Annual Report, 1997

Uncertainty in Earthquake Source Characterization for the Los Angeles Basin

Bill Foxall

A major source of uncertainty in seismic hazard estimates for the Los Angeles Basin is uncertainty in the geometry, and in some cases the existence, of sign)ficant active faults. This epistemic uncertainty stems largely from the fundamentally different tectonic deformation models -specifically, "thin-skinned" and "thick-skinned" models -- that have been proposed for the region. Thin-skinned models, in addition to containing major active blind thrust ramps, imply sign)ficantly different sub-surface geometries for some major surface faults from those in the thick-skinned model. These uncertainties are not systematically incorporated in the CDMG/USGS earthquake source model, which is the geological source model used in SCEC Phase III. The objective of my 1997 SCEC project, in collaboration with Norm Abrahamson and Allin Cornell, was to develop a set of internally consistent earthquake source characterizations for the LA Basin and the surrounding region that captures the range of viable alternative tectonic interpretations. This set of alternative characterizations will enable the source uncertainties to be propagated through probabilistic seismic hazard analyses to obtain rigorous estimates of uncertainties in the hazard using a "logic tree" approach (see, for example, Senior Seismic Hazards Analysis Committee [SSHAC], 1997). One of the most important steps in constructing the logic tree is assigning appropriate relative weights to the alternative source descriptions, based on careful consideration of the weight of evidence for each of the alternatives.

Development of the alternative source characterizations was centered on a workshop held at USC on September 18-19, 1997, which 54 SCEC scientists attended. The workshop was structured, following a simple form of the procedures recommended by SSHAC (1997), to: (1) Ensure that development of the source characterizations is based on the latest state of knowledge; (2) through the exchange of current information and directed debate of the evidence for against each of the alternative tectonic interpretations, to lay the foundation for determining relative weights for the source characterizations; and (3) ensure that the source characterizations and associated weights represent the consensus view of the SCEC community. The workshop included 18 invited talks given by proponents of the alternative tectonic models and by other researchers having recent results that have direct bearing on the models. Three knowledgeable "evaluators", Kathryn Hanson (Geomatrix), Lee Silver (Caltech) and Bob Smith (Univ. Utah), have the role of evaluating the tectonic models and source characterizations, and, in conjunction with the "Technical Integrator" team (Abrahamson, Cornell and Foxall), of weighting the alternatives. A set of "strawmen" source characterizations developed before the workshop, based on a thorough review of current data and interpretations, served to focus the discussion. Hazard estimates resulting from the strawmen were compared to examine the sensitivity of the hazard to differences among the source characterizations.

The workshop went a long way towards achieving the above objectives. However, we did not reach our goal of completing the logic tree and assigning a preliminary set of weights on which the workshop participants could give their feedback. This was because, in the time available, we were unable to incorporate all of the new information that was presented, particularly that relating to new regional detachment and crustal shortening models.

A sign)ficant outcome with respect to the central thin- versus thick-skinned issue was widespread (although not unanimous) agreement that evidence presented by Karl
Mueller strongly suggests that the Compton-Los Alamitos thrust of Shaw and Suppe (1996) has been inactive since at least 1 5ka. The existence and activity of the Compton thrust also has a direct bearing on the geometry and rupture characteristics of the Palos Verdes fault. This is the most complex source characterization issue in the southwestern Basin, since in the most plausible of the alternative models proposed by Shaw and Suppe the thrust offsets the Palos Verdes (and Newport-lnglewood) fault, allowing the possibility of vertical segmentation of the fault. There was similar agreement that, based on evidence summarized by Jim Dolan, the Santa Monica Mountains thrust is also either inactive or slips at a much lower rate (<0.5 mm/yr) than originally proposed by Davies et al. (1989).

The workshop also focused on the mechanics of crustal shortening in and surrounding the N and NE Basin, including assimilation to the NW of right-lateral displacement on the Elsinore-Whittier system. Data on Holocene deformatiopn rates of fold and fault structures in this area are still sparse. Given geodetic results showing a concentration of deformation in the N Basin and the apparently low slip rate on the central Sierra Madre fault, our poor understanding of this area is an important source of uncertainty in source characterization. The "extrusion" model of Walls et al. ( 1997) accounts for the geodetically observed rates by postulating slip rates in the range 0.5-3 mm/yr on prominent but poorly studied faults, including the Chino, San Jose and Verdugo, but these rates remain to be validated. Researchers studying deformation along the Elysian Park anticlinorium (Yeats, Oskin, Shaw) agreed that sources of active folding NW and south of the Whittier fault are most likely short (10-20 km), en echelon, steeply-dipping reverse fault segments. These faults, the sources of moderate events like the 1987 Whittier Narrows earthquake, are probably best modeled as a zone of distributed faulting. Other major uncertainty issues related to shortening in the N. Basin that are being incorporated into the source characterization include:

Implications for the activity and existence of the Elysian Park thrust ramp (and central Basin decollement) of Shaw and Suppe (1996) of the apparent inactivity of the Compton-Los Alamitos thrust.

Viability and implications of the regional decollement and associated low-angle splay faults under the San Gabriel Mountains and Valley interpreted by Ryberg and Fuis (1997) from LARSE seismic reflection Line 1. This interpretation proposes thinskinned deformation under the San Gabriel Valley, and presumably it can be viewed as new evidence for thin-skinned deformation under the NE LA Basin also. However, the splays, as interpreted by Ryberg and Fuis, would intersect a steeply-dipping Sierra Madre fault at shallow depths, making it difficult to reconcile this interpretation with surface geological evidence for the continuing activity of the fault. The single 2D reflection profile does not allow estimates of the dips or lateral extent of the proposed structures.

Viability and implications of the new seismotectonic model for the Los Angeles region proposed by Seeber and Geiser (Ms in preparation, 1997). This model postulates that a mid crustal ramp is forming under the LA Basin, which the authors associate with the reflection interpretation of Ryberg and Fuis, but also supports the deep, high-angle faults of current thick-skinned models.

I am presently completing development of the alternative source characterizations and the logic tree, based upon the workshop results and further discussions with the evaluator panel and with SCEC scientists. This work includes developing "limbs" on the tree to deal with the new issues uncovered at the workshop. The complete source characterization model will be sent to the evaluator panel, accompanied by summaries of the main issues and interpretations, and the evaluators will individually assign their weights to
the alternatives. The evaluator panel and Tl team will then meet to discuss the model and weights in preparation for a "mini" workshop involving 10-12 key SCEC researchers. At this workshop we will elicit the feedback on the source model and weights that will enable us to finalize the source characterization.

REFERENCES

Davis, T.L., J. Namson, and R.F. Yerkes, A cross section of the Los Angeles area: Seismically active fold and thrust belt, the 1987 Whittier Narrows earthquake, and earthquake hazard, Jour. Geophys. Res., 94, 9644-9664, 1989

Ryberg, T, and G. S. Fuis, The San Gabriel Mountains bright reflective zone: Possible evidence of mid-crustal thrust faulting, submitted to Tectonophysics, 1997.

Senior Seismic Hazards Analysis Committee, Recommendations for probabilistic seismic hazard analysis: Guidance on uncertatinty and the use of experts, NUREG/CR6372, Nuclear Regulatory Commission, Washington, 256 p., 1997.

Shaw, J.H., and J. Suppe, Earthquake hazards of active blind-thrust faults under the central Los Angeles basin, California, Jour. Geophys. Res., 101, 8623-8642, 1996.

Walls, C., T. Rockwell, K. Mueller, Y. Bock, S. Williams, J. Pfanner, J. Dolan, and P. Fang, Escape from L.A.: Extrusion tectonics in southern California and implications for seismic risk; geology and GPS agree, submitted to Nature, 1997.