From California to British Columbia, the Pacific Northwest coast bears an omnipresent earthquake and tsunami hazard from the Cascadia subduction zone. Multiple lines of evidence suggests that magnitude eight and greater megathrust earthquakes have occurred ‐ the most recent being 321 years ago (i.e., 1700 A.D.). Outstanding questions for the next great megathrust event include where it will initiate, what conditions are favorable for rupture to span the convergent margin, and how much slip may be expected. We develop the first 3‐D fully dynamic rupture simulations for the Cascadia subduction zone that are driven by fault stress, strength and friction to address these questions. The initial dynamic stress drop distribution in our simulations is constrained by geodetic coupling models, with segment locations taken from geologic analyses. We document the sensitivity of nucleation location and stress drop to the final seismic moment and coseismic subsidence amplitudes. We find that the final earthquake size strongly depends on the amount of slip deficit in the central Cascadia region, which is inferred to be creeping interseismically, for a given initiation location in southern or northern Cascadia. Several simulations are also presented here that can closely approximate recorded coastal subsidence from the 1700 A.D. event without invoking localized high‐stress asperities along the down‐dip locked region of the megathrust. These results can be used to inform earthquake and tsunami hazards for not only Cascadia, but other subduction zones that have limited seismic observations but a wealth of geodetic inference.
- Award ID(s):
- 1802364
- NSF-PAR ID:
- 10145603
- Date Published:
- Journal Name:
- Geophysical Journal International
- Volume:
- 219
- Issue:
- 1
- ISSN:
- 0956-540X
- Page Range / eLocation ID:
- 645 to 661
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
more » « less
Abstract -
So far in this century, six very large–magnitude earthquakes ( M W ≥ 7.8) have ruptured separate portions of the subduction zone plate boundary of western South America along Ecuador, Peru, and Chile. Each source region had last experienced a very large earthquake from 74 to 261 y earlier. This history led to their designation in advance as seismic gaps with potential to host future large earthquakes. Deployments of geodetic and seismic monitoring instruments in several of the seismic gaps enhanced resolution of the subsequent faulting processes, revealing preevent patterns of geodetic slip deficit accumulation and heterogeneous coseismic slip on the megathrust fault. Localized regions of large slip, or asperities, appear to have influenced variability in how each source region ruptured relative to prior events, as repeated ruptures have had similar, but not identical slip distributions. We consider updated perspectives of seismic gaps, asperities, and geodetic locking to assess current very large earthquake hazard along the South American subduction zone, noting regions of particular concern in northern Ecuador and Colombia (1958/1906 rupture zone), southeastern Peru (southeasternmost 1868 rupture zone), north Chile (1877 rupture zone), and north-central Chile (1922 rupture zone) that have large geodetic slip deficit measurements and long intervals (from 64 to 154 y) since prior large events have struck those regions. Expanded geophysical measurements onshore and offshore in these seismic gaps may provide critical information about the strain cycle and fault stress buildup late in the seismic cycle in advance of the future great earthquakes that will eventually strike each region.more » « less
-
Abstract On 19 September 2022, a major earthquake struck the northwestern Michoacán segment along the Mexican subduction zone. A slip model is obtained that satisfactorily explains geodetic, teleseismic, and tsunami observations of the 2022 event. The preferred model has a compact large-slip patch that extends up-dip and northwestward from the hypocenter and directly overlaps a 1973 Mw 7.6 rupture. Slip is concentrated offshore and below the coast at depths from 10 to 30 km with a peak value of ∼2.9 m, and there is no detected coseismic slip near the trench. The total seismic moment is 3.1×1020 N·m (Mw 7.6), 72% of which is concentrated in the first 30 s. Most aftershocks are distributed in an up-dip area of the mainshock that has small coseismic slip, suggesting near-complete strain release in the large-slip patch. Teleseismic P waveforms of the 2022 and 1973 earthquakes are similar in duration and complexity with high cross-correlation coefficients of 0.68–0.98 for long P to PP signal time windows, indicating that the 2022 earthquake is a quasi-repeat of the 1973 earthquake, possibly indicating persistent frictional properties. Both the events produced more complex P waveforms than comparable size events along Guerrero and Oaxaca, reflecting differences in patchy locking of the Mexican megathrust.more » « less
-
more » « less
Abstract We demonstrate the efficacy of a Bayesian statistical inversion framework for reconstructing the likely characteristics of large pre‐instrumentation earthquakes from historical records of tsunami observations. Our framework is designed and implemented for the estimation of the location and magnitude of seismic events from anecdotal accounts of tsunamis including shoreline wave arrival times, heights, and inundation lengths over a variety of spatially separated observation locations. The primary advantage of this approach is that all of the assumptions made in the inversion process are incorporated explicitly into the mathematical framework. As an initial test case we use our framework to reconstruct the great 1852 earthquake and tsunami of eastern Indonesia. Relying on the assumption that these observations were produced by a subducting thrust event, the posterior distribution indicates that the observables were the result of a massive mega‐thrust event with magnitude near 8.8 Mw and a likely rupture zone in the north‐eastern Banda arc. The distribution of predicted epicentral locations overlaps with the largest major seismic gap in the region as indicated by instrumentally recorded seismic events. These results provide a geologic and seismic context for hazard risk assessment in coastal communities experiencing growing population and urbanization in Indonesia. In addition, the methodology demonstrated here highlights the potential for applying a Bayesian approach to enhance understanding of the seismic history of other subduction zones around the world.
-
more » « less
SUMMARY Observations of historical tsunami earthquakes reveal that ruptures of these earthquakes propagate slowly at shallow depth with longer duration, depletion in high-frequency radiation and larger discrepancy of Mw–Ms than ordinary megathrust earthquakes. They can effectively generate tsunami and lead to huge damage to regional populated areas near the coast. In this study, we use a recently developed dynamic earthquake simulator to explore tsunami earthquake generation from a physics-based modelling point of view. We build a shallow-dipping subduction zone model in which locally locked, unstable patches (asperities) are distributed on a conditionally stable subduction interface at shallow depth. The dynamic earthquake simulator captures both quasi-static and dynamic processes of earthquake cycles. We find that earthquakes can nucleate on these asperities and propagate into the surrounding conditionally stable zone at slow speeds, generating tsunami earthquakes. A high normal stress asperity, representing a subducted seamount, can act as an asperity in some events but as a barrier in other events over multiple earthquake cycles. Low normal stress asperities typically act as asperities in tsunami earthquakes. The degree of velocity-weakening in the conditionally stable zone, which may sustain rupture at different speeds or stop rupture, is critical for tsunami earthquake generation and affects its recurrence interval. Distributed asperities may rupture in isolated events separated by tens of years, or in a sequence of events separated by hours to days, or in one large event in a cascade fashion, demonstrating complex interactions among them. The recurrence interval on a high normal stress asperity is much larger than that on low normal stress asperities. These modelling results shed lights on the observations from historical tsunami earthquakes, including the 1994 and 2006 Java tsunami earthquakes and 2010 Mentawai tsunami earthquake.
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.1093/gji/ggz302
