REGIONAL SEISMICITY AND THE SAN ANDREAS TRANSFORM BOUNDARY

See http://plate-tectonic.narod.ru/califcrustseismphotoalbum.html

REGIONAL SEISMICITY AND THE SAN ANDREAS TRANSFORM BOUNDARY
Dickinson (1981), among others, emphasized that the San Andreas fault system and the San Andreas transform boundary are not strictly equivalent structures. The latter is more general, incorporating, for example, the concept of temporal evolution of Pacific-North American plate interaction and the recognition that the faults accommodating most of the plate motion have changed over time. In this section, we emphasize that, although great earthquakes along the San Andreas fault system currently accommodate most of the relative plate motion, plate interactions along the transform boundary influence deformation of the brittle crust over a much broader region.

The breadth of the seismicity pattern in California and western Nevada (fig. 5.4) suggests the lateral extent of deformation associated with the San Andreas transform boundary. Indeed, it corresponds closely to the zone of distributed shear between those plates as interpreted by Ward (1988) from more than 5 years of very long baseline interferometry (VLBI) observations at 20 Western United States stations from 1982 through 1987. All M��1.5 events recorded by the telemetered seismic networks in figure 5.2 during the 7-yr interval 1980-86, as well as events recorded by adjacent telemetered networks in Nevada (see Rogers and others, in press). Although details within this seismicity pattern fluctuate from year to year, the broad aspects of the pattern have remained stable for the entire historical record of earthquake occurrence in California (see chap. 6; Ellsworth and others, 1981; Hill and others, in press; Hutton and others, in press).
In outline, the seismicity pattern for California and western Nevada forms a hollow ellipse with its long axis nearly coincident with the transform boundary. This pattern is widest across central California, where it approaches nearly a third of the 1,100-km length of the transform boundary, from the Mendocino triple junction in the north to the head of the Gulf of California at the south. Extended alignments of epicenters within this pattern suggest a coarse structural fabric linking the broad distribution of earthquakes to the transform boundary. Seismicity along the San Andreas fault system itself stands out as a series of subparallel, northwest-trending lineations extending the length of coastal California. The alignment of epicenter clusters along the east side of the Sierra Nevada in eastern California branches northward from the south end of the San Andreas fault system in the Salton Trough only to bend back toward the north terminus of the San Andreas fault system at the Mendocino triple junction in northern California. The Sierra N evada-Great Valley and western Mojave Desert blocks form a broad quiescent region between the San Andreas and eastern California seismicity bands. In southern California, pronounced transverse seismicity belts coincident with the southern margin of the Sierra Nevada and Transverse Ranges, respectively, span this otherwise-quiescent region. A weaker, more diffuse seismicity belt near lat 37�� N. spans the Sierra Nevada-Great Valley block in central California, forming a visual, if not structural, link between the San Andreas fault system and the dense cluster of epicenters in eastern California. This major cluster in the eastern Sierra Nevada represents an episode of intense earthquake activity in Long Valley caldera and vicinity that began in 1978 and has persisted to the present (Van Wormer and Ryall, 1980; Hill and others, 1985b).

The displacement pattern associated with earthquakes throughout California and western Nevada is broadly consistent with deformation under a coherent, regional stress field dominated by plate-boundary interaction along a northwest-striking, dextral transform boundary (Hill, 1982). Strike-slip focal mechanisms with right-lateral slip on northwest- to north-north west-striking planes, for example, are common through most of the region. Regional variations within this pattern include a tendency toward normal slip on northerly-striking planes along the western margin of the extensional Basin and Range province in eastern California, and toward reverse slip on easterly-striking planes in the Transverse Ranges of southern California. Compressional deformation perpendicular to the San Andreas fault within the Coast Ranges, however, represents an important deviation from this regional pattern.
SEISMICITY ALONG THE SAN ANDREAS FAULT SYSTEM
Sections of the San Andreas fault system stand out on seismicity maps as a network of northward-branching alignments of epicenters through the central and northern Coast Ranges and as subparallel lineations of clusters of epicenters that branch northward from the south end of the Imperial fault toward the Transverse Ranges in southern California. One of the most remarkable aspects of the seismicity pattern associated with the fault system, however, is the nearly complete absence of earthquake activity down to even the smallest magnitudes (M��1.5) along those sections of the fault that have ruptured with the largest historical earthquakes, the great (M��8) 1857 and 1906 earthquakes. The southernmost section of the San Andreas fault, from Indio to the Salton Sea, also lacks microseismicity, although no large earthquake has ruptured this section in the past 200 yr. These quiescent ("locked") segments of the fault stand in sharp contrast to the segments marked by persistent linear concentrations of small to moderate earthquakes.

This dual expression of the fault system evident on current seismicity maps apparently reflects fundamental differences in the long-term behavior of the respective segments. In particular, seismic activity along the "locked" segments of the main trace of the San Andreas fault may be limited to the recurrence of major earthquakes at intervals of 100 to 300 yr accompanied by immediate fore shock and aftershock sequences, and these segments may remain quiescent for most of the interevent time associated with the cycle between great earthquakes (Ellsworth and others, 1981). In contrast, those segments with persistent microearthquake activity probably seldom, if ever, rupture with great earthquakes, although they may be capable of generating earthquakes as large as M��6.

Aseismic creep also characterizes and is largely confined to those fault segments along the San Andreas fault system that show persistent micro earthquake activity (Wesson and others, 1973; Schulz and others, 1982). Creep is most pronounced along the central California segments of the fault system, where average creep rates match the long-term displacement rates of 32 to 34 mm/yr. Louie and others (1985) documented creep along sections of the seismically active fault segments in the Salton Trough, and Astiz and Allen (1983) documented creep along a section of the Garlock fault that is marked by microearthquake activity. The creep rates in these two areas, however, are more than an order of magnitude less than the long-term deformation rates.
In the following subsections, we consider the 1980-86 seismicity along and adjacent to the major sections of the San Andreas fault system in more detail. We begin with the Mendocino triple junction in the north and move southward, generalizing slightly Allen''s (1968) subdivision of the fault system into four major sections of contrasting seismic behavior: (1) the quiescent 1906 break and subparallel branches, (2) branches forming the central California active (creeping) section, (3) the quiescent 1857 break, and (4) branches forming the southern California active section south of the Transverse Ranges.