from California Geology, January 1981, Vol. 34, No. 1.


The 1906 event was the first major earthquake in the United States to be recorded on scientific instruments, and it produced the first recognized evidence for the association of faulting with earthquakes, as well as providing the first impetus for earthquake-resistant designs in this country. The possibility of a quake of similar magnitude occurring on the San Andreas fault or elsewhere on some other fault is often the controlling factor in the design of major engineering projects in California.

The Richter, or local magnitude, scale used in reporting magnitudes of California temblors was devised in 1935 for southern California by Charles F. Richter, now Caltech professor of seismology, emeritus. The local magnitude measures the size of an earthquake on a range of vibration highly damaging to structures, and is of particular importance to earthquake engineers. Richter and Caltech seismologist Beno Gutenberg later extended the scale to worldwide earthquakes. The scale has since gained wide acceptance and is currently the most commonly used measure of earthquake size.

Richter's original concept was elaborated upon, so that several types of magnitude scales, such as the surface-wave magnitude scale, are now in existence. The surface wave magnitude measures the overall size of the quake, including the length of faulting. It does not necessarily indicate the strength of ground shaking near the fault. These scales provide earthquake specialists with a more complex variety of information about the different types of ground motions produced by a specific quake than is available from the original Richter local magnitude scale.

Paul C. Jennings, professor of civil engineering and applied mechanics, and Hiroo Kanamori, professor of geophysics, have developed a new mathematical formula for interpreting old seismoscope records of the 1906 quake. The formula enables experts to derive the local magnitude of temblors from the records of seismoscopes---instruments designed to yield a different type of information about ground wave motions than that given by today's seismograph readings. Using the new formula Jennings and Kanamori have derived a local magnitude (ML) of 6.9 for the 1906 quake. This contrasts with 8.25, the surface wave magnitude (Ms) that the temblor is traditionally assigned. The newly obtained value, 6.9, is approximately comparable to that of many recent large earthquakes in California (for example, the 1971 San Fernando earthquake, 6.3 to 6.4; the 1968 Borrego Mountain earthquake, 6.4 to 6.9; the 1953 Kern County earthquake, 7.2). These figures indicate that the 1906 San Francisco quake is comparable to these temblors in strength of ground shaking near the fault, although the area affected in 1906 was much larger because of the extensive length of the San Andreas fault. Jennings and Kanamori were particularly interested in determining the local magnitude of the San Francisco quake because this temblor serves as a prototype for the disaster-causing potential of such an event.

In their calculation of the San Francisco quake's local magnitude, Jennings and Kanamori evaluated data from two instruments that recorded the event---a "simple pendulum" device located during the temblor at Yountville, California, 65 kilometers from the epicenter near the Golden Gate Bridge, and a Ewing Duplex Pendulum seismoscope that recorded the quake at Carson City, Nevada, 291 kilometers from the epicenter (see photo). The seismoscope at Carson City was retired from service in 1916; it recently was located in storage at the Reno campus of the University of Nevada.


Paul Jennings adjusts an antique seismoscope. In 1906 this seismoscope was located at Carson City, Nevada, and recorded the San Francisco earthquake. The 1906 records from this instrument and another recording device located at Yountville, California, were used to develop a formula for calculating the local magnitude of the 1906 event. Photo courtesy of California Institute of Technology.

Both instruments are similar to today's more sophisticated seismoscopes---devices that record earth motions via a conical pendulum suspended by a fine wire from a horizontal beam. The record is scribed by the pendulum on an inverted smoke watch glass, and the strength of the temblor is calculated through mathematical analyses of the extent of the scratch marks. Engineers use seismoscope records to calculate the effects of earth motions on buildings and other structures.

Their method of calculation should prove particularly important in determining the local magnitude of large earthquakes, because the more sensitive seismographs near an earthquake epicenter are generally thrown off-scale by very large temblors.

The two professors also have devised a technique to determine the local magnitude of a quake from the recordings of strong-motion accelerograms---other instruments used to record data about seismic waves produced by earthquakes. The popularly accepted 8.25 surface-wave magnitude of the San Francisco quake was derived mainly from stations in Europe. Information from instruments in the immediate area was so severely distorted by the shaking as to be of limited usefulness.

Before evaluating the record of the Carson City instrument, the scientists needed to know its inherent responses to seismic waves---its period, damping, and gain. When the instrument was recovered, it was damaged and some of its original parts were missing. At Caltech, it was repaired by Ivar Sedleniek of the Institute's Seismological Laboratory and Raul Relles of the Earthquake Engineering Laboratory.

To check the accuracy of their reconstruction, and to help evaluate the seismoscope's functioning, Kanamori and Jennings compared test results from their instrument with those from identical tests done on a similar instrument in London. The London instrument, which is in the London Science Museum, was tested by their colleague N. N. Ambraseys. After these tests, they calculated the magnitude of the quake from the records of both the Carson City and the Yountville seismoscopes, and arrived at an approximate local magnitude of 6.9.

In arriving at this figure they also examined records from Ewing Duplex Pendulum seismographs from Mt. Hamilton, Alameda, San Jose, Oakland, and Berkeley, California---all located from 20 to 36 kilometers from the earthquake fault. Records from those sites were seriously distorted by the strong motion, which drove the instruments off-scale. However, the scientists estimated that, even if the instruments had gone off-scale by a factor of 20, which was unlikely, the local magnitude of the San Francisco quake would not have been more than 6.9.

In addition to calculating the local magnitude of the San Francisco quake, Kanamori and Jennings applied their method for determining local magnitude to the Guatemala earthquake of February 4, 1976. A seismoscope in Guatemala City produced the only seismic recording obtained near the fault. Based on this limited record, the quake's local magnitude was calculated at 6.9. The surface-wave magnitude of this earthquake had been estimated at 7.5. . . . Caltech News Release.