Supershear earthquakes, propagating faster than the Earth's shear wave velocity, can generate strong ground motion at distances far from the ruptured fault. Despite the hazards associated with these earthquakes, the exact fault properties that govern their occurrence are not well constrained. Although early studies associated supershear ruptures with simple fault geometries, recent dynamic rupture models have revealed a supershear transition mechanism over complex fault geometries such as fault stepovers. Here we present the first observation of a supershear transition on a fault stepover system during the 2017 M_w 7.7 Komandorsky Islands earthquake. Using a high-resolution back-projection technique, we find that the earthquake's rupture velocity accelerates from 2.1 to 5.0 km/s between two offset faults, demonstrating the viability of a new supershear transition mechanism occurring in nature. Given the fault complexity of the Earth's transform plate boundaries, this result may improve our understanding of supershear rupture processes and their associated hazards.

Key Points

A supershear transition across a fault stepover is observed in nature for the first time during the 2017 Komandorsky Islands earthquake
These results have implications for the assessment of seismic hazard on fault stepover systems at all transform plate boundaries

Plain Language Summary

Earthquakes traveling faster than the shear wave velocity of the earth are occasionally observed in large strike-slip events. These so-called supershear earthquakes generate a potentially destructive shock wave analogous to the sonic boom of a supersonic aircraft. Although many supershear earthquakes are coincident with straight, continuous fault sections, we find that the 2017 Komandorsky Islands earthquake reached supershear speeds following a jump in rupture across two fault segments. The results of this study confirm that supershear ruptures can occur on complex fault systems, allowing the seismic hazard on similar fault systems to be better evaluated.

1 Introduction

The tectonic environment surrounding the Kamchatka-Aleutian triple junction is complex, in large part due to the westward transition from normal subduction to trench-parallel shearing along the Aleutian Trench (Figure 1a). Upper plate deformation west of 172°E is controlled by back-arc strike-slip faulting resulting from the increase in slip partitioning as the relative motion between the Pacific and North American plates becomes more oblique (Avé Lallemant, 1996; Newberry et al., 1986). Near the Komandorsky Islands (Figure 1c), back-arc strike-slip faults form the Bering Fracture Zone (BFZ), and the Komandorsky Sliver is bounded to the north and south by the BFZ and Aleutian Trench, respectively. The BFZ accommodates approximately two thirds (51 mm/yr) of the relative shearing motion between the Pacific and North American plates (78 mm/yr) and has a locking depth of 12 km, making the region a prime tectonic environment for large strike-slip events (Kogan et al., 2017).

Details are in the caption following the image — **Figure 1**
Open in figure viewer PowerPoint

Geographic map of the study area and the back-projection station distribution. (a) Subduction zones are shown as a black line with white triangles on the overriding plate. Relative motion of the Pacific Plate with respect to the North American plate along the Aleutian Trench is shown with black arrows (DeMets et al., 2010). The centroid-moment tensor shows the epicenter of the 2017 M_W 7.7 mainshock. The red inverted triangles show the 116 seismic stations removed by the genetic algorithm and the blue inverted triangles show the 107 stations kept by the genetic algorithm (optimized network). The black box outlines the location of the map area in (c). (b) The histogram shows the longitudinal distribution (bin widths of 0.1°) of 138 aftershocks above magnitude 4 for 3 weeks after the mainshock. The white star and black dashed line show the longitude of the mainshock epicenter with respect to the aftershock distribution. The red dashed line shows the beginning of the supershear rupture segment. (c) The Aleutian Trench is shown as a black line with white triangles on the overriding plate. The Bering Fracture Zone (BFZ) is shown as a black line with right-lateral strike-slip arrows. The area between the Aleutian Trench, and BFZ is the Komandorsky Sliver. The two dark gray islands are the Komandorsky Islands. The red and white stars show the M_W 6.3 foreshock and M_W 7.7 mainshock epicenters, respectively. The centroid-moment tensor of each earthquake is shown by their respective colors. The magnitude-scaled red and white dots are earthquakes above magnitude 4 for the day before (foreshocks) and 3 weeks after (aftershocks) the mainshock, respectively.

On 17 July 2017, a M_w 7.7 earthquake occurred 335 km east-southeast of the Kamchatka-Aleutian triple junction, near the Komandorsky Islands (Figure 1c). The National Earthquake Information Center (NEIC) hypocenter and the Global Centroid-Moment Tensor (GCMT) solution (Dziewonski et al., 1981; Ekström et al., 2012) indicate that this earthquake occurred on the BFZ. This is the largest event recorded on this portion of the BFZ by modern instruments. Approximately 12 hours before the mainshock, a large M_w 6.3 foreshock occurred 20.4 km northwest of the mainshock epicenter, also on the BFZ (Figure 1c). A sequence of seven additional smaller events occurred near the BFZ between the foreshock and the mainshock, and a significant cluster of aftershocks occurred immediately southeast of the mainshock (Figures 1b and 1c). The BFZ is expected to be structurally complex, especially southeast of the mainshock epicenter. Geodetic measurements indicate a probable migration of strike-slip faults from the back-arc to the fore-arc (Cross & Freymueller, 2008; Kogan et al., 2017), and the distribution of focal mechanisms reveals an abrupt change in the stress field near 170°E (Lutikov et al., 2019).

In this study, we image the rupture process of the 2017 Komandorsky Islands earthquake using a network of seismic stations in Alaska and the back-projection method, which stacks time-shifted waveforms on a grid of potential source locations to determine the spatiotemporal evolution of energy release during an earthquake (Ishii et al., 2005). We show that the earthquake rupture accelerated to supershear velocities across a fault stepover system with a transition mechanism that has previously only been observed in numerical models (Hu et al., 2016; Ryan & Oglesby, 2014).

2 Data and Methods

The back-projection method, benefiting from the emergence of dense seismic networks, has been used by several studies to image the rupture processes of earthquakes (e.g., Kiser & Ishii, 2017). The method does not require a priori information such as fault geometry and rupture velocity and is able to resolve complex features such as earthquake frequency dependence and rupture segmentation (e.g., Kiser & Ishii, 2011; Koper et al., 2011; Meng et al., 2011; Wang & Mori, 2011). Although any seismic phase and distribution of stations may be used, back-projection studies are typically limited to P waves recorded at teleseismic (30°–90°) distance windows to minimize interference with other seismic phases. Back-projection studies at regional (< 30°) distance windows may capture greater rupture complexity than those at teleseismic distance windows due to the increased relative aperture of the seismic network. However, these results are susceptible to artifacts caused by triplications from the mantle transition zone and other unwanted seismic phases (e.g., depth phases). Here we apply a genetic algorithm-based station selection method (Kehoe et al., 2019) to three networks of stations in Alaska within a distance of 35° (USArray Transportable Array: https://doi.org/10.7914/SN/TA; Alaska Regional Network: https://doi.org/10.7914/SN/AK; Alaska Volcano Observatory: https://doi.org/10.7914/SN/AV). This process removes stations producing artifacts in the source image, forming an optimized back-projection network (Figure 1a) and revealing the previously unobserved rupture complexity of the 2017 Komandorsky Islands earthquake.

Seismic data used in the back-projection analysis are obtained from the Incorporated Research Institutions for Seismology (IRIS) Consortium and are band-pass filtered to two frequency bands, 0.5 to 1 Hz (low-frequency) and 0.8 to 2 Hz (high-frequency). For all data, a cross-correlation procedure (Ishii et al., 2007) is used to align waveforms from a small event near the mainshock hypocenter. The alignment event used in this study had a similar GCMT solution to the mainshock, occurred 32.8 km from the mainshock hypocenter on 28 July 2017, and had a moment magnitude of 5.5 (02:39:15 UTC, 54.303°N 169.301°E; NEIC: https://earthquake.usgs.gov/earthquakes/eventpage/us2000a21g/executive). This process accounts for three-dimensional heterogeneity at grid points near the source by making empirical travel time corrections to P wave arrival times such that they agree with the arrival times calculated using a one-dimensional Earth model, iasp91 (Kennett & Engdahl, 1991). This cross-correlation procedure provides further adjustments such that P wave amplitudes are normalized, and waveform polarities are corrected with respect to a reference stack of waveforms.

Following the travel time corrections, a genetic algorithm-based station selection method (Text S1) is used to remove stations that produce artifacts in the back-projection image. For this step, a second small event near the mainshock hypocenter is used to find a distribution of stations that image the event as a point source. The point source event used in this study had a similar GCMT solution to the mainshock, occurred 19.2 km from the mainshock hypocenter on 17 July 2017, and had a body wave magnitude of 5.1 (11:23:01 UTC, 54.596°N 168.721°E; NEIC: https://earthquake.usgs.gov/earthquakes/eventpage/us20009wwq/executive). This approach is established on the principle that small events should be imaged by the back-projection method as point sources, and any additional imaged energy is likely the result of artifacts in the back-projection result. Given a seismic network of M stations, an optimal subset of stations that image a small event as a point source can be found in 2^M attempts. A genetic algorithm (Holland, 1992), typically used to find solutions to optimization problems with large and complex fitness landscapes, is used to significantly reduce this number of attempts and find a near-optimal solution.

Once the optimized network is determined, the performance of the genetic algorithm-based station selection method is quantified. Figure S1 shows how the fitness value of the point source event increases as a function of iteration, and Figure S2 shows the source image improvement of the point source event between the full and optimized seismic networks. The final back-projection result is obtained by first aligning the mainshock waveforms recorded at the optimized network independently of the previous alignment event before using a coherence-based back-projection approach (Ishii, 2011) to image the mainshock rupture. This alignment procedure robustly corrects for three-dimensional heterogeneity along the length of the southeastern rupture (Figure S6). The mainshock waveforms recorded at the complete distribution of stations are also aligned independently and back-projected using the same coherence-based approach to determine the improvement in spatiotemporal resolution between the optimized and full networks (Figure S3).

3 Regional Back-Projection Result

Back-projection results using high-frequency (0.8 to 2 Hz) data recorded at the optimized network reveal a complex southeastern rupture containing three distinct segments while results using low-frequency (0.5 to 1 Hz) data image a simple northwestern rupture (Figure 2a, Movies S1 and S2, Data Sets S1 and S2 ). Portions of the southeastern rupture are also imaged using low-frequency data (Figure S3, Movie S3, Data Set S3). The complete distribution of regional stations resolves some portions of the southeastern rupture at low and high frequencies but does not resolve the northwestern rupture, which is dominated by an eastward moving artifact (Figure S3, Movies S4–S6, Data Sets S4–S6). Consequently, we use the optimized network (Figure 1a) and present the high-frequency southeastern rupture and low-frequency northwestern rupture as our main back-projection result (Figure 2a).

The locations of energy release northwest and southeast of the reported epicenter suggest that this earthquake ruptured bilaterally. The high-frequency southeastern rupture begins with Segment 1, which originates near the reported epicenter and propagates 94.3 km southeast for 45 s at 2.1 km/s. During Segment 2, the rupture appears to turn due south and travel 18.8 km for 7 s at 2.7 km/s. As the rupture resumes propagation towards the southeast during Segment 3, the rupture speed quickly accelerates to 5.0 km/s, traversing 50.2 km in just 10 s. The rupture speed of Segment 3 exceeds the shear wave velocity of the crust, estimated by Crust1.0 to be 3.4–3.9 km/s in this region (Laske et al., 2013). The imaged low-frequency northwestern rupture begins 135.8 km northwest of the reported epicenter and propagates 41.6 km northwest for 31 s at 1.3 km/s. The dominant southeastern rupture obscures the first 79 s of the northwestern rupture, preventing this portion of the rupture from being imaged. Assuming a continuous northwestern rupture between the epicenter and the first location of imaged energy at 79 s, a rupture velocity of 1.7 km/s over 135.8 km is required. Previous models incorporating GPS observations require slip northwest of the epicenter, but back-projection results prior to this study have not robustly imaged the northwestern propagation of this earthquake (Lay et al., 2017; IRIS: https://doi.org/10.17611/DP/13610282). A teleseismic back-projection analysis using North American data band-pass filtered between 0.5 and 2 Hz resolves the general southeastern and northwestern ruptures (Figure S4, Movies S7 and S8, Data Sets S7 and S8), although the regional back-projection results offer superior spatiotemporal resolution. The complexity of the southeastern rupture can be seen in all back-projection results (Figures S3–S4), which is further illustrated by evaluating the epicentral distance of each source location as a function of time for the regional back-projection results (Figure S5).

The normalized energy release as a function of time is determined by applying a linear back-projection analysis to data band-passed filtered between 0.5 and 2 Hz (Figure 2b). The largest amplitude of energy release occurs during Segment 3 of the southeastern rupture. The total duration of this earthquake is estimated to be 110 s. Following the earthquake, there is an imaged episode of energy release southeast of Segment 3 on the BFZ, which may be an early aftershock (Movie S1).

4 Independent Supershear Validation

Supershear rupture velocities have been observed using a variety of different methods, with studies exploiting strong ground motion records (Archuleta, 1984; Bouchon et al., 2000, 2001; Dunham & Archuleta, 2004), finite fault model inversions (Konca et al., 2010; Yue et al., 2013), and far-field surface wave observations (Bao et al., 2019; Vallée & Dunham, 2012). Although the back-projection method is capable of resolving supershear ruptures (Bao et al., 2019; Walker & Shearer, 2009; Wang et al., 2012), here we use far-field Rayleigh wave observations to validate the supershear rupture segment imaged in the regional back-projection result.

Rayleigh waves generated by the 2017 M_w 7.7 Komandorsky Islands mainshock and a smaller M_w 6.3 foreshock are compared in this section. The two events share similar focal mechanisms (Figure 1c) and are 20.4 km apart. Seismic theory (Dunham & Bhat, 2008; Vallée & Dunham, 2012) predicts that within a far-field Mach cone, surface wave seismograms generated along the length of a supershear rupture will be identical to those created by smaller events with the same focal mechanism, and the surface wave amplitude ratio of the two events will be equal to their moment ratio. The opening angles of the Mach cones with respect to the supershear rupture segment are a function of the surface wave phase velocity and the rupture velocity. Due to the frequency-dependent and globally heterogenous nature of surface wave phase velocities, values are estimated for each Mach cone using GDM52, a global surface wave phase velocity model (Ekström, 2011). The northeastern and southeastern Mach cone locations are calculated using estimated Rayleigh wave phase velocities of 3.3 ± 0.4 km/s and 4.1 ± 0.3 km/s, respectively. Using constraints from the back-projection result, both Mach cone calculations use an estimated rupture velocity of 5.0 ± 0.5 km/s, and the supershear rupture segment is approximated as a straight line that is 50.2 km long, striking 126.0° (Figure S3a).

Broadband vertical component seismic data recorded within a distance window of 60° of the mainshock hypocenter are retrieved from the IRIS Consortium, the Northern California Earthquake Data Center (NCEDC: https://www.doi.org/10.7932/NCEDC), and the Southern California Earthquake Data Center (SCEDC: https://doi.org/10.7909/C3WD3xH1) and are band-pass filtered to periods of 15–25 s to minimize the effects of dispersion. Rayleigh wave first arrivals are manually picked in the mainshock data, and the Rayleigh wave train is manually cut from the foreshock data. The foreshock Rayleigh wave train is cross-correlated against the mainshock seismogram (beginning at the Rayleigh wave first arrival), and the maximum cross-correlation value at each station is used to compare waveform similarity inside and outside the Mach cones. There is a distinct pattern of high correlation values (> 0.8) within the Mach cones (Figure 3). The amplitude ratios of the recorded Rayleigh waves from the M_w 7.7 mainshock and the M_w 6.3 foreshock also agree with theoretical predictions. The amplitude ratio of recorded Rayleigh waves within the Mach cones is a factor of approximately 25–40. Had the supershear segment composed the entire mainshock rupture, the expected moment ratio between the M_w 7.7 mainshock and the M_w 6.3 foreshock would be a factor of approximately 126. Rayleigh waves from the subshear segment (Segment 1) of the mainshock rupture may interfere destructively with Rayleigh waves from the supershear segment (Segment 3) but have little effect on the overall waveforms recorded by stations within the Mach cones (Vallée & Dunham, 2012). We estimate the supershear segment released 20–30% of the total moment by assuming the rupture length is proportional to moment release. This reduces the expected moment ratio between the two events to the observed amplitude ratios within the Mach cones. The high cross-correlation values and the observed amplitude ratios within the Mach cones indicate that this segment of the mainshock ruptured at supershear speeds. Due to poor station coverage in the Pacific Ocean, the northeastern Mach cone is a more reliable indicator of the supershear rupture than the southeastern Mach cone. The nonzero time shifts associated with the maximum cross-correlation values of data recorded within the Mach cones are due to the initial subshear segment (Segment 1) of the mainshock rupture, which delays the arrival of the Rayleigh waves generated by the supershear segment (Segment 3).

5 Discussion and Conclusions

Earthquake ruptures that transition to supershear rupture velocities were initially explained by mechanisms on planar mode II (e.g., strike-slip) cracks (Andrews, 1976; Burridge, 1973), and many of the first observations of supershear ruptures were sustained on simple, continuous sections of strike-slip faults (Bouchon et al., 2010). Although supershear propagation may preferentially occur on simple fault geometries, numerical studies have shown that stress and frictional strength heterogeneities, likely related to fault complexity, may cause a rupture to transition to supershear speeds (e.g., Bruhat et al., 2016; Dunham, 2007; Liu & Lapusta, 2008). Dynamic rupture models have directly shown that supershear transitions on complex systems of faults such as fault stepovers may also occur (Hu et al., 2016; Ryan & Oglesby, 2014). Ryan and Oglesby (2014) explain that in this type of scenario, the sudden termination of an earthquake rupture at the end of the primary fault changes the stress field across the fault system, enabling the secondary fault to rupture at supershear speeds by increasing the shear stress and decreasing the normal stress at the beginning of the secondary fault. This type of supershear transition mechanism can occur on both compressional and extensional fault stepover systems, although high initial stresses are required. Consequently, high initial stresses also allow ruptures to jump across larger stepover widths.

The regional back-projection result and the independent Rayleigh wave validation show that the 2017 Komandorsky Islands earthquake accelerated from a sub-Rayleigh velocity of 2.1 km/s on Segment 1 to a supershear velocity of 5.0 km/s on Segment 3. The end of Segment 1 is associated with high amplitudes of high-frequency seismic radiation, consistent with stopping phases from the sudden termination of a rupture (Madariaga, 1976). Segment 2 is associated with a decrease in source amplitude (Figure 2b), indicative of a discontinuous rupture (Kiser & Ishii, 2013). Upon renucleation at the start of Segment 3, the rupture immediately begins to propagate at supershear speeds. This feature differentiates the 2017 Komandorsky Islands supershear transition mechanism from the classical Burridge-Andrews mechanism, which requires some sub-Rayleigh propagation distance prior to the onset of a supershear rupture (Andrews, 1976). The southeastern portion of this earthquake likely achieved supershear speeds via the mechanism described by Ryan and Oglesby (2014), jumping between two offset faults of a fault stepover system (Segments 1 and 3) and accelerating to supershear speeds immediately following renucleation on the secondary fault (Segment 3). Local bathymetric features (Tozer et al., 2019) illustrate the complexity of this fault system, revealing a basin between Segments 1 and 3 that is characteristic of an extensional basin in a right-stepping right-lateral fault system (Figures 2a and Figure 4).

The complexity of this fault system is further delineated by the distribution of aftershocks associated with the 2017 Komandorsky Islands earthquake. The majority of these aftershocks occur immediately southeast of the mainshock epicenter and decrease near 170°E (Figure 1b). This observation may be explained by the supershear segment of the mainshock rupture, which began near 170°E and ruptured southeast. Faults that achieve supershear speeds are often associated with a distinct lack of aftershocks (Bouchon & Karabulut, 2008), which instead occur on faults adjacent to the main rupture plane. This phenomenon is thought to be caused by the complete release of stress on the main fault by the supershear rupture and the activation of fault splays by the expansive supershear shock wave.

Rupture speed plays a critical role in the damage caused by earthquakes (Das, 2007). In particular, supershear earthquakes have the potential to transfer large stresses and generate strong ground motion at far distances from the ruptured fault (Dunham & Bhat, 2008). The majority of supershear earthquakes have ruptured on smooth, continuous sections of strike-slip faults (Bouchon et al., 2010), but supershear transitions tend to coincide with increased fault complexity. The 2001 M_w 7.8 Kokoxili (Kunlun) earthquake reached supershear speeds in an area exhibiting a change in rupture azimuth and a large push-up structure (Vallée et al., 2008). The 2015 M_w 7.2 Murghab, Tajikistan earthquake temporarily slowed down through a restraining bend connecting two continuous sections of a fault rupturing at supershear speeds (Sangha et al., 2017). The 2018 M_w 7.5 Palu earthquake reached supershear speeds immediately following nucleation, likely to due to fault roughness near the epicenter (Bao et al., 2019). Like these three earthquakes, the 2017 M_w 7.7 Komandorsky Islands earthquake achieved supershear speeds over a geometrically complex fault system. However, the observed transition mechanism across a fault stepover is expected to facilitate supershear rupture speeds within an initial stress field that precludes supershear transitions via the standard Burridge-Andrews mechanism (Hu et al., 2016; Ryan & Oglesby, 2014). The majority of strike-slip fault systems contain complexity in the form of fault stepovers (Wesnousky, 2006), and many of these areas are situated near large population centers (e.g., the San Andreas Fault, the North Anatolian Fault). The ability for a rupture to not only jump large stepover widths but also transition to supershear speeds across a fault stepover system is of great importance to the evaluation of global seismic hazard.

Acknowledgments

The authors would like to thank two anonymous reviewers and the editor, Gavin Hayes, for their constructive comments and suggestions. Additional conversations with Han Bao and Colton Lynner have helped improve the results presented in this manuscript. Seismic data used in this study were downloaded from the IRIS Consortium (http://ds.iris.edu/wilber3/find_event), the NCEDC (https://www.doi.org/10.7932/NCEDC), and the SCEDC (https://doi.org/10.7909/C3WD3xH1). The Generic Mapping Tools (GMT) were used to create figures presented in this manuscript. The Python software package ObsPy and a cross-correlation code written by Tom Eulenfeld were used in the Rayleigh wave cross-correlation analysis. This research was supported by National Science Foundation grant EAR-1802441. The back-projection results are available in the supporting information.

Supporting Information

References

Andrews, D. J. (1976). Rupture velocity of plane strain shear cracks. Journal of Geophysical Research, 81(32), 5679–5687. https://doi.org/10.1029/JB081i032p05679
10.1029/JB081i032p05679
ADSWeb of Science®Google Scholar
Archuleta, R. J. (1984). A faulting model for the 1979 Imperial Valley earthquake. Journal of Geophysical Research, 89(B6), 4559–4585. https://doi.org/10.1029/JB089iB06p04559
10.1029/JB089iB06p04559
ADSWeb of Science®Google Scholar
Avé Lallemant, H. G. (1996). Displacement partitioning and arc-parallel extension in the Aleutian volcanic island arc. Tectonophysics, 256(1–4), 279–293. https://doi.org/10.1016/0040-1951(95)00171-9
10.1016/0040-1951(95)00171-9
Web of Science®Google Scholar
Bao, H., Ampuero, J.-P., Meng, L., Fielding, E. J., Liang, C., Milliner, C. W. D., Feng, T., & Huang, H. (2019). Early and persistent supershear rupture of the 2018 magnitude 7.5 Palu earthquake. Nature Geoscience, 12(3), 200–205. https://doi.org/10.1038/s41561-018-0297-z
10.1038/s41561-018-0297-z
CASADSWeb of Science®Google Scholar
Bouchon, M., Bouin, M.-P., Karabulut, H., Toksöz, M. N., Dietrich, M., & Rosakis, A. J. (2001). How fast is rupture during an earthquake? New insights from the 1999 Turkey earthquakes. Geophysical Research Letters, 28(14), 2723–2726. https://doi.org/10.1029/2001GL013112
10.1029/2001GL013112
ADSWeb of Science®Google Scholar
Bouchon, M., & Karabulut, H. (2008). The aftershock signature of Supershear earthquakes. Science, 320(5881), 1323–1325. https://doi.org/10.1126/science.1155030
10.1126/science.1155030
CASADSPubMedWeb of Science®Google Scholar
Bouchon, M., Karabulut, H., Bouin, M. P., Schmittbuhl, J., Vallée, M., Archuleta, R., Das, S., Renard, F., & Marsan, D. (2010). Faulting characteristics of supershear earthquakes. Tectonophysics, 493(3–4), 244–253. https://doi.org/10.1016/j.tecto.2010.06.011
10.1016/j.tecto.2010.06.011
ADSWeb of Science®Google Scholar
Bouchon, M., Toksöz, N., Karabulut, H., Bouin, M.-P., Dietrich, M., Aktar, M., & Edie, M. (2000). Seismic imaging of the 1999 Izmit (Turkey) rupture inferred from the near-fault recordings. Geophysical Research Letters, 27(18), 3013–3016. https://doi.org/10.1029/2000GL011761
10.1029/2000GL011761
ADSWeb of Science®Google Scholar
Bruhat, L., Fang, Z., & Dunham, E. M. (2016). Rupture complexity and the supershear transition on rough faults. Journal of Geophysical Research: Solid Earth, 121, 210–224. https://doi.org/10.1002/2015JB012512
10.1002/2015JB012512
ADSWeb of Science®Google Scholar
Burridge, R. (1973). Admissible speeds for plane-strain self-similar shear cracks with friction but lacking cohesion. Geophysical Journal International, 35(4), 439–455. https://doi.org/10.1111/j.1365-246X.1973.tb00608.x
10.1111/j.1365-246X.1973.tb00608.x
ADSWeb of Science®Google Scholar
Cross, R. S., & Freymueller, J. T. (2008). Evidence for and implications of a Bering plate based on geodetic measurements from the Aleutians and western Alaska. Journal of Geophysical Research, 113, B07405. https://doi.org/10.1029/2007JB005136
10.1029/2007JB005136
Web of Science®Google Scholar
Das, S. (2007). The need to study speed. Science, 317(5840), 905–906.
10.1126/science.1142143
CASPubMedWeb of Science®Google Scholar
DeMets, C., Gordon, R. G., & Argus, D. F. (2010). Geologically current plate motions. Geophysical Journal International, 181(1), 1–80. https://doi.org/10.1111/j.1365-246X.2009.04491.x
10.1111/j.1365-246X.2009.04491.x
ADSWeb of Science®Google Scholar
Dunham, E. M. (2007). Conditions governing the occurrence of supershear ruptures under slip-weakening friction. Journal of Geophysical Research, 112, B07302. https://doi.org/10.1029/2006JB004717
10.1029/2006JB004717
ADSWeb of Science®Google Scholar
Dunham, E. M., & Archuleta, R. J. (2004). Evidence for a Supershear transient during the 2002 Denali Fault earthquake. Bulletin of the Seismological Society of America, 94(6B), S256–S268. https://doi.org/10.1785/0120040616
10.1785/0120040616
ADSWeb of Science®Google Scholar
Dunham, E. M., & Bhat, H. S. (2008). Attenuation of radiated ground motion and stresses from three-dimensional supershear ruptures. Journal of Geophysical Research, 113, B08319. https://doi.org/10.1029/2007JB005182
10.1029/2007JB005182
Web of Science®Google Scholar
Dziewonski, A. M., Chou, T.-A., & Woodhouse, J. H. (1981). Determination of earthquake source parameters from waveform data for studies of global and regional seismicity. Journal of Geophysical Research, 86(B4), 2825–2852. https://doi.org/10.1029/JB086iB04p02825
10.1029/JB086iB04p02825
ADSWeb of Science®Google Scholar
Ekström, G., Nettles, M., & Dziewoński, A. M. (2012). The global CMT project 2004–2010: Centroid-moment tensors for 13,017 earthquakes. Physics of the Earth and Planetary Interiors, 200–201, 1–9. https://doi.org/10.1016/J.PEPI.2012.04.002
10.1016/j.pepi.2012.04.002
ADSWeb of Science®Google Scholar
Ekström, G. (2011). A global model of Love and Rayleigh surface wave dispersion and anisotropy, 25-250 s. Geophysical Journal International, 187(3), 1668–1686. https://doi.org/10.1111/j.1365-246X.2011.05225.x
10.1111/j.1365-246X.2011.05225.x
ADSWeb of Science®Google Scholar
Holland, J. H. (1992). Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control, and Artificial Intelligence. Cambridge, MA: MIT Press.
10.7551/mitpress/1090.001.0001
Google Scholar
Hu, F., Xu, J., Zhang, Z., & Chen, X. (2016). Supershear transition mechanism induced by step over geometry. Journal of Geophysical Research: Solid Earth, 121, 8738–8749. https://doi.org/10.1002/2016JB013333
10.1002/2016JB013333
Web of Science®Google Scholar
Ishii, M. (2011). High-frequency rupture properties of the Mw 9.0 off the Pacific coast of Tohoku earthquake. Earth, Planets and Space, 63(7), 609–614. https://doi.org/10.5047/eps.2011.07.009
10.5047/eps.2011.07.009
ADSWeb of Science®Google Scholar
Ishii, M., Shearer, P. M., Houston, H., & Vidale, J. E. (2005). Extent, duration and speed of the 2004 Sumatra-Andaman earthquake imaged by the Hi-Net array. Nature, 435(7044), 933–936. https://doi.org/10.1038/nature03675
10.1038/nature03675
CASADSPubMedWeb of Science®Google Scholar
Ishii, M., Shearer, P. M., Houston, H., & Vidale, J. E. (2007). Teleseismic P wave imaging of the 26 December 2004 Sumatra-Andaman and 28 March 2005 Sumatra earthquake ruptures using the Hi-net array. Journal of Geophysical Research, 112, B11307. https://doi.org/10.1029/2006JB004700
10.1029/2006JB004700
ADSWeb of Science®Google Scholar
Kehoe, H. L., Kiser, E. D., & Okubo, P. G. (2019). The rupture process of the 2018 Mw 6.9 Hawaiʻi earthquake as imaged by a genetic algorithm-based back-projection technique. Geophysical Research Letters, 46, 2467–2474. https://doi.org/10.1029/2018GL080397
10.1029/2018GL080397
ADSWeb of Science®Google Scholar
Kennett, B. L. N., & Engdahl, E. R. (1991). Traveltimes for global earthquake location and phase identification. Geophysical Journal International, 105(2), 429–465. https://doi.org/10.1111/j.1365-246X.1991.tb06724.x
10.1111/j.1365-246X.1991.tb06724.x
ADSWeb of Science®Google Scholar
Kiser, E., & Ishii, M. (2011). The 2010 Mw 8.8 Chile earthquake: Triggering on multiple segments and frequency-dependent rupture behavior. Geophysical Research Letters, 38, L07301. https://doi.org/10.1029/2011GL047140
10.1029/2011GL047140
ADSWeb of Science®Google Scholar
Kiser, E., & Ishii, M. (2013, April). The 2010 Maule, Chile, coseismic gap and its relationship to the 25 march 2012 MW 7.1 earthquake. Bulletin of the Seismological Society of America, 103(2A), 1148–1153. https://doi.org/10.1785/0120120209
10.1785/0120120209
Web of Science®Google Scholar
Kiser, E., & Ishii, M. (2017). Back-projection imaging of earthquakes. Annual Review of Earth and Planetary Sciences, 45(1), 271–299. https://doi.org/10.1146/annurev-earth-063016-015801
10.1146/annurev-earth-063016-015801
CASADSWeb of Science®Google Scholar
Kogan, M. G., Frolov, D. I., Vasilenko, N. F., Freymueller, J. T., Steblov, G. M., Ekström, G., Titkov, N. N., & Prytkov, A. S. (2017). Plate coupling and strain in the far western Aleutian arc modeled from GPS data. Geophysical Research Letters, 44, 3176–3183. https://doi.org/10.1002/2017GL072735
10.1002/2017GL072735
ADSWeb of Science®Google Scholar
Konca, A. O., Leprince, S., Avouac, J.-P., & Helmberger, D. V. (2010). Rupture process of the 1999 Mw 7.1 Duzce earthquake from joint analysis of SPOT, GPS, InSAR, strong-motion, and Teleseismic data: A Supershear rupture with variable rupture velocity. Bulletin of the Seismological Society of America, 100(1), 267–288. https://doi.org/10.1785/0120090072
10.1785/0120090072
ADSWeb of Science®Google Scholar
Koper, K. D., Hutko, A. R., Lay, T., Ammon, C. J., & Kanamori, H. (2011). Frequency-dependent rupture process of the 2011 Mw 9.0 Tohoku earthquake: Comparison of short-period P wave backprojection images and broadband seismic rupture models. Earth, Planets and Space, 63(7), 599–602. https://doi.org/10.5047/eps.2011.05.026
10.5047/eps.2011.05.026
ADSWeb of Science®Google Scholar
Laske, G., Masters, G., Ma, Z., & Pasyanos, M. (2013). Update on CRUST1.0—A 1-degree global model of Earth's crust. Geophysical Research Abstracts, 15.
Google Scholar
Lay, T., Ye, L., Bai, Y., Cheung, K. F., Kanamori, H., Freymueller, J., Steblov, G. M., & Kogan, M. G. (2017). Rupture along 400 km of the Bering Fracture Zone in the Komandorsky Islands earthquake (MW 7.8) of 17 July 2017. Geophysical Research Letters, 44, 12,161–12,169. https://doi.org/10.1002/2017GL076148
10.1002/2017GL076148
Web of Science®Google Scholar
Liu, Y., & Lapusta, N. (2008). Transition of mode II cracks from sub-Rayleigh to intersonic speeds in the presence of favorable heterogeneity. Journal of the Mechanics and Physics of Solids, 56(1), 25–50. https://doi.org/10.1016/j.jmps.2007.06.005
10.1016/j.jmps.2007.06.005
ADSWeb of Science®Google Scholar
Lutikov, A. I., Rogozhin, E. A., Dontsova, G. Y., & Zhukovets, V. N. (2019). The MW 7.8 earthquake of July 17, 2017 off the Commander Islands and other large earthquakes at the Western segment of the Aleutian Island arc. Journal of Volcanology and Seismology, 13(2), 112–123. https://doi.org/10.1134/S0742046319020052
10.1134/S0742046319020052
Web of Science®Google Scholar
Madariaga, R. (1976). Dynamics of an expanding circular fault. Bulletin of the Seismological Society of America, 66(3), 639–666. https://doi.org/10.2307/1779766
10.1785/BSSA0660030639
Web of Science®Google Scholar
Meng, L., Inbal, A., & Ampuero, J.-P. (2011). A window into the complexity of the dynamic rupture of the 2011 mw 9 Tohoku-Oki earthquake. Geophysical Research Letters, 38, L00G07. https://doi.org/10.1029/2011GL048118
10.1029/2011GL048118
ADSWeb of Science®Google Scholar
Newberry, J. T., Laclair, D. L., & Fujita, K. (1986). Seismicity and tectonics of the far Western Aleutian Islands. Journal of Geodynamics, 6(1–4), 13–32. https://doi.org/10.1016/0264-3707(86)90030-X
10.1016/0264-3707(86)90030-X
ADSWeb of Science®Google Scholar
Ryan, K. J., & Oglesby, D. D. (2014). Dynamically modeling fault step overs using various friction laws. Journal of Geophysical Research: Solid Earth, 119, 5814–5829. https://doi.org/10.1002/2014JB011151
10.1002/2014JB011151
ADSWeb of Science®Google Scholar
Sangha, S., Peltzer, G., Zhang, A., Meng, L., Liang, C., Lundgren, P., & Fielding, E. (2017). Fault geometry of 2015, Mw7.2 Murghab, Tajikistan earthquake controls rupture propagation: Insights from InSAR and seismological data. Earth and Planetary Science Letters, 462, 132–141. https://doi.org/10.1016/j.epsl.2017.01.018
10.1016/j.epsl.2017.01.018
CASADSWeb of Science®Google Scholar
Tozer, B., Sandwell, D. T., Smith, W. H. F., Olson, C., Beale, J. R., & Wessel, P. (2019). Global bathymetry and topography at 15 arc seconds: SRTM15+. Earth and Space Science, 6. https://doi.org/10.1029/2019ea000658
10.1029/2019EA000658
Web of Science®Google Scholar
Vallée, M., & Dunham, E. M. (2012). Observation of far-field Mach waves generated by the 2001 Kokoxili supershear earthquake. Geophysical Research Letters, 39, L05311. https://doi.org/10.1029/2011GL050725
10.1029/2011GL050725
ADSWeb of Science®Google Scholar
Vallée, M., Landès, M., Shapiro, N. M., & Klinger, Y. (2008). The 14 November 2001 Kokoxili (Tibet) earthquake: High-frequency seismic radiation originating from the transitions between sub-Rayleigh and supershear rupture velocity regimes. Journal of Geophysical Research, 113, B07305. https://doi.org/10.1029/2007JB005520
10.1029/2007JB005520
ADSWeb of Science®Google Scholar
Walker, K. T., & Shearer, P. M. (2009). Illuminating the near-sonic rupture velocities of the intracontinental Kokoxili Mw 7.8 and Denali fault Mw 7.9 strike-slip earthquakes with global P wave back projection imaging. Journal of Geophysical Research, 114, B02304. https://doi.org/10.1029/2008JB005738
10.1029/2008JB005738
ADSWeb of Science®Google Scholar
Wang, D., & Mori, J. (2011). Frequency-dependent energy radiation and fault coupling for the 2010 Mw8.8 Maule, Chile, and 2011 Mw9.0 Tohoku, Japan, earthquakes. Geophysical Research Letters, 38, L22308. https://doi.org/10.1029/2011GL049652
10.1029/2011GL049652
ADSWeb of Science®Google Scholar
Wang, D., Mori, J., & Uchide, T. (2012). Supershear rupture on multiple faults for the Mw 8.6 off northern Sumatra, Indonesia earthquake of April 11, 2012. Geophysical Research Letters, 39, L21307. https://doi.org/10.1029/2012GL053622
10.1029/2012GL053622
ADSWeb of Science®Google Scholar
Wesnousky, S. G. (2006). Predicting the endpoints of earthquake ruptures. Nature, 444(7117), 358–360. https://doi.org/10.1038/nature05275
10.1038/nature05275
CASADSPubMedWeb of Science®Google Scholar
Yue, H., Lay, T., Freymueller, J. T., Ding, K., Rivera, L., Ruppert, N. A., & Koper, K. D. (2013). Supershear rupture of the 5 January 2013 Craig, Alaska (Mw 7.5) earthquake. Journal of Geophysical Research: Solid Earth, 118, 5903–5919. https://doi.org/10.1002/2013JB010594
10.1002/2013JB010594
ADSWeb of Science®Google Scholar

Citing Literature

Volume47, Issue10

28 May 2020

e2020GL087400

Filename	Description
grl60480-sup-0015-2020GL087400-SI.docxWord 2007 document , 7.1 MB	Supporting Information S1
grl60480-sup-0001-2020GL087400-ds01.txtplain text document, 1.6 KB	Data Set S1
grl60480-sup-0002-2020GL087400-ds02.txtplain text document, 843 B	Data Set S2
grl60480-sup-0003-2020GL087400-ds03.txtplain text document, 1.6 KB	Data Set S3
grl60480-sup-0004-2020GL087400-ds04.txtplain text document, 1.6 KB	Data Set S4
grl60480-sup-0005-2020GL087400-ds05.txtplain text document, 843 B	Data Set S5
grl60480-sup-0006-2020GL087400-ds06.txtplain text document, 1.6 KB	Data Set S6
grl60480-sup-0007-2020GL087400-ds07.txtplain text document, 1.7 KB	Data Set S7
grl60480-sup-0008-2020GL087400-ds08.txtplain text document, 942 B	Data Set S8
grl60480-sup-0009-2020GL087400-fs01.pdfPDF document, 7.5 KB	Figure S1
grl60480-sup-0010-2020GL087400-fs02.pdfPDF document, 55.9 KB	Figure S2
grl60480-sup-0011-2020GL087400-fs03.pdfPDF document, 68.9 KB	Figure S3
grl60480-sup-0012-2020GL087400-fs04.pdfPDF document, 3.1 MB	Figure S4
grl60480-sup-0013-2020GL087400-fs05.pdfPDF document, 15.4 KB	Figure S5
grl60480-sup-0014-2020GL087400-fs06.pdfPDF document, 159.7 KB	Figure S6
grl60480-sup-0016-2020GL087400-ms01.mp4MPEG-4 video, 48.2 MB	Movie S1
grl60480-sup-0017-2020GL087400-ms02.mp4MPEG-4 video, 23.9 MB	Movie S2
grl60480-sup-0018-2020GL087400-ms03.mp4MPEG-4 video, 24.5 MB	Movie S3
grl60480-sup-0019-2020GL087400-ms04.mp4MPEG-4 video, 26.2 MB	Movie S4
grl60480-sup-0020-2020GL087400-ms05.mp4MPEG-4 video, 23.9 MB	Movie S5
grl60480-sup-0021-2020GL087400-ms06.mp4MPEG-4 video, 27.5 MB	Movie S6
grl60480-sup-0022-2020GL087400-ms07.mp4MPEG-4 video, 74 MB	Movie S7
grl60480-sup-0023-2020GL087400-ms08.mp4MPEG-4 video, 23.9 MB	Movie S8

Evidence of a Supershear Transition Across a Fault Stepover

Abstract

Key Points

Plain Language Summary

1 Introduction

2 Data and Methods

3 Regional Back-Projection Result

4 Independent Supershear Validation

5 Discussion and Conclusions

Acknowledgments

Supporting Information

References

Citing Literature

Figures

References

Information

RESOURCES

PUBLICATION INFO

About Wiley Online Library

Help & Support

Opportunities

Connect with Wiley

Evidence of a Supershear Transition Across a Fault Stepover

Abstract

Key Points

Plain Language Summary

1 Introduction

2 Data and Methods

3 Regional Back-Projection Result

4 Independent Supershear Validation

5 Discussion and Conclusions

Acknowledgments

Supporting Information

References

Citing Literature

Figures

References

Related

Information

RESOURCES

PUBLICATION INFO