The TriNet project, launched in 1997, created an improved, real-time seismic monitoring network in Southern California. Planning of the network began in 1995 (e.g., Heaton et al., 1996), building on the success of the earlier TERRAscope network, which included 24 digital broadband and strong-motion instruments throughout Southern California (e.g., Kanamori et al., 1993). At the end of the five-year TriNet project the network comprised 150 real-time digital broadband stations and another 400 strong-motion sensors, 50 of which were also real-time. This network is now recording digital broadband data for Southern California earthquakes at an unprecedented rate, data that are already proving valuable for investigations of earthquake sources and regional wave propagation, as well as earthquake response.

In this report we describe an ongoing effort aimed at a full geotechnical characterization of the newly installed TriNet sets. Shallow geologic structure is known to play a substantial role in controlling the ground motions recorded at any site. Documentation of amplified ground motion at soft-sediment sites can be found among even early macroseismic observations of strong ground motions (e.g., Drake, 1815). Seismic waves are also now known to be strongly affected by deep basin structure as well (e.g., Frankel et al., 1991; Field, 2000; Joyner, 2000). It will be necessary to understand these effects to exploit fully the rich data set being recorded at TriNet sites. Clearly, the nature of site conditions and site response at the recording sites must be understood for studies focused on ground motions and hazard from future large earthquakes, but it is also necessary to understand these effects to conduct earthquake source studies of both large and small earthquakes.

Our multifaceted site characterization project involves geological/geotechnical site investigations, database development, and investigation of empirical amplification factors determined from broadband and strong-motion data recorded to date. Our goal is to complete a first-order geologic site characterization by 2005 and then employ appropriate methods to obtain direct constraint on shallow shear-wave velocity structure at each site. In the latter effort we will endeavor to find the most accurate and cost-effective methods to quantify geotechnical parameters at sites that have ranges of geologic site conditions and cultural settings.

Our long-term goal is to obtain direct estimates of the average shear-wave velocity in the upper 30 m, Vs30, at each site. If velocity information is available to greater depths, for example at sites characterized under the ROSRINE project (http://geoinfo.usc.edu/rosrine/), this information will be included in the database as well. The original motivation for using Vs30 to characterize near-surface velocity was pragmatic, determined by such factors as the typical reach of a drill rig in a single day. The parameter has become the accepted standard with which many seismic recording sites are characterized, however. In a recent study using mainshock and aftershock recordings at sites in the Los Angeles region, Wald and Mori (2000) observed good correlation between Vs30 and amplification at 1-7 Hz, albeit with significant scatter.

The purpose of this report is threefold. First, we describe our ongoing efforts and present site characterization results collected to date for 62 broadband and strong-motion stations in and around the greater Los Angeles metropolitan region. Second, we present preliminary results that illustrate how improved geologic site characterizations can improve the correlation between site conditions and site response. Finally, we invite feedback from the community to guide our future investigations, in particular with respect to our ongoing database development efforts.