Crustal faults that produce most of their slip aseismically typically generate large numbers of small earthquakes. These events have generally been interpreted as coming from localized patches of the fault that undergo unstable (stick±slip) sliding, surrounded by larger regions of stable sliding (creep). In published catalogues the microearthquakes often appear to be distributed over large portions of the fault surface. By accurately locating large numbers of microearthquakes from faults of different orientations in California and Hawaii, we show here that instead the locations de®ne highly concentrated streaks that are characteristically aligned in the direction of fault slip. The underlying cause of this structural organization of the fault surface remains to be determined.

Abstract. Earthquakes of magnitude 1 and greater seem to be ubiquitous features of dike propagation, but their origin is not well understood. We examine the elastic stress field surrounding propagating fluid-filled cracks, with an emphasis on assessing the ambient stress required to produce earthquakes with linear dimensions of-100 m near dikes with linear dimensions of a few kilometers. An important feature of the solutions is the dike "tip cavity," a low-pressure region where magma cannot penetrate and where the stress field differs most from the classical near-tip stress field. Two regions are considered: near the dike tip but away from the tip cavity and near the tip cavity. The stress state most conducive to failure occurs near the tip cavity when the cavity pressure is maintained by influx of host rock pore fluids rather than by exsolution of magmatic volatiles. Even in this case, however, shear fracture of previously intact rock seems unlikely. Thus most dike-induced seismicity with a frequency content typical of "tectonic" earthquakes should be interpreted as resulting from slip along suitably aligned existing fractures. Production of magnitude 1 earthquakes appears to require either large ambient differential stresses or low ambient confining pressures; in the latter case, the effective normal stress on prospective faults may be low enough for slip to be aseismic. We conclude that the distribution of (recorded) dike-induced seismicity reflects the distribution of ambient stresses that are near to failure and does not necessarily reflect the extent of the dike. This result is consistent with recent images of the seismicity associated with the 1983 dike intrusion at Kilauea.

aftershocks occurring closest to the mode II edges of the prior rupture, more than twice as many occur to the northwest than to the southeast. We interpret this asymmetry as resulting from the large contrast in material properties across the fault. Models of dynamic rupture between dissimilar media predict that ruptures in this region may run preferentially to the southeast, in the direction of motion of the lower-velocity material. If so, then the barriers that stop rupture fronts moving to the southeast should initially be farther from failure, on average, than the barriers that stop rupture fronts moving to the northwest. Once the rupture stops, the induced stress change is more symmetric but the fault remains farther from failure (on average) to the southeast. This interpretation receives some support from pulse width measurements on a localized set of 72 magnitude 0.6 to 3.6 earthquakes.

Abstract. In January 1983, a dike intrusion/fissure eruption generated a swarm of 375 magnitude 1 to 3 earthquakes along a 16-km segment of Kilauea' s Middle East Rift Zone. We searched the Hawaiian Volcano Observatory catalog for multiplets of similar events from this region from 1980 through 1985 and obtained precise relative locations by waveform cross correlation. Over 150 of the intrusion earthquakes could be grouped into 14 multiplets of five or more events with sufficient similarity for accurate relocation. Some multiplets were active for only a few minutes during the downrift migration phase of the seismic swarm, consistent with generation near the propagating dike tip, while others were active for several days. The two multiplets nearest the origin of the seismic swarm include events from the preceding days and months. Most multiplets span only 50 to 100 m following relocation, are located at about 3 to 4 km depth, and appear to deepen downrift. The catalog depths of those earthquakes in multiplets and those not in multiplets are similar, suggesting that most of the recorded seismicity may have come from a very limited depth interval despite the fact that the dike breached the surface. By analogy with a mechanical model used to explain a similar clustering of background seismicity in the Upper East Rift in 1991, we infer that the earthquakes are generated in regions of high stress concentration immediately above Kilauea's deforming deep rift body. This conclusion is consistent with the depth of the top of the deep rift body inferred from geodetic data and with numerical calculations suggesting that a significant ambient differential stress is required for dikes to produce earthquakes larger than magnitude 1.