Dense microearthquake swarms occur in the upper south flank of Kilauea, providing multiplets composed of hundreds of events. The similarity of their waveforms and the quality of the data have been sufficient to provide accurate relative relocations of their hypocenters. A simple and efficient method has been developed which allowed the relative relocation of more than 250 events with an average precision of about 50 m horizontally and 75 m vertically. Relocation of these events greatly improves the definition of the seismic image of the fault that generates them. Indeed, relative relocations define a plane dipping about 6° northward, although corresponding absolute locations are widely dispersed in the swarm. A composite focal mechanism, built from events providing a correct spatial sampling of the multiplet, also gives a well‐constrained northward dip of about 5° to the near‐horizontal plane. This technique thus collapses the clouds of hypocenters of single‐event locations to a plane coinciding with the slip plane revealed by previous focal mechanism studies. We cannot conclude that all south flank earthquakes collapse to a single plane. There may locally be several planes, perhaps with different dips and depths throughout the south flank volume. The 6° northward‐dipping plane we found is too steep to represent the overall flexure of the oceanic crust under the load of the island of Hawaii. This plane is probably an important feature that characterizes the basal slip layer below the upper south flank of Kilauea volcano. Differences in seismicity rate and surface deformations between the upper and lower south flank could be related to the geometry of this deep fault plane. The present work illustrates how high precision relative relocations of similar events in dense swarms, combined with the analysis of geodetic measurements, can help to describe deep fault plane geometry. Systematic selection and extensive relative relocation of similar earthquakes could be attempted in other well‐instrumented, highly seismic areas to provide reliable basic information, especially useful for understanding of earthquake generation processes.

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.

S U M M A R YA new three-dimensional delay traveltime tomography is performed to image the intermediate structure of the western Gulf of Corinth. A large data set, collected in 1991 during a two-month passive tomographic experiment, has been reanalysed for the reconstruction of detailed Vp and Vs images. An improved tomography method, based on an accurate traveltime computation, is applied to invert simultaneously delayed P and S first-arrival traveltimes for both velocity and hypocentre parameters. We perform different synthetic tests to analyse the sensitivity of tomography results to the model parametrization and to the starting 1-D model selection. The analysis of the retrieved Vp and Vs models as well as deduced Vp/Vs and Vp · Vs images allows us to interpret and delineate the distribution of lithological variation, porosity/crack content and fluid saturation in the upper 9-11 km of the crust beneath the gulf. The tomographic models image a rather complex crustal structure, which is characterized by a vertical change in both velocity features and seismicity distribution. We identify a shallower zone of the crust (0-5 km depth), in which velocity distributions seem to be controlled by the still active N-S extensional regime and a deeper zone (7-11 km depth), which matches the seismogenic zone. The correlation between this latter and a specific unit of the Hellenic mountain structure (the Pyllite-Quartzite series) allows us to suggest a possible explanation for seismicity concentration in a narrow band at 7-9 km depth. Finally, the occurrence of clusters showing low-angle normal fault mechanisms in areas characterized by high Vp/Vs values indicates a possible role of fluids in triggering brittle creep along the identified low-angle normal faults.

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.