ASSOCIATED FAULTS [c5, p133-136]

Although the southernmost section of the San Andreas fault is almost completely aseismic, associated subparallel faults are extremely active. These faults are marked by the three bold north-south- to northwest-trending alignments of epicenters that dominate the seismicity pattern within the San Andreas fault system south of the Transverse Ranges (Figure 5.10A), from east to west: (1) the Brawley seismic zone (Johnson, 1979), defined by a dense, spindle-shaped cluster of epicenters connecting the north end of the Imperial fault and the south end of the Indio segment of the San Andreas fault; (2) the northwestward alignment of densely clustered epicenters along the San Jacinto fault zone, which appears to branch from the northern section of the Imperial fault; and (3) the northwestward alignment of more diffusely clustered epicenters along the Elsinore fault, which appears to branch from somewhere near the south end of the Imperial fault.

The Brawley seismic zone and the cluster of epicenters at the south end of the Imperial fault (coincident with the Cerro Prieto volcanic-geothermal field in Mexico) represent the two northernmost in the series of small spreading centers offset by right-lateral transform faults that characterize oblique spreading in the Gulf of California (Lomnitz and others, 1970; Johnson and Hill, 1982). The Imperial fault itself, which is marked by a scattered alignment of epicenters, serves as the transform fault between these two small spreading centers. The M=7.1 El Centro earthquake ruptured the entire length of the Imperial fault in 1940, and the M=6.6 Imperial Valley earthquake of 1979 ruptured the north two-thirds of the fault; intensity data suggest that moderate earthquakes (5.5<M<6.3) in 1906, 1915, 1917, and 1927 may also have been located on the Imperial fault (Johnson and Hill, 1982). Most of the aftershocks associated with the 1979 Imperial Valley earthquake were concentrated in the south half of the Brawley seismic zone, which was first recognized because of the many earthquake swarms it produced from 1973 through mid-1979 (Hill and others, 1975; Johnson 1979; Johnson and Hutton, 1982). Many of the individual swarm sequences, as well as individual clusters of events in the aftershock sequence, defined lineations transverse to the strike of the Imperial fault and the long axis of the Brawley seismic zone. Most earthquakes within the Brawley seismic zone have strike-slip focal mechanisms; thus, kinematically, these transverse lineations represent conjugate structures to the dominant north-northwestward trend of the Imperial-Brawley fault system.

Irregular clusters of epicenters mark the San Jacinto fault zone, which runs along the southwest base of the Santa Rosa and San Jacinto Mountains. These clusters tend to be concentrated near bends and junctions within the complex set of multiple fault strands that form the surface expression of this fault zone. In several places, particularly within the southern and northern sections of the fault zone, epicenters define linear concentrations that tend to be closely aligned with mapped fault traces. The San Jacinto fault zone has produced at least 10 earthquakes of M=6.0-6.6 since 1890, the most recent of which were the M=6.2 earthquake of 1954, the M=6.6 Borrego Mountain earthquake of 1968, and the M=6.6 Superstition Hills earthquake of 1987. Thatcher and others (1975) pointed out that this series of historical M>6 earthquakes along the San Jacinto fault zone has left two seismic gaps: one along the northern 40 km of the fault, and the other along a 20-km-long stretch of the central section of the fault zone (the Anza gap). The Anza gap shows up in figure 5.10A as a relatively quiescent stretch of the fault zone between two dense clusters, with a third cluster located off the fault zone some 20 km southwest of the gap (see Fletcher and others, 1987; Sanders and Kanamori, 1984).

The Elsinore fault zone is defined not so much by a coincident alignment of epicenters as by the loci of western end points for clusters of epicenters elongate northeastward between the Elsinore and San Jacinto fault zones. This pattern is most pronounced along the southeast half of the fault; the northwest half, which defines the northeast scarp of the Elsinore Mountains, is marked by scattered clusters of epicenters. As the Elsinore fault enters the Los Angeles Basin to the north, it splays into the Whittier and Chino faults. Historical seismicity levels are considerably lower along the Elsinore fault than either the San Jacinto fault zone or the Imperial fault/Brawley seismic zone. The largest historical earthquake on the Elsinore fault was an M=6 event in 1910 in the central section. The Whittier Narrows earthquake (ML=5.9) of 1987, which caused over $300 million in damage, was located at the north end of the Elsinore-Whittier fault. Because its mechanism was thrust faulting on an east-west-striking plane with a shallow dip, however, it does not appear to be simply related to the Elsinore system.

Seismicity in the relatively quiescent southwestern corner of California between the Elsinore fault and the coast shows up in figure 5.10A as small, sparsely scattered clusters of epicenters. Activity picks up again, however, in the vicinity of the major northwest-striking faults along the coast (the Rose Canyon fault through San Diego and the Newport-Inglewood and Palos Verdes faults along the western margin of the Los Angeles Basin). Except for weak alignments along the Newport-Inglewood fault, which ruptured with an M=6.3 earthquake in 1933 (Richter, 1958), the seismicity patterns associated with these faults show little tendency to align along mapped fault traces.

The cross sections in figure 5.10B emphasize that, except in the immediate vicinity of the Salton Sea, maximum focal depths associated with earthquakes aligned along the principal branches of the southern section of the San Andreas fault system are systematically deeper than those aligned along its central and northern sections. Maximum focal depths, for example, decrease from 15 to 18 km beneath the central section of the Imperial fault near the United States-Mexican border to less than 10 km beneath the north end of the Brawley seismic zone at the southeast tip of the Salton Sea ( cross sec. Q-Q', Figure 5. 10B). The focal depths associated with earthquakes along the Coyote Creek, Superstition Hills, and Superstition Mountain faults forming the southwestern section of the San Jacinto fault zone adjacent to the Salton Sea are concentrated in the upper 10 km of the crust ( cross sec. P-P', Figure 5.10B).

Those segments of the San Andreas fault system in southern California with maximum focal depths shallower than 12 to 15 km are also those that show evidence of aseismic creep (see Louie and others, 1985). Indeed, actively creeping segments of the San Andreas fault system throughout California seem to be confined to those along which microearthquakes are concentrated in the shallow crust (focal depths of less than 12-15 km).

Moving northwestward along the San Jacinto fault zone, the base of the seismogenic crust deepens systematically to a maximum of 20 km beneath the stretch adjacent to San Jacinto Mountain (which at 3,293 m, is the second highest point in southern California) midway along the fault zone ( cross sec. P-P', Figure 5.10B). The base of the seismogenic crust maintains this 20-km depth farther northwestward along the fault zone to its junction with the Banning fault just south of San Bernardino ( Figure 5.10A), beyond which it begins to shallow again. Note, in particular, that earthquakes tend to be concentrated between 10- and 20-km depth beneath the San Jacinto fault zone, leaving the upper 10 km of the crust relatively quiescent along the middle stretch of the fault zone. The dense knot of hypocenters in the upper 5 km of the crust midway along cross section P-P' corresponds to the cluster of epicenters 15 km southwest of the fault zone near the Anza gap ( Figure 5.10A). The Anza gap itself shows up between A=120 and 140 km in cross section P-P' as a quiescent zone below and southeast of the shallow cluster of hypocenters (Fletcher and others, 1987; Sanders, 1987). The distribution of hypocenters beneath the Elsinore fault zone (cross sec. R-R', Figure 5.10B) is in many ways similar to that beneath the San Jacinto fault zone. Maximum focal depths increase northwestward from 12-15 km at the southeast end of the fault near the United States-Mexican border to about 20 km midway along the fault zone (generally coincident with the highest topography in this section of the Peninsular Ranges) and then gradually decrease farther northwestward toward the Los Angeles Basin. Maximum focal depths show evidence of increasing again at the northwest end of the fault as it approaches the Transverse Ranges and branches into the Whittier and Chino faults. The hypocenters along the south half of the Elsinore fault also tend to concentrate in the lower 10 km of the seismogenic crust, although this pattern is not as well defined in the diffuse seismicity of the Elsinore fault zone as in the dense clustering along the San Jacinto fault zone.