Understanding Volcanic Sources Through Improved Inversion of Deformation and Gravity Data

Stanford University, Stanford CA

Investigators

Abstract

An important goal in studying volcanoes is to determine the size and shape of magma reservoirs. How deep are they? Are they stock-like, or roughly equi-dimensional bodies, sills, or lacoliths? It is also important to determine the size, shapes, and locations of conduits through which magma is transported to the surface, and how precursory seismicity and deformation are controlled by magma volume and chemistry. Ideally, it would be possible to distinguish between periods of unrest that lead to eruptions versus those that end with intrusion, as well as the size and explosivity of future eruptions. Distinguishing between unrest caused by shallow hydrothermal activity from that generated by magmatic processes is critical in some silicic systems such as Long Valley Caldera and Yellowstone. The increasing amount and quality of geodetic data that Earthscope will make available in the next decade will help address these questions, however improved methods are needed to analyze these data. This project is addressing these issues by developing new inverse methods to determine the location and shape of magmatic reservoirs from deformation data. Rather than assume a particular source shape, the researchers are developing methods to solve for the magma chamber geometry from the data. Previous work utilized distributed point sources to model magma chambers without physical motivation. The present study looks instead for a closed surface with uniform pressure inside an elastic earth as a representation of a magma chamber. They invert for the best fitting set of point sources and then seek an equivalent pressurized cavity that causes the same surface deformation. This yields a physically motivated model, based on observed deformation that is in principal completely general. Additionally, this project is employing time dependent inversion techniques to estimate the temporal and spatial evolution of the magmatic system. These methods, collectively referred to as Network Inversion Filters, are capable of imaging the spatio-temporal evolution of dikes, and faults. Simultaneous inversion of deformation data and repeated gravity measurements permits researchers to estimate the density of subsurface fluids. This can allow one to discriminate reliably between magmatic and aqueous fluids as causes of deformation.