I use ``historical earthquakes'' to refer to events that occurred after the deployment of early seismographs during the last decade of the 19th century, but before the deployment of the World-Wide Standard Seismograph Network (WWSSN) in the early 1960's. Before the WWSSN existed, the distribution of seismographs around the world was not comprehensive and there were many different types of instruments. Nevertheless, because earthquakes have recurrence times much longer than the lifetime of seismographic networks such as the WWSSN, many important problems in earthquake seismology require the use of historical data. With the ongoing deployment of the digital Global Seismic Network (GSN) network, the day is not far off when analog WWSSN data will be regarded as ``historical'' An increased understanding of seismic hazard in areas of infrequent earthquakes often depends on studies of historical earthquakes. Over the last 4 years a number of studies of historical earthquakes have increased our understanding of both earthquakes and tectonics.
Wald et al. [1993] completed a source study of the 1906 San Francisco, California earthquake that showed two large subevents for this earthquake, one near San Francisco and another approximately 110 km to the northwest. A particularly interesting aspect of their model is that they found little slip in areas where geodetic data indicate slip was large. Helmberger et al. [1992] modeled regional and teleseismic waveforms of the 1927 Lompoc, California earthquake and found a primarily reverse faulting mechanism with a location farther to the south than previous investigators. Several studies of a total of 6 events occurring from 1918 to 1954 in the San Jacinto and Imperial fault systems of southern California [ Doser, 1990 1992; Bent and Helmberger, 1992] provided improved locations and indicated that the earthquakes all had strike-slip mechanisms. Doser and Yarwood [1991] studied the 1928 Sabukia earthquake in the East African rift and demonstrated that many large earthquakes, in this and other active rifts, have a surprisingly large component of strike-slip motion. Doser and Rodriguez [1993] found that the 1928 Parral earthquake in Chihuahua, Mexico had a mechanism similar to other events in the Basin and Range farther to the north. Suleiman et al. [1993] studied historical earthquakes along the coastal margin of western Africa. They concluded that the varied nature of these events precludes generalizations to other continental margins, such as the eastern coast of North America, where significant earthquakes have occurred in the pre-instrumental period.
Tsunami data have also been used to study historical earthquakes. Tsunami records were used to study both the 1938 Alaskan earthquake [ Johnson and Satake, 1993a] and the 1957 Aleutian earthquake [ Johnson and Satake, 1993b].