| 英文摘要: | The San Andreas Fault is the largest and arguably the most dangerous source of earthquakes in Southern California. It has produced very large (magnitude greater than 7.5) earthquakes in the past, and it is likely to do so again, with essentially no warning. One of the key tools that seismologists can use to help anticipate and mitigate the effects of such future disasters is to make numerical (computer) models of potential earthquakes. Current numerical models indicate that an earthquake that propagates from Indio toward Los Angeles, through the San Gorgonio Pass (SGP), can channel ground shaking energy directly toward the Los Angeles Basin, where it is amplified by basin sediments and can lead to very high ground motion. Thus, it is extremely important to determine whether such an earthquake that passes through the SGP is likely (or even possible). This is not a trivial question, as the fault structure in the San Gorgonio Pass is quite complex, with the main San Andreas splitting into a number of smaller faults with different orientations and directions of slip. We will use the most accurate available modeling techniques to address the question of whether an earthquake can propagate through this region, leading to a large, damaging event for the region. We will focus on the eastern SGP, where the Coachella Valley strand of the SAF branches into the Mission Creek, Banning, and Garnet Hill fault strands. Each of these strands has very different structure and apparent activity levels, so an understanding of through-going earthquake potential in the region may hinge on whether there is a preferred rupture path in this region. This project implements numerical models of potential earthquakes in this region to determine which of these strands is the most likely path of an earthquake. The models use fault stresses derived from long-term simulations of fault slip and Earth surface deformation in the region, and give information on potential earthquake rupture path, slip distribution, earthquake size, and ground motion. These innovative earthquake models will also include the effects of rock failure away from the faults, which may have a significant effect on the energy budget of earthquakes. This work constitutes an important advance in the science of earthquake modeling, and it has important broader impacts, including estimates of earthquake size and ground motion in the region, with further implications for engineering, zoning, and emergency response.
The question of whether an earthquake can propagate through the San Gorgonio Pass (SGP) region of the San Andreas Fault (SAF) is of tremendous scientific and societal importance. The SGP is a ?pinch point? along this fault system, in which the SAF splits into multiple non-coplanar segments, including both strike-slip and thrust; there may well be no through-going surface to support continuous earthquake rupture in this region [e.g., Yule, 2009]. Numerical models [e.g., Olsen et al., 2008] indicate that an earthquake rupture that propagates through the SGP from southeast to northwest can channel seismic radiation directly toward the Los Angeles Basin, where it is amplified by basin sediments and can lead to very high ground motion. Thus, the question of whether the San Gorgonio Pass serves as a significant barrier to earthquake rupture propagation is of great practical as well as theoretical interest, and it affects ground motion not just locally but in the most populated regions of Southern California. This work aims to help answer this question by performing numerical models of potential earthquakes in this region. Specifically, we will focus on the eastern SGP, where the Coachella Valley segment of the SAF branches into the Mission Creek, Banning, and Garnet Hill fault strands. Each of these strands has very different structure and apparent activity levels, so an understanding of through-going earthquake potential in the region may hinge on whether there is a preferred rupture path in this region. We will use a combination of methods to address this question: coupled long-term and interseismic quasi-static modeling method to determine the on- and off-fault stresses in the region, and a 3D dynamic rupture model that uses these stresses to model potential earthquakes, including rupture propagation, slip, and near-source ground motion. Each of these methods has been well tested individually, but this is the first time that they have been combined in this way. The work will is a significant advance in the field of realistic fault dynamics, in that it is (to the PIs knowledge) the first study to combine a tectonically- and geometrically-consistent heterogeneous stress field with off-fault failure; prior models have had either but not both of these abilities. Earlier homogenous-stress models that incorporate such failure mechanisms [Andrews, 2005; DeDontney et al., 2012; Duan and Day, 2008; Dunham et al., 2011; Gabriel et al., 2013] indicate that off-fault plasticity can have a very strong effect on the energy budget and rupture propagation during an earthquake; we expect the inclusion of these effects in our heterogeneous stress models to be at least as significant. Our work will advance the science of fault dynamics in regions of geometrical complexity, especially fault branches. The methods developed in this project are readily applicable to other earthquake-prone regions with significant fault complexity, such as Northern California. The results have important implications for earthquake occurrence in the SGP, as well as for seismic hazard throughout Southern California. In particular, it has important implications for estimates of potential earthquake size in Southern California, and consequently the generation of strong ground motion and estimation of seismic hazard. This information can be useful for engineering, zoning, and emergency response purposes. In addition, the investigators perform outreach activities to the local Morongo Band Of Mission Indians in San Gorgonio Pass. This funding supports the Ph.D. research of a talented young postdoctoral researcher of Afro-Caribbean ancestry, a significantly under-represented group within the Geosciences. |