Seismic hazard characterization of the eastern United States: Comparative evaluation of the LLNL and EPRI studies

In 1982, the Lawrence Livermore National Laboratory (LLNL developed a methodology to characterize the seismic hazard for all sites of the eastern United States (EUS) east of the Rocky Mountains. The utility sponsored Electric Power Research Institute (EPRI) followed suit in late 1983 with a similar study. The purpose of this study was to help in understanding the reasons for differences in results between the LLNL and the EPRI study. We first investigated possible differences in the theories and assumptions used to develop the hazard models and concluded that all inputs being equal, the two methods were essentially equivalent. Then we analyzed the various input parameters, their values and the way they were collected. And finally we performed some sensitivity analyses, in particular with the seismicity parameters, the ground motion models and the value of the lower bound magnitude of integration (m/sub o/), to estimate the importance and evaluate the relative contributions to the total differences between the two studies. The three main differences were found to be: (1) the value of m/sub o/, (2) the ground motion models, and (3) the fact that LLNL accounted for a local site correction and EPRI did not