Part I: Structure of Central and Southern Mexico from Velocity and Attenuation Tomography. Part II: Physics of Small Repeating Earthquakes

Citation

Chen, Ting (2012) Part I: Structure of Central and Southern Mexico from Velocity and Attenuation Tomography. Part II: Physics of Small Repeating Earthquakes. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/REJW-YJ88. https://resolver.caltech.edu/CaltechTHESIS:03262012-175814763

Abstract

In part I, the 3D velocity and attenuation structure of the Cocos subduction zone in Mexico is imaged using earthquakes recorded by two temporary seismic arrays and local stations. Inversion results reveal low-attenuation and high-velocity Cocos slab. The slab dip angle increases from almost flat in central Mexico near Mexico City to about 30 degrees in southern Mexico near the Isthmus of Tehuantepec. High attenuation and low velocity in the crust beneath the Trans-Mexico Volcanic Belt correlate with low resistivity, and are probably related to dehydration and melting process. The most pronounced high-attenuation, low-Vp and high-Vp/Vs anomaly is found in the crust beneath the Veracruz Basin. A high-velocity structure dipping southward from the Gulf of Mexico near the Isthmus of Tehuantepec coincides with a discontinuity from a receiver functions study, and provides an evidence for the collision between the Yucatan Block and Mexico in the Miocene.

In part II, we show that a model of small repeating earthquakes based on laboratory-derived rate and state friction laws reproduces the observed scaling between the recurrence time and seismic moment. In the model, a small fault patch governed by velocity-weakening friction is surrounded by a much larger velocity-strengthening region. For a fixed set of friction parameters, the observed scaling is reproduced by varying the size of the velocity-weakening patch. We further investigate the behavior of small repeating earthquakes in related models under different scenarios, including several forms of the state evolution equations in rate- and state-dependent friction laws, rectangular velocity-weakening patch geometries, quasi-dynamic vs. fully dynamic representation of inertial effects, and 2D vs. 3D simulations. We find that the simulated scalings between the recurrence time and seismic moment for these different scenarios are similar while differences do exist. We propose a theoretical model for the scaling between the recurrence time and seismic moment of small repeating earthquakes. The obtained theoretical insight is used to find the combinations of fault properties that allow the model to fit the observed scaling and range of the seismic moment and recurrence time.

Thesis Files