Slow slip events are well documented in global subduction zones at depths of 30-50 km. Tectonic (non-volcanic) tremor is considered to be the seismic manifestation of slow slip and is spatiotemporally correlated with slip events in most regions. Along the northern Hikurangi Margin, New Zealand, where a seamount studded igneous plateau subducts beneath the North Island, slow slip occurs shallowly at depths <15 km. Here, slow slip is associated with increases in microseismicity levels and has previously been weakly linked to tectonic tremor. Over a six-year period, the spatiotemporal progression of slow slip events along the northern Hikurangi Margin with respect to tremor, earthquake occurrence, and Coulomb failure stress changes imparted on the megathrust is analyzed. In this study, the first comprehensive tremor catalog is presented for 2010-2015. The catalog demonstrates that tremor is temporally associated with shallow slow slip events and deep tremor episodes may indicate the occurrence of previously undetected long duration slip events. A slow slip event in 2014 was recorded by the Hikurangi Ocean Bottom Investigation of Tremor and Slow Slip (HOBITSS) experiment, resulting in a detailed spatiotemporal analysis of various interplate slip processes with respect to a local subducted seamount. Coulomb failure stress change analysis of this event suggests that seamount subduction plays a dominant role in the stress state of the shallow megathrust, and that the northern Hikurangi Margin is weakly coupled and largely releases strain through slow slip events. A detailed analysis of the Coulomb failure stress change imparted on the shallow megathrust by seven slow slip events along the northern Hikurangi Margin between 2010-2014 demonstrates that stress changes from these events influence the along strike migration of slow slip event sequences. Additionally, the stress changes dictate the spatial relationship between tectonic tremor and slow slip, with onshore tectonic tremor occurring almost wholly within regions of stress increase. Over multiple slow slip events, the shallow-most part of the plate interface experiences a net stress increase and may promote failure in future shallow earthquakes, such as tsunami earthquakes, which impact seismic hazards along the east coast of New Zealand.

Earthquake stress drop, strong ground shaking, recurrence interval and the rupture characteristics of large and great earthquakes are controlled by fault strength and the frictional properties of the plate interface. Monitoring the fault strength and its variability during the earthquake cycle is critical for earthquake hazard assessment. Subduction zones in particular, generate the largest and potentially more catastrophic earthquakes around the world. The rapid convergence (85-90 mm/yr) of the Cocos plate underneath the Caribbean plate along the Nicoya Peninsula, the northwestern margin of the Costa Rica subduction zone has generated M7+ earthquakes every 50 years (1853, 1900, 1950), with the last one of these events, a moment magnitude (Mw) 7.6 earthquake, occurring on 5 September 2012. Because of the advantageous location of the Nicoya Peninsula, extending seaward, over the seismogenic zone, a very dense seismic and geodetic network has recorded with incredible detail the megathrust activity throughout the last decade. Geodetic observations prior to the occurrence of the M7.6, 2012 earthquake, identified regions of strain accumulation that delineated the rupture area of this event. Regions of the plate interface up-dip and down-dip of this patch accommodate relative plate motions aseismically, hosting numerous Slow Slip events (SSEs) that have been well recorded since 2007, suggesting a partition of the plate interface into regions of frictionally stable and unstable or conditionally stable sliding behavior. This dissertation focuses on the study of fault zone heterogeneity and the frictional strength variability along the plate interface and upper plate of the Nicoya Peninsula subduction zone margin and its plausible relationship with the previously documented seismic and aseismic frictional slip behaviors.

The use of long-term cross-correlation of ambient seismic noise has become widely adopted at several seismological observatories for a) studying the susceptibility to crustal velocity variations along volcanic areas and the Earth’s subsurface due to transient events like volcanic eruptions, shallow earthquakes, landslides, etc., and b) seismic imaging. The use of random wave fields (coda waves or ambient seismic noise) for seismic imaging has proved to overcome certain limitations so far presented by the traditional tomography methods. This dissertation extends the use of the long-term cross-correlation of ambient noise vibrations to monitor variations in elastic properties and fault heterogeneity located along the overriding plate; before, during and after the 5 September, 2012 M7.6, Nicoya Peninsula megathrust earthquake. This large event was followed by thousands of aftershocks spatially localized in four clusters distributed along strike, as well as some geodetically detected post-seismic slip.

Earthquake focal mechanisms for nearly 300 of the Nicoya earthquake aftershocks were estimated using regional full waveform moment tensor inversions and P wave first motion polarities. The finite fault model of this large event generated by Liu et al., [2015] was used to compute the Coulomb static stress change imposed by the main shock on the nodal planes of the aftershock sequence in order to make a first evaluation of the static versus dynamic earthquake triggering hypotheses.

In addition, a total of 5 repeating earthquake families, earthquakes with nearly identical waveforms and thought to represent the rupture of the same fault asperity at a relatively constant repeat interval due to continuous loading by surrounding aseismic slip, were recognized using waveform matching techniques from 2007 through 2016. These earthquake families locate along the up-dip margin of the fault that ruptured during 5 September 2012, Mw 7.6 Nicoya megathrust. Due to the repeating nature of these events they do sense the temporal changes in the physical properties of the medium, including variations in the frictional strength of the plate interface. This dissertation extends the use of repeating earthquakes to derive stress drop or fault yield strength before and after the occurrence of the Nicoya peninsula earthquake.

Seismology is a geophysical tool that can probe the earth and provide insight into processes that are otherwise difficult to observe. I am primarily interested in using seismological methods to detect and study ubiquitous processes that are challenging to observe due to noise sources and other observational limitations. This dissertation encompasses two seemingly distinct areas: aseismic fault slip and oceanic internal waves. These areas of study can ultimately be instrumentally and scientifically connected in offshore regions using ocean-bottom seismometers (OBS) for dual-purpose research goals. In particular, as the title of this dissertation suggests, we make use of the often-discarded portion of seismic signals (i.e., “noise”) to study earth processes.

Slip on faults occurs as a spectrum, ranging from stably sliding (aseismic creep) to rapid ruptures (earthquakes). Within the fault slip spectrum are transient events of slow, aseismic slip known as slow slip events (SSEs). SSEs may precede and even trigger large megathrust earthquakes and are therefore critical to detect and understand. SSEs are too slow to generate seismic waves and are generally detected geodetically. However, the shallow portion (<15 km) of many subduction zones is typically located offshore, out of the range of land geodetic networks. This presents an observational limitation of shallow SSEs that may be improved using related secondary seismicity, such as tremor and microearthquakes, to infer the location of shallow offshore SSEs.

Chapter 1 and Chapter 2 study a specific form of microseismicity, burst-type or near-repeating earthquakes, that can be used to improve the detection of transient aseismic slip. Near-repeating or burst-type repeating earthquakes are families of closely spaced earthquakes (within 200 m) with highly-correlated waveforms, aperiodic repeat intervals and variable magnitudes. We test the hypothesis that sequences of burst-type or near-repeating earthquake families concentrated in time and space are driven to failure by transient aseismic slip on the surrounding fault and can therefore be used as a proxy for aseismic slip transients. Although not included in this dissertation, I am additionally a coauthor on a study that analyzed a cluster of highly similar earthquakes in the southern Cascadia subduction zone likely related to an aseismic slip transient or changes in plate interface coupling (Alongi et al., 2021).

Chapter 1 analyzes burst-type repeating earthquakes related to a well-recorded shallow SSE offshore of Gisborne, New Zealand in September/October 2014 using OBS data. (Shaddox et al., 2019). These burst-type repeating earthquakes are coincident with tectonic tremor and are located within an upper plate fracture network above a subducted seamount after the 2014 Gisborne SSE. I was additionally a coauthor on a study of the microseismicity during the entire OBS deployment (Yarce et al., 2019). We have also performed a brief follow-up study during a similarly located SSE in 2019, and once again find burst-type repeating earthquakes within the upper plate fracture network after the plate boundary SSE. We propose that the SSEs caused fluid migration from over-pressured sediments down-dip of the seamount into the upper plate fracture network, triggering further slow slip on preexisting faults. This is strong evidence that burst-type repeating earthquakes are a promising proxy for transient aseismic slip. These findings also demonstrate the need for offshore seismic observations to detect microseismicity related to shallow, offshore transient aseismic slip. However, although OBS instrumentation is necessary to detect small offshore earthquakes not detected by land seismometers, they generally have low signal-to-noise ratios below ~3 Hz due to oceanographic processes. This is a current limitation in OBS networks.

In Chapter 2 we further demonstrate the utility of near-repeating earthquakes as a proxy for transient aseismic slip. We investigate the occurrence of near-repeating earthquake families during aseismic transients independently detected by borehole strainmeter data but beneath the noise level of continuous Global Positioning System (cGPS) stations in the trifurcation area of the San Jacinto fault zone near Anza in southern California (Shaddox, Schwartz et al., 2021). We find that all moderate-sized earthquakes occurring in this region during the time studied have afterslip signals on borehole strainmeter data and are accompanied by near-repeating earthquakes and elevated seismicity rates. Afterslip geometries defined by the near-repeating earthquake families are consistent with strain change observations. We conclude that families of near-repeating earthquakes, similar to low-frequency earthquakes within tremor, can be useful indicators of aseismic slip transients and can reveal faulting complexities during aseismic slip.

Oceanic internal gravity waves propagate along density stratification within the water column and are ubiquitous. They can propagate thousands of kilometers before breaking in shoaling bathymetry and the ensuing turbulent mixing affects coastal processes and climate feedbacks. Despite their importance, internal waves are intrinsically difficult to detect as they result in only minor amplitude deflection of the sea surface; the need for global detection and long time series of internal waves motivates a search for geophysical detection methods. The pressure coupling of a propagating internal wave with the sloping seafloor provides a potential mechanism to generate seismically observable signals.

In Chapter 3 we use data from the South China Sea where exceptional oceanographic and satellite time series are available for comparison to identify internal wave signals in an onshore passive seismic dataset for the first time (Shaddox, Brodsky et al., 2021). We analyze potential seismic signals on broadband seismometers in the context of corroborating oceanographic and satellite data available near Dongsha Atoll in May-June 2019 and find a promising correlation between long-period transient seismic tilt signals and internal wave arrivals and collisions in oceanic and satellite data. It appears that we have successfully detected oceanic internal waves using a subaerial seismometer. This initial detection suggests that the onshore seismic detection and amplitude determination of oceanic internal waves is possible and can potentially be used to expand the historical record by capitalizing on existing island and coastal seismic stations.

After the completion of the terrestrial study, we acquired data from a broadband OBS station that was deployed offshore of Dongsha Atoll from November 2019 – December 2020, providing a great opportunity to detect and quantify the signal of internal waves on an OBS. We were able to successfully identify internal waves on this OBS, opening up the possibility of a year-long study of internal wave activity near Dongsha Atoll. Further, this may allow us to ultimately remove the long-period “noise” from internal waves on this OBS to improve the detection of microseismicity.

There is important information within the “noise” of geophysical data that can help identify and characterize processes that are difficult to otherwise observe. We have successfully identified offshore aseismic slip on minor faults using burst-type or near-repeating earthquakes on noisy OBS instruments, identified and modeled small aseismic slip transients beneath the noise threshold of cGPS stations with near-repeating earthquakes and borehole strainmeter data, and have performed the first subaerial seismic detection of oceanic internal waves as well as observed internal waves on an OBS at periods that are generally considered “noise” for microseismic investigations. The following dissertation provides in depth details of the methods and implications of these findings.

Over nearly a century, seismology has provided valuable earthquake observation and information. The locations of earthquake hypocenters have helped define the spatial distribution of faults that host earthquakes and helped identify regions that may have large and damaging events in the future. However, exactly how aseismic slip, small earthquakes, and smaller faults are related to large events and the seismic cycle remains unresolved. Recent advances in instrumentation have extended research into the oceanic realm, which is important for at least two reasons. First, it enables the detection of ever smaller earthquakes that provide information about slip behavior on offshore portions of previously unresolved faults. Second, with ever-increasing resolution, marine seismic reflection images provide increasingly precise depictions of fault zones that are unattainable in the terrestrial environment. This dissertation is dedicated to better understanding the spatial distribution of faults and their potential to influence slip behavior and seismic cycle through the use of marine seismic methods. It encompasses two distinct facets of marine seismology: passive recording of earthquakes and the interrogation of subsurface faulting through seismic reflection imaging. These dual areas of investigation offer unique insights into the subsurface structure and the interplay of seismic and aseismic processes. Chapter 1 explores the enigmatic plate interface in southernmost Cascadia with the application of a dense array of ocean-bottom seismometers complementing traditional terrestrial seismic stations. A high-quality seismic catalog is created through the meticulous analysis of continuous waveform data and advanced earthquake location techniques. The results reveal a conspicuous absence of seismic activity at seismogenic depths and the shallowest up dip section, indicative of high coupling or locking and strain accumulation increasing the magnitude potential associated with this fault. Notably, a cluster of low-magnitude earthquakes (M < 3) is identified near the plate interface, exhibiting a response to nearby strain transient observed in prior studies. These strain transients are interpreted as either the tail end of a slow slip event spanning the preceding one and a half years or a rapid change in coupling. Template matching of these plate interface earthquakes demonstrates their uniqueness, with no recurrence over the observed decade. The correlation between the sudden onset of clustered earthquake activity and strain transient suggests that the two are related, and the interpretation is made that the cluster of earthquakes is a response to local stressing rate changes. Chapter 1 provides critical insights into the southern Cascadia plate interface, shedding light on the complex interactions between seismic behavior and slow slip events. In Chapters 2 and 3, this dissertation examines the shallow near-surface portion of the fault zone, with a specific focus on the spatial distribution of secondary faults surrounding the primary fault. This damaged area around the main fault provides an important window into comprehending the inelastic response of the Earth’s crust to strain, the allocation of fracture energy in the earthquake energy budget, near-fault hydrogeology, and for near-field hazards. The dimensions, both in width and depth of the damage zone are important in addressing these science questions. The primary objective of these two studies is to gain insight on the in-situ expression of the fault damage with marine controlled source seismic reflection images, an approach which had not been fully explored prior. The advantage of this data type lies in its ability to provide rich and dense sampling of the subsurface and the capability to directly image fault offsets over substantial distances that are often unattainable through other means. To achieve a comprehensive understanding of the damage zone, ranging from the seafloor to the basement, roughly 2 kilometers below, marine active source reflection surveys of varying resolutions are employed. Both Chapters 2 and 3 focus on the San Pedro Shelf region of the Palos Verdes Fault offshore southern California because of the richness in available reflection datasets and the opportunity to address unanswered questions about the generation of fault damage. Chapter 2 focuses on examining the damage zone of the Palos Verdes Fault using 2 overlapping 3D seismic volumes and demonstrates the development of a workflow for automating fault detections. Automation is achieved through the implementation of multi-trace waveform similarity or semblance-based attribute called thinned fault likelihood (TFL). These results reveal peak fault likelihoods at the location of mapped fault strands, with fault likelihood exponentially decaying with distance from the fault. Importantly, this decay with distance intersects a relatively undamaged region or background at 2 kilometers from the fault across all depths (ranging from 450 m to 2.2 km). Lithological constraints are provided by well tied 3D horizons, and damage decay trends are for each geologic unit. Notably, the findings indicate an overall increase in background damage with increasing depth. In addition to background damage increasing with depth, it is demonstrated that the decay of damage also decreases with increasing depth. This surprisingly, results in a consistent damage zone width of 2 kilometers regardless of the variability in background damage and decay trends. Chapter 3 investigates the shallowest portion of the Palos Verdes Fault damage zone which could not be fully explored in Chapter 2. This section leverages newly collected 2D high-resolution sparker multichannel seismic lines and sub-bottom profiles (chirp). Resolving the challenges of imaging in shallow water profiles is achieved through the development of custom seismic processing workflows designed to eliminate the ringing and overprinting of seismic waves trapped in the water column. Similar to Chapter 2, a semblance-based technique is used to identify discontinuities in the seismic images associated with faults and fractures. The analysis of fault perpendicular TFL profiles yields a diverse pattern of damage in the vicinity of the fault on adjacent profiles. In this study we define and map the active fault strand determined by the offset of the near seafloor sediment beds. An average pattern of damage decreasing with distance from the fault is shown by stacking fault perpendicular profiles with respect to the active fault. The stacked profiles reveal that damage decreases with distance both east and west of the fault. Notably, the peak of stacked fault damage occurs within roughly 50 meters of the active fault strand, with splay faults manifesting as secondary peaks in the stacked results. The width of the damage zone along the slope appears to align with power law displacement scaling relationships established by previous compilation studies. However, the significant reduction in damage zone width and intensity towards the south may be better attributed to the reduced obliquity of fault in that direction. In contrast to canonical strike-slip models, the presented data suggests compelling evidence that the damage zone does not widen as it approaches the surface, a characteristic that may distinguish syndepositional submarine fault’s structure. Furthermore, there is an apparent correlation between fault damage and seafloor seeps that are visibly evident in the water column. This correlation strongly implies that damage plays a role in controlling fluid flow around the fault.

Conditions at the base of an ice stream control the ability of basal material to resist ice sliding, which affects ice stream mass balance. Yet, basal properties are notoriously hard to constrain. Tiny magnitude ~-2 to -1 stick-slip basal icequakes occurring near or on the basal sliding surface hold valuable information about this basal environment. In this dissertation, I investigate spatiotemporal patterns in these basal icequakes occurring beneath the Whillans Ice Plain (WIP), in West Antarctica, and interpret these patterns to gain insight into basal material heterogeneity and temporally evolving basal conditions.

In Chapter 2, I determine where basal icequakes happen beneath the entire WIP. I find spatially variable seismicity rates, with basal seismicity most common in a ~40 km wide area surrounding a dynamically important region where ice-plain-wide unstable slip nucleates. This result implicates icequake-generating bed conditions in large-scale ice stream stick-slip. Additionally, I propose that basal icequakes occur where basal erosion exposes over-consolidated till to the ice base.

In Chapter 3, I use back-projection to detect basal icequakes beneath a small seismic network near the nucleation region. Here, basal icequakes occur in streaks elongated along ice flow and in conjunction with low-amplitude (~2m) undulating basal topography. These patterns suggest the presence of mega-scale glacial lineations (MSGL), elongate bedforms common on paleo-ice stream beds. One icequake streak may occur in a shallow trough beside an MSGL, suggesting that these icequakes occur in erosion-impacted lows between MSGL where over-consolidated till or stiff sediment outcrops contact the ice base.

In Chapter 4, I analyze an improved basal icequake catalog generated by cross correlating icequakes detected in Chapter 3. I estimate icequake moment magnitudes of Mw = -2.1 to -1.2 and fault rupture areas of 1-100 m^2 for several large basal icequakes. Families of nearly-identical repeating basal icequakes continue for ice sliding distances of typically <0.5 m, and most <0.2 m. If this distance also represents the approximate size of an icequake-generating fault, then basal icequake faults have dimensions of cm to m. I explore evidence for ice-bed interface healing between unstable slip events. Lastly, I discuss four possible mechanisms for basal seismicity.

The Red Sea Rift, an archetype of a newly formed ocean basin, is an ideal environment in which to study the controversial processes associated with continental rifting. Different models have been proposed to explain how rifting in the Red Sea evolved; however, accurate constraints on lithospheric structure have not been available to discriminate rifting models. We use the S-wave receiver function technique to produce the first images of the lithosphere-asthenosphere boundary (LAB) structure along the Red Sea and throughout the Arabian Peninsula. Lithospheric thickness varies considerably, with thin lithosphere centered on the rift axis, thickening toward the Arabian interior. Gravity data are well fit by our structural model and indicate that high surface topography along the rift flank is not in isostatic equilibrium, requiring dynamic compensation for its support. While our derived structure is consistent with active rifting processes, previous studies demonstrated that the Red Sea initiated as a passive rift. Therefore, our results suggest a two-stage rifting history, where extension and erosion by flow in the underlying asthenosphere are responsible for variations in LAB depth. LAB topography guides asthenospheric flow beneath western Arabia and the Red Sea, demonstrating the important role lithospheric variations play in the thermal modification of tectonic environments. (c) 2007 Elsevier B.V. All rights reserved.