A Decade of Regional Moment Tensor Analysis at UC Berkeley

Several magnitude estimates are routinely determined by regional networks in California. The issue of magnitude reporting hierachies, particularly the question of when moment magnitude (M_w) should be preferred over local magnitude (M_L), is being addressed in the framework of the development of a statewide earthquake monitoring system in California through the CISN, and is also faced by the ANSS. The Berkeley Seismological Lab (BSL) has made routine determinations of the seismic moment tensor and moment magnitude for over 10 years. These determinations are made automatically for events with local magnitude of 3.5 and higher in northern California and typically reviewed for events of M4.0 and higher. The moment tensor catalog includes nearly 400 events in northern and central California, southern Oregon, and eastern Nevada, as well as events of interest in Washington, Idaho, Utah, and Arizona. We compare the regional estimates of M_w with estimates of local and duration magnitude. In general, both the automated and reviewed estimates of M_w compare well with other magnitudes, particularly in the M3.5-5.0 range. In some areas of Northern California, there are systematic differences, particularly between M_L and M_w. In the Cape Mendocino area, estimates of M_w are consistently 0.5-1.0 magnitude units higher than M_L for offshore events in the transform. Wurman et al.(this meeting) have found that these events have low Orowan stress drops and low corner frequencies. Other areas showing systematic difference are the Geysers and certain regions in eastern California. We also compare the regional estimates of M_w with global estimates from the Harvard CMT Project and the NEIC for events in the western United States. The regional estimates of M_w agree extremely well with the global estimates, with the exeception of an intriguing systematic shift of \~0.08 magnitude units. We investigate possible reasons for this shift.