properties that are necessary for accurate models of volcano deformation. This study investigates how uncertainties in seismic tomography models propagate into variations in the estimates of volcano deformation source parameters inverted from geodetic data. We conduct finite element model-based nonlinear inverse analyses of interferometric synthetic aperture radar (InSAR) data for Okmok volcano, Alaska, as an example. We then analyze the estimated parameters and their uncertainties to characterize the magma chamber. Analyses are performed separately for models simulating a pressurized chamber embedded in a homogeneous domain as well as for a domain having a heterogeneous distribution of material properties according to seismic tomography. The estimated depth of the source is sensitive to the distribution of material properties. The estimated depths for the homogeneous and heterogeneous domains are 2666 ± 42 and 3527 ± 56 m below mean sea level, respectively (99% confidence). A Monte Carlo analysis indicates that uncertainties of the seismic tomography cannot account for this discrepancy at the 99% confidence level. Accounting for the spatial distribution of elastic properties according to seismic tomography significantly improves the fit of the deformation model predictions and significantly influences estimates for parameters that describe the location of a pressurized magma chamber.