Small interseismic asperities and widespread aseismic creep on the northern Japan subduction interface

Johnson, K. M.; Mavrommatis, A. P.; Segall, P.

doi:10.1002/2015gl066707

Cited by 23 publications

(33 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A single fault might slip either seismically or aseismically in different circumstances. As a salient example, much of the fault that slipped coseismically in the 2011 Tohoku earthquake slipped stably both before and after the M 9 earthquake, as inferred from both repeating earthquake sequences [e.g., Uchida and Matsuzawa , ] and geodetic inversions of afterslip [ Ito et al ., ; Johnson et al ., , ; Perfettini and Avouac , ]. This bimodal character is also observed elsewhere [e.g., Bürgmann et al ., ; Pritchard and Simons , ; Bedford et al ., ; Lin et al ., ; Barnhart et al ., ] and on deeper fault sections that produce tremor [e.g., Wech and Bartlow , ].…”

Section: Introductionmentioning

confidence: 99%

Slow and fast ruptures on a laboratory fault controlled by loading characteristics

2017

View full text Add to dashboard Cite

Recent geodetic observations indicate that a single fault may slip slowly and silently or fast and seismically in different circumstances. We report laboratory experiments that demonstrate how the mode of faulting can alternate between fast “stick‐slip” events (70 to ~500 mm/s sliding rates) and slow silent events (~0.001 to 30 mm/s) as a result of loading conditions rather than friction properties or stiffness of the loading machine. The 760 mm long granite sample is close to the critical nucleation length scale for unstable slip, so small variations in nucleation properties result in measurable differences in slip events. Slow events occur when instability cannot fully nucleate before reaching the sample ends. Dynamic events occur after long healing times or abrupt increases in loading rate which suggests that these factors shrink the spatial and temporal extents of the nucleation zone. Arrays of slip, strain, and ground motion sensors installed on the sample allow us to quantify seismic coupling and study details of premonitory slip and afterslip. We find that seismic coupling decreases when slip rates fall below about 70 mm/s. The slow slip events we observe are primarily aseismic (less than 1% of the seismic coupling of faster events) and produce swarms of very small M −6 to M −8 events. These mechanical and seismic interactions suggest that faults with transitional behavior—where creep, small earthquakes, and tremor are often observed—could become seismically coupled if loaded rapidly, either by a slow slip front or dynamic rupture of an earthquake that nucleated elsewhere.

show abstract

Section: Introductionmentioning

confidence: 99%

Slow and fast ruptures on a laboratory fault controlled by loading characteristics

2017

View full text Add to dashboard Cite

show abstract

“…This change was therefore interpreted as a gradual decrease in the area of the megathrust that was locked. Following up on this hypothesis, Johnson et al () modeled the time‐dependent interseismic velocity field by assigning full coupling within historical M w 7–8 rupture zones on the Japan Trench (Figure a), surrounded by creep at constant stress (as in Bürgmann et al, ). They showed that in 2009, interseismic coupling on the megathrust, as well as afterslip following M w ∼7 events, was consistent with locked asperities being only ∼25% the size of historical rupture areas determined by seismic source inversions (Figure c), remarkably consistent with Bürgmann et al's () results.…”

Section: Introductionmentioning

confidence: 99%

“…They showed that in 2009, interseismic coupling on the megathrust, as well as afterslip following M w ∼7 events, was consistent with locked asperities being only ∼25% the size of historical rupture areas determined by seismic source inversions (Figure c), remarkably consistent with Bürgmann et al's () results. Following Johnson et al (), we modeled the velocity field in 1998 by assigning full coupling in historical rupture areas. We find the data are well explained by a model with full‐sized asperities (Figure b).…”

Section: Introductionmentioning

confidence: 99%

“…(a) Historical (1938–2010) rupture areas of M 7–8 earthquakes on the Japan Trench (after Johnson et al, ). (b) Interseismic velocities (vectors; observed in blue and modeled in green) and inferred coupling (background colors) in 1998.…”

Section: Introductionmentioning

confidence: 99%

“…(b) Interseismic velocities (vectors; observed in blue and modeled in green) and inferred coupling (background colors) in 1998. (c) Interseismic velocities and inferred coupling in 2009 from Johnson et al ().…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A Physical Model for Interseismic Erosion of Locked Fault Asperities

2017

Self Cite

View full text Add to dashboard Cite

The conventional “asperity model” posits that faults are partitioned into fixed velocity‐weakening (VW) patches (asperities) that are locked interseismically and velocity‐strengthening (VS) regions that creep stably without accumulating stress. However, studies of GPS‐derived deformation in northern Japan have shown that interseismic strain in the Tohoku region did not accumulate at a constant rate (as expected) but gradually decreased from 1996 to 2011. This change in strain rate is consistent with locked asperities shrinking by ∼75% in area during this period. Here we consider a modification to the conventional asperity model, such that thermal pressurization (TP) is active over an area that encompasses a VW region and part of the surrounding VS region. In our quasi‐dynamic simulations, TP causes shear stress during rapid slip to decrease to very low levels. During the interseismic period, stress gradually recovers to steady state friction at the plate rate, at which point stable creep initiates. The creep front propagates inward, effectively eroding the locked asperity. For uniform properties, the locked area shrinks roughly linearly in time through the VS region. Locked asperities shrink more slowly with higher nominal friction coefficient or background effective normal stress in the VS region, lower hydraulic diffusivity, and larger TP zones. Lateral heterogeneity in properties can give rise to nonlinear erosion. Predictions from this model can be compared against GPS data to test whether the model can explain the observed changes in interseismic strain rate in Tohoku.

show abstract

Coseismic slip and early afterslip of the 2015 Illapel, Chile, earthquake: Implications for frictional heterogeneity and coastal uplift

et al. 2016

View full text Add to dashboard Cite

Great subduction earthquakes are thought to rupture portions of the megathrust, where interseismic coupling is high and velocity‐weakening frictional behavior is dominant, releasing elastic deformation accrued over a seismic cycle. Conversely, postseismic afterslip is assumed to occur primarily in regions of velocity‐strengthening frictional characteristics that may correlate with lower interseismic coupling. However, it remains unclear if fixed frictional properties of the subduction interface, coseismic or aftershock‐induced stress redistribution, or other factors control the spatial distribution of afterslip. Here we use interferometric synthetic aperture radar and Global Position System observations to map the distribution of coseismic slip of the 2015 Mw 8.3 Illapel, Chile, earthquake and afterslip within the first 38 days following the earthquake. We find that afterslip overlaps the coseismic slip area and propagates along‐strike into regions of both high and moderate interseismic coupling. The significance of these observations, however, is tempered by the limited resolution of geodetic inversions for both slip and coupling. Additional afterslip imaged deeper on the fault surface bounds a discrete region of deep coseismic slip, and both contribute to net uplift of the Chilean Coastal Cordillera. A simple partitioning of the subduction interface into regions of fixed frictional properties cannot reconcile our geodetic observations. Instead, stress heterogeneities, either preexisting or induced by the earthquake, likely provide the primary control on the afterslip distribution for this subduction zone earthquake. We also explore the occurrence of coseismic and postseismic coastal uplift in this sequence and its implications for recent hypotheses concerning the source of permanent coastal uplift along subduction zones.

show abstract

Small interseismic asperities and widespread aseismic creep on the northern Japan subduction interface

Cited by 23 publications

References 38 publications

Slow and fast ruptures on a laboratory fault controlled by loading characteristics

Slow and fast ruptures on a laboratory fault controlled by loading characteristics

A Physical Model for Interseismic Erosion of Locked Fault Asperities

Coseismic slip and early afterslip of the 2015 Illapel, Chile, earthquake: Implications for frictional heterogeneity and coastal uplift

Contact Info

Product

Resources

About