Raúl W. Valenzuela (Instituto de Geofísica; Universidad Nacional Autónoma de México; 04510. México, D. F., Mexico; ph. (52) (5) 622-4138; e-mail: raul@ollin.igeofcu.unam.mx); Michael E. Wysession (Department of Earth and Planetary Sciences; Washington University; St. Louis, MO 63130-4899; ph. (314) 935-5625; e-mail: michael@wucore.wustl.edu)

The differential travel times of SKKS-SKS and Sdiff-SKS phases are used to simultaneously constrain the velocity structure on both sides of the core-mantle boundary (CMB). Differential times are determined with the method of Wysession et al. [1995] and are obtained from the displacement records through a cross correlation of SKS with SKKS and with Sdiff. In order to obtain a meaningful cross correlation of SKS with SKKS, the following corrections are applied. The SKS pulse is Hilbert-transformed to account for the phase shift introduced upon reflection off the underside of the CMB and it is convolved with an attenuation factor to correct for the difference in paths. It is also necessary to add a travel time correction to account for the change in wave shape. For comparison with Sdiff, the SKS attenuation factor is determined from reflectivity synthetic seismograms. The travel time residuals are computed relative to reflectivity synthetic seismograms. Since the outer core is fluid and does not sustain any lateral heterogenities, the Sdiff-SKS anomalies are interpreted to arise in D". The residual is mapped onto the Sdiff Fresnel zone to quantify the S-wave velocity anomaly. Since the bottoming depth for SKS is deeper than for SKKS, and since the bottoming depths of both increase with distance, the SKKS-SKS residuals can be used to determine the radial velocity structure of the outer core.

We present results from the Tonga earthquake of April 7, 1995 and the Marianas earthquake of August 23, 1995. Seismic records are taken from the Missouri to Massachusetts (MOMA) broadband array, which operated from March 1995 to March 1996 [Wysession et al., 1996]. The array consisted of 18 portable broadband seismometers deployed in a line between the permanent stations Cathedral Caves, Missouri (CCM) and Harvard, Massachusetts (HRV). Station spacing was roughly 90 km and the network spanned a total distance of 1740 km. MOMA was strategically located to sample the CMB using earthquakes in the southwest Pacific.