EXPERIMENTAL AND NUMERICAL STUDY OF THE PRESSURE SURGE EFFECT AND MAGMA CHAMBER RESONANCE

This dissertation is devoted to study elastodyanmic responses of a fluid-filled crack and a magma chamber in the subsurface using laboratory experiment and numerical modeling. I have made two important findings: i) a fluid-filled fracture can dynamically amplify the fluid pressure inside the fracture; ii) the presence of free surface is critical in generating long-period long-duration (LPLD) signals in volcano seismology. To be specific, the first finding is on laboratory verification of the Pressure Surge effect, where the fluid pressure in a fracture can be dynamically amplified relative to the incident wave pressure. I designed and built a low-frequency underwater system including a low-frequency acoustic source (Xfrac-S) using electromagnets that can generate low-frequency acoustic waves in the frequency range of 12~70Hz. I also built flexural transducers (Xfrac-H) that can measure the fluid pressure in the fracture directly. After investigating the effects of different frequencies and fracture apertures (0.2mm to 9.2mm), I achieved a maximum amplification factor about 25.2 using a 1.2-meter block. The Pressure Surge effect may be the underlying mechanism for multiple natural phenomena such as dynamic earthquake trigging, hydrogeological permeability change, and triggered mud volcano eruptions. The second finding is on the magma chamber resonance using a half-space boundary element modeling. I created a new 3D boundary element code to model the elastodynamic response of a magma chamber below the earth surface. I found the presence of the free surface can enhance the resonance energy of the magma chamber and generate LPLD signal much more efficiently. This modeling work provides a basis to constrain the magma chamber geometry and its depth using sparsely observed seismic data. In developing the BEM codes, I introduced a novel regularization of half-space Green’s functions in BEM which can reduce the integration time significantly. I also found a new expression for hypersingular Green’s functions which are critical in solving wave scattering by empty cavities in elastic media.