EXTRACTING THE GREEN’S FUNCTION BETWEEN TWO STATIONS USING AMBIENT NOISE DATA

LIANG, C., and LANGSTON, C.A, Center for Earthquake Research and Information, the University of Memphis, TN, 38152, cliang@memphis.edu, clangstn@memphis.edu.

We cross-correlate 10 hours of ambient noise data recorded by station pairs and stack to extract the Green's function between two stations. Those signals on the correlated seismograms have particle motions and dispersion characteristics close to those of Rayleigh waves observed from earthquake sources. Clear Rayleigh wave trains with a group velocity about 3.0 km/s are observed for most station pairs. Based on the relative strength of surface wave trains on the correlated seismograms, we conclude that the dominant ambient noise sources in the central and eastern United States are oceanic micro-tremors from the Atlantic Ocean. The Great Lakes are also found to contribute a significant part of the noise field for stations close to them. The extraction of the Green's function between two stations by cross-correlation can be simply demonstrated by a wave field division method based on acoustic wave propagation. Numerical experiments and mathematical derivations are presented to support this schema.