of critical engineering structures. But they also reveal much concerning the nature of crustal deformation and contribute to our understanding of the strength of the crust and the way structural flaws or ancient weak zones influence where faulting and folding take place. Local studies reveal the style, rates, episodicity, or clustering of deformational events. They document the style and distribution of strain features such as faults, folds, warpings, rotations, translations, and earthquakes. This information needs then to be dovetailed with regional investigations focused on arriving at better plate-tectonic generalizations. The latter research aids in arriving at improved geologic models, and these in turn will aid in understanding what is controlling deformation at local sites. To the extent that the models are applicable they will aid in extrapolating concepts to regions where engineering structures are needed and planned but where direct information on active deformation is missing because of the piecemeal and incomplete nature of the pertinent geologic record.
The question arises: How far back into the geologic past are deformational data useful in understanding the ongoing flexings and breakings of the crust that affect human enterprises? A first answer to this question is: as far back as the deformational style and strain pattern are essentially the same as those operating today. Except within limited areas, the plate-tectonic control of deformation has been the same during the past million years or so. The limited areas where strain patterns and styles of deformation have changed significantly are at plate boundaries, and especially at junctures such as near the Mendocino triple junction. Locally within major plates, as along braided fault zones, deformation styles evolve through time but across limited areas.
In general, the active-tectonic realms correspond to the physiographic provinces of the western conterminous United States because tectonic activity is primarily responsible for mountains and valleys, the shape of coastlines, and the boundaries of regions such as the Basin and Range province. An active tectonic realm is, therefore, defined as a region where the tectonic deformation in progress at present and for the past half million years or so has the same style and pattern. This means that the orientation of folds and faults undergoing growth and the locations and type of earthquakes and volcanic centers are nearly the same throughout the realm. Several maps have already been published that show features such as earthquake distribution, stress patterns, and elevation changes (e.g., Buchanan-Banks et al., 1978; Zoback and Zoback, 1980, 1981; Sbar, 1982; Gable and Hatton, 1983). The boundaries between some realms are sharp, but most are transitional. In addition, the concept involves scale. In working with an area the size of a city or county, small difference may be significant and may warrant separation into different subrealms. When dealing with a region as large as the three Pacific Coast states and adjacent portions of Baja California and British Columbia, however, generalizations are appropriate and local differences are smoothed.
In Figure 1.1, only the main provinces are demarcated. They are described briefly from south to north because of the convenience of picturing the sliding of western North America obliquely away from the divergent plate boundary in the Gulf of California. The main splintered boundary is the San Andreas transform system, upon which most of this sliding takes place. The Basin and Range physiographic province is primarily a broad region of stretching in this scheme and is the region north of the Mendocino triple junction where oblique plate convergence is taking place. The floor of the Pacific Ocean west of northernmost California, Oregon, and Washington is both moving relatively northwestward with respect to the continent (Pacific plate) and also eastward beneath it (Juan de Fuca plate). All these movements are relative, however, with respect to the North American lithospheric plate, which is by no means fixed. For example, as the Atlantic Ocean widens, North America may be viewed as moving relatively westward.
The peninsula of Baja California, as it moves northwestward, is in the process of being rifted from the mainland of Mexico, thereby opening the Gulf of California. The crust flooring the Gulf is expanding as new seafloor is formed at depth along a series of spreading segments defined by a pattern of oblique transform faults (Figure 1.1). The peninsula broke away from the Mexican mainland on the east about 4 million years (m.y.) ago (Larson et al., 1968; Moore and Buffington, 1968), and the spreading process is still continuing. This movement pattern results in several superimposed styles at a local scale within the region: strike-slip displacements near the major transform faults, sagging and warping over pull-apart basins where the rift floor is stretching, and downslope displacements along the margins of the Gulf of California where high-standing terrain of the old continent is relatively unsupported at the edge of the rift.
The Salton Trough lies at the northwestern end of the