times (Luyendyk et al., 1980). Although these tectonic rotations are not known to be going on today, they have contributed to the heterogeneity of the bedrock, and so affect strain patterns now under development.
The San Fernando earthquake of February 9, 1971, was located within the Transverse Ranges on a fault dipping northward beneath the San Gabriel Mountains (Grantz et al., 1971; Oakeshott, 1975a). The fault was one of many known to have broken young deposits and with fault landforms along it, such as the Raymond, Cucamonga, Sierra Madre, San Andreas, and San Jacinto. But it was not known beforehand that this fault, nor the segment where the ground was broken, was most likely to be the site of a disruptive earthquake. Most of the active faults in this region, and elsewhere around the Los Angeles metropolitan region, have now been demarcated, but tectonic understanding has not yet progressed far enough for geoscientists to pinpoint with confidence which fault is most likely to break next. Faults known to be recently active are most likely to rupture again, but more research to learn which are the most dangerous is required.
The geologic situation in the vicinity of the San Fernando earthquake serves to illustrate problems in appraising future earthquake hazards. Young and datable sedimentary deposits, such as alluvium and terrace deposits, are needed to ascertain the time of movement of tectonic features that disrupt them. Faults are commonly known to be active where they cut deposits only a few thousand years old. Faults are suspected of being active where the youngest strata or geomorphic features cut are several tens of thousand or a hundred thousand years old. In the absence of datable young beds that are clearly cut by faults or warped and distorted by folding, there is at present no satisfactory way to determine whether a fault should be labeled as active and is likely to break again in the near future. Many faults are mappable in older rocks, but they may have been formed by geologically ancient deformations under tectonic regimes and stress situations long abandoned.
The difficulties are illustrated in Figure 1.2, a cross section through the region of the San Fernando earthquake and extending northeastward into ancient rocks, including those of Precambrian age. Along the edge of the valley, transected young deposits prove that the faulting is young, but faults within the higher mountains underlain only by basement rocks may have been active at intervals any time since the rocks were consolidated over a billion years ago. These ancient rocks have also been tilted and folded as shown by layering in igneous rocks formed by settling of crystals on the flat floor of a magma chamber in Precambrian time. Whether the tilting of the layering, which was originally nearly hori
zontal, took place not long after crystallization or but recently cannot be determined from examining local outcrops. Regional understanding of the tectonic history, however, may provide a basis for useful inferences.
That the basement rocks in this region have been folded and displaced recently is also shown by the shape of the unconformity between young deposits and the basement (Figure 1.2; Oakeshott, 1975b). The syncline beneath the San Fernando earthquake fault is accompanied by an anticline above it, even though along the plane of the cross section there are no outcrops of the unconformity for about 8 km northward. The folded unconformity, dated as young because the deposits just above it are geologically young, proves that the latest deformation of the basement block took place in recent times. But this interpretation does not preclude many other deformational events, including several episodes of folding and faulting in the geologically remote past. During the earthquake the San Gabriel Fault did not rupture the ground surface so far as is known; this major ancient fault was ignored in the modern stress regime. Old faults may well become reactivated if they are oriented so that the resolved shear stress on them is sufficient for slip within the present-day tectonic situation. For this reason it is desirable for investigators to learn as much as they can concerning the style of deformation so that they can advance reasonable inferences concerning which faults are dangerous. Models of crustal structure and behavior that are well substantiated may be helpful in this task.