LAUSR

S U M M A R Y Beamforming of ambient noise recorded by regional arrays of seismometers is presented as an alternative imaging approach to cross-correlations between pairs of sensors. The method is used to obtain phase velocities and propagation directions of Rayleigh surface waves around the first and second microseism peaks in southern California. The derived velocity maps and propagation directions correlate with major geological structures and changes of the coastal shape in the region. The results are consistent with and complementary to those obtained using cross-correlations of long-duration data between pairs of sensors. Significant advantages of the presented high-resolution adaptive beamforming method over point-to-point noise crosscorrelations are the short time interval of required data (hours to days compared to a year) and robust performance with directive (rather than omnidirectional) noise propagation. Given the recent trend toward dense and large seismic arrays at various scales, the combination of beamforming and noise-correlation processing may provide an optimal strategy for performing noise-based tomography.