Benchun Duan
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Benchun Duan has a total of 45 co-authors. Here are the top co-authors by number of shared publications:
- Dunyu Liu
The University of Texas at Austin
6 shared publications - Alice-Agnes Gabriel
Scripps Institution of Oceanography, UC San Diego, La Jolla, California, U.S.A.
3 shared publications - Brittany A. Erickson
University of Oregon
3 shared publications - Camilla Cattania
Department of Earth, Atmospheric, and Planetary Sciences Massachusetts Institute of Technology Cambridge MA USA
3 shared publications - Casper Pranger
Department of Earth and Environmental Sciences Ludwig‐Maximilians‐Universität München Munich Germany
3 shared publications - Duo Li
Department of Earth and Environmental Sciences Ludwig‐Maximilians‐Universität München Munich Germany
3 shared publications
Latest Publications
- 5 October 2023
- Published by Seismological Society of America (SSA) in Bulletin of the Seismological Society of America
- Vol. 114 (2) , 710-725
Dynamic Rupture Models of the 2016 ML 5.8 Gyeongju, South Korea, Earthquake, Constrained by a Kinematic Rupture Model and Seismic Waveform Data
The 5.8 earthquake that jolted Gyeongju in southeastern Korea in 2016 was the country’s largest inland event since instrumental seismic monitoring began in 1978. We developed dynamic rupture models of the Gyeongju event constrained by near‐source ground‐motion data using full 3D spontaneous dynamic rupture modeling with the slip‐weakening friction law. Based on our results, we propose two simple dynamic rupture models with constant strength excess (SE) and slip‐weakening distance ( ) that produce near‐source ground‐motion waveforms compatible with recorded ones in the low‐frequency band. Both dynamic models exhibit relatively large stress‐drop values, consistent with previous estimates constrained by source spectrum analyses. The fracture energy estimates were also larger than those predicted by a scaling relationship with the seismic moment. The dynamic features constrained in this study by spontaneous rupture modeling and waveform comparison may help understand the source and ground‐motion characteristics of future large events in southeastern Korea and thus the seismic hazard of the region.- 5 October 2023
- Published by Seismological Society of America (SSA) in Bulletin of the Seismological Society of America
- Vol. 114 (2) , 710-725
Dynamic Rupture Models of the 2016 ML 5.8 Gyeongju, South Korea, Earthquake, Constrained by a Kinematic Rupture Model and Seismic Waveform Data
The 5.8 earthquake that jolted Gyeongju in southeastern Korea in 2016 was the country’s largest inland event since instrumental seismic monitoring began in 1978. We developed dynamic rupture models of the Gyeongju event constrained by near‐source ground‐motion data using full 3D spontaneous dynamic rupture modeling with the slip‐weakening friction law. Based on our results, we propose two simple dynamic rupture models with constant strength excess (SE) and slip‐weakening distance ( ) that produce near‐source ground‐motion waveforms compatible with recorded ones in the low‐frequency band. Both dynamic models exhibit relatively large stress‐drop values, consistent with previous estimates constrained by source spectrum analyses. The fracture energy estimates were also larger than those predicted by a scaling relationship with the seismic moment. The dynamic features constrained in this study by spontaneous rupture modeling and waveform comparison may help understand the source and ground‐motion characteristics of future large events in southeastern Korea and thus the seismic hazard of the region.- Open Access
- 9 November 2022
- Published by Wiley in ESS Open Archive
Inverting for dynamic stress evolution on earthquake faults directly from seismic recordings
Dynamic stress evolution during earthquake rupture contains information of fault frictional behavior that governs dynamic rupture propagation. Most of earthquake stress drop and evolution studies are based on kinematic slip inversions. Several dynamic inversion methods in the literature require dynamic rupture modeling that makes them cumbersome with limited applicability. In this study, we develop a fault-stress model of earthquake sources in the framework of the representation theorem. We then propose a dynamic stress inversion method based on the fault-stress model to directly invert for dynamic stress evolution process on the fault plane by fitting seismic data. In this inversion method, we calculate numerical Green’s function once only, using an explicit finite element method EQdyna with a unit change of shear or normal stress on each subfault patch. A linear least-squares procedure is used to invert for stress evolution history on the fault. To stabilize the inversion process, we apply several constraints including zero normal slip (no separation or penetration of the fault), non-negative shear slip, and moment constraint. The method performs well and reliably on a synthetic model, a checkerboard model and the 2016 M w 5.0 Cushing (Oklahoma) earthquake. The proposed fault-stress model of earthquake sources with inversion techniques such as one presented in this study provides a new paradigm for earthquake source studies using seismic data, with a potential of deciphering more physics from seismic recordings of earthquakes.
- 9 November 2022
- Published by Wiley in ESS Open Archive
Open AccessInverting for dynamic stress evolution on earthquake faults directly from seismic recordings
Dynamic stress evolution during earthquake rupture contains information of fault frictional behavior that governs dynamic rupture propagation. Most of earthquake stress drop and evolution studies are based on kinematic slip inversions. Several dynamic inversion methods in the literature require dynamic rupture modeling that makes them cumbersome with limited applicability. In this study, we develop a fault-stress model of earthquake sources in the framework of the representation theorem. We then propose a dynamic stress inversion method based on the fault-stress model to directly invert for dynamic stress evolution process on the fault plane by fitting seismic data. In this inversion method, we calculate numerical Green’s function once only, using an explicit finite element method EQdyna with a unit change of shear or normal stress on each subfault patch. A linear least-squares procedure is used to invert for stress evolution history on the fault. To stabilize the inversion process, we apply several constraints including zero normal slip (no separation or penetration of the fault), non-negative shear slip, and moment constraint. The method performs well and reliably on a synthetic model, a checkerboard model and the 2016 M w 5.0 Cushing (Oklahoma) earthquake. The proposed fault-stress model of earthquake sources with inversion techniques such as one presented in this study provides a new paradigm for earthquake source studies using seismic data, with a potential of deciphering more physics from seismic recordings of earthquakes.
- Open Access
- 9 October 2022
- Published by Wiley in ESS Open Archive
Hydraulic Fracturing-driven Infrasound Signals -A New Class of Signal for Subsurface Engineering
Fluid injection into subsurface causes rock deformations, which give rise to mechanical waves in the surrounding rock. This article focuses on the infrasound signals (2-80 Hz) recorded by hydrophones during a meso-scale (~10 meter) hydraulic fracturing experiment at depth of 1.5 kilometer. We present a full-waveform-based data-driven workflow to map the spatiotemporal evolution of the infrasound sources produced during hydraulic fracturing. The infrasound source locations are compared against the simultaneously created microseismic source locations. Orientation of the infrasound source point cloud strongly agrees with natural fracture orientation, as inferred from the discrete fracture-network modelling. Finally, we arrive at a conceptual model of fluid-injection driven infrasound generation in subsurface and posit that the reopening of natural fractures is the main mechanism of the infrasound generation. A joint analysis of signals from microseismicity and infrasound sources can improve subsurface fracture imaging.
- 9 October 2022
- Published by Wiley in ESS Open Archive
Open AccessHydraulic Fracturing-driven Infrasound Signals -A New Class of Signal for Subsurface Engineering
Fluid injection into subsurface causes rock deformations, which give rise to mechanical waves in the surrounding rock. This article focuses on the infrasound signals (2-80 Hz) recorded by hydrophones during a meso-scale (~10 meter) hydraulic fracturing experiment at depth of 1.5 kilometer. We present a full-waveform-based data-driven workflow to map the spatiotemporal evolution of the infrasound sources produced during hydraulic fracturing. The infrasound source locations are compared against the simultaneously created microseismic source locations. Orientation of the infrasound source point cloud strongly agrees with natural fracture orientation, as inferred from the discrete fracture-network modelling. Finally, we arrive at a conceptual model of fluid-injection driven infrasound generation in subsurface and posit that the reopening of natural fractures is the main mechanism of the infrasound generation. A joint analysis of signals from microseismicity and infrasound sources can improve subsurface fracture imaging.
- 20 September 2022
- Published by Springer Nature in Pure and Applied Geophysics
- Vol. 179 (10) , 3543-3568
A Workflow to Integrate Numerical Simulation, Machine Learning Regression and Bayesian Inversion for Induced Seismicity Study: Principles and a Case Study
- 20 September 2022
- Published by Springer Nature in Pure and Applied Geophysics
- Vol. 179 (10) , 3543-3568
A Workflow to Integrate Numerical Simulation, Machine Learning Regression and Bayesian Inversion for Induced Seismicity Study: Principles and a Case Study
- Open Access
- 14 June 2022
- Published by Wiley in ESS Open Archive
Do upper-plate material properties or fault frictional properties dominate tsunami earthquake characteristics?
Tsunami earthquakes are a type of shallow subduction zone events that rupture slowly (<1.5 km/s) with exceptionally long duration and depleted high frequency radiation, resulting in a large discrepancy of Mw and Ms magnitudes and abnormally large tsunami along coastal areas. Heterogeneous fault frictional properties at shallow depth have been thought to dominate tsunami earthquake generation. Some recent studies propose heterogeneous upper-plate material properties determine rupture behavior of megathrust earthquakes, including characteristics of tsunami earthquakes. In this study, we use a recently developed dynamic earthquake simulator to explore tsunami earthquake generation and systematically examine roles of upper-plate material properties and fault frictional properties in tsunami earthquake characteristics in a physics-based framework. For heterogeneous fault friction, we consider isolated asperities with strongly velocity-weakening properties embedded in a conditionally stable zone with weakly velocity-weakening properties. For heterogeneous upper-plate properties, we consider a generic depth profile of seismic velocity and rigidity constrained from seismic surveys. We design a set of models to explore their effects on tsunami earthquake generation and characteristics. We find that the conditionally stable zone can significantly slow down rupture speeds of earthquakes that nucleate on asperities to be < 1.5 km/s over a large depth range (1-20 km), while heterogeneous upper-plate properties can only reduce rupture speeds to be ~1.5-2.0 km/s over a narrow depth range (1-3km). Nevertheless, heterogeneous upper-plate properties promote cascading rupture over multiple isolated asperities on the shallow subduction plane, contributing to large tsunami earthquake generation. We also find that heterogeneous friction dominates normalized duration and high-frequency depletion in tsunami earthquakes. In addition, the effective normal stress on the subduction plane, which affects fault frictional strength, significantly influences the characteristics of tsunami earthquakes, including long normalized duration and low stress drop.
- 14 June 2022
- Published by Wiley in ESS Open Archive
Open AccessDo upper-plate material properties or fault frictional properties dominate tsunami earthquake characteristics?
Tsunami earthquakes are a type of shallow subduction zone events that rupture slowly (<1.5 km/s) with exceptionally long duration and depleted high frequency radiation, resulting in a large discrepancy of Mw and Ms magnitudes and abnormally large tsunami along coastal areas. Heterogeneous fault frictional properties at shallow depth have been thought to dominate tsunami earthquake generation. Some recent studies propose heterogeneous upper-plate material properties determine rupture behavior of megathrust earthquakes, including characteristics of tsunami earthquakes. In this study, we use a recently developed dynamic earthquake simulator to explore tsunami earthquake generation and systematically examine roles of upper-plate material properties and fault frictional properties in tsunami earthquake characteristics in a physics-based framework. For heterogeneous fault friction, we consider isolated asperities with strongly velocity-weakening properties embedded in a conditionally stable zone with weakly velocity-weakening properties. For heterogeneous upper-plate properties, we consider a generic depth profile of seismic velocity and rigidity constrained from seismic surveys. We design a set of models to explore their effects on tsunami earthquake generation and characteristics. We find that the conditionally stable zone can significantly slow down rupture speeds of earthquakes that nucleate on asperities to be < 1.5 km/s over a large depth range (1-20 km), while heterogeneous upper-plate properties can only reduce rupture speeds to be ~1.5-2.0 km/s over a narrow depth range (1-3km). Nevertheless, heterogeneous upper-plate properties promote cascading rupture over multiple isolated asperities on the shallow subduction plane, contributing to large tsunami earthquake generation. We also find that heterogeneous friction dominates normalized duration and high-frequency depletion in tsunami earthquakes. In addition, the effective normal stress on the subduction plane, which affects fault frictional strength, significantly influences the characteristics of tsunami earthquakes, including long normalized duration and low stress drop.
- Open Access
- 14 June 2022
- Published by Wiley in ESS Open Archive
Comparing Roles of Fault Friction and Upper-Plate Rigidity in Depth-Dependent Rupture Characteristics of Megathrust Earthquakes
Megathrust earthquakes with a wide along-dip rupture extent show clearly depth-dependent variations in rupture characteristics such as rupture velocity, frequency contents of seismic radiation and slip distribution. Some recent studies propose that heterogeneous upper-plate rigidity determines this phenomenon, though along-dip variations in fault friction have long been thought to play a dominant role. In this study, we use dynamic rupture modeling to explore and compare roles of these two factors in depth-dependent rupture characteristics of megathrust earthquakes along a shallow-dipping subduction plane that is governed by the rate- and state- dependent friction. We find that an updip transition from velocity-weakening behavior downdip to velocity-strengthening behavior near the trench suppresses rupture propagation toward the trench and a thicker transition zone results in a more confined slip at depth. The updip transition in velocity-dependent frictional property also dominates high-frequency depletion in seismic radiation at shallow depth. With an addition of a conditionally stable zone at shallow depth, rupture velocity significantly decreases, resulting in longer rupture duration as the thickness of the conditionally stable zone increases. The low-velocity layers in the upper plate at shallow depth lead to a more compliant prism and thus significantly higher total slip near the trench. Although they place some limits to rupture velocity at shallow depth, they enhance high-frequency radiation and thus do not contribute to high-frequency depletion observed in recent megathrust earthquakes. We conclude that fault friction plays more important roles than upper-plate rigidity in determining depth-dependent rupture characteristics of megathrust earthquakes.
- 14 June 2022
- Published by Wiley in ESS Open Archive
Open AccessComparing Roles of Fault Friction and Upper-Plate Rigidity in Depth-Dependent Rupture Characteristics of Megathrust Earthquakes
Megathrust earthquakes with a wide along-dip rupture extent show clearly depth-dependent variations in rupture characteristics such as rupture velocity, frequency contents of seismic radiation and slip distribution. Some recent studies propose that heterogeneous upper-plate rigidity determines this phenomenon, though along-dip variations in fault friction have long been thought to play a dominant role. In this study, we use dynamic rupture modeling to explore and compare roles of these two factors in depth-dependent rupture characteristics of megathrust earthquakes along a shallow-dipping subduction plane that is governed by the rate- and state- dependent friction. We find that an updip transition from velocity-weakening behavior downdip to velocity-strengthening behavior near the trench suppresses rupture propagation toward the trench and a thicker transition zone results in a more confined slip at depth. The updip transition in velocity-dependent frictional property also dominates high-frequency depletion in seismic radiation at shallow depth. With an addition of a conditionally stable zone at shallow depth, rupture velocity significantly decreases, resulting in longer rupture duration as the thickness of the conditionally stable zone increases. The low-velocity layers in the upper plate at shallow depth lead to a more compliant prism and thus significantly higher total slip near the trench. Although they place some limits to rupture velocity at shallow depth, they enhance high-frequency radiation and thus do not contribute to high-frequency depletion observed in recent megathrust earthquakes. We conclude that fault friction plays more important roles than upper-plate rigidity in determining depth-dependent rupture characteristics of megathrust earthquakes.
- Open Access
- 10 March 2022
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Solid Earth
- Vol. 127 (3)
Community‐Driven Code Comparisons for Three‐Dimensional Dynamic Modeling of Sequences of Earthquakes and Aseismic Slip
Dynamic modeling of sequences of earthquakes and aseismic slip (SEAS) provides a self‐consistent, physics‐based framework to connect, interpret, and predict diverse geophysical observations across spatial and temporal scales. Amid growing applications of SEAS models, numerical code verification is essential to ensure reliable simulation results but is often infeasible due to the lack of analytical solutions. Here, we develop two benchmarks for three‐dimensional (3D) SEAS problems to compare and verify numerical codes based on boundary‐element, finite‐element, and finite‐difference methods, in a community initiative. Our benchmarks consider a planar vertical strike‐slip fault obeying a rate‐ and state‐dependent friction law, in a 3D homogeneous, linear elastic whole‐space or half‐space, where spontaneous earthquakes and slow slip arise due to tectonic‐like loading. We use a suite of quasi‐dynamic simulations from 10 modeling groups to assess the agreement during all phases of multiple seismic cycles. We find excellent quantitative agreement among simulated outputs for sufficiently large model domains and small grid spacings. However, discrepancies in rupture fronts of the initial event are influenced by the free surface and various computational factors. The recurrence intervals and nucleation phase of later earthquakes are particularly sensitive to numerical resolution and domain‐size‐dependent loading. Despite such variability, key properties of individual earthquakes, including rupture style, duration, total slip, peak slip rate, and stress drop, are comparable among even marginally resolved simulations. Our benchmark efforts offer a community‐based example to improve numerical simulations and reveal sensitivities of model observables, which are important for advancing SEAS models to better understand earthquake system dynamics.
- 10 March 2022
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Solid Earth
- Vol. 127 (3)
Open AccessCommunity‐Driven Code Comparisons for Three‐Dimensional Dynamic Modeling of Sequences of Earthquakes and Aseismic Slip
Dynamic modeling of sequences of earthquakes and aseismic slip (SEAS) provides a self‐consistent, physics‐based framework to connect, interpret, and predict diverse geophysical observations across spatial and temporal scales. Amid growing applications of SEAS models, numerical code verification is essential to ensure reliable simulation results but is often infeasible due to the lack of analytical solutions. Here, we develop two benchmarks for three‐dimensional (3D) SEAS problems to compare and verify numerical codes based on boundary‐element, finite‐element, and finite‐difference methods, in a community initiative. Our benchmarks consider a planar vertical strike‐slip fault obeying a rate‐ and state‐dependent friction law, in a 3D homogeneous, linear elastic whole‐space or half‐space, where spontaneous earthquakes and slow slip arise due to tectonic‐like loading. We use a suite of quasi‐dynamic simulations from 10 modeling groups to assess the agreement during all phases of multiple seismic cycles. We find excellent quantitative agreement among simulated outputs for sufficiently large model domains and small grid spacings. However, discrepancies in rupture fronts of the initial event are influenced by the free surface and various computational factors. The recurrence intervals and nucleation phase of later earthquakes are particularly sensitive to numerical resolution and domain‐size‐dependent loading. Despite such variability, key properties of individual earthquakes, including rupture style, duration, total slip, peak slip rate, and stress drop, are comparable among even marginally resolved simulations. Our benchmark efforts offer a community‐based example to improve numerical simulations and reveal sensitivities of model observables, which are important for advancing SEAS models to better understand earthquake system dynamics.
- Open Access
- 20 February 2022
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Solid Earth
- Vol. 127 (2)
Observation‐Constrained Multicycle Dynamic Models of the Southern San Andreas and the Northern San Jacinto Faults: Addressing Complexity in Paleoearthquake Extent and Recurrence With Realistic 2D Fault Geometry
Understanding mechanical conditions that lead to complexity in earthquakes is important to seismic hazard analysis. In this study, we simulate physics‐based multicycle dynamic models of the San Andreas fault (Carrizo through San Bernardino sections) and the San Jacinto fault (Claremont and Clark strands). We focus on a complex fault geometry based on the Southern California Earthquake Center Community Fault Model and its effect over multiple earthquake cycles. Using geodetically derived strain rates, we validate the models against geologic slip rates and recurrence intervals at various paleoseismic sites. We find that the interactions among fault geometry, dynamic rupture and interseismic stress accumulation produce stress heterogeneities, leading to rupture segmentation and variability in earthquake recurrence. Our models produce earthquakes with rupture extents similar to a recent comprehensive paleoseismic catalog. The “earthquake gates” of the Big Bend and the Cajon Pass occasionally impede dynamic ruptures. The angle of compression, which is the subtraction of the maximum shear strain rate direction from the local fault strike, can better determine the likelihood of the impedance of restraining bends to dynamic ruptures. Because the Big Bend has an angle of compression of ∼20°, ruptures that traverse the Big Bend, like the 1857 Fort Tejon earthquake, are more frequent than expected based on empirical relations which predict the ∼40° restraining bend to terminate most ruptures. Our models indicate that large ruptures tend to initiate north of the Big Bend and propagate southwards, similar to the 1857 earthquake, providing critical information for ground shaking assessment in the region.
- 20 February 2022
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Solid Earth
- Vol. 127 (2)
Open AccessObservation‐Constrained Multicycle Dynamic Models of the Southern San Andreas and the Northern San Jacinto Faults: Addressing Complexity in Paleoearthquake Extent and Recurrence With Realistic 2D Fault Geometry
Understanding mechanical conditions that lead to complexity in earthquakes is important to seismic hazard analysis. In this study, we simulate physics‐based multicycle dynamic models of the San Andreas fault (Carrizo through San Bernardino sections) and the San Jacinto fault (Claremont and Clark strands). We focus on a complex fault geometry based on the Southern California Earthquake Center Community Fault Model and its effect over multiple earthquake cycles. Using geodetically derived strain rates, we validate the models against geologic slip rates and recurrence intervals at various paleoseismic sites. We find that the interactions among fault geometry, dynamic rupture and interseismic stress accumulation produce stress heterogeneities, leading to rupture segmentation and variability in earthquake recurrence. Our models produce earthquakes with rupture extents similar to a recent comprehensive paleoseismic catalog. The “earthquake gates” of the Big Bend and the Cajon Pass occasionally impede dynamic ruptures. The angle of compression, which is the subtraction of the maximum shear strain rate direction from the local fault strike, can better determine the likelihood of the impedance of restraining bends to dynamic ruptures. Because the Big Bend has an angle of compression of ∼20°, ruptures that traverse the Big Bend, like the 1857 Fort Tejon earthquake, are more frequent than expected based on empirical relations which predict the ∼40° restraining bend to terminate most ruptures. Our models indicate that large ruptures tend to initiate north of the Big Bend and propagate southwards, similar to the 1857 earthquake, providing critical information for ground shaking assessment in the region.
- Open Access
- 3 February 2022
- Published by Wiley in ESS Open Archive
Community-Driven Code Comparisons for Three-Dimensional Dynamic Modeling of Sequences of Earthquakes and Aseismic Slip (SEAS)
Dynamic modeling of sequences of earthquakes and aseismic slip (SEAS) provides a self-consistent, physics-based framework to connect, interpret, and predict diverse geophysical observations across spatial and temporal scales. Amid growing applications of SEAS models, numerical code verification is essential to ensure reliable simulation results but is often infeasible due to the lack of analytical solutions. Here, we develop two benchmarks for three-dimensional (3D) SEAS problems to compare and verify numerical codes based on boundary-element, finite-element, and finite-difference methods, in a community initiative. Our benchmarks consider a planar vertical strike-slip fault obeying a rate- and state-dependent friction law, in a 3D homogeneous, linear elastic whole-space or half-space, where spontaneous earthquakes and slow slip arise due to tectonic-like loading. We use a suite of quasi-dynamic simulations from 10 modeling groups to assess the agreement during all phases of multiple seismic cycles. We find excellent quantitative agreement among simulated outputs for sufficiently large model domains and small grid spacings. However, discrepancies in rupture fronts of the initial event are influenced by the free surface and various computational factors. The recurrence intervals and nucleation phase of later earthquakes are particularly sensitive to numerical resolution and domain-size-dependent loading. Despite such variability, key properties of individual earthquakes, including rupture style, duration, total slip, peak slip rate, and stress drop, are comparable among even marginally resolved simulations. Our benchmark efforts offer a community-based example to improve numerical simulations and reveal sensitivities of model observables, which are important for advancing SEAS models to better understand earthquake system dynamics.
- 3 February 2022
- Published by Wiley in ESS Open Archive
Open AccessCommunity-Driven Code Comparisons for Three-Dimensional Dynamic Modeling of Sequences of Earthquakes and Aseismic Slip (SEAS)
Dynamic modeling of sequences of earthquakes and aseismic slip (SEAS) provides a self-consistent, physics-based framework to connect, interpret, and predict diverse geophysical observations across spatial and temporal scales. Amid growing applications of SEAS models, numerical code verification is essential to ensure reliable simulation results but is often infeasible due to the lack of analytical solutions. Here, we develop two benchmarks for three-dimensional (3D) SEAS problems to compare and verify numerical codes based on boundary-element, finite-element, and finite-difference methods, in a community initiative. Our benchmarks consider a planar vertical strike-slip fault obeying a rate- and state-dependent friction law, in a 3D homogeneous, linear elastic whole-space or half-space, where spontaneous earthquakes and slow slip arise due to tectonic-like loading. We use a suite of quasi-dynamic simulations from 10 modeling groups to assess the agreement during all phases of multiple seismic cycles. We find excellent quantitative agreement among simulated outputs for sufficiently large model domains and small grid spacings. However, discrepancies in rupture fronts of the initial event are influenced by the free surface and various computational factors. The recurrence intervals and nucleation phase of later earthquakes are particularly sensitive to numerical resolution and domain-size-dependent loading. Despite such variability, key properties of individual earthquakes, including rupture style, duration, total slip, peak slip rate, and stress drop, are comparable among even marginally resolved simulations. Our benchmark efforts offer a community-based example to improve numerical simulations and reveal sensitivities of model observables, which are important for advancing SEAS models to better understand earthquake system dynamics.
- Open Access
- 3 November 2021
- Published by Wiley in ESS Open Archive
Community-Driven Code Comparisons for Three-Dimensional Dynamic Modeling of Sequences of Earthquakes and Aseismic Slip (SEAS)
Dynamic modeling of sequences of earthquakes and aseismic slip (SEAS) provides a self-consistent, physics-based framework to connect, interpret, and predict diverse geophysical observations across spatial and temporal scales. Amid growing applications of SEAS models, numerical code verification is essential to ensure reliable simulation results but is often infeasible due to the lack of analytical solutions. Here, we develop two benchmarks for three-dimensional (3D) SEAS problems to compare and verify numerical codes based on boundary-element, finite-element, and finite-difference methods, in a community initiative. Our benchmarks consider a planar vertical strike-slip fault obeying a rate- and state-dependent friction law, in a 3D homogeneous, linear elastic whole-space or half-space, where spontaneous earthquakes and slow slip arise due to tectonic-like loading. We use a suite of quasi-dynamic simulations from 10 modeling groups to assess the agreement during all phases of multiple seismic cycles. We found excellent quantitative agreement among simulated outputs for sufficiently large model domains and fine grid spacings. However, discrepancies in rupture fronts of the initial event are influenced by the free surface and various computational factors. The recurrence intervals and nucleation phase of later earthquakes are particularly sensitive to numerical resolution and domain-size-dependent loading. Despite such variability, key properties of individual earthquakes, including rupture style, duration, total slip, peak slip rate, and stress drop, are comparable among even marginally resolved simulations. Our benchmark efforts offer a community-based example to improve numerical simulations and reveal sensitivities of model observables, which are important for advancing SEAS models to better understand earthquake system dynamics.
- 3 November 2021
- Published by Wiley in ESS Open Archive
Open AccessCommunity-Driven Code Comparisons for Three-Dimensional Dynamic Modeling of Sequences of Earthquakes and Aseismic Slip (SEAS)
Dynamic modeling of sequences of earthquakes and aseismic slip (SEAS) provides a self-consistent, physics-based framework to connect, interpret, and predict diverse geophysical observations across spatial and temporal scales. Amid growing applications of SEAS models, numerical code verification is essential to ensure reliable simulation results but is often infeasible due to the lack of analytical solutions. Here, we develop two benchmarks for three-dimensional (3D) SEAS problems to compare and verify numerical codes based on boundary-element, finite-element, and finite-difference methods, in a community initiative. Our benchmarks consider a planar vertical strike-slip fault obeying a rate- and state-dependent friction law, in a 3D homogeneous, linear elastic whole-space or half-space, where spontaneous earthquakes and slow slip arise due to tectonic-like loading. We use a suite of quasi-dynamic simulations from 10 modeling groups to assess the agreement during all phases of multiple seismic cycles. We found excellent quantitative agreement among simulated outputs for sufficiently large model domains and fine grid spacings. However, discrepancies in rupture fronts of the initial event are influenced by the free surface and various computational factors. The recurrence intervals and nucleation phase of later earthquakes are particularly sensitive to numerical resolution and domain-size-dependent loading. Despite such variability, key properties of individual earthquakes, including rupture style, duration, total slip, peak slip rate, and stress drop, are comparable among even marginally resolved simulations. Our benchmark efforts offer a community-based example to improve numerical simulations and reveal sensitivities of model observables, which are important for advancing SEAS models to better understand earthquake system dynamics.