2013 SCEC Project Report

Enhancements to the SCEC Unified Structural Representation: CFM, SCFM, & CVMH

Principal Investigator: John H. Shaw

Professor of Structural & Economic Geology Harvard University

Dept. of Earth & Planetary Sciences 20 Oxford St., Cambridge, MA 02138

shaw@eps.harvard.edu //(617) 495-8008

co-Investigator: Andreas Plesch (Harvard)

Proposal Categories: A: Data gathering and products Primary Focus Area: USR

Primary Discipline Group: Geology Science Objectives: 4A, 4C, 6A

Abstract Our efforts this past year focused on making a series of specific improvements to the SCEC Community Fault (CFM) and Velocity (CVMH) models. These include completing peer evaluation of the CFM in southern California, and initiating a review of the Northern California fault model. Together, these fault models include more than 300 faults, with alternative representations of many sources, and comprise the SCEC Statewide Community Fault Model (SCFM). In addition, we completed a series of improvements to the CFM fault representations in the Transverse Ranges using more precise surface traces from the USGS Quaternary Fault and Fold Database and refined earthquake hypocenter and focal mechanism catalogs. Finally, we developed and implemented a new representation of the Santa Maria basin in the CVM-H. These activities represent the primary effort to support and develop the SCEC Community Fault and Velocity models (CFM, SCFM, CVMH), which are used in a wide range of efforts in source characterization, fault system modeling, strong ground motion simulation, and hazard assessment, including UCERF3. Technical Report Community Fault Model This past year we focused on making a series of specific improvements to the CFM and facilitating peer review of these models. This included completing the CFM evaluation in southern California, a process that involved a virtual workshop in which participants assessed the completeness and accuracy of each fault representation, in part by assigning a quality ranking (1-5) to each fault representation. These rankings, which largely reflect the amount of data available to directly constrain the interpretations, were used to construct a preferred set of representations that comprise a new model release (CFM 4.1). The workshop was facilitated using the SCEC VDO software, which was customized and bundled with the fault representations and their primary data constraints, including the USGS Quaternary Fault & Fold (Qfault) database traces and seismicity. In addition, we initiated a similar review of the CFM faults in northern California. This evaluation will be completed in 2014. In addition, we made a series of improvements to the CFM fault representations in the Transverse Ranges working in collaboration with Craig Nicholson (UCSB). This represents part of a systematic effort to improve the precision of CFM representations by incorporating detailed surface fault traces from the Qfault database and constraints on subsurface geometry from new focal mechanism and relocated hypocenter catalogs produced by SCEC investigators (e.g., Shearer et al., 2005; Lin et al., 2007; Hauksson et al., 2011). At the outset of the effort to develop the CFM, the fault trace map of Jennings (1994) was chosen as the primary database for defining surface fault traces, due to its widespread use in hazard assessment efforts. Since the early days of the CFM, however, a major effort has since been undertaken by the USGS and CGS to compile a more accurate and comprehensive catalog of surface fault traces, and to reflect this information in the Quaternary Fault & Fold database. These fault traces are generally more precise than Jennings (1994), and often include more fault splays than the original compilation. In addition, SCEC investigators have also developed a series of new, relocated

earthquake catalogs that offer much improved hypocentral locations that can be used to constrain subsurface fault geometries (Shearer et al., 2005; Lin et al., 2007; Waldhauser and Schaff, 2008; Hauksson et al., 2011). In collaboration with Craig Nicholson (UCSB), and in coordination with the USGS and CGS, we continued our efforts to systematically revise fault representations in the CFM to be more compatible with the Qfault traces and the relocated earthquake catalogs. This involves redefining the interpolated patches of the CFM faults based on the CFM traces, and re-interpolating the 3D fault surfaces to be consistent with both surface and subsurface constraints. Our focus this past year was on major faults in the western Transverse Ranges (e.g., Pine Mountain, Little Pine, Santa Ynez) and the Los Angeles basin (Figure 1). This analysis resulted in much more highly detailed and segmented representations of the faults. Care was given to defining the termination and linkage of fault segments, and to generating subsurface representations that are compatible with earthquake locations. These more precise representations are designed to inform assessments of single and multi-segment earthquake scenarios based on geometric criteria (e.g., distance between fault segments), and to serve as a basis for more detailed fault representations in geodetic block model and dynamic rupture simulations. These improved fault representations, along with similar updates for other regions, will be incorporated into a new release of the CFM.

Figure 1: Perspective view of the western Transverse Ranges in the CFM. Faults representations highlighted in purple have been revised as part of this study using relocated earthquake catalogs (yellow) and detailed traces from the U.S. Geological Survey Fault and Fold database (red).

Finally, we also developed a new set of regular, quality controlled meshes for all of the

CFM fault representations to facilitate their use in a wide range of modeling and simulation studies. T-surfaces (Tsurfs) were chosen as the native format for CFM faults because they provide for accurate representations of complex, curviplanar surfaces that vary their geometries in depth or along strike. To date, however, many fault system models require evenly meshed tsurfs to represent faults (e.g., Thomas, 1993; Kuriyama and Mizuta, 1993). Native tsurfs in the CFM, however, generally do not satisfy these criteria as triangles are not equilateral and the mesh density varies considerably as a function of data constraints. Thus, this past year we developed a set of regularly meshed tsurfs for use in these applications. We generated regular meshes for each fault at a resolution of about 1500m (Figure 2) using Gocad’s finite element mesh workflow. Care was taken to ensure that the new meshes treat many fault junctions in ways (such as having common nodes) that help facilitate their use in fault system modeling applications. These remeshed fault representations will be made available to the SCEC Community through the SCEC Community Modeling Environment (CME).

Figure 2: View of the original (left) and remeshed (right) representations of the San Jacinto fault in the CFM.

CVM-H We also implemented a new representation of the Santa Maria Basin into the SCEC Community Velocity Model (CVM-H). This new basin model was developed with support of the USGS NEHRP program and included an improved representation of the sediment-basement interface that incorporates offsets of major fault systems in the region (Plesch et al., 2007). The velocity structure at shallow depths is characterized by velocities around 1900 m/s, with slow velocities reflecting thick Quaternary sediments. Small velocity (~150 m/s) inversions occur in the basin between 400 m and 1000 m depth, particularly in regions characterized by intrusive igneous rocks at or near the surface. At depth and along the basin edges the velocity structure that is characterized by a strong contrast between basin sediments and basement. Basin velocities range from 1900 to 4500 m/s, generally increase with depth. In contrast, basement velocities just beneath the basin range from 5000 to 5500 m/s. Velocities variability also increases with depth, yielding velocity inversions within the deepest parts of the basin (~3300 m depth to 3800 m depth) of up to ~500 m/s. This local model was embedded into the CVM-H 11.9 model in a manner that ensured consistency in horizon representations and smooth lateral velocity gradients between the new basin representation and the surrounding model. Before being made part of a new CVM-H model release, such local model improvements must go through a process that involves new tomographic inversions of crust and upper mantle structure to ensure the

consistency of model components. Thus, we released the modified CVM-H including the Santa Maria Basin updates as a parallel model track, making it available to users while we perform the iterations necessary to incorporate this into a formal CVM-H model release.

Figure 3: Contour map (500 m interval) of the basement surface. The bold black line is the outcrop of the surface. The red line is the hangingwall cut-off at the blind Orcutt reverse fault.

Intellectual Merit This project represents the primary effort to develop and maintain the SCEC Community Fault (CFM) and Velocity (CVM-H) models, which help define the inventory of active fault systems and the properties of the surrounding crust and upper mantle in southern California. These models are used in a wide range of source characterization, fault system modeling, earthquake simulation, and wave propagation studies. Broader Impacts The CFM and CVM-H models represent an important contribution to SCEC’s efforts to assess the seismic hazards in southern California through strong ground motion simulations and the USGS led Uniform California Earthquake Rupture Forecast, Version 3 (UCERF3). Publications Nicholson, Craig, Andreas Plesch, Christopher Sorlien, John Shaw, and Egill Hauksson, 2013, Updating the 3D fault set for the SCEC Community Fault Model (CFM-v4) and revising its associated fault database, SCEC Annual Meeting, Palm Springs, CA.

References Hauksson, E., W. Yang, and P. Shearer, 2011, Waveform Relocated Earthquake Catalog for Southern California (1981 to 2011), SCEC Annual Meeting, Palm Springs, CA. Jennings, C. W. (Compiler) (1994). Fault activity map of California and adjacent areas, California Department of Conservation, Division of Mines and Geology, Geologic Data Map No. 6, scale 1:750,000. Kuriyama, K., and Y. Mizuta (1993). Three-dimensional elastic analysis by the displacement discontinuity method with boundary division into triangular leaf elements, Int. J. Rock Mech. Min. Sci. & Geomech. Abstr. 30, 111–123. Lin, G., P. M. Shearer, and E. Hauksson, 2007, Applying a three-dimensional velocity model, waveform crosscorrelation, and cluster analysis to locate southern California seismicity from 1981 to 2005, J. Geophys. Res., v.112, n.B12, 14 pp, B12309, doi:10.1029/2007JB004986. Plesch, A., J. H. Shaw, C. Benson, W. A. Bryant, S. Carena, M. Cooke, J. Dolan, G. Fuis, E. Gath, L. Grant, E. Hauksson, T. Jordan, M. Kamerling, M. Legg, S. Lindvall, H. Magistrale, C. Nicholson, N. Niemi, M. Oskin, S. Perry, G. Planansky, T. Rockwell, P. Shearer, C. Sorlien, M. P. Süss, J. Suppe, J. Treiman, and R. Yeats, 2007, Community Fault Model (CFM) for Southern California, Bulletin of the Seismological Society of America, Vol. 97, No. 6, doi: 10.1785/012005021 Shearer, P., E. Hauksson, and G. Lin, Southern California hypocenter relocation with waveform cross correlation: Part 2. Results using source-specific station terms and cluster analysis, Bull. Seismol. Soc. Am., 95, 904-915, doi: 10.1785/0120040168, 2005. Thomas, A. (1993). POLY3D: A three-dimensional, polygonal element, displacement discontinuity boundary element computer program with applications to fractures, faults, and cavities in the Earths crust, Master’s Thesis, Stanford University, Stanford, California. Waldhauser, F. and D.P. Schaff (2008), Large-scale relocation of two decades of Northern California seismicity using cross-correlation and double-difference methods, J. Geophys. Res., 113, B08311, doi:10.1029/2007JB005479.