Local magnitudes and apparent variations in seismicity rates in Southern California

Abstract

Redetermination of local magnitudes for moderate earthquakes recorded by the Southern California Seismographic Network (SCSN) from 1932 to 1990 has shown that the magnitudes have not been consistently determined over that time period. The amplitudes of ground velocities recorded on Wood-Anderson instruments were systematically overestimated prior to 1944 compared to present reading procedures, leading to a significant overestimation of local magnitudes. In addition, the change from human to computerized estimation of event magnitude from a suite of amplitudes in 1975 led to slightly lower event magnitudes for the time after 1975 compared to the time before. These changes contribute to an apparently higher rate of seismicity in the 1930s and 1940s than later in the catalog, which had been interpreted as a decrease in seismicity rate after the 1952 Kern County (M_w 7.5) earthquake. Wood-Anderson amplitudes have been reread and consistent magnitudes recalculated using uniform procedures for all earthquakes with a catalog magnitude of 4.5 and greater within the SCSN from 1932 to 1943 and those with a catalog magnitude of 4.8 and greater from 1944 to 1990 so as to create a complete list of all earthquakes with a modern local magnitude of 5.0 or greater. Using these new magnitudes, we find that the rate of M_L 5.0 and greater earthquakes in southern California over this 59-year period to be Poissonian, with no changes in rate significant above the 90% level. From this rate, in any 30-year period, the Poissonian probability of a M ≧ 6 earthquake is 99.7%, the probability of an M ≧ 7 earthquake is 65%, and the probability of an M ≧ 8 event is 18%.

Files

Additional details