SCEC Award Number 20199 View PDF
Proposal Category Collaborative Proposal (Integration and Theory)
Proposal Title Modeling the Rupture Dynamics of Strong Ground Acceleration (>1g) in Fault Stepovers
Investigator(s)
Name Organization
Julian Lozos California State University, Northridge Sinan Akciz California State University, Fullerton
Other Participants
SCEC Priorities 4a, 4c, 4d SCEC Groups FARM, GM, CS
Report Due Date 03/15/2021 Date Report Submitted 03/15/2024
Project Abstract
 Following the 2019 Ridgecrest earthquakes, field investigators noted that pebble- to boulder-sized rocks had been displaced from their place in the desert pavement. This implies localized ground motions above 1 g, contrasting the instrumental maximum of ~0.5 g. These rocks are concentrated within an extensional stepover within the M7.1 rupture. Similar rock displacement occurred in a stepover in the 2010 M7.2 El Mayor-Cucapah earthquake. This suggests that some aspect of how earthquake rupture negotiates a strike-slip fault stepover produces extremely localized strong ground acceleration. We use the 3D finite element method to investigate how rupture through strike-slip stepovers influences strong ground acceleration. For subshear ruptures, the presence of a stepover in general matters more than its dimensions; the strongest ground acceleration always occurs at the end of the first fault. For supershear ruptures, the stepover is effectively irrelevant, since the strongest particle acceleration occurs at the point of the supershear transition on the first fault. Our subshear and supershear ruptures both produce particle acceleration above 1 g, but over a region so close to the fault (< 1 km) that a seismic network may not catch it. We suggest that the physics of rupture through a stepover may be responsible for the displaced rocks in the Ridgecrest and El Mayor-Cucapah earthquakes, and that stepovers may have particularly high ground motion hazard. Ground motion predictions and local hazard assessments should therefore account for much stronger shaking in the immediate near field of active faults, especially around geometrical discontinuities.
Intellectual Merit Seismic hazard is quantified in terms of ground acceleration. This study can deepen our understanding of what physical conditions generate high accelerations, what may cause them to localize around fault stepovers, and what pattern of strong shaking occurs within that local zone.

This study will also introduce displaced rocks like the ones associated with the Ridgecrest and El Mayor-Cucapah earthquakes as a form of quantifiable fragile geologic feature. Classic precariously balanced rocks have been used as constraints on model ground motions before, but smaller and subtler features like these displaced rocks have not previously been used in this way. We also hope that this study may inspire others to consider how to use more unorthodox fragile geologic feature data (or other unusual datasets) to quantify and constrain ground motion from historic or modeled earthquakes.
Broader Impacts From an educational standpoint, this project has introduced CSUN undergraduate student Holland Ladage to the concepts and process of dynamic rupture modeling, as well as to the SCEC community. She ran about half of the models herself, and has been very actively engaged in interpretation as well. She also presented our preliminary work at the 2020 SCEC meeting.

From a societal standpoint, this work will help quantify how complex fault geometries affect strong ground motion. This in turn may help refine seismic hazard maps around stepovers in strike-slip faults. While this is very pertinent in southern California (the cities of San Jacinto and Hemet lie within a major stepover on the San Jacinto Fault, and the city of Lake Elsinore contains a stepover on the Elsinore Fault), these results will have applications to anywhere with discontinuous strike-slip faults.
Exemplary Figure Figure 2 is most representative, since it shows ground motion values across our geometrical parameter space.
Caption: Peak horizontal and vertical particle accelerations (G) for all of our simulations. Note the breaks in the Y-axis. Also note that we had to use Y-axis increments of only 0.02 G to show what little variation there was between these models.