HR: 0800h
AN: S31A-0275    [Abstracts]
TI: Locating the Source of the Twenty-six-second Microseism by Cross-correlating Records of Ambient Seismic Noise
AU: * Shapiro, N M
EM: nshapiro@merckx.colorado.edu
AF: Center for Imaging the Earth's Interior, Department of Physics, University of Colorado, Boulder, Campus Box 390, Boulder, CO 80309 United States
AU: Ritzwoller, M H
EM: ritzwoll@merckx.colorado.edu
AF: Center for Imaging the Earth's Interior, Department of Physics, University of Colorado, Boulder, Campus Box 390, Boulder, CO 80309 United States
AU: Bensen, G D
EM: gbensen@colorado.edu
AF: Center for Imaging the Earth's Interior, Department of Physics, University of Colorado, Boulder, Campus Box 390, Boulder, CO 80309 United States
AB: The surface-wave part of the Green function can be extracted by cross-correlating records of the ambient seismic noise between pairs of stations. The principal assumption of this method is that noise sources over the Earth's surface are randomly distributed, leading to constructive interference for sources located near the great circle connecting two stations. This condition breaks down if spatially localized, near-permanent noise sources are located off the great-circle line. The most prominent example discovered so far of such noise is the narrow band spectral peak centered on 26 sec period in the seismic background noise observed in seismograms during calm local conditions. Earlier studies identified this signal as Rayleigh waves originating in the equatorial Atlantic near the African coast (e.g., Oliver, 1962, 1963; Holcomb, 1980, 1998). Holcomb (1998) showed that peak amplitudes are strongly seasonal, maximizing during the southern hemisphere winter, suggesting an atmospheric or an oceanic origin of this narrow band microseism. Very strong, narrow band arrivals peaking around 26 sec are observed in cross-correlations of noise records for US, European, and African stations with apparent arrival times that are often smaller than expected for propagation along the great circle between the stations. Inversion of these apparent travel times shows that the narrow band 26 sec arrivals are waves originating off the west African coast in the Gulf of Guinea and propagate with an average speed close to 3.5 km/s; i.e., close to the expected group speed of Rayleigh waves at this period. Inversions using data from different months demonstrate that the source location is temporally stable with amplitudes maximizing during the southern hemisphere winter. A microseism with similar features but with a slightly broader spectral peak is observed at stations in East Asia and in the western Pacific, and originates in the North Fiji Basin. The physical cause or causes of the microseisms originating in the Gulf of Guinea and the North Fiji Basin remain(s) unclear. We believe the most likely cause to be long period oceanic waves that interact with the continental shelf, perhaps reflecting from it, and constructively interfere in a narrow frequency band in the deep ocean. Detailed physical modeling is required before the source mechanism is understood.
DE: 7255 Surface waves and free oscillations
DE: 7270 Tomography (6982, 8180)
SC: Seismology [S]
MN: Fall Meeting 2005