Modeling of Geodetic Crustal Motion Velocities in Southern California: Undergraduate Research

With funding from the National Science Foundation's Opportunities for Enhancing Diversity in the Geosciences, we have undertaken a project with two primary goals: (1) to introduce undergraduate students and K-14 educators to research in geology/geophysics, and (2) to use GPS to monitor deformation across the boundary between the Pacific and North American plates in southern California, and to model the slip on specific faults that could be responsible for that deformation. Starting in July 2002, we collected campaign-style GPS data twice a year from 13 sites along a line across the San Andreas and San Jacinto faults from Norco through San Bernardino to Lucerne Valley. We are also modeling data from the SCEC Crustal Deformation Velocity Map 2.0 [http://www.scecdc.scec.org/group_e/release.v2/]. Our initial approach has been to use a one-dimensional model of dislocations in an elastic half-space. We are studying the portion of the plate boundary from San Bernardino southward to the U.S.-Mexico border. We have divided this region into seven transects that are perpendicular to the plate boundary. We used a spreadsheet macro to systematically model a range of slip rates and locking depths for each fault. Out of hundreds or thousands of possible combinations for each transect, we sorted the models according to their goodness of fit, using the sum of the squares of the residuals as a criterion. We are also beginning to use the program Simplex (G. Lyzenga. J. Parker) to model the velocity data from all transects simultaneously. This will allow us to take into account the complex fault geometry of the region. Our preliminary results from the one-dimensional modeling suggest that the best-fitting slip rate of the San Andreas fault is 26 mm/yr for the section from Indio to Durmid. However, slip rates in the range of 20-30 mm/yr also fit the geodetic data relatively well. Slip rates of 15 or 35 mm/yr do not fit well. For the San Jacinto fault, the best-fitting slip rate is 13 mm/yr for the section from Anza to Borrego Mountain and 15 mm/yr for the section farther south, which ruptured in 1968. However, slip rates within the range 10-20 mm/yr also fit these data relatively well. The best-fitting rate for the Superstition Hills fault is 15 mm/yr, with rates of 10-15 mm/yr fitting reasonably well, whereas a rate of 20 mm/yr does not fit well. The best-fitting slip rate for the Elsinore fault was 5 mm/yr for the section near Julian, but rates ranging from 2-8 mm/yr also fit relatively well. For the southernmost section of the Elsinore fault, from Agua Caliente Springs to the Coyote Mountains, the best-fitting slip rate was 2 mm/yr, but rates from 2-4 mm/yr fit relatively well. The best-fitting rate for the Laguna Salada fault was 4 mm/yr, with rates from 2-6 mm/yr also fitting fairly well. These results generally agree with geologic estimates of the Holocene slip rates for these faults. There has been considerable debate as to whether the San Andreas and San Jacinto faults contribute approximately equally to the plate boundary deformation in southern California or whether the San Andreas fault contributes substantially more than the San Jacinto fault. Our preliminary results suggest that the San Andreas fault most likely is contributing more to the plate boundary deformation than is the San Jacinto fault, but we cannot rule out the possibility that they are equal contributors.