SCEC Award Number 12223 View PDF
Proposal Category Individual Proposal (Integration and Theory)
Proposal Title Long-term behavior of faults with heterogeneous strength due to fault non-planarity
Investigator(s)
Name Organization
Nadia Lapusta California Institute of Technology
Other Participants Graduate student Junle Jiang
SCEC Priorities 3c, 3e, 6b SCEC Groups FARM, GMP, Seismology
Report Due Date 03/15/2013 Date Report Submitted N/A
Project Abstract
 Dynamic rupture simulations suggest that both variations in fault strength and fault prestress can strongly influence dynamic ruptures. The distributions of the two are typically assumed independently, but are related due to prior fault slip. To study the effect of stress redistribution due to prior slip history, we simulate long-term slip on 2D and 3D fault segments with strength heterogeneity due to normal stress variations, motivated by local deviations from planar fault geometry, over many earthquake cycles. These simulations incorporate laboratory-derived friction laws, including enhanced co-seismic weakening, and reveal how prior slip affects earthquakes in the presence of fault heterogeneity. We find that stress redistribution indeed significantly reduces the effect of heterogeneity. Furthermore, enhanced coseismic weakening interacts with strength heterogeneity during seismic slip, resulting in more dynamic weakening in statically stronger spots, further reducing the effect of the heterogeneity. We compare the effect of 1D and 2D strength heterogeneity and find that the difference is quite significant, emphasizing the need for 3D simulations of long-term fault slip to understand realistic fault behavior in the presence of heterogeneity, including fault stress state, seismicity, and seismic radiation. We have also started to study the relationship between the depth extent of enhanced dynamic weakening and seismicity in 3D fault models. We find that large earthquakes can penetrate below the conventionally-defined seismogenic zone due to enhanced dynamic weakening in the deeper creeping region, eliminating concentrated microseismicity at depth, which would be expected otherwise.
Intellectual Merit The main goal of this work is to study how dynamic rupture behavior and earthquake patterns evolve in the presence of fault heterogeneity over many earthquake cycles, using laboratory-derived friction laws including enhanced coseismic weakening. Dynamic rupture simulations suggest that both variations in fault strength and fault prestress can strongly influence the development of dynamic ruptures, e.g. induce or suppress supershear rupture speeds (e.g., Day, 1982; Madariaga et al., 1998; Madariaga and Olsen, 2000; Fukuyama and Olsen, 2002; Dunham et al., 2003; Liu and Lapusta, 2008). The two distributions – of fault strength and shear stress – are typically assumed independently, but they are related due to prior fault slip. Our previous studies supported by SCEC had shown that accounting for stress redistribution due to prior slip history can be crucial for determining the true effect of the assumed fault heterogeneity on dynamic rupture propagation and hence on ground motion. Here we simulate long-term slip on a fault segment that includes earthquake cycles and (i) resolves all the stages of every single earthquake in detail, including earthquake nucleation, dynamic rupture propagation and arrest, (ii) reproduces post-seismic slippage and interseismic creep and (iii) in some cases, incorporates thermal weakening mechanism including thermal pressurization of pore fluids (2D and 3D) and flash heating (2D only) (Lapusta et al., 2000; Lapusta and Liu, 2009; Noda and Lapusta, 2011; based on Dieterich, 1979, 1981; Ruina, 1983; Geubelle and Rice, 1995; Ben-Zion and Rice, 1997). We considered (a) the effect of enhanced co-seismic weakening at high slip rates, (b) the comparison between 1D and 2D distributions of fault heterogeneity, respectively, and (c) the effect of dynamic weakening extending deeper than the seismogenic zone as defined by microseismicity. We find that enhanced co-seismic weakening interacts with heterogeneous fault strength, affecting the average stress levels and conditions for nucleation, with important potential implications for mature faults such as San Andreas Fault. Comparing the effect of 1D and 2D strength heterogeneity on fault plane in 3D fault models, we find that the effects of heterogeneity is quite reduced for 2D strength heterogeneity, emphasizes the need for 3D simulations of long-term seismic behaviors of faults using laboratory-validated friction laws to understand realistic fault behavior and seismic radiation due to fault heterogeneity. We also find that enhanced dynamic weakening can enable events to penetrate below the seismogenic zone defined based on microseismicity, leading to larger events, of which one aspect that could be studied observationally is the microseismicity patterns at the base of seismogenic zone.
Broader Impacts Large-scale dynamic rupture simulations carried out by SCEC teams have the potential to provide novel and critical information for the assessment of seismic hazard in Southern California. The results of this project, when further developed, would (a) suggest ways of assigning initial conditions for these simulations consistent with other assumed ingredients such as the fault strength heterogeneity, and (b) provide better understanding of the long-term behavior of faults, including nucleation conditions and seismicity at rheological boundaries. A student and a postdoctoral fellow have gained valuable research experience by participating in the project and interacting with the SCEC community.
Exemplary Figure Figure 3. (Top) 3D model setup for a rate-and-state fault with non-uniform distribution of normal stress. Second row: Characteristic distributions of normal stress on the fault plane, including (A) the case of 1D along-strike heterogeneity and (B) the case of 2D along-strike-and-dip heterogeneity. (Rows I-III) The long-term behavior of faults with the heterogeneous strength of cases (A) and (B). Row I: The spatial distribution of rupture speed for a typical earthquake in the long-term fault slip history. 1D heterogeneity induces much larger variation in local rupture speed than 2D heterogeneity, with the latter allowing a dynamic event to rupture through local variations relatively easily. Row II and III: the spatial distribution of static stress drop and slip, respectively, both of which are smoother in the case of 2D heterogeneity. Row IV: The mismatch between the representative static strength and typical prestress before large events (blue) and static stress drop (red) along the horizontal middle line of the fault. Overall, the earthquake scenarios produced with 2D along-strike-and-dip heterogeneity have smaller variation in rupture speed, stress drop, slip, and stress mismatch than those produced by only assuming along-strike heterogeneity of fault strength. Jiang and Lapusta, manuscript in preparation, 2013.
Linked Publications

Add missing publication or edit citation shown. Enter the SCEC project ID to link publication.