Earthquake prediction is inherently statistical. Although some people continue to think of earthquake prediction as the specification of the time, place, and magnitude of a future earthquake, it has been clear for at least a decade that this is an unrealistic and unreasonable definition. the reality is that earthquake prediction starts from the long-term forecasts of place and magnitude, with very approximate time constraints, and progresses, at least in principle, to a gradual narrowing of the time window as data and understanding permit. Primitive long-term forecasts are clearly possible at this time on a few well-characterized fault systems. Tightly focuses monitoring experiments aimed at short-term prediction are already underway in Parkfield, California, and in the Tokai region in Japan; only time will tell how much progress will be possible.

Earthquakes & Volcanoes (USGS)↗

USGS1987

Temporal changes in microseismicity and creep near Parkfield, California

The 25-km-long section of the San Andreas fault near Parkfield, California ruptured in similar magnitude-6 earthquakes in 1881, 1901, 1922, 1934 and 1966. On the basis of a number of lines of seismological evidence, a section of the San Andreas fault, now termed the Parkfield preparation zone, has been identified as the locus of the next Parkfield earthquake 1,3 . Here we describe coincident changes in surface creep rates and deep seismicity near the Parkfield preparation zone following the 2 May 1983 Coalinga earthquake, and suggest that both respond to the same stimuli. These changes were concentrated near the point of initiation of the magnitude-6 characteristic Parkfield earthquakes, lending credence to the hypothesis that this section of the San Andreas fault is characterized by a unique set of physical properties 4–6 which make it ideal for earthquake prediction studies and which may also be useful for identifying hypocentral regions in other areas.

California↗

USGS1986

Parkfield earthquake prediction scenarios and response plans

Explore the source record for details and available documents.

Open-File Report↗

USGS1984

The 1984 Morgan Hill, California, earthquake

The Morgan Hill, California, earthquake (magnitude 6.1) of 24 April 1984 ruptured a 30-kilometer-long segment of the Calaveras fault zone to the east of San Jose. Although it was recognized in 1980 that an earthquake of magnitude 6 occurred on this segment in 1911 and that a repeat of this event might reasonably be expected, no short-term precursors were noted and so the time of the 1984 earthquake was not predicted. Unilateral rupture propagation toward the south-southeast and an energetic late source of seismic radiation located near the southeast end of the rupture zone contributed to the highly focused pattern of strong motion, including an exceptionally large horizontal acceleration of 1.29g at a site on a dam abutment near the southeast end of the rupture zone.

California↗

USGS1982

Seismicity of the Parkfield, California, region, 1969 to 1979

Explore the source record for details and available documents.

Open-File Report↗

USGS1979

A change in fault-plane orientation between foreshocks and aftershocks of the Galway Lake earthquake, ML = 5.2, 1975, Mojave Desert, California

A marked change is observed in P / SV amplitude ratios, measured at station TPC, from foreshocks to aftershocks of the Galway Lake earthquake. This change is interpreted to be the result of a change in fault-plane orientation occurring between foreshocks and aftershocks. The Galway Lake earthquake, M L = 5.2, occurred on June 1, 1975. The first-motion fault-plane solutions for the main shock and most foreshocks and aftershocks indicate chiefly right-lateral strike-slip on NNW-striking planes that dip steeply, 70–90°, to the WSW. The main event was preceded by nine located foreshocks, ranging in magnitude from 1.9 to 3.4, over a period of 12 weeks, starting on March 9, 1975. All of the foreshocks form a tight cluster approximately 1 km in diameter. This cluster includes the main shock. Aftershocks are distributed over a 6-km-long fault zone, but only those that occurred inside the foreshock cluster are used in this study. Seismograms recorded at TPC ( Δ = 61 km ), PEC ( Δ = 93 km ), and CSP ( Δ = 83 km ) are the data used here. The seismograms recorded at TPC show very consistent P / SV amplitude ratios for foreshocks. For aftershocks the P / SV ratios are scattered, but generally quite different from foreshock ratios. Most of the scatter for the aftershocks is confined to the two days following the main shock. Thereafter, however, the P / SV ratios are consistently half as large as for foreshocks. More subtle (and questionable) changes in the P / SV ratios are observed at PEC and CSP. Using theoretical P / SV amplitude ratios, one can reproduce the observations at TPC, PEC and CSP by invoking a 5–12° counterclockwise change in fault strike between foreshocks and aftershocks. This interpretation is not unique, but it fits the data better than invoking, for example, changes in dip or slip angle. First-motion data cannot resolve this small change, but they permit it. Attenuation changes would appear to be ruled out by the fact that changes in the amplitude ratios, P TPC / P PEC and p tpc / p csp , are observed, and these changes accompany the changes in P / SV . Observations for the Galway Lake earthquake are similar to observations for the Oroville, California, earthquake ( M L = 5.7) of August 1, 1975, and the Brianes Hills, California, earthquake ( M L = 4.3) of January 8, 1977 (Lindh et al., Science Vol. 201, pp. 56–59). A change in fault-plane orientation between foreshocks and aftershocks may be understandable in terms of early en-echelon cracking (foreshocks) giving way to shear on the main fault plane (main shock plus aftershocks). Recent laboratory data (Byerlee et al., Tectonophysics, Vol. 44, pp. 161–171) tend to support this view.

California↗

USGS1978

Amplitude of foreshocks as a possible seismic precursor to earthquakes

In recent years, we have made significant progress in being able to recognize the long-range pattern of events that precede large earthquakes. For example, in a recent issue of the Earthquake Information Bulletin, we saw how the pioneering work of S.A. Fedotov of the U.S.S.R in the Kamchatka-Kurile Islands region has been applied worldwide to forecast where large, shallow earthquakes might occur in the next decades. Indeed, such a "seismic gap" off the coast of Alaska was filled by the 1972 Sitka earthquake. Promising results are slowly accumulating from other techniques that suggest that intermediate-term precursors might also be seen: among these are tilt and geomagnetic anomalies and anomalous land uplift. But the crucial point remains that short-term precursors (days to hours) will be needed in many cases if there is to be a significant saving of lives.

Earthquake Information Bulletin (USGS)↗