No abstract

We have developed regression relationships between Modified Mercalli Intensity ( Imm) and peak ground acceleration (PGA) and velocity (PGV) by comparing horizontal peak ground motions to observed intensities for eight significant California earthquakes. For the limited range of Modified Mercalli intensities ( Imm), we find that for peak acceleration with V ≤ Imm ≤ VIII, Imm = 3.66 log( PGA) − 1.66, and for peak velocity with V ≤ Imm ≤ IX, Imm = 3.47 log( PGV) + 2.35. From comparison with observed intensity maps, we find that a combined regression based on peak velocity for intensity > VII and on peak acceleration for intensity < VII is most suitable for reproducing observed Imm patterns, consistent with high intensities being related to damage (proportional to ground velocity) and with lower intensities determined by felt accounts (most sensitive to higher-frequency ground acceleration). These new Imm relationships are significantly different from the Trifunac and Brady (1975) correlations, which have been used extensively in loss estimation.

Occurrence of large earthquakes close to cities in California is inevitable. The resulting ground shaking will subject buildings in the near-source region to large, rapid displacement pulses which are not represented in design codes. The simulated Mw7.0 earthquake on a blind-thrust fault used in this study produces peak ground displacement and velocity of 200 cm and 180 cm/sec, respectively. Over an area of several hundred square kilometers in the near-source region, flexible frame and base-isolated buildings would experience severe nonlinear behavior including the possibility of collapse at some locations. The susceptibility of welded connections to fracture significantly increases the collapse potential of steel-frame buildings under strong ground motions of the type resulting from the Mw7.0 simulation. Because collapse of a building depends on many factors which are poorly understood, the results presented here regarding collapse should be interpreted carefully.

Rapid (3-5 minutes) generation of maps of instrumental ground-motion and shaking intensity is accomplished through advances in real-time seismographic data acquisition combined with newly developed relationships between recorded ground-motion parameters and expected shaking intensity values. Estimation of shaking over the entire regional extent of southern California is obtained by the spatial interpolation of the measured ground motions with geologically based frequency and amplitude-dependent site corrections. Production of the maps is automatic, triggered by any significant earthquake in southern California. Maps are now made available within several minutes of the earthquake for public and scientific consumption via the World Wide Web; they will be made available with dedicated communications for emergency response agencies and critical users.

The Landers earthquake, which had a moment magnitude (M(w)) of 7.3, was the largest earthquake to strike the contiguous United States in 40 years. This earthquake resulted from the rupture of five major and many minor right-lateral faults near the southern end of the eastern California shear zone, just north of the San Andreas fault. Its M(w) 6.1 preshock and M(w) 6.2 aftershock had their own aftershocks and foreshocks. Surficial geological observations are consistent with local and far-field seismologic observations of the earthquake. Large surficial offsets (as great as 6 meters) and a relatively short rupture length (85 kilometers) are consistent with seismological calculations of a high stress drop (200 bars), which is in turn consistent with an apparently long recurrence interval for these faults.