Because many major population centers are located near active fault zones, such as the San Andreas, millions of people have suffered personal and economic losses as a result of destructive earthquakes, and even more have experienced earthquake motions. Not surprisingly, some people believe that, when the “Big One” hits, California will suddenly “break off” and “fall into the Pacific,” or that the Earth will “open up” along the fault and “swallow” people, cars, and houses. Such beliefs have no scientific basis whatsoever. Although ground slippage commonly takes place in a large earthquake, the Earth will not open up. Nor will California fall into the sea, because the fault zone only extends about 15 km deep, which is only about a quarter of the thickness of the continental crust. Furthermore, California is composed of continental crust, whose relatively low density keeps it riding high, like an iceberg above the ocean.
Aerial view, looking north toward San Francisco, of Crystal Springs Reservoir, which follows the San Andreas fault zone. (Photograph by Robert E. Wallace, USGS.)
Like all transform plate boundaries, the San Andreas is a strike-slip fault, movement along which is dominantly horizontal. Specifically, the San Andreas fault zone separates the Pacific and North American Plates, which are slowly grinding past each other in a roughly north-south direction. The Pacific Plate (western side of the fault) is moving horizontally in a northerly direction relative to the North American Plate (eastern side of the fault). Evidence of the sideways shift of these two landmasses can be found all along the fault zone, as seen from the differences in topography, geologic structures, and, sometimes, vegetation of the terrain from one side of the fault to the other. For example, the San Andreas runs directly along Crystal Springs Reservoir on the San Francisco Peninsula. Topographically, this reservoir fills a long, straight, narrow valley that was formed by erosion of the easily erodible rocks mashed within the fault zone.
Movement along the San Andreas can occur either in sudden jolts or in a slow, steady motion called creep. Fault segments that are actively creeping experience many small to moderate earthquakes that cause little or no damage. These creeping segments are separated by segments of infrequent earthquake activity (called seismic gaps), areas that are stuck or locked in place within the fault zone. Locked segments of the fault store a tremendous amount of energy that can build up for decades, or even centuries, before being unleashed in devastating earthquakes. For example, the Great San Francisco Earthquake (8.3-magnitude) in 1906 ruptured along a previously locked 430 km-long segment of the San Andreas, extending from Cape Men-docino south to San Juan Bautista.
Map of the San Andreas and a few of the other faults in California, segments of which display different behavior: locked or creeping (see text). (Simplified from USGS Professional Paper 1515.)
The stresses that accumulate along a locked segment of the fault and the sudden release can be visualized by bending a stick until it breaks. The stick will bend fairly easily, up to a certain point, until the stress becomes too great and it snaps. The vibrations felt when the stick breaks represent the sudden release of the stored-up energy. Similarly, the seismic vibrations produced when the ground suddenly ruptures radiate out through the Earth’s interior from the rupture point, called the earthquake focus. The geographic point directly above the focus is called the earthquake epicenter. In a major earthquake, the energy released can cause damage hundreds to thousands of kilometers away from the epicenter.
A dramatic photograph of horses killed by falling debris during the Great San Francisco Earthquake of 1906, when a locked segment of the San Andreas fault suddenly lurched, causing a devastating magnitude-8.3 earthquake. (Photograph by Edith Irvine, courtesy of Brigham Young University Library, Provo, Utah.)
The magnitude-7.1 Loma Prieta earthquake of October 1989 occurred along a segment of the San Andreas Fault which had been locked since the great 1906 San Francisco earthquake. Even though the earthquake’s focus (approximately 80 km south of San Francisco) was centered in a sparsely populated part of the Santa Cruz Mountains, the earthquake still caused 62 deaths and nearly $6 billion in damage. Following the Loma Prieta earthquake, the fault remains locked from Pt. Arena, where it enters California from the ocean, south through San Francisco and the peninsula west of San Francisco Bay, thus posing the threat of a potential destructive earthquake occurring in a much more densely populated area.
The lesser known Hayward Fault running east of San Francisco Bay, however, may pose a potential threat as great as, or perhaps even greater than, the San Andreas. From the televised scenes of the damage caused by the 7.2-magnitude earthquake that struck Kobe, Japan, on 16 January 1995, Bay Area residents saw the possible devastation that could occur if a comparable size earthquake were to strike along the Hayward Fault. This is because the Hayward and the Nojima fault that produced the Kobe earthquake are quite similar in several ways. Not only are they of the same type (strike-slip), they are also about the same length (60p;80 km) and both cut through densely populated urban areas, with many buildings, freeways, and other structures built on unstable bay landfill.
On 17 January 1994, one of the costliest natural disasters in United States history struck southern California. A magnitude-6.6 earthquake hit near Northridge, a city located in the populous San Fernando Valley just north of Los Angeles, California. This disaster, which killed more than 60 people, caused an estimated $30 billion in damage, nearly five times that resulting from the Loma Prieta earthquake. The Northridge earthquake did not directly involve movement along one of the strands of the San Andreas Fault system. It instead occurred along the Santa Monica Mountains Thrust Fault, one of several smaller, concealed faults (called blind thrust faults) south of the San Andreas Fault zone where it bends to the east, roughly paralleling the Transverse Mountain Range. With a thrust fault, whose plane is inclined to the Earth’s surface, one side moves upward over the other. Movement along a blind thrust fault does not break the ground surface, thus making it difficult or impossible to map these hidden but potentially dangerous faults. Although scientists have found measurable uplift at several places in the Transverse Range, they have not found any conclusive evidence of ground rupture from the 1994 Northridge earthquake. Similar earthquakes struck the region in 1971 and 1987; the San Fernando earthquake (1971) caused substantial damage, including the collapse of a hospital and several freeway overpasses.
Not all fault movement is as violent and destructive. Near the city of Hollister in central California, the Calaveras Fault bends toward the San Andreas. Here, the Calaveras fault creeps at a slow, steady pace, posing little danger. Much of the Calaveras fault creeps at an average rate of 5 to 6 mm/yr. On average, Hollister has some 20,000 earthquakes a year, most of which are too small to be felt by residents. It is rare for an area undergoing creep to experience an earthquake with a magnitude greater than 6.0 because stress is continually being relieved and, therefore, does not accumulate. Fault-creep movement generally is non-threatening, resulting only in gradual offset of roads, fences, sidewalks, pipelines, and other structures that cross the fault. However, the persistence of fault creep does pose a costly nuisance in terms of maintenance and repair.
Mid-plate earthquakes — those occurring in the interiors of plates — are much less frequent than those along plate boundaries and more difficult to explain. Earthquakes along the Atlantic seaboard of the United States are most likely related in some way to the westward movement of the North American Plate away from the Mid-Atlantic Ridge, a continuing process begun with the break-up of Pangaea. However, the causes of these infrequent earthquakes are still not understood.
East Coast earthquakes, such as the one that struck Charleston, South Carolina, in 1886 are felt over a much larger area than earthquakes occurring on the West Coast, because the eastern half of the country is mainly composed of older rock that has not been fractured and cracked by frequent earthquake activity in the recent geologic past. Rock that is highly fractured and crushed absorbs more seismic energy than rock that is less fractured. The Charleston earthquake, with an estimated magnitude of about 7.0, was felt as far away as Chicago, more than 1,300 km to the northwest, whereas the 7.1-magnitude Loma Prieta earthquakes was felt no farther than Los Angeles, about 500 km south. The most widely felt earthquakes ever to strike the United States were centered near the town of New Madrid, Missouri, in 1811 and 1812. Three earthquakes, felt as far away as Washington D.C., were each estimated to be above 8.0 in magnitude. Most of us do not associate earthquakes with New York City, but beneath Manhattan is a network of intersecting faults, a few of which are capable of causing earthquakes. The most recent earthquake to strike New York City occurred in 1985 and measured 4.0 in magnitude, and a pair of earthquakes (magnitude 4.0 and 4.5) shook Reading, Pennsylvania, in January 1994 causing minor damage.
Left: Creeping along the Calaveras fault has bent the retaining wall and offset the sidewalk along 5th Street in Hollister, California (about 75 km south-southeast of San Jose). Right: Close-up of the offset of the curb. (Photographs by W. Jacquelyne Kious.)
We know in general how most earthquakes occur, but can we predict when they will strike? This question has challenged and frustrated scientists studying likely precursors to moderate and large earthquakes. Since the early 1980s, geologists and seismologists have been intensively studying a segment of the San Andreas near the small town of Parkfield, located about halfway between San Francisco and Los Angeles, to try to detect the physical and chemical changes that might take place — both above and below ground — before an earthquake strikes. The USGS and State and local agencies have blanketed Parkfield and the surrounding countryside with seismographs, creep meters, stress meters, and other ground-motion measurement devices.
The Parkfield segment has experienced earthquakes measuring magnitude 6.0 about every 22 years on average since 1881. During the most recent two earthquakes (1934, 1966), the same section of the fault slipped and the amount of slippage was about the same. In 1983, this evidence, in addition to the earlier recorded history of earthquake activity, led the USGS to predict that there was a 95 percent chance of a 6.0 earthquake striking Parkfield before 1993. But the anticipated earthquake of magnitude 6.0 or greater did not materialize. The Parkfield experiment is continuing, and its primary goals remain unchanged: to issue a short-term prediction; to monitor and analyze geophysical and geochemical effects before, during, and after the anticipated earthquake; and to develop effective communications between scientists, emergency-management officials, and the public in responding to earthquake hazards.
While scientists are studying and identifying possible precursors leading to the next Parkfield earthquake, they also are looking at these same precursors to see if they may be occurring along other segments of the fault. Studies of past earthquakes, together with data and experience gained from the Parkfield experiment, have been used by geoscientists to estimate the probabilities of major earthquakes occurring along the entire San Andreas Fault system. In 1988, the USGS identified six segments of the San Andreas as most likely to be hit by a magnitude 6.5 or larger earthquake within the next thirty years (1988-2018). The Loma Prieta earthquake in 1989 occurred along one of these six segments. The Parkfield experiment and other studies carried out by the USGS as part of the National Earthquake Hazards Reduction Program have led to an increased official and public awareness of the inevitability of future earthquake activity in California. Consequently, residents and State and local officials have become more diligent in planning and preparing for the next big earthquake.