Most of the data were collected from November 13, 1993 to December 11, 1993. During the four weeks of continuous recording, stations recorded signals from several hundred teleseismic, regional, and local events. The Preliminary Determination of Epicenters (PDE) catalog produced by the USGS National Earthquake Information Center (NEIC) contains information for the approximately 150 teleseismic events that occurred during LARSE93 (Fig. 4; Table 5). The regional and teleseismic events include the December 4, 1993 aftershock (Mw=5.4) in Klamath Falls, Oregon as well as a number of intermediate magnitude earthquakes with epicenters in the Aleutian Island, Kamchatka, Kuril, mid-Atlantic Ridge, Solomon Island, Japan, Fiji Island, Peru, and Chile regions. According to catalog data supplied by the Southern California Earthquake Center Data Center (SCEC_DC), approximately 450 local events with ML greater than or equal to 2.0 occurred during the same time window (Fig. 5; Table 6). The local events include aftershocks of recent Southern California earthquakes.

The sensors recorded velocity as voltage. Since the sensors were different, the instrument responses were also different. The software packages (e.g., SAC routines produced by PASSCAL) which translate the data into ground velocity require the input of three parameters that define velocity sensitivity: the free period of the sensor, the observed generator constant, and the observed damping ratio (Rodgers et al., 1995). The values for these parameters were obtained by Aaron Martin at the Institute for Crustal Studies at the University of California at Santa Barbara, by calibrating each sensor after use in LARSE93. The resulting free period, generator constant, and damping ratio values for each sensor component are given in Table 7 (Table 7a: Vertical; 7b: North-South; 7c: East-West).

In general, stations in the San Gabriel Mountains and Mojave Desert display higher signal-to-noise ratios than stations in the Los Angeles basin (which includes the San Gabriel Valley). For most teleseismic events, the initial impulsive arrival is the direct P wave. Because all stations recorded the vertical component, the initial P wave is often the best recorded phase. Ten high-quality teleseismic record-section profiles are shown in Fig. 6. These plots contain the time-corrected, vertical-component velocity signals for events which occurred in the northwest Pacific, south Pacific, South America, and mid-Atlantic. Seismograms which contain high levels of noise and severe, uncorrected timing problems were removed. The largest gaps in these plots correspond to regions in the Los Angeles basin where no stations were deployed. The seismograms have been bandpass filtered for frequencies between 0.1 and 1.0 Hz. In some teleseismic sections, the initial P wave is followed by the larger-amplitude pP and sP phases. The quiet mountain stations show subtle variations in crustal phase amplitudes. The noiser Los Angeles basin stations are usually too noisy to obtain arrival times.

Profiles of time-corrected, vertical-component record sections from local and regional events are shown in Fig. 7. These figures show velocity records for events which occurred in Southern California less than 200 km from the array. They include events which occurred in the Mojave Desert, Los Angeles basin, and the San Gabriel and San Bernardino Mountains. As before, seismograms which contain high levels of noise and severe timing problems were removed from these figures and the largest gaps in these plots correspond to regions in the Los Angeles basin. The seismograms have been bandpass filtered for frequencies between 0.1 and 10.0 Hz. The major arrivals are the crustal phases Pg, Sg, and other crustal reverberations. Data, such as those shown in Figs. 6 and 7, were corrected for timing errors wherever possible.

P-wave travel-time residual curves for stations along the array have been determined for 17 teleseismic events. These events fell into several distinct back-azimuth ranges with distances between sources and receivers ranging from 30 degrees-90 degrees. P-wave travel-time residuals were determined for each station by subtracting one-dimensional Earth model IASP91 travel times (Kennett and Engdahl, 1991). Within each back-azimuth range, the resulting demeaned travel-time residual curves display consistent patterns. Residuals increase from negative values in the northernmost San Gabriel Valley and southern San Gabriel Mountain foothills to positive values across most of the central and northern San Gabriel Mountains, including the San Andreas fault (Fig. 8). The most drastic difference in residuals, occurring for raypaths from the northwest (Kamchatka, Unimak Island, and Alaska), is about one second. The residual curves display only small variations for different back azimuths and incidence angles, but show almost no spatial (parallax) shift of residual peaks and troughs, indicating that the source of the large residual difference is most likely shallow. The patterns of residuals suggest that a sharp gradient in shallow velocities is required between the Los Angeles basin and the San Gabriel Mountains over a horizontal distance of less than 50 km. A relatively shallow, low-velocity anomaly is required to explain the larger San Gabriel Mountain station residuals.

Previous Page Table of Contents Next Page