On October 17, 1989, at 5:04 p.m. PDT, the Loma Prieta earthquake occurred on the Central Coast of California. The shock was centered in The Forest of Nisene Marks State Park in Santa Cruz County, approximately 10 miles (16 km) northeast of Santa Cruz on a section of the San Andreas Fault System and was named for the nearby Loma Prieta Peak in the Santa Cruz Mountains. With an magnitude of 6.9 and a maximum Modified Mercalli intensity of IX (Violent), the shock was responsible for 63 deaths and 3,757 injuries. The Loma Prieta segment of the San Andreas Fault System had been relatively inactive since the 1906 San Francisco earthquake (to the degree that it was designated a seismic gap) until two moderate foreshocks occurred in June 1988 and again in August 1989.
Damage was heavy in Santa Cruz County and less so to the south in Monterey County, but effects extended well to the north into the San Francisco Bay Area, both on the San Francisco Peninsula and across the bay in Oakland. No surface faulting occurred, though many other ground failures and landslides were present, especially in the Summit area of the Santa Cruz Mountains. Liquefaction was also a significant issue, especially in the heavily damaged Marina District of San Francisco, but its effects were also seen in the East Bay, and near the shore of Monterey Bay, where a non-destructive tsunami was also observed.
Because it happened during a national live broadcast of the 1989 World Series, the annual championship series of Major League Baseball, taking place between Bay Area teams San Francisco Giants and the Oakland Athletics, it is sometimes referred to as the "World Series earthquake", with the championship games of the year being referred to as the "Earthquake Series". Rush-hour traffic on the Bay Area freeways was much lighter than normal because the game, being played at Candlestick Park in San Francisco, was about to begin, and this almost certainly prevented a larger loss of life, as several of the Bay Area's major transportation structures suffered catastrophic failures. The collapse of a section of the double-deck Nimitz Freeway in Oakland was the site of the largest number of casualties for the event, but the collapse of human-made structures and other related accidents contributed to casualties occurring in San Francisco, Los Gatos, and Santa Cruz.
Background
The history of earthquake investigations in California has been largely focused on the San Andreas Fault system because of its strong influence in the state as the boundary between the Pacific plate and the North American plate; it is the most studied fault on Earth. Andrew Lawson, a geologist from the University of California, Berkeley, had named the fault after the San Andreas Lake (prior to the occurrence of the 1906 San Francisco earthquake) and later led an investigation into that event. The San Andreas Fault ruptured for a length of during the 1906 shock, both to the north of San Francisco and to the south in the Santa Cruz Mountains region. Several long term forecasts for a large shock along the San Andreas Fault in that area had been made public prior to 1989 (the event and its aftershocks occurred within a recognized seismic gap) but the earthquake that transpired was not what had been anticipated. The 1989 Loma Prieta event originated on an undiscovered oblique-slip reverse fault that is located adjacent to the San Andreas Fault.
Forecasts
Since many forecasts had been presented for the region near Loma Prieta, seismologists were not taken by surprise by the October 1989 event. Between 1910 and 1989 there were 20 widely varying forecasts that were announced, with some that were highly specific, covering multiple aspects of an event, while others were less complete and vague. With a M6.5 event on the San Juan Bautista segment, or an M7 event on the San Francisco Peninsula segment, United States Geological Survey (USGS) seismologist Allan Lindh's 1983 forecasted rupture length of (starting near Pajaro Gap, and continuing to the northwest) for the San Juan Bautista segment nearly matched the actual rupture length of the 1989 event. An updated forecast was presented in 1988, at which time Lindh took the opportunity to assign a new name to the San Juan Bautista segment – the Loma Prieta segment.
In early 1988, the Working Group for California Earthquake Probabilities (WGCEP) made several statements regarding their forecasts for the northern San Andreas Fault segment, the San Francisco Peninsula segment, and a portion of that segment which was referred to as the southern Santa Cruz Mountains segment. The thirty year probability for one or more M7 earthquakes in the study area was given as 50%, but because of a lack of information and low confidence, a 30% probability was assigned to the Southern Santa Cruz Mountains segment. Two moderate shocks, referred to as the Lake Elsman earthquakes by the USGS, occurred in the Santa Cruz Mountains region in June 1988 and again in August 1989. Following each event, the State office of Emergency Services issued (for the first time in Bay Area history) short term advisories for a possible large earthquake, which meant there was "a slightly increased likelihood of an M6.5 event on the Santa Cruz Mountains segment of the San Andreas fault". The advisories following the two Lake Elsman events were issued in part because of the statements made by WGCEP and because they were two of the three largest shocks to occur along the 1906 earthquake's rupture zone since 1914.
Foreshocks
The 5.3 June 1988 and the 5.4 August 1989 events also occurred on previously unknown oblique reverse faults and were within of the M6.9 Loma Prieta mainshock epicenter, near the intersection of the San Andreas and Sargent faults. Total displacement for these shocks was relatively small (approximately of strike-slip and substantially less reverse-slip) and although they occurred on separate faults and well before the mainshock, a group of seismologists considered these to be foreshocks because of when and where they occurred relative to the main event. Each event's aftershock sequence and effect on stress drop was closely examined, and their study indicated that the shocks affected the mainshock's rupture process. Following the August 8, 1989, shock, in anticipation of an upcoming large earthquake, staff at the University of California, Santa Cruz deployed four accelerometers in the area, which were positioned at the UCSC campus, two residences in Santa Cruz, and a home in Los Gatos. Unlike other nearby (high gain) seismographs that were overwhelmed (driven off scale) by the large magnitude mainshock, the four accelerometers captured a useful record of the main event and more than half an hour of the early aftershock activity.
The June 27, 1988, shock occurred with a maximum intensity of VI (Strong). Its effects included broken windows in Los Gatos, and other light damage in Holy City, where increased flow was observed at a water well. Farther away from the Santa Cruz Mountains, pieces of concrete fell from a parking structure at the Sunnyvale Town Center, a two-level shopping mall in Santa Clara County. More moderate damage resulted from the August 8, 1989, shock (intensity VII, Very strong) when chimneys were toppled in Cupertino, Los Gatos, and Redwood Estates. Other damage included cracked walls and foundations and broken underground pipes. At the office of the Los Gatos City Manager, a window that was cracked had also been broken in the earlier shock. Also in Los Gatos, one man died when he fell or jumped through a window and impacted the ground five stories below.
Earthquake
The Loma Prieta earthquake was named for Loma Prieta Peak in the Santa Cruz Mountains, which lies just to the east of the mainshock epicenter. The duration of the heaviest shaking in the Santa Cruz Mountains was about 15 seconds, but strong ground motion recordings revealed that the duration of shaking was not uniform throughout the affected area (because of different types and thicknesses of soil). At sites with rocky terrain, the duration was shorter and the shaking was much less intense, and at locations with unconsolidated soil (like the Marina District in San Francisco or the Cypress Street Viaduct in Oakland) the intensity of the shaking was more severe and lasted longer. The strong motion records also allowed for the causative fault to be determined – the rupture was related to the San Andreas Fault System.
While a Mercalli Intensity of VIII (Severe) covered a large swath of territory relatively close to the epicenter (including the cities of Los Gatos, Santa Cruz, and Watsonville) farther to the north, portions of San Francisco were assessed at intensity IX (Violent). At more than distant, the San Francisco Bay Area recorded peak horizontal accelerations that were as high as 0.26 g, and close to the epicenter they peaked at more than 0.6 g. In a general way, the location of aftershocks of the event delineated the extent of the faulting, which (according to seismologist Bruce Bolt) extended about in length. Because the rupture took place bilaterally, the duration of strong shaking was about half of what it would have been had it ruptured in one direction only. The duration of a typical M6.9 shock with a comparable rupture length would have been about twice as long.
Characteristics
Gregory Beroza, a seismologist with Stanford University, made several distinctions regarding the 1906 and 1989 events. Near Loma Prieta, the 1906 rupture was more shallow, had more strike-slip, and occurred on a fault that was near vertical. The 1989 event's oblique-slip rupture was at 10 km and below on a fault plane that dipped 70° to the southwest. Because much of the slip in 1989 occurred at depth and the rupture propagated up dip, Beroza proposed that the overlying San Andreas Fault actually inhibited further rupture and also maintains that the occurrence of an event at the location that was forecast by the WGCEP in 1988 was coincidental.
The contrasting characteristics of the 1906 and 1989 events were examined by seismologists Hiroo Kanamori and Kenji Satake. The significant amount of vertical displacement in 1989 was a key aspect to consider because a long-term sequence of 1989-type events (with an 80–100-year recurrence interval) normally result in regions with high topographic relief, which is not seen in the Santa Cruz Mountains. Three scenarios were presented that might explain this disparity. The first is that the geometry of the San Andreas Fault goes through a transition every several thousand years. Secondly, slip type could vary from event to event. And lastly, the 1989 event did not occur on the San Andreas Fault.
Ground effects While the effects of a four-year drought limited the potential of landslides, the steep terrain near the epicenter was prone to movement, and up to 4,000 landslides may have occurred during the event. The majority of landslides occurred to the southwest of the epicenter, especially along road cuts in the Santa Cruz Mountains and in the Summit Road area, but also along the bluffs of the Pacific Coast, and as far north as the Marin Peninsula. Highway 17 was blocked for several weeks by a large slide and one person was killed by a rockfall along the coast. Other areas with certain soil conditions were susceptible to site amplification due to the effects of liquefaction, especially near the shore of San Francisco Bay (where its effects were severe in the Marina District) and to the west of the epicenter near rivers and other bodies of water. Minor lateral spreading was also seen along the shores of San Francisco Bay and to the south near Monterey Bay. Other ground effects included downslope movement, slumps, and ground cracks.
Injuries and fatalities