U.S. Geological Survey, Menlo Park, CA, U.S.A.

We collected teleseismic waveforms from a digital microseismic network deployed by the Istituto Nazionale di Geofisica (ING) in collaboration with the U.S. Geological Survey (USGS), on the Alban Hills Quaternary volcano during the 1989-1990 seismic swann. About 50 events were recorded by the network, 30 of them by at least 4 stations. We analysed the data in order to image crustal heterogeneities beneath the volcano. The results show large delay time residuals up to - 1 second for stations located on the volcano with respect to station CP9 of the National Seismic Network located about 20 km to the east, on the Apennines. This suggests that the whole area overlies a broad low-velocity region. Although the ray coverage is not very dense, we model the gross seismic structure beneath the volcano by inverting the teleseismic relative residuals with the ACH technique. The main features detected by tbc inversion are a low-velocity zone beneath the southwestern fiank of tbc volcano, and a high-velocity region beneath the center. The depth extension of these anomalous zones ranges between 5 and 16 km. The correspondence between the low-velocity region and the most recent activity of the volcano (- 0.027 Ma) leads us to infer the presence of a still hot magmatic body in the crust beneath the southwestern side of the volcano, whereas the central part overlies the older and colder high-velocity volcanic roots related to the previous central activity (0.7 to 0.3 Ma).

We investigate shear wave polarization in the Hayward fault zone near Niles Canyon, Fremont, CA. Waveforms of 12 earthquakes recorded by a seven-accelerometer seismic array around the fault are analysed to clarify directional site effects in the fault damage zone. The analysis is performed in the frequency domain through H/V spectral ratios with horizontal components rotated from 0◦ to 180◦, and in the time domain using the eigenvectors and eigenvalues of the covariance matrix method employing three component records. The near-fault ground motion tends to be polarized in the horizontal plane. At two on-fault stations where the local strike is N160◦, ground motion polarization is oriented N88 ± 19◦ and N83 ± 32◦, respectively. At a third on-fault station, the motion is more complex with horizontal polarization varying in different frequency bands. However, a polarization of N86 ± 7◦, similar to the results at the other two on-fault stations, is found in the frequency band 6–8 Hz. The predominantly high-angle polarization from the fault strike at the Hayward Fault is consistent with similar results at the Parkfield section of the San Andreas Fault and the Val d’Agri area (a Quaternary extensional basin) in Italy. In all these cases, comparisons of the observed polarization directions with models of fracture orientation based on the fault movement indicate that the dominant horizontal polarization is near-orthogonal to the orientation of the expected predominant cracking direction. The results help to develop improved connections between fault mechanics and near-fault ground motion.

A model of stress transfer implies that earthquakes in 1933 and 1952 increased the Conlomb stress at the site of the 1971 San Fernando earthquake. The 1971 earthquake in turn raised stress and produced aftershocks at the site of the 1987 Whittier Narrows and 1994 Northridge ruptures. The Northridge main shock raised stress in areas where its aftershocks and surface faulting occurred. Together, M ? 6 earthquakes near Los Angeles since 1933 have stressed parts of the Oak Ridge, Sierra Madre, Santa Monica Mountains, Elysian Park, and Newport-Inglewood faults by > 1 bar. While too small to cause earthquakes, these stress changes can trigger events if the crust is already near failure, or advance future earthquake occurrence if it is not.

Earthquake mechanics may be determined by the geometry of a fault system. Slip on a fractal branching fault surface can explain: 1) regeneration of stress irregularities in an earthquake; 2) the concentration of stress drop in an earthquake into asperities; 3) starting and stopping of earthquake slip at fault junctions, and 4) self-similar scaling of earthquakes. Slip at fault junctions provides a natural realization of barrier and asperity models without appealing to variations of fault strength. Fault systems are observed to have a branching fractal structure, and slip may occur at many fault junctions in an earthquake. Consider the mechanics of slip at one fault junction. In order to avoid a stress singularity of order 1/r, an intersection of faults must be a triple junction and the Burgers vectors on the three fault segments at the junction must sum to zero. In other words, to lowest order the deformation consists of rigid block displacement, which ensures that the local stress due to the dislocations is zero. The elastic dislocation solution, however, ignores the fact that the configuration of the blocks changes at the scale of the displacement. A volume change occurs at the junction; either a void opens or intense local deformation is required to avoid material overlap. The volume change is proportional to the product of the slip increment and the total slip since the formation of the junction. Energy absorbed at the junction, equal to confining pressure times the volume change, is not large enongh to prevent slip at a new junction. The ratio of energy absorbed at a new junction to elastic energy released in an earthquake is no larger than P/µ where P is confining pressure and µ is the shear modulus. At a depth of 10 km this dimensionless ratio has th value P/µ= 0.01. As slip accumulates at a fault junction in a number of earthquakes, the fault segments are displaced such that they no longer meet at a single point. For this reason the volume increment for a given slip increment becomes larger. A juction with past accumulated slip ??0 is a strong barrier to earthquakes with maximum slip um < 2 (P/µ) u0 = u0/50. As slip continues to occur elsewhere in the fault system, a stress concentration will grow at the old junction. A fresh fracture may occur in the stress concentration, establishing a new triple junction, and allowing continuity of slip in the fault system. The fresh fracture could provide the instability needed to explain earthquakes. Perhaps a small fraction (on the order of P/µ) of the surface that slips in any earthquake is fresh fracture. Stress drop occurs only on this small fraction of the rupture surface, the asperities. Strain change in the asperities is on the order of P/µ. Therefore this model predicts average strais change in an earthquake to be on the order of (P/µ)2 = 0.0001, as is observed.

he M 7.4 Landers earthquake triggered widespread seismicity in the Western U.S. Because the transient dynamic stresses induced at regional distances by the Landers surface waves are much larger than the expected static stresses, the magnitude and the characteristics of the dynamic stresses may bear upon the earthquake triggering mechanism. The Landers earthquake was recorded on the UPSAR array, a group of 14 triaxial accelerometers located within a 1-square-km region 10 km southwest of the town of Parkfield, California, 412 km northwest of the Landers epicenter. We used a standard geodetic inversion procedure to determine the surface strain and stress tensors as functions of time from the observed dynamic displacements. Peak dynamic strains and stresses at the Earth's surface are about 7 microstrain and 0.035 MPa, respectively, and they have a flat amplitude spectrum between 2 s and 15 s period. These stresses agree well with stresses predicted from a simple rule of thumb based upon the ground velocity spectrum observed at a single station. Peak stresses ranged from about 0.035 MPa at the surface to about 0.12 MPa between 2 and 14 km depth, with the sharp increase of stress away from the surface resulting from the rapid increase of rigidity with depth and from the influence of surface wave mode shapes. Comparison of Landers-induced static and dynamic stresses at the hypocenter of the Big Bear aftershock provides a clear example that faults are stronger on time scales of tens of seconds than on time scales of hours or longer.