Abstract
Three-component digital array recordings of a midcrustal microearthquake cluster beneath the southeast flank of Kilauea volcano have been analyzed at four nearby portable stations for the presence of S-wave reflections from the top of the old oceanic crust and from the M-discontinuity beneath Hawaii. Cross-spectral correlation was applied to both P- and S-wave arrivals to obtain very precise relative hypocentral locations for the 29 events in the cluster. Reciprocal array analysis, in which the plane of earthquakes is treated as an observing array and the station as a source (Spudich and Bostwick, 1987), was applied to the S-wave coda for record sections at each of the four stations used in the joint hypocenter determination. Results of apparent velocity measurements across the earthquake (observing) array show that the early S-wave coda (up to about 1 to 1.5 sec after direct S) is dominated by near-station reverberations. None of the large amplitude arrivals that are seen on the records immediately following direct S, including those which were previously interpreted by us to be possible reflections from the upper part of the oceanic crust beneath Hawaii, are in fact deep reflections. Array slownesses for two of the stations studied reveal a weak component of backscattered energy at coda arrival times greater than about 1.5 sec that may be due to reflections from lower crustal reflectors and from the Moho. Those signals, however, originate at reflector depths greater than those postulated for the upper part of the old oceanic crust beneath Hawaii. The crustal reflections indicate reflector depths around 10 km or 11 km, about the predicted depth for the boundary between upper (layer 2) and the lower (layer 3) sections of the old oceanic crust. Reflections from the Moho give two-way travel times indicative of total crustal thickness under that part of Hawaii of about 13 to 14 km. When the seismograms are source migrated (by normalizing the travel times and epicentral distances to remove the effect of varying focal depth), both deep crustal and Moho reflections are seen as low-frequency signals that are roughly aligned on the record sections. The apparent absence of strong reflections from the the top of the old oceanic crust beneath Hawaii means either that the boundary is not marked by a very strong seismic discontinuity or that the earthquakes studied are so close to the boundary that even large amplitude reflections are effectively masked by the dominating effect of near station reverberations.
