For usarray_calif_cub.pdf:
-------------------------

Ambient noise Rayleigh wave tomography across Southern
California using 62 stations (shown with triangles)
from the Transportable Array (TA)
component of USArray. Surface wave Green functions are
computed from raw data by cross-correlating 1 month
time series (Aug 2004) for each pair of TA stations. 
Rayleigh wave group speed as a function of period is 
measured on each Green function. (a) - (c) Tomographic 
maps produced at three periods by ambient noise tomography. 
Near 7.5 sec, Rayleigh waves are most sensitive to the 
existence of sediments and the major sedimentary basins
in Southern California are seen (San Joaquin Valley,
Ventura Basin, LA Basin, Imperial Valley/Salton Trough).
Wave sensitivity deepens as period increases so that
by 18 sec period the sensitivity is predominantly to the
middle and deep crust. The deep crustal substructure of the
Sierras and Peninsular Range is revealed.
(d) The 18 sec group speed map from the University
of Colorado global 3-D model of the crust and upper
mantle is shown for comparison.


Note: This work has been published. The reference is:

Shapiro, N.M. M. Campillo, L. Stehly, and M.H. Ritzwoller,  
High resolution surface wave tomography from ambient seismic 
noise, Science, 307(5715), 1615-1618, 11 March 2005.


For backbone_us_cub.pdf:
------------------------

Ambient noise Rayleigh wave tomography across the US
using 126 broad-band stations (shown with triangles) acquired
from the IRIS/DMC. Surface wave Green functions are
computed from raw data by cross-correlating 4 month
time series (Nov 2003 - Feb 2004)) for each pair of stations. 
Rayleigh wave group speed as a function of period is 
measured on each Green function. Tomography here is
for the 16 sec Rayleigh wave which is sensitive
predominantly to shear wave speeds in the middle and
lower crust. (a) The 4508 inter-station paths with 
acceptable group speed measurements. (b) The
starting model, the 16 sec group speed map from the 
University of Colorado global 3-D model of the crust 
and upper mantle. (c) The Rayleigh wave group speed map
that emerges from ambient noise tomography, producing
a 65% variance reduction relative to the starting
model. The map demonstrates finer scale features across 
the US than imaged by traditional methods of surface 
wave tomography based on teleseismic earthquakes.


Note: This work has not been published yet or even
submitted for publication. It remains preliminary
in nature.