The Mechanisms and Triggering of Earthquakes in the Ridge-Transform Environment
Danielle Frances Sumy
Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Graduate School of Arts and Sciences
COLUMBIA UNIVERSITY
2011
© 2011 Danielle Frances Sumy All rights reserved ABSTRACT
The Mechanisms and Triggering of Earthquakes in the Ridge-Transform Environment
Danielle Frances Sumy
The theory of plate tectonics introduced a paradigm shift in the way we view and study our planet. Many of the world’s plate boundaries, however, are beneath our oceans making data collection, the key to furthering our knowledge about these zones, difficult. Recent advances in research vessels, seismic data acquisition techniques, and equipment built to withstand the temperatures and pressures within the oceans and on the seafloor, have made a huge impact in helping us understand the complicated structure and dynamics of our Earth.
In the field of seismology, precise earthquake locations can illuminate regions of active seismic deformation, and help us better understand the orientation, mechanics, and kinematics of plate boundary zones. Although ocean bottom seismometers have been in use since the late
1930s, the instruments did not have the recording capacity and endurance to withstand being placed on the seafloor for a large span of time. Today, ocean bottom seismometers are deployed in densely spaced arrays that record seismic signals for approximately a year. The high- precision seismic data now available can help us redefine plate boundaries and further our understanding of the internal processes and deformation within these zones. In this dissertation,
I aim to use ocean bottom seismometer data to explore the Pacific-North America plate boundary within the Gulf of California, and the internal workings of the 9º50’N East Pacific Rise high- temperature hydrothermal system.
The first chapter of my dissertation uses data collected from an ocean bottom seismometer array deployed along the plate boundary within the Gulf of California from October
2005 to October 2006. In this study, I detect and locate ~700 earthquakes mainly located on the NW-SE striking oceanic transform faults that delineate the plate boundary. In addition, we calculate regional moment tensors for ~30 of these events, and find that the majority are right- lateral strike-slip events consistent with observed transtensional plate motion.
Chapter 2 investigates the relationship between tides and ~3500 microearthquakes recorded on six ocean bottom seismometers deployed in the vicinity of the 9º50’N East Pacific
Rise high-temperature hydrothermal vent system from October 2003 to April 2004. I find unequivocal evidence for tidal triggering of microearthquakes with maximum extensional stresses induced by the solid Earth tide at this site. Although tides are not the underlying cause of earthquake nucleation within the region, the modulation of microearthquakes by these small amplitude tidal stresses indicates that the hydrothermal system is a high-stress environment that is maintained at a critical state of failure due to on-going tectonic and magmatic processes.
In Chapter 3, I further investigate the role of tides in triggering microearthquake activity at the 9º50’N East Pacific Rise high-temperature hydrothermal vent site, and observe systematic along-axis variations between peak microearthquake activity and maximum predicted tidal extension. I interpret this systematic triggering to result from pore-pressure perturbations propagating laterally through the hydrothermal system, and from this result and a one- dimensional poroelastic model, I provide an estimate of bulk permeability at this site. This observation may allow for more sophisticated investigations into the heat and chemical exchange between the newly formed oceanic crust and hydrothermal fluids, and may provide insight into the plumbing supporting the subsurface biosphere.
Table of Contents
List of Figures...... ………………………………………………………………ii
List of Tables…………………………………………………………………………………...... iii
Acknowledgements……………………………………………………………………………….iv
Dedication…………………………………………………………………………………………v
Introduction………………………………………………………………………………………..1
Chapter 1: The Mechanics of Earthquakes and Faulting in the southern Gulf of California…………………………………………………………………....5
Chapter 2: Pulse of the Seafloor: Tidal triggering of microearthquakes at 9º50’N East Pacific Rise………………………………………………………...... 64
Chapter 3: Systematic along-axis tidal triggering of microearthquakes observed at 9º50’N East Pacific Rise...... 74
References………………………………………………………………………………………..86
i List of Figures
Chapter 1
Figure 1: Map of Gulf of California and Baja California...... 44 Figure 2: Global catalogued events within the Gulf of California...... 45 Figure 3: Map showing Hypoinverse relocated events...... 46 Figure 4: Velocity-depth profile...... 47 Figure 5: Double-difference relocated hypocenters and RMTs...... 48 Figure 6: Regional Moment Tensor Solution Examples...... 49 Figure 7: Moment Magnitude vs. Local Magnitude...... 50 Figure 8: Northern Guaymas Transform Events...... 51 Figure 9: Southern Guaymas Transform and Guaymas Basin Events...... 52 Figure 10: Cross-sections of southern Guaymas Basin and Transform events...... 53 Figure 11: Carmen Transform events...... 54 Figure 12: Magnitude Plots and Spatial Sequences on the Carmen Transform...... 55 Figure 13: Farallon Transform Events...... 56 Figure 14: Alarcon Ridge-Transform events...... 57 Figure 15: August 2006 Baja California earthquake sequence...... 58 Figure 16: Map comparison between our study and Castro et al., 2010...... 59
Chapter 2 Figure 1: Bathymetric Map of 9º50’N East Pacific Rise...... 66 Figure 2: Relationship between Earthquakes andTides...... 67 Figure 3: Definition of Tidal Phase...... 68 Figure 4: Percentage of Excess Events during times of Encouraging Stress...... 72
Chapter 3 Figure 1: Bathymetric Map with Double-Difference Relocated Epicenters...... 76 Figure 2: Along-Axis Observations of Tidal Triggering...... 77 Figure 3: Mean Tidal Phase Angle Against Along-Axis Latitude...... 78 Figure 4: Permeability as a Function of Porosity and Temperature...... 78 Figure 5: Volumetric Tidal Stress over Time...... 85
ii List of Tables
Chapter 1 Table 1: The 695 Hypoinverse events divided into qualitative bins...... 60 Table 2: Focal mechanism solutions for the 31 earthquakes...... 61 Table 3: Parameters for total effective seismic stress drop calculation...... 63
Chapter 2 Table 1: Statistical Parameters Associated with the Tidal Components...... 71 Table 2: Relationship between Peak-to-Peak Tidal Stresses and Microearthquakes...... 73
iii Acknowledgements
Thank you to my friends and family who were gracious enough to see me through this process, especially Jennifer Arbuszewski, Adrienne Block, Anna Foster, Lauren
Friedman-Way, Heather and Paul Garcia, Kate and Jeff Hoops, Robert and Elizabeth
Kearns, Jennifer Levy-Varon, Kylene and Jeff Lopo, Jacqueline and Steve Phillips,
Sandy Rivera, Ashley Shuler, Erika and Jonathan Thornton, and Andrew Way. I would especially like to thank my parents, Chuck and Jean Stroup, who nurtured my love of earth science. Thank you for letting me choose Charles Richter as my role model!
I would also like to thank my advisor committee, Jim Gaherty, Bill Menke, and
Maya Tolstoy. Thank you for letting me tirelessly ask questions and for never treating me badly because of it. Thank you for always considering me as a whole complicated person, and not just a graduate student. Your thoughts, encouragement, and constant advice have made me not only a better seismologist, but also a better person.
Thank you to Del Bohnenstiehl, who encouraged me to apply to Lamont, and even though he left me in the dust after my first year, still let me call and nag him frequently. Thank you for continuing to serve on my committee, and for taking the time and energy to help me understand my science more thoroughly. Thank you to Jeff
McGuire for serving on my defense committee, and for your encouragement. I cannot honestly read an entire oceanic transform paper without seeing your work cited, and your work has transformed my own personal knowledge of oceanic faults.
Finally, two of the chapters in this thesis were previously published in
Geophysical Research Letters, and are printed here with permission granted by the
American Geophysical Union.
iv Dedication
To my husband, Nathan Sumy
You are a perfect reflection of God’s love for me, and you have pushed me to complete the work I started and thrive while doing it. Thank you for being my hero, my best friend, and for being a man that I can look up to, love, and respect. Thank you for constantly telling me that you are proud of me, no matter what. Your love and encouragement pushes me beyond my boundaries, and into the realm of possibility. This is only the beginning, and I look forward to sharing any and all success with you. I love you!
v 1
Introduction
2
Introduction
In the last fifty years, the discovery of plate tectonics has revolutionized the field of geophysics, and what we know about the way our Earth works. The study of past plate motions, magnetics, seismology, gravity, and heat flow are the tools with which the theory was illuminated and explained. A key contribution of seismology was the discovery that the geographical distribution of earthquake activity is non-uniform, and highlights the world’s active plate boundary zones. Earthquakes occur due to release of strain, built-up in rocks due to the relative motion between adjacent plates, and the mechanisms of earthquakes allow the direction of displacement along the plate boundary to be determined.
In the oceans, earthquakes illuminate mid-ocean ridges, the world’s longest volcanic chain extending over ~60,000 km. Much of the heat loss from the interior of the Earth results from the creation of basaltic crust and cooling of oceanic lithosphere at mid-ocean ridges.
Hydrothermal circulation in the vicinity of the ridge crest facilitates the exchange of heat and chemicals between the ocean and the solid Earth, and supports chemosynthetic microbial and macro-faunal communities. The loss of heat by hydrothermal circulation accounts for approximately one-third of the oceanic heat flux, or approximately one-quarter of the global heat flux [e.g. Williams and Von Herzen, 1974; Sclater et al., 1980; Stein and Stein, 1994].
Temperature studies in the ocean show locally high temperatures (on the order of 350º to 400º C) around high-temperature hydrothermal vents [e.g. Speiss et al., 1980], and lows where cold seawater enters the crust. Earthquake seismology can illuminate zones of thermal cracking and shed light on the internal structure of high-temperature hydrothermal sites [Tolstoy et al., 2008].
Connecting the ridge segments together are the oceanic transform faults, which are conservative plate boundaries where oceanic crust is neither created nor destroyed. Plate 3 boundary motion along these transform boundaries is accommodated by horizontal shear along- strike of these faults. In contrast to continental strike-slip faults, such as the San Andreas Fault in California, much of the slip along oceanic transforms is thought to occur aseismically [e.g.
Brune, 1968; Bird et al., 2002; Boettcher and Jordan, 2004]. In addition, the focal depths of earthquakes and laboratory friction experiments suggest that seismic rupture along oceanic transform faults is largely temperature dependent, with the base of the seismogenic zone bounded by the 600ºC isotherm [e.g. Abercrombie and Ekström, 2001; Braunmiller and Nábělek,
2008; Boettcher et al., 2007].
Precise earthquake locations along mid-ocean ridges, transform faults, and within hydrothermal systems can illuminate regions of active seismic deformation, and help us better understand the mechanics and kinematics of these plate boundaries. Recent advances in research vessels, seismic data acquisition techniques, and equipment built to withstand the temperatures and pressures within the oceans and on the seafloor, have made huge strides in helping us further understand the complicated structure and dynamics of our Earth. For instance, ocean bottom seismographs have significantly improved detection and recording capabilities at the world’s active plate margins, since we can place the instruments directly on the seafloor and close to seismic activity.
In this dissertation, I use ocean bottom seismograph data obtained from two separate deployments along the Gulf of California and 9º50’N East Pacific Rise high-temperature hydrothermal vent site, respectively. The Pacific-North America plate boundary in the Gulf of
California is a newly-formed oceanic rift (~6 Ma) [Atwater and Stock, 1998; Axen, 2000], characterized by long oceanic transform faults and short nascent spreading centers, and connects to the fast-spreading (~110 mm/yr) East Pacific Rise in the mouth of the gulf. The high- 4 temperature hydrothermal vent site at 9º50’N East Pacific Rise has been the subject of much interdisciplinary research since an eruption was discovered in 1991 [Haymon et al., 1993].
Hydrothermal fluid circulation at this site hosts black-smoker chimneys and chemosynthetic microbial and macro-faunal communities, and is driven by the presence of a shallow ~1.4 km deep axial magma chamber [Kent et al., 1993]. Earthquake hypocentral information collected from both of these studies shed light on the internal deformation processes at work within these oceanic settings, and is the subject of this dissertation.
5
Chapter 1:
The Mechanics of Earthquakes and Faulting in the southern Gulf of California
This chapter is in preparation for publication with co-authors J. B. Gaherty, W.-Y. Kim, and T. Diehl. 6
Abstract
Accurate earthquake locations and their focal mechanisms can illuminate the distribution and mode of deformation at rifted continental margins. The Pacific-North America plate boundary within the Gulf of California (GoC) provides an excellent opportunity to explore such rifting, as continental extension in the north transitions to seafloor spreading in the south. From
October 2005 to October 2006, an array of fourteen four-component ocean-bottom seismographs were deployed in the GoC to record earthquakes and other natural seismic signals as part of the
Sea of Cortez Ocean-Bottom Array (SCOOBA) experiment. By combining the data from this deployment with that from the on-shore NARS-Baja array, we detect and locate ~700 earthquakes (Mw 2.5-6.6) mainly located on the NW-SE striking transform faults that delineate the plate boundary. The earthquakes occur mainly during short-term swarm events that cluster predominantly around the inside corner of ridge-transform intersections, with events occurring on both the strike-slip and normal faults within the system. We use the waveforms of 31 events with high signal-to-noise long-period Rayleigh waves (Mw 3.45-4.92) to calculate regional deviatoric moment tensors and determine the focal mechanisms of these earthquakes. Most of the focal mechanisms exhibit right-lateral strike-slip focal mechanisms along the transform faults; however, several oblique and rotated mechanisms reveal a more complex structure within the gulf. In addition, we capture a swarm of events on Baja California along an apparent right- lateral NW-SE striking fault, which may be evidence of an extension of the Carmen fracture zone across the Baja California peninsula. The combination of high-resolution earthquake locations, with the broad distribution of event focal mechanisms improves our understanding of the distribution of seismic deformation within the greater extensional zone in the southern GoC.
7
1. Introduction
The Pacific-North America plate boundary within the Gulf of California (GoC) extends
~1300 km from the San Andreas Fault to the East Pacific Rise, and provides a modern-day example of the transition between continental extension to seafloor spreading. The current style of rifting within the gulf initiated ~6 Ma [Atwater and Stock, 1998; Axen, 2000], and can be described as an oblique rift system composed of an en echelon array of long right-lateral transform faults connected by short nascent ocean spreading centers. Due to changes in the amount of magmatism, sedimentation, and heat flow between the individual transform segments, distances between unextended continental crust and normal oceanic crust vary from segment to segment, and the system is characterized by both narrow and wide rift types [Lizarralde et al.,
2007; Páramo et al., 2008].
Localization of the plate boundary within the gulf has largely resulted from the transfer of the Baja California peninsula to the Pacific plate. This process was short-lived, lasting ~1 Ma
[Oskin et al., 2001], and magnetic anomalies at the Alarcon Basin in the mouth of the gulf indicate transfer was complete by ~3.6 Ma [Lonsdale, 1989; DeMets, 1995]. However, only
~92% of the Pacific-North America plate motion is accommodated within the gulf, suggesting that Baja California is not completely attached to the Pacific plate and may act as a separate block or microplate [e.g. DeMets and Dixon, 1999; Dixon et al., 2000; Fletcher and Munguía,
2000; Gonzalez-Garcia et al., 2003; Michaud et al., 2006; Plattner et al., 2007]. The additional motion may thus be accommodated by off-axis plate boundary faults, either on Baja California or within the gulf itself.
Precise earthquake locations can illuminate regions of active faulting, and help delineate the plate boundary and regions of off-axis faulting. In addition, the focal mechanisms of these 8 earthquakes can help accurately determine the direction of plate motion and the orientation of fault zones. An investigation into the temporal and spatial distribution of seismicity along each of the transform or basin segments provides insight into whether strain is differentially accommodated along the plate boundary. The fault pattern, orientation and kinematics of the plate boundary and other off-axis faults can thus be determined through earthquake locations and mechanisms, and allows for a more detailed understanding of how the Pacific-North America plate motion is accommodated within the gulf.
In this study, we accurately identify and relocate ~700 earthquakes recorded by a coupled on-shore/off-shore experiment deployed from October 2005 to October 2006 in the southern
GoC. We present a detailed earthquake relocation analysis of these events, and show that the majority of earthquakes locate along the plate boundary and the main bathymetric features of the
GoC. In addition, we perform regional deviatoric moment tensor (RMT) inversion solutions for
~30 events that exhibit clear Rayleigh waves in the 11-20 s period. The GoC provides an excellent place to study rifting processes because both sides of the rift are easily accessible and the history of plate interactions in the area is fairly well known.
2. Tectonic Background of the Gulf of California
The Pacific-North America plate boundary has undergone a dramatic reorganization over the past ~12 Ma, when the subduction of the now-extinct Farallon plate ceased beneath the North
America plate [e.g. Stock and Hodges, 1989; Atwater and Stock, 1998]. As the former Pacific-
Farallon ridge approached the paleotrench west of Baja California, Farallon spreading and subduction stalled due to a clockwise change in the spreading direction, with motion becoming more parallel with Pacific-North America plate motion [Michaud et al., 2006]. As the subducted
Farallon slab detached, the remnants of the Farallon plate fragmented into the Magdelena and 9
Guadalupe microplates. Magnetic anomalies show that spreading continued between the Pacific-
Magdelena and Pacific-Guadalupe plates from ~12 to ~7-8 Ma [e.g. Lonsdale, 1991; Michaud et al., 2006]. Extinction of the spreading centers ~7-8 Ma may be concurrent with the onset of dextral strike-slip motion along the Tosco-Abreojos Fault west of the Baja California peninsula along the former paleotrench [Spencer and Normark, 1989]. The duration and amount of motion along the shear zone, and whether the shear zone accommodates some Pacific-North America plate motion today, is still debated [Dixon et al., 2000; Marsaglia, 2004].
Around ~8-6 Ma, the Pacific-North America plate boundary moved inland to the former
Farallon subduction back-arc, transferring the Baja California peninsula to the Pacific plate over a period of ~1 Ma and opening the present-day GoC [Atwater, 1989; Oskin et al., 2001]. Rifting in the GoC is generally thought to have occurred in two distinct phases, an early ‘protogulf’ phase of continental rifting within the former Farallon subduction back-arc spreading zone, followed by a period of transtensional plate motion accommodated by ridge-transform structures still active in the GoC today [e.g. Karig and Jensky, 1972; Fletcher and Munguía, 2000]. The transition between these two phases is thought to have initiated ~5-6 Ma, and concluded with the formation of new oceanic crust at the Alarcon basin ~3.6 Ma [e.g. Lonsdale, 1989; Umhoefer et al., 1994]. The transfer of the Baja California peninsula to the Pacific plate may not be complete, however, since recent evidence shows that Baja California-North America plate motion is ~3-8 mm yr-1 slower than Pacific-North America plate motion [Dixon et al., 2000], moving at a rate of ~43-47 mm yr-1 that increases from north to south along the plate boundary
[Plattner et al., 2007].
Continental extension in the northern GoC is characterized by a complex system of multiple oblique-slip faults that interact with one another [Persaud et al., 2003], and transitions 10 to classic ridge-transform fault (RTF) geometry around the Guaymas basin, marking the beginning of the southern GoC (Figure 1) [eg. Einsele, 1982; Fabriol et al., 1999]. The
Guaymas basin is a narrow rift (conjugate margin width < 200 km), blanketed by ~1-2 km of organic-rich sediment, and is an observed site of vigorous hydrothermal circulation and robust magmatism which has accreted ~6-8 km of crust since continental breakup ~6 Ma [Lizarralde et al., 2007; Lizarralde et al., 2011]. Additionally, an unknown volume of magma has intruded into the overlying sediments, and is observed to release carbon up to 50 km away from the plate boundary, a much greater distance than that observed at unsedimented mid-ocean ridges, and promotes methane-hosted seafloor biological communities [Lizarralde et al., 2011].
The Carmen, Farallon, and Pescadero basins directly south of Guaymas (Figure 1) are in the embryonic stages of seafloor spreading, and are characterized as magma-poor sites with narrow bathymetric deep rift zones [Lizarralde et al., 2007]. In contrast, the Alarcon basin in the mouth of the gulf is a wide rift (conjugate margin width >200 km) that has evolved into a
‘normal’ mid-ocean ridge, where magnetic anomalies show that oceanic crust has been accreting since ~3.6 Ma, with a mean spreading rate of 50 mm yr-1 over the last 1 Ma [DeMets, 1995].
Additionally, the Alarcon basin is covered by <40 m of sediment and was magma-poor during rifting [Lizarralde et al., 2007]. Like the Guaymas basin, the Alarcon basin supports hydrothermal circulation, although heat flow at Alarcon is suppressed well below typical oceanic lithosphere values (~15-55% of heat input at the base of the lithosphere for crust of this age)
[Fisher et al., 2001]. The Guaymas and Alarcon basins delineate the north and south ends, respectively, of the southern Gulf of California, which is the focus of our study.
11
3. NARS-Baja and SCOOBA Seismograph Deployments
In the spring of 2002, the onshore Network of Autonomously Recording Seismographs
(NARS)-Baja array [Trampert et al., 2003; Clayton et al., 2004] was established as a cooperative effort between Utrecht University (the Netherlands), Centro de Investigación Científica y de
Educación Superior de Enseñada (CICESE, Mexico), and the California Institute of Technology.
Fifteen three-component broadband seismometers recording at 20 samples/s were deployed on
Baja California and the Sonora and Sinaloa coasts of Mexico, with a typical station spacing of
~100-150 km (blue squares in Figure 2). Each seismic station is composed of a Streikeisen STS-
2 broadband sensor, a global positioning system (GPS) receiver, and a 24-bit data logger. The stations remained active until the end of 2008, however, we encountered timing errors for station
NE76 (green square in Figure 2) on the order of ~10-15 seconds, thus inhibiting use of this station for our earthquake relocation analysis (described in the section below).
From October 2005 to October 2006, eight ocean-bottom seismographs (OBSs) were deployed in the southern Gulf of California as part of the Sea of Cortez Ocean Bottom Array
(SCOOBA). Small sub-arrays were deployed in the Guaymas and Alarcon basins to supplement wide-angle and multi-channel seismic reflection data collected by Lizarralde et al., 2007, with an
~20 km station spacing. Additionally, two independent stations were deployed between the two subarrays, with an ~100 km station spacing (red inverted triangles in Figure 2). The OBSs used in SCOOBA are four-component broadband instruments recording at 32 samples/s, and are operated by Scripps Institution of Oceanography. During the SCOOBA deployment, fourteen
NARS-Baja stations were operational; therefore, a total of twenty-two seismographs compose the coupled on-shore/off-shore array used for this study. Data from both the NARS-Baja and 12
SCOOBA deployments are widely available for download from the Incorporated Research
Institutions for Seismology (IRIS) Data Management Center (DMC).
The NARS-Baja and SCOOBA deployments greatly augment the existing global array, and allow for more comprehensive insight into the seismic activity within the GoC. From
October 2005 to October 2006, 47 earthquakes from the GoC and Baja California were reported in the International Seismological Center (ISC) catalog, and seven of these events were large enough magnitude (M > 4.5) to calculate centroid moment tensors (CMTs) reported in the global catalog [http://www.globalcmt.org/] (black and white dots, respectively, in Figure 2). Many of these events occur as foreshocks to a larger magnitude mainshock (e.g. 4 January 2006 sequence), and on 28 May 2006, an earthquake doublet occurred on the southern Guaymas basin; these behaviors are typical of the mid-ocean ridge environment [McGuire et al., 2005; McGuire,
2008]. However, many of these earthquakes do not locate on the plate boundary from Bird,
2003, and detection of seismicity in the southern GoC is sparse.
4. Initial Earthquake Detection and Location Techniques
After removing the instrument response from the SCOOBA and NARS-Baja stations, we automatically detect P-waves on the vertical-component only, with an algorithm that compares the short-term average (STA) to long-term average (LTA) amplitudes with a 5-15 Hz
Butterworth filter applied. When four or more P-wave picks are detected, an earthquake is located based on the iasp91 global velocity model [Kennett and Engdahl, 1991] and the TauP travel-time grid [Crotwell et al., 1999]. We manually inspect the P-wave picks for the automatically-detected events, and due to the emergent nature of the P-wave arrivals on the
OBSs, we had to either move badly-located picks to the correct arrival time or remove mispicks altogether. Additionally, since the STA/LTA automatic detection scheme is mainly sensitive to 13 changes in amplitude, high-frequency spikes are sometimes detected and picked inadvertently, resulting in the location of spurious ‘events’. After phase picking is complete, we determine the local magnitude (ML) of each event by averaging the zero-to-peak amplitude of the largest arrival on the vertical component for all available stations, with the Wood-Anderson seismometer filter applied [e.g. Richter, 1935].
The manually inspected P-wave pick information is then used to locate earthquakes with the Hypoinverse algorithm [Klein, 2002]. Hypoinverse uses a single-event procedure to locate the individual picks to an earthquake, with station location information and a typical four-layer one- dimensional (1D) velocity model that is appropriate for the GoC. Due to timing errors found for station NE76 and the emergent nature of the P-wave arrivals on the SCOOBA stations, picks recorded on the NARS-Baja stations are weighted at 100%, the SCOOBA stations at 75%, and the station NE76 at 25%. Furthermore, pick information on stations ≥500 km away from a detected event is downweighted compared to close station information. An earthquake is detected when four or more P-wave picks are correspondingly identified, and then located with a starting depth of 5 km. During the location process, however, depth is allowed to vary as a free parameter and can therefore deviate away from the initial 5 km starting depth. With this procedure, 695 earthquakes are identified, mainly locating on the plate boundary with depths <10 km (Figure 3a). Mean absolute location errors are ~6.29 km in the horizontal, and the root- mean-square (RMS) error in timing is ~0.84 s.
5. Earthquake Relocation Procedure
In order to gain a high-resolution picture of seismic deformation along the plate boundary, we apply a hierarchical analysis to the seismicity catalog for earthquake relocation purposes. First, we focus on events that meet stringent criteria based on the station distribution 14
(≤180º azimuthal station gap), and RMS and horizontal errors (≤0.3 s and ≤6.29 km, respectively). Out of the 695 events located with Hypoinverse, 228 earthquakes meet these criteria, and are mainly located on or near the major NW-SE strike-slip faults that delineate the plate boundary (Figure 3b). We assign these events to category ‘A’, and further subdivide the remaining 467 events into B-D categories by relaxing the horizontal error cutoff, and constraining based on station distribution and RMS errors only (Table 1).
The first step in our relocation procedure is to invert for the best-fitting one-dimensional
(1D) velocity model based on the hypocentral parameters and the station array configuration using the VELEST algorithm [Kissling, 1988; Kissling et al., 1994]. In general, the travel time of a seismic wave is a nonlinear function of the earthquake’s hypocentral parameters and the seismic velocities sampled along the raypath between the hypocenter and station, thus termed the joint velocity-hypocenter problem. We apply VELEST to the best-determined ‘A’ category earthquakes to find a 1D velocity model that minimizes the RMS timing errors of the individual earthquakes and the station corrections. Station corrections are influenced by unmodeled near- surface velocity heterogeneities and station elevations, and are used to account for lateral variations in the shallow subsurface velocity structure during the inversion process. This procedure has proven successful for relocating earthquakes using a coupled on-shore/off-shore array in northern Chile [Husen et al., 1999], and thus is the preferred technique to use for our array configuration.
For the VELEST inversion, we begin with the tectonic North America S-wave velocity model of Grand and Helmberger, 1984 and convert it to a P-wave velocity model by varying a vp/vs ratio for the crust and the mantle [B. Savage, pers. comm.]. Using this as an a priori starting velocity model, VELEST traces rays from the initial hypocenters to the stations and 15 calculates theoretical arrival times. The optimal 1D velocity model is then chosen through a trial-and-error process that leads to an average minimum of root-mean-squared (RMS) errors between the observed and theoretical arrival times for all earthquakes [Kissling et al., 1994].
However, three-dimensional structure is hard to account for by a 1D velocity model, especially in a region like the gulf where young oceanic crust is abutting against continental crust. More specifically, the Moho or crust-mantle boundary is marked by a large velocity increase ~12-15 km deep for the GoC, while the Moho in the tectonic North America model of Grand and
Helmberger, 1984 (where most of our seismometers are located) is 35 km deep. To account for these variations in Moho depths, we first try modifying the tectonic North America model by moving the Moho to a range between 10-35 km deep, and find that the minimum station corrections are attained with two smaller velocity jumps at 20 and 35 km, respectively. We use this modified tectonic North America model as the starting velocity model in the VELEST inversion procedure, and after two inversions, the final velocity model is shown in Figure 4. The final velocity model is marked by a gradual increase in velocity, rather than any large velocity step indicative of a Moho, and represents a 1D average of the 3D structure sampled by the raypaths.
We revise the hypocentral parameters of the B-D quality earthquakes by applying this velocity model in the single-event mode with VELEST. On average, the VELEST locations differ from the Hypoinverse locations by ~0.36 s in timing and ~6 km in the horizontal; additionally, the average RMS errors are reduced in half from ~0.84 s in Hypoinverse to ~0.39 s in VELEST.
To further improve the event locations of all of the A-D quality events, we subdivide the earthquakes into clusters and relocate using the hypoDD double-difference algorithm 16
[Waldhauser and Ellsworth, 2000]. The hypoDD algorithm utilizes differential travel-times between closely-spaced pairs of earthquakes recorded on shared stations to minimize location errors based on unmodeled 3D velocity structure, and requires the hypocentral separation between two earthquakes to be small compared to the event-station distance and the scale length of the velocity heterogeneity. The rays therefore travel similar paths from the sources to a common station, and their travel time difference will not be affected by the unmodeled structure for the parts of the path they share, reducing the dependence of the velocity model for relocation.
Thus, the difference in travel times between the two events observed at a particular station can be attributed to the spatial offset between them with high accuracy.
We apply the hypoDD double-difference algorithm to the clusters of earthquakes using differential times calculated from our P-wave picks. Tests are performed in hypoDD by initially setting all of the earthquake focal depths (the least constrained hypocentral parameter) to 5, 10, and 15 km, respectively, and the maximum distance between events to 10, 20, and 30 km, respectively, for a total of nine different trials examined. The first set of iterations (iterations 1-
5) is performed without reweighting the data, followed by another set of iterations (iterations 6-
10) with distance and misfit weighting applied to the data to remove outliers. The minimum
RMS timing errors resulted when the earthquake depths are set to 10 km initially, and when the maximum inter-event distance is 20 km or less. Overall, the double-difference locations with the catalog picks form tighter clusters of earthquakes than the relocations from VELEST as shown in
Figure 5, and results from this analysis will be discussed in a later section.
6. Focal Mechanisms from Regional Waveform Inversion
Focal mechanisms allow us to constrain the type, orientation, and kinematics of faulting, and provide useful information in understanding the seismotectonic setting of the GoC. We 17 examine the entire catalog of events from October 2005 to October 2006, and find 31 events that have clear Rayleigh waves in the 11-20 s period range and did not occur temporally close to any teleseismic earthquakes reported in the Global CMT catalog that could perturb the waveform.
The events chosen are geographically diverse, and range in magnitude from ML 3 to 4.8.
For each of these events, we select clear records from the 14 NARS-Baja and 8 SCOOBA stations to determine the source mechanisms of these events by regional waveform inversion.
We select stations that record waveforms with the largest signal-to-noise ratio (SNR) and have minimal timing errors. More specifically, we make group velocity plots of each of the three- components recorded by a particular station to assess the SNR, and to examine whether station timing errors are an issue. In this way, we modify station NE76 by ~10-15 seconds to use for waveform inversion techniques, even though we heavily downweighted NE76 for location purposes. For each event, the number of stations chosen that match these criteria varies from 3 to 15 stations. In cases where the number of stations chosen for inversion is low, the P-wave first motions recorded on the vertical component provide additional information to constrain the focal mechanisms.
We first calculate the Green’s functions of the observed displacement waveforms using the frequency-wavenumber (f-k) integration method of Saikia, 1994, which models the earthquake as a point source embedded in a 1D velocity model. For our inversion, we use the 1D minimum P-wave velocity model determined in VELEST, and an S-wave velocity model calculated using a vp/vs ratio of 1.73 (Figure 4). In addition, a moderate level of anelastic attenuation is included in the modeling. We model the complete waveform, including body and surface waves, over a period range from 10 to 33 s for most of the regional records at distances greater than 100 km, and we use a period range from 5 to 20 s for records from stations at shorter 18 distances, following the procedure described by Kim et al., 2010. Long-period body waves, Pnl, and fundamental mode Rayleigh and Love waves are most evident in the filtered displacement records. The long-period Pnl phase in the 10 to 33 second period is robust in the presence of heterogeneous structure like that of the gulf [e.g. Dreger and Helmberger, 1993], and variations in the relative amplitudes of the initially up- or down-going long-period Pnl phase at the source are sensitive to source depth [Kim et al., 2010].
We calculate synthetics for the deviatoric component of the moment tensor only (no volume change), thus the source is modeled as a double-couple that can be described by its strike, dip, and rake. The inversion method matches the observed and synthetic seismograms over discrete wave trains, and allows relative time shifts between pairs of observed and synthetic waveforms to find the best fit [Zhao and Helmberger, 1994]. The focal mechanism is found by a grid-search procedure through the entire parameter space of strike, dip, and rake, and we estimate seismic moment by taking the ratio between the peak amplitude of the observed record and the synthetic seismogram, respectively, for the best-fit time window [e.g. Du et al., 2003;
Kim, 2003].
Figures 6a and 6b compare the observed and synthetic waveforms for the preferred solution to a typical right-lateral strike-slip event most commonly observed within the GoC, and for an unusual oblique normal faulting event observed within the Guaymas Basin, respectively.
A trapezoidal source time function is used to calculate synthetics with a rise time of 1 sec, hold time 1 sec, and a fall time of 1 s; however, this source-time function is too short to be determined in the inversion because of the long-period pass band used of 10 to 33 s [e.g. Kim et al., 2010].
To minimize the fitting error, we estimate the cross-correlation function between the observed and synthetic waveforms, and allow for a relative time shift between them, where 19 positive values indicate that the model is too fast and vice versa. The time shifts required to align the traces are mostly positive, ranging from -1.41 to 4.06 seconds and average ~0.98 seconds (Table 2). The positive time shifts indicate that the velocity model derived from the
VELEST inversion may be slightly faster than the observed values for most of the raypaths. A range of velocity models were tested during the inversion process, and overall, the VELEST model fit the synthetics better and provides better constraints on source depth. Thus, we conclude that this is an acceptable velocity model for regional waveform modeling and source mechanism inversion for the GoC.
We run the inversion for a range of source depths for each earthquake, seeking the global minimum misfit and the preferred source depth. We measure the quality of fit between the observed record and the synthetic seismograms (Figure 6a and 6b) by calculating the variance reduction and the fitting error function described by Zhao and Helmberger, 1994, and choose the preferred source depth for each earthquake that maximizes variance reduction and reduces fitting error. Figure 6c and 6d show the range of trial source depths for the typical right-lateral strike- slip event most common in the GoC, and for the oblique normal faulting event within the
Guaymas Basin, respectively.
The preferred focal mechanisms for the 31 earthquakes constrained by regional moment tensor inversion are shown in Figure 5, and their associated focal parameters are shown in Table
2, broken up into geographic clusters following the relocation methods described above. The earthquakes examined have fairly well-constrained centroid depths between 1-12 km, and are determined relative to the seafloor. We do not account for a water or sediment layer in our one- dimensional velocity model derived with VELEST (Figure 4), since these layers vary dramatically along the length of the GoC; however, the long-period body waves that are sensitive 20 to source depth are well-fit by the VELEST model and suggest small centroid depth inaccuracies on the order of ~1-2 km. For the faulting parameters and centroid depth, the range of appropriate mechanisms and depths is defined as fitting error values below 1.05 times the minimum fitting error (i.e. 1.05Emin), as shown in Figure 6c and 6d.
The seismic moment of the earthquake mechanisms determined with the regional centroid moment tensor inversion range from 2 x1014 to 2.72 x1016 N-m. Further, we use the standard deviation in the values calculated from each trace to characterize the uncertainty in the estimate of the seismic moment [e.g. Du et al., 2003], which in this study averages ~0.35 x1016
N-m. From the seismic moment (Mo), we calculate the moment magnitude (Mw) of each
7 earthquake, where Mw = 2/3*log10Mo (dyne-cm) – 10.7 [Kanamori, 1977], where 10 dyne-cm equals 1 N-m. Figure 7 shows the relationship between the local magnitude (ML) and moment magnitude (Mw) for the 31 earthquakes examined with regional waveform modeling. Although there is significant scatter in the results, the data are best fit by Mw ≈ 0.62*ML + 1.8. In general, this result indicates that the ML underestimates the actual size of smaller earthquakes, and is not calibrated well for the GoC. Thus, we transform the magnitudes for each of the earthquakes from ML to Mw, and use this to examine the clusters of earthquakes described in the following section.
7. Results
In this section, we describe the results from each of the clusters relocated through the hypoDD double-difference analysis (Figure 5). We note that only the P-wave picks are used in the relocation procedure, therefore, we have poor constraint on the focal depth of the earthquakes. The centroid depths for events with regional moment tensor solutions are well constrained, however, so we keep the depths of these events constant and correct the depths of 21 the remaining earthquakes accordingly. In addition, many of the relocated clusters of earthquake activity are associated with Global CMT (GCMT) events recorded during the SCOOBA deployment. In the following discussion, we explore the spatial and temporal relationship between the earthquakes within the clusters and the large-magnitude events recorded within the
GCMT catalog.
4 January 2006 Mw 6.6 Guaymas Transform event: The northernmost cluster of earthquakes relocated by the hypoDD double-difference procedure surrounds the Mw 6.6 mainshock of 4 January 2006, recorded in the GCMT catalog (Figure 8a). In addition, an Mw 5.2 foreshock of this event also was recorded in the GCMT catalog, and two foreshocks of the 4
January 2006 mainshock and one other event on 20 April 2006 are constrained by regional moment tensor (RMT) inversion (Figure 8a). All of the focal mechanisms are right-lateral strike- slip, consistent with fault motion along the Guaymas transform. The ~130º strike of the nodal plane for the Mw 6.6 mainshock from the GCMT catalog agrees well with mechanisms found by
RMT inversion and the bathymetry of the Guaymas transform. Note that the bathymetry and the plate boundary model of Bird, 2003 do not always coincide with the bathymetry, implying a more complex geometry for the plate boundary than that incorporated by Bird, 2003.
We examine the 67 events relocated by hypoDD both along-strike (~130º; profile A-B), and across-strike (profile C-D from east to west), and observe shallow earthquakes (<10 km focal depth) and potentially two clusters of earthquakes coinciding with bathymetry on either side of the inferred plate boundary of Bird, 2003 (Figure 8b-e). The earthquakes examined in this cluster are well-recorded, with picks on a combination of 10 stations. Average RMS timing errors after relocation are ~0.29 s (Figure 8b), and bootstrap error analysis of this cluster reveals location errors on the order of ~3 km. Although we set initial focal depths to 10 km in hypoDD, 22 many of the depths become shallow after relocation (<10 km initial starting depth), consistent with the 3 km focal depths of the earthquakes constrained by RMT inversion (Figures 8d and
8e).
In addition, 34 events occur in the ±3 days surrounding the mainshock, with 12 foreshocks and 21 aftershocks occurring mainly with the northeast cluster of activity at the site of the 4 January 2006 Mw 6.6 mainshock (Figure 8f). The foreshocks and aftershocks occur within an average of ~16 km away from the mainshock, though the foreshocks are slightly more spatially clustered together than the aftershocks. Further, much of the foreshock activity occurs within hours of the mainshock (Figure 8c), with a large ramp-up in seismic activity over ~1 day, while the aftershock activity is more temporally diffuse occurring over ~3 days.
The foreshocks also release much more moment than the aftershocks, with larger magnitude foreshocks (Mw 3.4 – 5.2) releasing almost twice as much total moment compared to
17 the aftershocks (Mw 3.4 - 4.8); or more specifically, ~1.19 x 10 N-m total moment released from the foreshocks compared to the ~5.24 x 1016 N-m released from the aftershocks (Figure 8f).
Overall, the 4 January 2006 Mw 6.6 mainshock, and foreshock/aftershock sequence of earthquakes, is responsible for releasing almost all of the moment (~99%) for the northern
Guaymas transform cluster of earthquakes, and is the most energetic sequence recorded during our year deployment.
Southern Guaymas transform and 11 May 2006 Guaymas basin sequence: The
Guaymas transform connects with the upper Guaymas basin at ~27.5ºN, and this ridge-transform intersection was very active during our year deployment, with 70 earthquakes relocated with the double-difference algorithm by P-wave picks on a combination of six stations (Figure 9a and
10a). Average RMS error of the relocated hypocenters is ~0.08 s, and relative location errors 23 along the Guaymas transform and basin are on the order of ~1.5 km and ~4.3 km, respectively.
Of these, eight were large enough magnitude, and had clear Rayleigh waves for regional moment tensor inversion (Figure 9a). The seven events on the southern Guaymas transform are all right- lateral strike-slip along ~300º-320º (or ~120º-140º strike) consistent with the direction of transform motion, including an event on 23 April 2006 that locates at the ridge-transform junction (Table 2). The preferred depths from the RMT analysis along the southern Guaymas transform range from 2-7 km, and cross-sections of the events along-strike and across-strike
(profile A-B and C-D; Figure 10c-d) indicate shallow depths of all of the earthquakes, with the majority <10 km deep. In addition, the lack of any apparent dipping feature across-strike (profile
C-D; Figure 10d) is consistent with the right-lateral strike-slip behavior along the transform.
On 11 May 2006, an event of Mw 3.94 occurred on the upper Guaymas basin, and exhibits an oblique normal faulting type mechanism, which is unique to our study (Figures 6b and 9a). More specifically, a large increase in the number of events at this time (day 200; Figure
9b) is from events along the northern Guaymas basin that occur within hours of the 11 May 2006 event, including one foreshock ~20 minutes before the event (Mw 3.4; light grey dot in Figure
9a), seven aftershocks within an hour after the event (Mw 3-3.4; grey dots in Figure 9a), and two additional aftershocks within twelve hours of the event (Mw 3.4 and Mw 3.4, respectively; dark grey dots in Figure 9a). The 11 May 2006 event along the Guaymas basin has a very shallow focal depth of 1 km, and the across-strike profile A-B (Figure 10f) indicates that these depths are shallow, and perhaps even more so than recovered by the hypoDD double-difference analysis.
The sediment layer is very thick along this section of the Guaymas Basin (~1-2 km thick;
Lizarralde et al., 2011], and the shallow centroid depth of 1 km may be in error due to structural effects not taken into account; however, the synthetics based on our velocity model fit the 24 observed waveforms well with small timing errors (Figure 6b), which indicates this error is most likely small. Although these events occurred very spatially and temporally close together, the events only account for ~7% of the total moment release in this cluster (Figure 9d), suggesting that much more of the seismic deformation is accommodated along the Guaymas transform rather than within the basin itself.
28 May 2006 doublet and 30 July 2006 Carmen transform sequence: Along the northern Carmen transform, we recorded over 100 hypoDD relocated earthquakes during our year deployment with differential time data derived from 9 stations (Figure 11a and b). Average
RMS timing errors of these events is ~0.02 s, and relative location errors determined by bootstrap error analysis are ~1 km. Three of these events were large enough to be recorded in the GCMT catalog, and two additional earthquakes were suitable for regional moment tensor inversion. Although the GCMT earthquakes are largely non-double couple, the focal mechanisms from the RMT inversion indicate right-lateral strike-slip motion along ~120-130º, consistent with plate motion in the region. Along- and across-strike cross-sections of the ~100 events relocated along the Carmen transform indicate shallow depths (<15 km) for all of the earthquakes, consistent with focal depths of 5 km derived from RMT inversion (Figure 11d and
11e). Further, the rupture properties of the 30 July 2006 GCMT event (Figure 11a) have been studied in detail by Roderíguez-Lozoya et al., 2010, who found a best-fit focal depth of 5.5 km for this earthquake.
Spatially, the bulk of the events form a broad cloud within the inner corner of the ridge- transform intersection. The GCMT and RMT fault planes suggest that much of this activity is on the transform southeast of the lower Guaymas spreading center; however, given the spatial 25 correlation with the extensional basin, we speculate that some of these events may be oblique or normal faulting events within the basin low.
Of the three GCMT earthquakes, a doublet, or two earthquakes with similar magnitudes occurring at approximately the same time and location, occurred on 28 May 2006 at 14:02 and
14:18 with magnitudes Mw 5.2 and Mw 5.1, respectively. Five days before the doublet, however, a particularly energetic sequence of 21 events was recorded in this same region (day 205 in
Figure 11f). Thirteen events occurred during the two hours leading up to largest event in this sequence, an Mw 4.1 event on 23 May 2006 at 21:23; however, the magnitudes of these events ranged from Mw 3-4, suggesting that this sequence is swarm-like in its behavior, rather than a foreshock-mainshock sequence (Figure 12a). Spatially, this swarm of events forms two streaks, one to the north and another to the northeast leading up the largest event in the sequence (white dot in Figure 12b).
During our deployment, moment release along the Carmen Transform largely results from events associated with the 23 May 2006 swarm sequence, the 28 May 2006 doublet and the
30 July 2006 Mw 5.9 event, the largest on the Carmen Transform (Figure 11f). Examination of the earthquakes temporally clustered around the 28 May 2006 doublet and the 30 July 2006 Mw
5.9 event display mainshock-aftershock behavior, rather than a swarm-type sequence (Figure
12c-f). The doublet on 28 May 2006 had one Mw 4.8 foreshock, and in the day following the mainshock, 20 aftershocks in the range of Mw 2.8-3.6 (Figure 12c). These events form an indistinct cluster of activity surrounding the mainshock with no apparent spatial and/or temporal pattern (Figure 12d). The 30 July 2006 Mw 5.9 on the other hand was completely devoid of foreshocks, and triggered ~15 aftershocks in the first 24 hours (Mw 3.4-4.4) that spread away 26 from the mainshock to the north-northwest, similar to the 23 May 2006 swarm sequence (Figure
12e-f).
Farallon Transform: Along the Farallon transform segment, 73 earthquakes were relocated with the hypoDD double-difference algorithm with P-wave picks recorded on a total of
10 stations, and average RMS timing errors of ~0.01 s (Figure 13a-b). Bootstrap error analysis reveals location inaccuracies on the order of ~0.3 km. Out of these events, three were suitable for RMT inversion, including a pair of earthquakes occurring on 27 November 2005 that are close enough in magnitude to potentially be considered a doublet, rather than a foreshock- mainshock sequence (Figure 13a). The results from the RMT inversion show right-lateral strike- slip focal mechanisms along an ~120º-130º azimuth (Table 2, Figure 13a); however, the 19
February 2006 event has approximately east-west oriented strike consistent with the observations and modeling of earthquake mechanisms located within the inside corner of ridge-transform intersections [Engeln et al., 1986; Wolfe et al., 1993; Behn et al., 2002]. Focal depths of these earthquakes are shallow, ranging from 2-5 km, and along-strike and across-strike cross-sections of the hypoDD relocated events indicate depths of <15 km (Figure 13c-e).
Spatially, this cluster of events appears very similar to the distribution of seismicity along the Carmen transform. The events here tend to cluster at either end of the Farallon transform, within the inside corner bathymetric lows where it connects to the Carmen and Farallon basins, respectively. Furthermore, the 19 February 2006 event may reflect the geometry within this inside corner, since its 180º strike is rotated 50º away from the two events on the Farallon transform, and its strike bisects the ridge-transform intersection [e.g Engeln et al., 1986].
The 27 November 2005 Mw 4.49 and 4.69 earthquakes release the greatest amount of moment (~28% of the total moment) along the Farallon transform during our year deployment 27
(Figure 13f). In addition, a sequence of four earthquakes beginning with the largest Mw 3.97 event on 3 November 2005 is responsible for a large jump in the seismic moment over time, but only contributes ~3% to the total moment released on this transform. No large swarm events or foreshock-mainshock-aftershock sequences were detected on the transform during our time period of study; however, a study conducted by Castro et al., 2010 relocated a cluster of earthquakes occurring with an Mw 5.5 right-lateral strike-slip event on 22 February 2005 (prior to the SCOOBA deployment) along the Farallon transform at ~25.9ºN, -110.1ºW, which most likely released much of the seismic slip along this portion of the plate boundary.
Alarcon Ridge-Transform System: In the southernmost portion of the GoC, we relocated over 100 events along the Alarcon ridge-transform system with average RMS timing errors of ~0.15 s (Figure 14a). Bootstrap location errors reveal uncertainties on the order of ~1.6 km; however, the earthquakes in this region are most likely the poorest located, since the P-wave picks come from a total of 5 stations that are not azimuthally well-distributed (Figure 14b). In fact, many of the earthquakes align with the subarray within the Alarcon basin, indicating that relocation draws the location to the nearest station rather than the earthquake’s actual location.
During our year deployment of OBSs, one GCMT event and three events suitable for
RMT inversion were relocated within or near this cluster of events, yet only the 9 August 2006 event was relocated using the double-difference algorithm. Focal mechanisms of two events on the Alarcon transform show right-lateral strike-slip mechanisms along ~120-140º azimuth. The centroid depths derived from RMT inversion range from 2-8 km in depth, consistent with the along-strike and across-strike cross-sections of the hypoDD relocated hypocenters, which are predominantly <10 km deep (Figures 14c-d). One additional event suitable for RMT inversion is situated within a cluster of ~20 events located on or near the Alarcon Seamounts (abbreviated 28
ASM in Figure 14a), and exhibits a highly-oblique right-lateral strike-slip mechanism with a reverse faulting component. The Alarcon Seamounts are located on ~1.0-1.5 Ma crust most likely sourced from mid-ocean ridge basalt (MORB) type mantle of the Alarcon Rise [Castillo et al., 2002]. The location of these events and the small location errors suggest that Alarcon
Seamounts are still active; however, additional evidence to support this hypothesis is lacking.
Examination of the cumulative number of earthquakes and the cumulative seismic moment release over the timeframe of our deployment indicates a steady rise of earthquake activity from day ~65-120 (26 December 2005 to 18 February 2006), with more than half of the earthquake activity occurring during this almost two month time period (Figure 14f). The largest
16 release in seismic moment (~1.21 x 10 N-m), however, occurs from the 5 March 2006 Mw 4.69
RMT event, which is outside of this time frame. This event is responsible for relieving half of the total seismic moment (3.62 x 1016 N-m) released along these faults during our deployment period. Further, there are no swarm sequences or foreshock-mainshock-aftershock sequences easily identifiable within the time series.
August 2006 Baja California sequence: Twelve out of the thirteen earthquakes relocated with hypoDD on Baja California occurred between 1 August 2006 and 22 August 2006
(Figure 5). Average RMS timing errors are on the order of ~0.05 s, and bootstrap location uncertainties are on the order of ~1.5 km. Focal mechanisms for seven of these earthquakes (Mw
3.4 – 4.3) indicate slip along a right-lateral strike-slip fault oriented approximately southeast- northwest (Figure 15a). The largest event in this sequence (Mw 4.9) occurred on 8 August 2006 at 04:42, however, the waveforms were not suitable for RMT analysis due to a high level of noise. First motions of this event on three stations (NE75, NE76, and N03) are consistent with the rest of the activity on Baja California, and thus the event most likely exhibits a right-lateral 29 strike-slip faulting mechanism. Due to the short duration (~22 days) of earthquake activity, the relatively small range in magnitudes, and the largest event temporally locating within the middle of the sequence [e.g. Roland and McGuire, 2009], we classify this sequence of events as a swarm. Examination of the cumulative number of events and the cumulative seismic moment release over the ~22 days of activity shows the largest jumps occur around the RMT events, which are shown as dotted lines in Figure 15b and 15c, respectively.
Although there are several strike-slip faults evident in this region, these fault are not easily accessible and are not very well mapped [R. Castro, pers. comm.]. Thus, these events may illuminate a previously unknown or unmapped fault structure. These earthquakes are large enough to have been felt, which would provide additional information and constraints on these earthquakes; however, the closest town of San Ignacio (white star in Figure 15a) has a population of only ~4,000, and whether this swarm of earthquakes was felt there is unknown [R. Castro, pers. comm.].
8. Discussion and Conclusions
In summary, we detect and locate ~700 events (Mw ~2.5-6.6) during a coupled on- shore/off-shore deployment of seismographs from October 2005 to October 2006. By adopting a hierarchical analysis to these events, we invert for a 1D P-wave velocity model based on 228 of the best-located events using the VELEST algorithm [Kissling et al., 1994] that accommodates our station distribution and the 3D velocity structure within the gulf. We then subdivide the
VELEST relocated events into clusters, and apply the hypoDD double-difference relative relocation procedure to the events using differential time data derived from P-wave picks
[Waldhauser and Ellsworth, 2000] to obtain a high-resolution picture of the distribution of seismic deformation within the GoC. Through this procedure, we locate 428 well-connected 30 events, with horizontal absolute location errors conservatively estimated at ~6.5 km, derived from the Hypoinverse analysis, and average RMS timing errors and bootstrap relative location errors of ~0.1 s and ~2 km, respectively, determined from the hypoDD double-difference analysis.
The 428 well-determined hypoDD relocated events, and the additional 267 VELEST relocated events, locate predominantly on the plate boundary and on off-axis faults in southern
Baja California. For 31 events with high signal-to-noise ratio surface waves, we apply regional deviatoric moment tensor (RMT) analysis to determine focal mechanism solutions, and find all of the events have shallow focal depths (1-12 km deep), and locate mainly along the NW-SE striking transform faults within the GoC. We compare the ~700 events relocated in our analysis to a study by Castro et al., 2010, who relocated earthquakes reported by the National Earthquake
Information Center (NEIC) using data from the NARS-Baja array from 2002-2006 (Figure 16).
Many of the events relocated by Castro et al., 2010 also locate on the plate boundary, and fill in spatial gaps along the transforms that were not seismically active during our OBS deployment.
Most notably, a cluster of sixteen earthquakes along the southern Alt transform fault fills in the transform with seismicity in a region where we did not locate many earthquakes during our deployment. This cluster of activity occurred mainly during January 2004, only twenty-one months before our deployment, most likely relieving the tectonic stress in the region.
Much of the relocated seismic activity occurs in swarm-like clusters, which are defined as occurring over a relatively large spatial area compared to the total release of seismic moment, have a long temporal duration of many similar magnitude events, typically lacks a large magnitude mainshock with the largest events occurring later in the sequence, and fails to decay with time following standard aftershock scaling laws [e.g. McGuire et al., 2005; Roland and 31
McGuire, 2009]. Additionally, swarm sequences have often been characterized as having anomalously low effective seismic stress drops (ratio between total seismic moment and fault area) that are an order of magnitude lower than typical mainshock-aftershock sequences on strike-slip faults [Forsyth et al., 2003; Vidale and Shearer, 2006]. Following the analysis of
Roland and McGuire, 2009, we calculate the effective stress drop of each of the relocated clusters of earthquakes to determine whether the cluster is swarm-like in behavior, and compare our results to that determined for swarms observed in southern California.
For a vertical strike-slip fault, which is a close approximation for the NW-SE striking transform faults along the plate boundary, the stress drop is estimated as: ∆σ = (2/π)µ(D/w), where D = Mo/(µS) [Kanamori and Anderson, 1975]. In this relationship, µ is the shear modulus
(which cancels out in the equation), D is the average slip, and S = wL is the fault area, where w is the fault width and L is the fault length. Here, we consider the fault width to be 10 km, consistent with the RMT analysis, and the fault length is determined by the earthquake locations within the individual clusters (Table 3).
The results of the calculation are reported in Table 3. We note that in the case where the clusters overlap a transform and basin segment, only the transform portion of seismic activity is considered. The largest effective stress drops are observed in regions where the clusters of activity surround large magnitude events reported in the GCMT catalog, like the 4 January 2006
Mw 6.6 Guaymas transform event, and the 28 May 2006 Mw 5.2 and Mw 5.1 doublet and 30 July
2006 Mw 5.9 Carmen transform events; however, these events may not be swarm-like in behavior, but rather follow a foreshock-mainshock-aftershock sequence. The rest of the clusters of activity have rather low effective seismic stress drops on the order of kPa, including the
August 2006 Baja California sequence. Our calculations for the clusters of activity not 32 associated with GCMT reported events are about an order of magnitude smaller compared to the seismic swarms observed along southern California and East Pacific Rise transform faults reported in Roland and McGuire, 2009, which may be due to the clusters of activity in our study covering much greater fault lengths.
The distribution of seismicity within the GoC indicates that deformation cannot be simply described as strike-slip behavior on vertical transform faults. Much of the seismicity along the transforms is localized within the inside corner bathymetric lows at the ridge-transform intersections. From these observations and the orientation of the strike-slip focal mechanisms along the transforms, deformation appears to remain oblique to the transform faults rather than fully separated into traditional ridge-transform structures. This behavior is most apparent along the Carmen and Farallon segments, in topographic lows centered roughly along plate boundary spreading centers, but where evidence of new oceanic crust is lacking [e.g. Lizarralde et al.,
2007]. This suggests that seismic deformation is not fully localized along the plate boundary.
Although the regional moment tensor catalog is dominated by strike-slip events, analysis of an oblique normal-faulting mechanism within the Guaymas basin may reflect bookshelf faulting in this region, where the fault along the basin is being rotated by simple shear in the direction of plate motion [e.g. Wetzel et al., 1993], rather than pure ridge-parallel extension.
Overall, the spatial distribution of earthquake activity along the plate boundary shows that much of the seismic activity is accommodated via the transform segments rather than along the ridges, even within the Guaymas and Alarcon basins where our OBS stations are located and the recording capability is best. The dearth of moderate-magnitude earthquake activity on the ridge segments is consistent with observations along fast-spreading (>80 mm yr-1) mid-ocean ridge systems [e.g. Fox et al., 2001; Bohnenstiehl et al., 2002; Bohnenstiehl and Dziak, 2008; Dziak et 33 al., 2009], and thus must accommodate displacement through aseismic slip events or by abundant microearthquake activity too small to be detected by our array.
In addition, regional moment tensor analysis of an event located on the inside corner of
Farallon ridge-transform intersection (RTI) shows an anomalous rotated strike-slip mechanism
(Figure 13). Unusual mechanisms deviating away from transform parallel motion have been previously observed at the inside-corners of RTIs [Engeln et al., 1986; Wolfe et al., 1993], and have been attributed to stress variations near these intersections or anomalous thermal structure
[e.g. Fox and Gallo, 1984].
Further, 3D modeling of oceanic transform faults reveals rotated strike-slip mechanisms at the inside corners of RTIs due to weak seismic coupling along these zones [Behn et al., 2002].
Seismic coupling, or the ratio between the observed (seismic) to predicted (tectonic) moment release [Scholz, 2002], across oceanic transform faults is predicted to be <10%, and may exhibit significant temporal variability during the seismic cycle, since seismic moment release is predominantly controlled by large magnitude events [Behn et al., 2002]. During our year deployment along the GoC, seismic coupling would be largest for the transforms that experience major GCMT events; however, one year of observations does not capture any significant portion of the seismic cycle, and may bias coupling statistics between the transforms. In contrast, a study of GoC transforms over a ~20 year period by Tajima and Tralli, 1992 indicates that seismic slip only accounts for ~17-30% of the tectonic slip predicted by the NUVEL-1 global plate model [DeMets et al., 1990], and interpret this as evidence that a large portion of plate boundary motion is accommodated aseismically, which is consistent with observations and modeling results of weakly coupled transforms [e.g. Engeln et al., 1986; Behn et al., 2002;
Boettcher and Jordan, 2004]. 34
In addition, seven earthquakes suitable for regional moment tensor inversion locate on
Baja California, and exhibit right-lateral strike-slip focal mechanisms along an approximately southeast-northwest striking fault (Figure 15). The location of these events in line with the
Carmen transform in a region with deep bathymetry, and the geometry of the focal mechanisms suggest an active continuation of the fault across Baja California. Many strike-slip faults are observed to diverge from the transform faults within the GoC and cut across the Baja California peninsula in southern California at the opening of the gulf and in the southern gulf from ~25ºN to the mouth of the gulf [e.g. Suarez-Vidal et al., 1991; Humphreys and Weldon, 1991; Fletcher and Munguía, 2000]; however, few active faults are known along the margins of the gulf between ~25ºN to ~30ºN [Umhoefer et al., 2002]. Further, in September 2007, almost a full year after our deployment, 800 Mw > 3 aftershocks of an Mw 6.1 event occurred on an off-axis left- lateral strike-slip fault, parallel to the Alt basin and perpendicular to the transforms [Ortega and
Quintanar, 2010]. Earthquake activity on this previously unrecognized fault, and on the inferred right-lateral strike-slip fault on Baja California and other off-axis faults within the gulf suggest that Baja California-North America plate motion is not completely accommodated along the plate boundary.
In conclusion, we find that the relocated seismicity and their associated focal mechanisms within the GoC fall predominantly along the NW-SE striking transform faults that accommodate a large amount of Pacific-North American plate boundary motion, which is consistent with findings from previous studies [e.g. Goff et al., 1987; Castro et al., 2010]. Depths of these earthquakes are shallow (1-12 km), with little variation between the transform segments, and are consistent with observations that the maximum depth of brittle failure is thermally-controlled in the oceanic setting [e.g. Abercrombie and Ekström, 2001; Boettcher and Jordan, 2004]. 35
Furthermore, investigation into the spatial and temporal behavior of the clusters of seismicity relocated through the hypoDD double-difference analysis suggest that many of the regions exhibit swarm behavior, rather than a typical foreshock-mainshock-aftershock sequence. A calculation of the effective seismic stress drop indicates an unusually low amount of seismic release along these transforms, and perhaps shows that much of the plate boundary deformation is accommodated aseismically, which is a well-documented phenomenon observed on oceanic transforms [Bird et al., 2002; Boettcher and Jordan, 2004].
36
Figure Captions
Figure 1: Map of the Gulf of California (GoC) and Baja California, with basins and faults on the Baja California peninsula marked. For purposes of our study, the southern Gulf of California extends from the Guaymas transform fault directly north of the Guaymas basin, to the Tamayo transform fault directly south of the Alarcon basin in the mouth of the gulf. The Tamayo transform fault terminates at the East Pacific Rise, marking the end of ridge-transform structures that compose the southern gulf. Note that the long transform fault north of each of the basin has the same name (e.g. Guaymas transform fault and Guaymas basin). The plate boundary (black line) down the GoC is from Bird, 2003, and the off-axis plate boundary faults on the southern
Baja California peninsula (black lines) are from Plattner et al., 2007. Major southern Baja
California faults are marked and shown in the legend. GoC bathymetry data is downloaded from
GeoMapApp.
Figure 2: Bathymetric map of the GoC with NARS-Baja (blue squares; Trampert et al., 2003) and SCOOBA (red inverted triangles) stations shown. Note that station NE76 (green square) had observed timing errors on the order of ~10-15 s, and could not be used for relocation purposes.
During our deployment from October 2005 to October 2006, 47 earthquakes from the
International Seismological Center (ISC) (black dots), and seven Global CMT (black dots) events were recorded in their respective catalogs. Many of the Global CMT events are largely non-double couple, yet they all exhibit right-lateral strike-slip mechanisms consistent with the
NW-SE striking transtensional plate boundary within the GoC.
Figure 3: Map showing the 695 Hypoinverse located events, color-coded based on (a) depth, and (b) quality. Of these events, ~70% are shallow (<10 km; black dots), while only ~12% are considered deep (>20 km; red dots). Note that many of these ‘deep’ earthquakes lie on the 37
Pacific-North America plate boundary, and may be in error. Additionally, the best-located ‘A’ quality earthquakes lie on the plate boundary model of Bird, 2003 (black dots), yet only one ‘A’ quality earthquake lies deeper than 20 km. The earthquakes are color-coded based on
Hypoinverse data in Figure 3b, but the locations and depths shown are after the VELEST relocation is performed. Off-axis plate boundary faults on Baja California are from Plattner et al., 2007. The NARS-Baja and SCOOBA stations are shown as grey boxes and inverted triangles, respectively.
Figure 4: Velocity-depth profile of P-wave velocity model (solid line) derived from the
VELEST joint-hypocenter determination method. The S-wave model (dashed line) is found from dividing the P-wave velocity model by a vp/vs ratio of 1.73. These models are both used in the regional moment tensor inversion.
Figure 5: The seven clusters of earthquakes (N = 428 events) relocated by the hypoDD double- difference algorithm [Waldhauser and Ellsworth, 2000], and color-coded based on their geographical location and annotated in the legend. Additionally, the focal mechanisms of the 31 earthquakes analyzed by regional deviatoric moment tensor (RMT) inversion are shown. Note that the majority of the events are characterized by right-lateral strike-slip focal mechanisms, except for an oblique normal-faulting event within the Guaymas basin. The NARS-Baja and
SCOOBA stations are shown as grey squares and inverted triangles, respectively.
Figure 6: An example of the regional deviatoric moment tensor (RMT) inversion results for (a) a typical right-lateral strike-slip faulting mechanism for an event along the Farallon transform, and (b) an oblique normal faulting mechanism for an event on the Guaymas basin. We used three-component data for clear waveforms observed on the NARS-Baja and SCOOBA stations.
Black waveforms are the data, while the red waveforms are the inversion results. Also shown 38 are the individual components of the moment tensor, the seismic moment, time shift, fault plane information, compression (P) and tension (T) axes, and the fitting error and variance reduction information. The results shown are for the best fitting focal depth that minimizes the fitting error and maximizes the variance reduction. (c) The range of focal depths tried for the Farallon transform event, and (d) for the Guaymas basin event. The focal depth space appropriate for each of the mechanisms is defined as 1.05Emin, where Emin is the minimum fitting error. Note for the Farallon transform event, the full range of focal depths tried from 4-18 km is reasonable, yet the preferred focal depth of 12 km is the global minimum for the fitting error.
Figure 7: A comparison between the local magnitude (ML) determined from the P-wave picks and the moment magnitude (Mw) derived from the regional deviatoric moment tensor inversion analysis. If the magnitudes were the same, they would fall along the one-to-one line (black line); however, the ML generally underestimates the Mw. The best-fit line through the data is defined by: Mw ≈ 0.62 ML + 1.8 (dotted line). The magnitudes reported in our study are converted from
ML to Mw, unless otherwise determined by regional moment tensor analysis or the GCMT catalog.
Figure 8: (a) Map of events relocated along the northern Guaymas transform, and associated focal mechanisms of events determined by either regional moment tensor (RMT) inversion or by the Global CMT catalog, color-coded by time and defined in the legend. Focal depths for the
RMT determined mechanisms are all 3 km. (b) Station map of the 10 stations (blue inverted triangles) with P-wave pick data, and (c) map view of all of the relocated earthquakes (red dots) compared to the original location (blue dots). (d) Along-axis cross-section of the relocated hypocenters (red dots) compared to the original locations (blue dots) along A-B shown in Figure
8c, and (e) across-axis cross-section of the relocated hypocenters (red dots) compared with the 39 original locations (blue dots) along C-D shown in Figure 8c. Note that the original depths are all set to 10 km, but after relocation, many of the hypocenters have shallow focal depths. (f) The moment magnitude (Mw), cumulative number of earthquakes, and the cumulative seismic moment release over the ± 3 days surrounding the 4 January 2006 Mw 6.6 event (red line and/or dot), the largest event recorded during our deployment. Note that much of the activity is foreshocks occurring before the event, which is a behavior common to the oceanic transform setting [McGuire et al., 2005].
Figure 9: (a) Map of relocated events along the southern Guaymas transform and Guaymas basin, with the focal mechanisms derived from the RMT analysis shown, color-coded by time and defined in the legend. Focal depths of these earthquakes range from 1-7 km. In addition, the foreshocks and aftershocks surrounding the 11 May 2006 are also shown as varying shades of grey dots described in the text. (b) The moment magnitude (Mw), cumulative number of earthquakes, and the cumulative seismic moment release over the entire year of recording. Note that a spike in the number of earthquakes around day 200 corresponds with activity along the
Guaymas basin on 11 May 2006. However, this cluster of activity does not correspond with a spike in the cumulative seismic moment, and therefore does not contribute much to plate boundary motion at this site.
Figure 10: (a) Station map of the 6 stations (blue inverted triangles) with P-wave pick data used in the relocation procedure, and (b) map view of all of the relocated earthquakes (red dots) compared to the original location (blue dots) for the Guaymas transform segment only. (c)
Along-axis cross-section of the relocated hypocenters (red dots) compared to the original locations (blue dots) along A-B shown in Figure 10b, and (d) across-axis cross-section of the relocated hypocenters (red dots) compared with the original locations (blue dots) along C-D 40 shown in Figure 10b. Note that the original depths are all set to 10 km, but after relocation, many of the hypocenters have shallow focal depths. (e) Map view of the relocated earthquakes
(red dots) compared to the original locations (blue dots) for the Guaymas basin. (f) Across-axis cross-section and (g) along-axis cross-section of the relocated hypocenters (red dots) compared to the original locations (blue dots) for the Guaymas basin.
Figure 11: (a) Map of relocated events along the northern Carmen transform, with the focal mechanisms derived from the RMT analysis and in the GCMT catalog shown, color-coded by time and defined in the legend. Focal depths of the earthquakes found through RMT analysis are
5 km. The cluster is predominantly comprised of events surrounding the 28 May 2006 Mw 5.2 and Mw 5.1 doublet and the 30 July 2006 Mw 5.9 earthquake. (b) Station map of the 9 stations
(blue inverted triangles) that had usable P-wave pick data for cluster of earthquakes shown as red dots. (c) Map view of the relocated earthquakes (red dots) compared to the original locations
(blue dots) for the Guaymas basin. (d) Along-axis cross-section from A-B and (e) across-axis cross-section from C-D of the relocated hypocenters (red dots) compared to the original locations
(blue dots) for the Carmen transform. Note that many of the hypocentral depths are shallow
(<10 km). (f) The cumulative seismic moment release over the entire year of recording. Dotted lines mark the events that are responsible for a large amount of seismic moment release.
Figure 12: (a) The moment magnitude (Mw) of events 2.5 hours before and 8 hours after the 23
May 2006 Mw 4.2 event. This sequence is most likely a swarm since the events are all of similar magnitude, and the largest event comes later in the sequence. (b) Map view of the earthquakes associated with the 23 May 2006 Mw 4.1 event (white dot). Note that the events spread bilaterally to the north and to the northeast, respectively. (c) The moment magnitude (Mw) of events during the 2.4 hours before and 24 hours after the 28 May 2006 Mw 5.2 and Mw 5.1 41 doublet of earthquakes. This sequence behaves more like a typical mainshock-aftershock sequence. (d) Map view of the earthquakes associated with the 28 May 2006 doublet, with the
Mw 5.2 event shown as a white dot. Note that the events surrounding the doublet align with the
NW-SE trending strike of the Carmen transform. (e) The moment magnitude (Mw) of events during the 2.4 hours before and 24 hours after the 30 July 2006 Mw 5.9 earthquake. This sequence behaves more like a typical mainshock-aftershock sequence, and is completely devoid of any foreshock activity. (f) Map view of the earthquakes associated with the 30 July 2006 Mw
5.9 earthquake (white dot). Note that the aftershocks spread to the northeast, and then trend along the lower Guaymas basin.
Figure 13: (a) Map of relocated events along the Farallon transform, and three focal mechanisms derived from RMT analysis, color-coded by time as described in the legend. The focal depths of these earthquakes range from 2-5 km, respectively. (b) Station map (blue inverted triangles) of the 6 stations with P-wave picks for these earthquakes (red dots). (c) Map view of the relocated events (red dots), (d) the along-strike cross-section A-B of the relocated events compared to the original locations (blue dots), and (e) the across-strike cross-section C-D of the relocated events compared to the original locations. Note that many of the hypocenters relocate to shallow focal depths (<10 km), consistent with the focal depths derived from RMT inversion. (f) The cumulative seismic moment released over the duration of the seismograph deployment. The largest releases in seismic moment are due to the 3 November 2005 event and 27 November
2005 events, shown as dotted lines and described in the text.
Figure 14: (a) Map of relocated events along the Alarcon transform, Alarcon basin, and Tamayo transform, including an additional set of events that occurs on the Alarcon Seamount (ASM).
The focal mechanisms of four events derived from RMT analysis are color-coded by time and 42 described in the legend. The focal depths of these earthquakes range from 2-8 km, respectively.
(b) Station map (blue inverted triangles) of the five stations with P-wave picks for these earthquakes (red dots). (c) Map view of the relocated events (red dots) compared to their original locations (blue dots). The blue box shown defines the region for the (d) along-strike cross- section A-B along the Alarcon transform, and (e) the across-strike cross-section C-D of the relocated events compared to the original locations. Note that many of the hypocenters relocate to depths between 0 and 20 km, the deepest of our study. (f) The cumulative number of earthquakes and the cumulative seismic moment released over the duration of our deployment.
The largest release in seismic moment is due to the 5 March 2006 RMT event, shown as dotted lines and described in the text.
Figure 15: (a) Map showing thirteen double-difference relocated events on Baja California during a ~22 day swarm in August 2006. The focal mechanisms of seven of these events were obtained through RMT analysis (color-coded by time as shown in the legend), and all have right- lateral strike-slip mechanisms along a fault trending east-west. The focal depths of these events range from 4-10 km. The earthquakes are large enough to have felt reports, yet the closest town of San Ignacio (white star) is very small, and any felt reports are unknown [R. Castro, pers. comm.]. (b) The cumulative number of earthquakes over the ~22 day swarm sequence, with the
RMTs shown as dotted lines. (c) The cumulative seismic moment released over the ~22 day swarm sequence, with RMTs shown as dotted lines. Note that the largest spikes in seismic moment surround these large magnitude events that were suitable for RMT analysis.
Figure 16: A map view comparison between the locations of our 695 relocated events (black dots) and the events relocated by Castro et al., 2010. The Castro et al., 2010 study uses data from the NARS-Baja stations (grey squares) deployed from 2002-2006, while we use the 43 combination of the NARS-Baja and SCOOBA (grey inverted triangles) for our study. Note that in regions where we observe a lack of seismicity, predominantly along the Alt transform, the
Castro et al., 2010 study fills them in, indicating that tectonic strain was released in this region before our deployment.
44 45
46 47
48 49
50 51
52
53 54 55 56 57
58
59
60
61 62
63
64
Chapter 2:
Pulse of the Seafloor: Tidal triggering of microearthquakes at 9º50’N East Pacific Rise
This chapter is published as an article in Geophysical Research Letters:
Stroup, D. F., D. R. Bohnenstiehl, M. Tolstoy, F. Waldhauser, and R. T. Weekly (2007). Pulse of the seafloor: Tidal triggering of microearthquakes at 9º50’N East Pacific Rise, Geophys. Res. Lett., 34(L15301), doi:10.1029/2007GL030088. 65
66
67
68
69
70 71
72
73
74
Chapter 3:
Systematic along-axis tidal triggering of microearthquakes observed at 9º50’N East Pacific Rise
This chapter is published as an article in Geophysical Research Letters:
Stroup, D. F., M. Tolstoy, T. J. Crone, D. R. Bohnenstiehl, and F. Waldhauser (2009). Systematic along-axis tidal triggering of microearthquakes observed at 9º50’N East Pacific Rise, Geophys. Res. Lett., 36(L18302), doi:10.1029/2009GL039493. 75
76
77
78
79
80
81
82
83
84
85
86
References
Abercrombie, R. E., and G. Ekström (2001). Earthquake slip on oceanic transform faults, Nature, 410 74-77, doi:10.1038/3506564.
Ambeh, W. B., and J. D. Fairhead (1991). Regular, deep seismicity beneath Mt. Cameroon volcano – Lack of evidence for tidal triggering, Geophys. J. Int., 106 287-291.
Anderko, A., and K.S. Pitzer (1993), Equation of state representation of phase equilibria and volumetric properties of the system NaCl-H20 above 573 K, Geochim. Cosmochim. Acta., 57, 1657-1680.
Atwater, T. (1989). Plate tectonic history of the northeast Pacific and western North America, in The Eastern Pacific Ocean and Hawaii Edward L. Winterer, Donald M. Hussong, and Robert W. Decker (Editors), Geological Society of America, Geology of North America, Boulder, Colorado, 499-521.
Atwater, T., and J. Stock (1998). Pacific-North America Plate Tectonics of the Neogene Southwestern United States: An Update, International Geology Review 40 375-402.
Axen, G. J., M. Grove, D. Stockli, O. M. Lovera, D. A. Rothstien, J. M. Fletcher, K. Farley, and P. L. Abbott (2000). Thermal evolution of Monte Blanco dome: Low-angle normal faulting during Gulf of California rifting and late Eocene denudation of the eastern Peninuslar Ranges, Tectonics 19(2) 197-212.
Beeler, N. M., and D. A. Lockner (2003). Why earthquakes correlate weakly with the solid Earth tides: Effects of periodic stress on the rate and probability of earthquake occurrence, J. Geophys. Res., 108(B8), 2391, doi:10.1029/2001GC000156.
Behn, M. D., J. Lin, and M. T. Zuber (2002). Evidence for weak oceanic transform faults, Geophys. Res. Lett., 29(4) doi:10.1029/2002GL015612.
Bird, P., Y. Y. Kagan, and D. D. Jackson (2002). Plate tectonics and Earthquake Potential of Spreading Ridges and Oceanic Transform Faults, in Plate Boundary Zones, S. Stein and J. T. Freymueller (Editors), AGU Monograph, Geodyn. Ser., 30 203-218.
Bird, P. (2003). A updated digital model of plate boundaries, Geochem. Geophys. Geosys. 4(3) doi:10.1029/2001GC000252.
Boettcher, M. S., and T. H. Jordan (2004). Earthquake scaling relations for mid-ocean ridge transform faults, J. Geophys. Res. 109(B12), B12302, doi:10.1029/2004JB003110.
Boettcher, M. S., G. Hirth, and B. Evans (2007). Olivine friction at the base of oceanic seismogenic zones, J. Geophys. Res., 112 B01205, doi:10.1029/2006JB004301.
87
Bohnenstiehl, D.R., and S.M. Carbotte (2001), Faulting patterns near 19º30’S on the East Pacific Rise: Fault formation and growth at a superfast spreading center, Geochem. Geophys. Geosyst., 2(9), doi:10.1029/2001GC000156.
Bohnenstiehl, D. R., M. Tolstoy, R. P. Dziak, C. G. Fox, and D. K. Smith (2002). Aftershock sequences in the mid-ocean ridge environment: An analysis using hydroacoustic data, Tectonophysics, 354 49-70.
Bohnenstiehl, D. R., and R. P. Dziak (2008). Mid-ocean ridge seismicity, in Encyclopedia of Ocean Sciences (2nd edition), John H. Steele, Karl K. Turekian, and Steve A. Thorpe (Editors), Elsevier, 837-851.
Braunmiller, J., and J. Nábělek (2008). Segmentation of the Blanco Transform Fault Zone from earthquake analysis: Complex tectonics of an oceanic transform fault, J. Geophys. Res., 113 B07108 doi:10.1029/2007JB005213.
Brodsky, E.E., V. Karakostas, and H. Kanamori (2000), A New Observation of Dynamically Triggrered Regional Seismicity: Earthquakes in Greece Following the August, 1999 Izmit, Turkey Earthquake, Geophys. Res. Lett., 27, 2741-2744.
Brune, J. N. (1968). Seismic moment, seismicity, and rate of slip along major fault zones, J. Geophys. Res., 73 777-784 doi:10.1029/2007JB005213.
Castillo, P. R., J. W. Hawkins, P. F. Lonsdale, D. R. Hilton, A. M. Shaw, and M. D. Glascock (2002). Petrology of Alarcon Rise lavas, Gulf of California: Nascent intracontinental ocean crust, J. Geophys. Res. 107(10), doi:10.1029/2001JB000666.
Castro, R. R., A. Perez-Vertti, I. Mendez, A. Mendoza, and L. Inzunza (2010). Location of Moderate-Sized Earthquakes Recorded by the NARS-Baja Array in the Gulf of California Region Between 2002 and 2006, Pure Appl. Geophys. doi: 10.1007/s00024-010-0177-y.
Clayton, R. W., J. Trampert, C. J. Rebollar, J. Ritsema, P. Persaud, H. Paulssen, X. Pérez- Campos, A. van Wettum, A. Pérez-Vertti, and F. Di Luccio (2004). The NARS-Baja array in the Gulf of California Rift Zone, Margins Newslett. 13 1-4.
Cochran, E. S., J. E. Vidale, and S. Tanaka (2004). Earth tides can trigger shallow thrust fault earthquakes, Science, 306, 1164-1166.
Crone, T.J., and W.S.D. Wilcock (2005), Modeling the effects of tidal loading on mid-ocean ridge hydrothermal systems, Geochem. Geophys. Geosyst., 6, Q07001, doi:10.1029/2005GC000905.
Crotwell, H. P., T. J. Owens, and J. Ritsema (1999). The TauP Toolkit: Flexible seismic travel- time and ray-path utilities, Seismol. Res. Lett. 70 154-160.
88
Davis, S. D., and C. Frohlich (1991). Single-link cluster-analysis, synthetic earthquake catalogs, and aftershock identification, Geophys. J. Int., 104, 289-306.
Davis, E.E., K. Wang, K. Becker, and R.E. Thomson (2000). Formation-scale hydraulic and mechanical properties of oceanic crust inferred from pore pressure response to periodic seafloor loading, J. Geophys. Res., 105(B6), 13,423-13,435.
DeMets, C., R. G. Gordon, D. F. Argus, and S. A. Stein (1990). Current plate motions, Geophys. J. Int., 101 425-478.
DeMets, C. (1995). A reappraisal of seafloor spreading lineations in the Gulf of California: Implications for transfer of Baja California to the Pacific plate and estimates of Pacific-North America motion, Geophys. Res. Lett. 22 3545-3548.
DeMets, C., and T. H. Dixon (1999). New kinematic models for Pacific-North America motion from 3 Ma to present, I: Evidence for steady motion and biases in the NUVEL-1A model, Geophys. Res. Lett. 26 1921-1924.
Dixon, T. H., F. Farina, C. DeMets, F. Suarez-Vidal, J. Fletcher, B. Margquez-Azua, M. Miller, O. Sanchez, P. Umhoefer (2000). New kinematic models for Pacific-North America motion from 3 Ma to present, II: Evidence for a Baja California shear zone, Geophys. Res. Lett. 27 3961-3964.
Dreger, D. S., and D. V. Helmberger (1993). Determination of source parameters at regional distances with single station or sparse network data, J. Geophys. Res. 98 8107-8125.
Du, W., W.-Y. Kim, and L. R. Sykes (2003). Earthquakes Source Parameters and State of Stress for Northeastern United States and Southeastern Canada from Analysis of Regional Seismograms, Bull. Seismo. Soc. Am. 93(4) 1633-1648.
Dziak, R. P., D. R. Bohnenstiehl, H. Matsumoto, M. J. Fowler, J. H. Haxel, M. Tolstoy, and F. Waldhauser (2009). January 2006 seafloor-spreading event at 9º50’N, East Pacific Rise: Ridge dike intrusion and transform fault interactions from regional hydroacoustic data, Geochem. Geophys. Geosys., 10(6), doi:10.1029/2009GC002388.
Einsele, G. (1982). Mechanism of sill intrusion into soft sediment and expulsion of pore water, Initial Rep. Deep Sea Drill. Proj. 64 1169-1176.
Emter, D. (1997). Tidal triggering of earthquakes and volcanic events, in Tidal Phenomena, edited by S. Bhattacharji et al., 293-309, Springer, New York.
Engeln, J. F., D. A. Wiens, and S. Stein (1986). Mechanisms and Depths of Atlantic Transform Earthquakes, J. Geophys. Res., 91(B1) 548-577.
Fabriol, H., L. A. Delgado-Argote, J. J. Dañobeitia, D. Córdoba, A. González, J. García- Abdeslem, R. Bartolomé, B. Martin-Atíenza, and V. Frías-Camacho (1999). Backscattering and 89 geophysical features of volcanic rifts offshore Santa Rosalia, Baja California Sur, Gulf of California, Mexico, J. Volcanol. Geotherm. Res. 93 75-92.
Fang, W.W., M.G. Langseth, and P.J. Schultheiss (1993), Analysis and Application of in Situ Pore Pressure Measurements in Marine Sediments, J. Geophys. Res., 98(B5), 7921-7938.
Fisher, A.T. (1998), Permeability within basaltic oceanic crust, Rev. Geophys., 36(2), 143-182.
Fisher, A. T., E. Giambalvo, J. Sclater, M. Kastner, B. Ransom, Y. Weinstein, and P. Lonsdale (2001). Heat flow, sediment, and pore fluid chemistry, and hydothermal circulation on the east flank of Alarcon Ridge, Gulf of California, Earth Planet. Sci. Lett. 188 521-534.
Fisher, A.T., E.E. Davis, and K. Becker (2008), Borehole-to-borehole hydrologic response across 2.4 km in the upper oceanic crust: Implications for crustal-scale properties, J. Geophys. Res., 113, B07106, doi:10.1029/2007JB005447.
Fletcher, J. M. and L. Munguía (2000). Active continental rifting in southern Baja California, Mexico: Implications for plate motion partitioning and the transition to seafloor spreading in the Gulf of California, Tectonics 19(6) 1107-1123.
Fornari, D.J., et al. (2004), Submarine lava flow emplacement at the East Pacific Rise 9º50’N: Implications for uppermost ocean crust stratigraphy and hydrothermal fluid circulation, in Mid- Ocean Ridges: Hydrothermal Interactions Between the Lithosphere and Oceans, Geophs. Mongr. Ser., vol. 148, edited by C. R. German, J. Lin, and L. M. Parson, pp. 187-217, AGU, Washington, D. C.
Forsyth, D. W., Y. Yang, M. D. Mangriotis, and Y. Shen (2003). Couple seismic slip on adjacent ocean transform faults, Geophys. Res. Lett. 30(12), 1618, doi:10.1029/2002GL016454.
Fox, P. J., and D. G. Gallo (1984). A tectonic model for ridge-transform-ridge plate boundaries: Implications for the structure of oceanic lithosphere, Tectonophysics, 104 205-242.
Fox, C. G., H. Matsumoto, and T.-K. A. Lau (2001). Monitoring Pacific Ocean seismicity from an autonomous hydrophone array, J. Geophys. Res., 106(B3) 4183-4206, doi:10.1029/2000JB900404.
Glasby, G.P., and J. Kasahara (2001), Influence of tidal effects on the periodicity of earthquake activity in diverse geological settings with particular emphasis on submarine hydrothermal systems. Earth Sci. Rev., 52, 261-297.
Goff, J. A., E. A. Bergman, and S. C. Solomon (1987). Earthquake Source Mechanisms and Transform Fault Tectonics in the Gulf of California, J. Geophys. Res. 92(B10), 10485-10510.
Good, P.I. (2005), Permutation, Parametric, and Bootstrap Tests of Hypotheses, 3rd ed., Springer, New York.
90
Gonzales-Garcia, J. J., L. Prawirodirdjo, Y. Bock, and D. Agnew (2003). Guadalupe Island, Mexico as a new constraint for Pacific plate motion, Geophys. Res. Lett. 20(16) 1872, doi:10.1029/2003GL017732.
Grand, S. P., and D. V. Helmberger (1984). Upper mantle shear structure of North America, Geophys. J. R. astr. Soc. 76 399-438.
Haymon, R. M., D. J. Fornari, M. H. Edwards, S. Carbotte, D. Wright, and K. C. MacDonald (1991). Hydrothermal vent distribution along the East Pacific Rise crest (9º09’ – 54’N) and its relationship to magmatic and tectonic processes on fast-spreading mid-ocean ridges, Earth Planet. Sci. Lett., 104 513-534, doi:10.1016/0012-821X(91)90226-8.
Holzbecher, E.O. (1998), Modeling Density-Driven Flow in Porous Media, Springer, New York.
Humphreys, E. D., and R. J. Weldon, II (1991). Kinematic constraints on the rifting of Baja California, in The Gulf and Peninsular Province of the Californias J. P. Dauphin and B. R. T. Simoneit (Editors), American Association of Petroleum Geologists Memoir 47, 217-229.
Husen, S., E. Kissling, E. Flueh, and G. Asch (1999). Accurate Hypocentre Determination in the Seismogenic Zone of the Subducting Nazca Plate in Northern Chile Using a Combined On- /Offshore Network, Geophys. J. Int. 138 687-701.
Johnston, M. J., and F. J. Mauk (1972). Earth tides and the triggering of eruptions from Mt. Stromboli, Italy, Nature, 239 266-267.
Jones, T. A. (2006). MATLAB function to analyze directional (azimuthal) data – I: Single- sample inference, Comput. Geosci., 32 166-175.
Jupp, T.E., and A. Schultz (2004), A poroelastic model for the tidal modulation of seafloor hydrothermal systems, J. Geophys. Res., 109, B03105, doi:10.1029/2003JB002583.
Kanamori, H. (1977). Energy release in great earthquakes, J. Geophys. Res. 82(20), 2981-2987.
Kanamori, H., and D. L. Anderson (1975). Theoretical basis of some empirical relations in seismology, Bull. Seism. Soc. Am. 65, 1073-1095.
Karig, D. E., and W. Jensky (1972). The protogulf of California, Earth Planet. Sci. Lett. 17 169- 174.
Kennett, B. L. N., and E. R. Engdahl (1991). Traveltimes for global earthquake location and phase identification, Geophys. J. Int. 105(2), 429-465.
Kent, G.M., A.J. Harding, and J.A. Orcutt (1993), Distribution of magma beneath the East Pacific Rise between the Clipperton Transform and the 9º17’N deval from forward modeling of common depth point data, J. Geophys. Res., 98, 13,971-13,995.
91
Kim, W.-Y. (2003). 18 June 2002 Caborn, Indiana Earthquake: Reactivation of Ancient Rift in the Wabash Valley Seismic Zone?, Bull. Seismol. Soc. Am. 93, 2201-2211.
Kim, W.-Y., H. Choi, and M. Noh (2010). The 20 January 2007 Odaesan, Korea, Earthquake Sequence: Reactivation of a Buried Strike-Slip Fault?, Bull. Seismol. Soc. Am. 100(3), 1120- 1137.
Kissling, E. (1988). Geotomography with Local Earthquake Data, Rev. Geophys. 26(4) 659-698.
Kissling, E., W. L. Ellsworth, D. Eberhart-Phillips, and U. Kradolfer (1994). Initial reference models in local earthquake tomography, J. Geophys. Res. 99(B10) 19635-19646.
Klein, F. W. (1976). Earthquake swarms and semidiurnal solid earth tide, Geophys. J. R. Astron. Soc., 45 245-295.
Klein, F. W. (2002). User’s Guide to HYPOINVERSE-2000, a Fortran Program to Solve for Earthquake Locations and Magnitudes, U. S. Geological Survey Open-File Rept. 02-171, 113 pp.
Lizarralde, D., G. J. Axen, H. E. Brown, J. M. Fletcher, A. González-Fernández, A. J. Harding, W. S. Holbrook, G. M. Kent, P. Paramo, F. Sutherland, and P. J. Umhoefer (2007). Variation in styles of rifting in the Gulf of California, Nature 448 466-469.
Lizarralde, D., S. A. Soule, J. S. Seewald, and G. Proskurowski (2011). Carbon release by off- axis magmatism in young sedimented spreading centre, Nature Geosci. 4 doi:10.1038/NGEO1006.
Lonsdale, P. (1989). Geology and tectonic history of the Gulf of California, in The Eastern Pacific Ocean and Hawaii Edward L. Winterer, Donald M. Hussong, and Robert W. Decker (Editors), Geological Society of America, Geology of North America, Boulder, Colorado, 499- 521.
Lonsdale, P. (1991). Structural patterns of the Pacific floor offshore of peninsular California, in The Gulf and Peninsular Province of the Californias J. P. Dauphin and B. R. T. Simoneit (Editors), American Association of Petroleum Geologists Memoir 47, 87-125.
Lowell, R. P., and L. N. Germanovich (1994). On the temporal evolution of high-temperature hydrothermal systems at ocean ridge crest, J. Geophys. Res., 99 565-575 doi:10.1029/93JB02568.
Marsaglia, K. M. (2004). Sandstone detrital modes support Magdalena Fan displacement from the mouth of the Gulf of California, Geology 32 45-48 doi:10.1130/G20099.1.
Matsumoto, K., T. Takenezawa, and M. Ooe (2000). Ocean tide models developed by assimilating TOPEX/POSEIDON altimeter data into hydrodynamical model: A global model and a regional model around Japan, J. Oceanogr., 56 567-581.
92
Matsumoto, K., T. Sato, T. Takenezawa, and M. Ooe (2001), GOTIC2: A program for computation of oceanic tidal loading effect, J. Geod. Soc. Jpn., 47, 243-248.
Mauk, F. J., and M. J. Johnston (1973). Triggering of volcanic eruptions by Earth tides, J. Geophys. Res., 78 3356-3362.
McGuire, J. J., M. S. Boettcher, and T. H. Jordan (2005). Foreshock sequence and short-term earthquake predictability on East Pacific Rise transform faults, Nature 434 457-461.
McGuire, J. J. (2008). Seismic cycles and earthquake predictability on East Pacific Rise Transform Faults, Bull. Seism. Soc. Am. 98(3) 1067-1084.
McNutt, S. R., and R. J. Beavan (1981). Volcanic earthquakes at Pavlof Volcano correlated with the solid earth tide, Nature, 291 615-618.
Menke, W. H. (1984) Geophysical Data Analysis: Discrete Inverse Theory, Academic Press, Orlando, Florida.
Michaud, F., J.Y. Royer, J. Bourgois, J. Dyment, T. Calmus, W. Bandy, M. Sosson, C. Mortera- Gutiérrez, B. Sichler, M. Rebolledo-Viera, and B. Pontoise (2006). Oceanic-ridge subduction vs. slab break off: Plate tectonic evolution along the Baja California Sur continental margin since 15 Ma, Geology 34(1) 13-16.
Pitzer, K.S., J.C. Peiper, and R.H. Busey (1984), Thermodynamic properties of aqueous sodium chloride solutions, J. Phys. Chem. Ref. Data, 13, 1-106.
Plattner, C. R. Malservisi, T. H. Dixon, P. LaFemina, G. F. Sella, J. Fletcher, and F. Suarez- Vidal (2007). New constraints on relative motion between the Pacific Plate and Baja California microplate (Mexico) from GPS measurements, Geophys. J. Int. 170 1373-1380.
Pros, Z., J. Vaněk, and K. Klíma (1962). The velocity of elastic waves in diabase and greywacke under pressure up to 4 kilobars, Stud. Geophys. Geod., 6 347-367 doi:10.1098/BF02585234.
Ortega, R., and L. Quintanar (2010). Seismic evidence of a ridge-parallel strike-slip fault off the transform system in the Gulf of California, Geophys. Res. Lett., 37(L06301), doi:10.1029/2009GL042208.
Oskin, M., J. Stock, and A. Martín-Barajas (2001). Rapid localization of Pacific-North America plate motion in the Gulf of California, Geology 29(5) 459-462.
Páramo, P., W.S. Holbrook, H.E. Brown, D. Lizarralde, J. Fletcher, P. Umhoefer, G. Kent, A. Harding, A. Gonzalez, and G. Axen (2008). Seismic structure of the southern Gulf of California from Los Cabos block to the East Pacific Rise, J. Geophys. Res. 113(B3) doi:10.1029/2007jb005113.
93
Persaud, P., J. M. Stock, M. S. Steckler, A. Martín-Barajas, J. B. Diebold, A. González- Fernández, and G. S. Mountain (2003). Active deformation and shallow structure of the Wagner, Consag, and Delfín Basins, northern Gulf of California, Mexico, J. Geophys. Res. 108(B7) doi:10.1029/2002JB001937.
Richter, C. F. (1935). An instrumental earthquake magnitude scale, Bull. Seism. Soc. Am 25 1- 32.
Roderíguez-Lozoya, H. E., L. Quintanar Robles, C. I. Huerta López, E. Bojórques Mora, and I. León Monzón (2010). Source parameters of the July 30, 2006 (Mw 5.5) Gulf of California earthquake and a comparison with other moderate earthquakes in the region, Geofís. Int. 49(3) 119-129.
Roland, E., and J. J. McGuire (2009). Earthquake swarms on transform faults, Geophys. J. Int. 178, 1677-1690.
Rydelek, P. A., P. M. Davis, and R. Y. Koyanagi (1988). Tidal triggering of earthquake swarms at Kilauea Volcano, Hawaii, J. Geophys. Res., 93 4401-4411.
Rydelek, P. A., I. S. Sacks, and R. Scarps (1992). On tidal triggering of earthquake at Campi Flegrei, Italy, Geophys. J. Int., 109 125-137.
Saikia, C. K. (1994). Modified frequency-wavenumber algorithm for regional seismograms using Filon’s quadrature: modeling of Lg waves in eastern North America, Geophys. J. Int. 118 142- 158.
Scholz, C. H. (2002). The Mechanics of Earthquakes and Faulting, 2nd ed., Cambridge Univ. Press, New York.
Schultz, A., and H. Elderfield (1997). Controls on the physics and chemistry of seafloor hydrothermal circulation, Philos. Trans. R. Soc., Ser. A, 355 387-425 doi:10.1098/rsta.1997.0014.
Schuster, A. (1897). On lunar and solar periodicities of earthquakes, Proc. R. Soc. London, 61, 455-465.
Sclater, J. G., C. Jaupart, and D. Galson (1980). The heat flow through oceanic and continental crust and the heat loss of the Earth, Rev. Geophys., 18 269-311.
Sohn, R. A., J. A. Hildebrand, and S. C. Webb (1999). A microearthquake survey of high- temperature vent field on the volcanically active East Pacific Rise (9º50’N), J. Geophys. Res., 104, 25367-25377.
Sohn, R. A., A. H. Barclay, and S. C. Webb (2004). Microearthquake patterns following the 1998 eruption of Axial Volcano, Juan de Fuca Ridge: Mechanical relaxation and thermal strain, J. Geophys. Res., 109 B01101 doi:10.1029/2003JB002499. 94
Sohn, R.A., S.C. Webb, and J.A. Hildebrand (2004), Fine-scale seismic structure of the shallow volcanic crust on the East Pacific Rise at 9º50’N, J. Geophys. Res., 109, B12104, doi:10.1029/2004JB003152.
Soule, S.A., et al. (2005), Channelized lava flows at the East Pacific Rise crest, 9–10°N; the importance of off-axis lava transport in developing the architecture of young oceanic crust, Geochem. Geophys. Geosyst., 6, Q08005, doi:10.1029/2005GC000912.
Speiss, F. N., et al. (1980). East Pacific Rise: Hot springs and geophysical experiments, Science, 207 1421-1433.
Spencer, J. E., and W. Normark (1989). Neogene plate-tectonic evolution of the Baja California Sur continental margin and the southern Gulf of California, Mexico, in The Eastern Pacific Ocean and Hawaii Edward L. Winterer, Donald M. Hussong, and Robert W. Decker (Editors), Geological Society of America, Geology of North America, Boulder, Colorado, 499-521.
Stein, C. A., and S. Stein (1994). Constraints on hydrothermal heat flux through the oceanic lithosphere from global heat flow, J. Geophys. Res., 99 3081-3095.
Stock, J. M., and K. V. Hodges (1989). Pre-Pliocene Extension Around the Gulf of California and the Transfer of Baja California to the Pacific Plate, Tectonics 8(1) 99-115.
Stroup, D.F., D.R. Bohnenstiehl, M. Tolstoy, F. Waldhauser, and R.T. Weekly (2007), Pulse of the seafloor: Tidal triggering of microearthquakes at 9º50’N East Pacific Rise, Geophys. Res. Lett., 34, L15301, doi:10.1029/2007GL030088.
Sturtevant, B., H. Kanamori, and E.E. Brodsky (1996), Seismic triggering by rectified diffusion in geothermal systems, J. Geophys. Res., 101(B11), 25,269-25,282.
Suarez-Vidal, F., R. Armijo, G. Morgan, P. Bodin, and R. G. Gastil (1991). Framework of recent and active faulting in northern Baja California, in The Gulf and Peninsular Province of the Californias J. P. Dauphin and B. R. T. Simoneit (Editors), American Association of Petroleum Geologists Memoir 47, 285-300.
Tajima, F., and D. M. Tralli (1992). Variation of Seismic Slip in the Gulf of California and Possible Effect on Geodetic Measurements of Pacific-North American Plate Motion, J. Geophys. Res., 97(B4) 4903-4913.
Takanezawa, T., K. Matsumoto, M. Ooe, and I. Naito (2001). Effects of the period ocean tides on Earth rotation, gravity and crustal deformation predicted by global barotropic model – Period from Mtm to Sa, J. Geod. Soc. Jpn., 47 545-550.
Tanaka, S. (2006a). Tidal triggering of earthquakes in the subducting Phillippine Sea plate beneath the locked zone of the plate interface in the Tokai region, Japan, Tectonophysics, 417 69-80. 95
Tanaka, S. (2006b). Tidal triggering of earthquakes precursory to the 2004 Mw = 9.0 off Sumatra earthquakes, paper presented at the 4th International Workshop on Statistical Seismology, Grad. Univ. for Adv. Stud., Kanagawa, Japan, 9-13 Jan.
Tanaka, S., M. Ohtake, and H. Soto (2002). Evidence for tidal triggering of earthquakes as revealed from statistical analysis of global data, J. Geophys. Res., 107(B10), 2211 doi:10.1029/2001JB001577.
Tolstoy, M., S. Constable, J.A. Orcutt, H. Staudigel, and F.K. Wyatt (1998), Short and long baseline tiltmeter measurements on Axial Seamount, Juan De Fuca Ridge, Phys. Earth Planet. Inter., 108, 129-141.
Tolstoy, M., F.L. Vernon, J.A. Orcutt, and F.K. Wyatt (2002), Breathing of the seafloor: Tidal correlations of seismicity at Axial volcano, Geology, 30, 503-506.
Tolstoy, M., et al. (2006). A seafloor spreading event captured by seismometers, Science, 314 1920-1922, doi:10.1038/science.1133950.
Tolstoy, M., F. Waldhauser, D.R. Bohnenstiehl, R.T. Weekly, and W.-Y. Kim (2008), Seismic identification of along-axis hydrothermal flow on the East Pacific Rise, Nature 451, 181-184, doi:10.1038/nature06424.
Trampert, J., H. Paulssen, A. van Wettum, J. Ritsema, R. Clayton, R. Castro, C. Rebollar, and A. Perez-Vertti (2003). New Array Monitors Seismic Activity near the Gulf of California in Mexico, EOS Trans. Am. Geophys. Union 84(4) 29-32.
Umhoefer, P. J., R. J. Dorsey, and P. Renne (1994). Tectonics of the Pliocene Loreto basin, Baja California Sur, Mexico, and evolution of the Gulf of California, Geology 22(7) 649-652.
Umhoefer, P. J., L. Mayer, and R. J. Dorsey (2002). Evolution of the margin of the Gulf of California near Loreto, Baja California Peninsula, Mexico, Geol. Soc. Am. Bull., 114(7) 849-868.
Vidale, J. E., and P. M. Shearer (2006). A survey of 71 earthquake bursts across southern California: exploring the role of pore fluid pressure fluctuations and asesismic slip as drivers, J. Geophys. Res. 111(B05312), doi:10.1029/2005JB004034.
Waldhauser, F., and W. L. Ellsworth (2000). A double-difference earthquake location algorithm: Method and application to the Northern Hayward Fault, California, Bull. Seism. Soc. Am. 90 1353-1368.
Wang, K., and E.E. Davis (1996), Theory for the propagation of tidally induced pore pressure variations in layered subseafloor formations, J. Geophys. Res., 101(B5), 11,483-11,495.
Wetzel, L. R., D. A. Wiens, and M. C. Kleinrock (1993). Evidence from earthquakes for bookshelf faulting at large non-transform ridge offsets, Nature, 362 235-237.
96
Wilcock, W. (2001), Tidal triggering of microearthquakes on the Juan de Fuca Ridge. Geophys. Res. Lett., 28, 3999-4002.
Wilcock, W.S.D., S.D. Archer, and G.M. Purdy (2002), Microearthquakes on the Endeavour segment of the Juan de Fuca Ridge, J. Geophys. Res., 107(B12), 2336, doi:10.1029/2001JB000505.
Wilcock, W.S.D., and A.T. Fisher (2004), Geophysical constraints on the sub-seafloor environment near mid-ocean ridges, in The Subseafloor Biosphere at Mid-Ocean Ridges, Geophys. Monogr. Ser., vol. 144, edited by W. S. D. Wilcock et al., pp. 51–74, AGU, Washington D. C.
Williams, D. L., and R. P. Von Hersen (1974). Heat loss from the Earth: New estimate, Geology, 2 327.
Wolfe, C., E. Bergman, and S. Solomon (1993). Oceanic transform earthquakes with unusual mechanisms or locations: Relation to fault geometry and state of stress in adjacent lithosphere, J. Geophys. Res., 98 16187-16211.
Wright, D. J., R. M. Haymon, and K. C. Macdonald (1995). Breaking new ground: Estimates of crack depth along the axial zone of the East Pacific Rise (9º12’-54’N), Earth Planet. Sci. Lett., 134 441-457, doi:10.1016/0012-821X(95)00081-M.
Zhao, L. S., and D. V. Helmberger (1994). Source estimation from broadband regional seismograms, Bull. Seism. Soc. Am. 84 91-104.