UC Berkeley

A number of geologic processes on the surface and in the interior of the Earth result in ground deformation on human timescales. These processes include the steady motion of tectonic plates, the internal deformation of heterogeneous regions in the crust, the response of the Earth to surface loads, and deformation at plate boundaries, all acting to produce millimeter-scale displacements each year. These same processes are also responsible for many natural hazards that cause harm to human communities: earthquakes and tsunamis are the most directly related, but sea level rise and extreme flooding are also pertinent natural threats exacerbated by the deformation of the ground. In order to understand these natural hazards and to mitigate their impacts, it is important to quantify surface deformation and to understand the physical mechanisms underlying it. The chapters of this dissertation address several questions in the field of tectonic deformation that relate to these processes. In particular, I focus on understanding earthquake hazards and fault zone properties using a variety of tools at the intersection of space geodesy and seismology.

Zones of weakness in the Earth’s crust at all scales are relevant for the study of natural hazards. They respond to external stresses and often host earthquake ruptures. Fault zones, as one small-scale example, are thought to be mechanically weaker than their surroundings because the accumulated near-fault damage results in lower mechanical strength, forming compliant fault zones. Importantly, the presence of a compliant fault zone around a fault affects the types of earthquake ruptures the fault can sustain. In order to study compliant fault zones around a major fault, I investigated the elastic properties of the crust near the San Andreas Fault Zone in northern California. I used GPS measurements to characterize compliant fault structures along strike and found that their distribution is heterogeneous even over short spatial scales.

Larger zones of weakness also impact present-day deformation. On continental scales, inherited zones of weakness from previous tectonic episodes can affect the seismic activity of a region even millions of years later. In the craton of southern Africa, I studied the rupture process of a rare M6.5 earthquake in Botswana that ruptured the lower crust in a continental plate interior. Using a joint analysis of teleseismic waveforms, InSAR data, and relocated aftershocks, I identified the rupture plane out of the two possible focal planes. The modeled strike matches with the boundary of an ancient collisional mountain belt that has been reactivated in the present day as a set of normal faults. Intraplate earthquakes such as this one are challenging to forecast because they occur in regions of low interseismic strain, but they can be especially damaging because they occur in places that aren’t expecting earthquakes. Understanding the structure of pre-existing weak zones in the crust may be key in such situations.

Clues to the internal structure of the crust may also come from the study of periodic hydrological loads on the Earth’s surface. Seasonal loads from snow, rivers, lakes, groundwater, the ocean, and the atmosphere cause surface deformation that depends on the structure of the underlying medium and the processes involved. In order to analyze these processes in a tectonically active region, I performed a study on hydrological loading in GPS data from South Asia and Southeast Asia, a region impacted by a strong yearly monsoon. The annual deformation varies across the region but is generally consistent with the elastic loading modeled from an independent gravity dataset. The appropriate modeling of the hydrological loading deformation in the future could help quantify the storage of aquifers, improve our understanding of earthquake cycle deformation, and aid in the accurate detection of transient fault behavior.

The remaining two chapters of this dissertation relate to transient fault behavior of the Mendocino Triple Junction in northern California. This region lies at the intersection of three major plate-bounding faults, and it produces some of the largest earthquakes in California. Several of the plate-bounding faults at the Mendocino Triple Junction produce a mix of seismic and aseismic moment release; this results in fascinating time-dependent slip behavior and an interplay between aseismic and seismic slip modes. In one study of this behavior, I used characteristically repeating earthquakes to identify regions of the Mendocino Fault Zone that are creeping aseismically. Using a dataset from 2008 to 2018, I found several dozen families of small-magnitude repeating earthquakes that show a high degree of active creep on the plate boundary fault. The creep rate is calculated to be about 65% of the overall slip budget, a significant amount but in line with previous estimates of aseismic slip on oceanic transform faults. During this time interval, the slip appears to be relatively steady, but the repeating earthquake catalog allows us to study any time-dependent variations of the creep rate in the future.

On the neighboring Cascadia Subduction Zone, I also find aseismic creep with time-dependent rate variations. Well-known slip transients called Episodic Tremor and Slip events occur on the southern Cascadia Subduction Zone margin every 7-10 months, resulting in several centimeters of slip at about 30 km depth. However, in the final chapter of this dissertation, I document changes in surface velocity that appear uncorrelated with the process of Episodic Tremor and Slip. I find that surface GPS velocities near the Mendocino Triple Junction show systematic variations in their east/west component that last for several years apiece. The timing of several velocity changes is coincident with the timing of large (M>6.5) offshore earthquakes. The spatial pattern and temporal pattern of the observations do not indicate more usual processes like afterslip, viscoelastic relaxation, or hydrological loading are sufficient to explain the observations. Instead, inversion of the velocity changes suggests that in a region slightly updip of the Episodic Tremor and Slip portion of the interface, interseismic coupling may both increase and decrease in connection with the offshore earthquakes. A speculative dynamic triggering mechanism is presented. Such observations suggest that understanding fault zone properties is of paramount importance in the study of earthquake hazards. The findings also show that in light of newer geodetic and seismological datasets, there is still much to be learned and many unanswered questions about the behavior of fault zones throughout the earthquake cycle.