Part I: A Contribution to the Study of the Seismicity of Southern California. Part II: Inversion of Phase Times for Hypocenters and Shallow Crustal Velocities

Citation

Hileman, James Alan (1978) Part I: A Contribution to the Study of the Seismicity of Southern California. Part II: Inversion of Phase Times for Hypocenters and Shallow Crustal Velocities. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/6805-xg34. https://resolver.caltech.edu/CaltechTHESIS:09202024-225628727

Abstract

Southern California seismicity data for the period 1932 through 1975 are summarized in a series of epicenter maps. These maps show the seismic activity for one-year periods, for five-year periods, and as summaries for all earthquakes with magnitudes equal to or greater than 4, 5, and 6 respectively. Enlarged epicenter maps are given for the Los Angeles area and for the more significant aftershock sequences in the region.

A regional value of b = 1.00 ± .02 (log N = a - bM) is found for forty years of data. For several smaller areas in the region, b-values have a range of 0.76 - 1.00. Temporal variations of b-values (maximum likelihood estimates) for the various areas studied do not show any strong correlation with the occurrence of large earthquakes. A slight increase of the regional b-value after the 1952 Kern County earthquake is suggested.

Seismicity in the Imperial Valley indicates that several faults there are susceptible to triggering, i.e. they are loosely coupled to motions of neighboring faults. Some earthquake swarms indicate this ease of triggering. A survey of the high level of swarm activity in the Imperial Valley is given. Two unusual aftershock sequences with periodic activity are described because the periodicity suggests sensi­tivity to some triggering phenomenon.

Some seismicity in Southern California seems to be aligned in weakly defined zones that are transverse to the general tectonic fabric. These zones are thought to reflect conditions in the lower crust or uppermost mantle. The cause of these zones is unknown, but their trends are similar to those for the early Paleozoic continental boundary and to a recently discovered upper-mantle velocity anomaly.

Depth is usually the least-certain hypocenter parameter because it depends critically upon the accuracy of the velocity model. With enough arrival-time data, velocity estimation is feasible in addition to the usual hypocenter determinations. Linear least-squares inversion theory is adapted for the simultaneous determination of hypocenters and local velocity structure. A maximum likelihood formulation is used so that the data are weighted according to their estimated vari­ances. A tradeoff parameter controls the relative importance of the RMS error and the amount by which the model is changed at each itera­tion. The inversion is also stabilized by specifying the allowed variances of each of the model parameters.

Arrival times for a set of 20 earthquakes in the central Mojave Desert were inverted to improve the local velocity model. Each of the trials indicated that shallow crustal velocities in the vicinity of Galway Lake are somewhat lower than those of the usual velocity models. The velocities were not strongly constrained by this data set. This study points out the need for several seismographic stations placed within an aftershock area for best control of velocity estimates.

Thesis Files