technique
could still not give any resolution of the depth distribution of anisotropy. In addition, it can only be
applied to the area with horizontal oriented hexagonally symmetric anisotropy (i.e., transverse isotropy
with a horizontal axis). Thus a new method is needed to give more details of anisotropy.
Because of its sensitivity to velocity discontinuities in the subsurface, receiver function technique
has become a routine procedure used to investigate interior structure of the Earth (LIU et al,
1997; Wu et al, 2007a, b; Owens et al, 1984). Information about Earth’s interior anisotropic structure
can be remained on the receiver functions with a pattern that varies for different back azimuths
on both radial and transverse components. However, the traditional methods, such as migration
imaging and waveform inversion techniques that are based on isotropic media assumption,
cannot effectively extract these information. Only by computing the response based on anisotropic
media assumption, e.g., a model including anisotropy with a symmetric axis of arbitrary orientation,
can we obtain the detailed anisotropy information presented on the radial and transverse receiver
functions. In recent years, plentiful achievements about crust anisotropy using receiver
functions were made by scholars in China and aboard. For example, Levin and Park (1997) studied
the crustal anisotropy in the Ural Mountains foredeep from teleseismic receiver functions and
the result showed that there are significant anisotropy in a low velocity layer at surface (~1 km)
and in the lower crust (33~40 km); Savage (1998) studied the lower crustal anisotropy and dipping
boundaries’ effects on receiver functions in New Zealand; Ozacar and Zand (2004) researched the
crustal seismic anisotropy at BNS station in central Tibet considering both anisotropy and dipping
interface, and showed the implications for deformational style and flow in the crust; waveform
modeling using the neighborhood algorithm (NA) applied to receiver functions computed from 11
PASSCAL stations spanning the eastern plateau of Tibet yielded a suit of crustal models that include
anisotropy by Sherrington et al (2004), and those models suggested that the Tibetan crust
contains 4%~14% anisotropy at different depths and the alignment of the symmetry axis of anisotropy
in the surface layer shows a well relationship with crustal fabrics associated with E-W
trending faults and sutures.
The method of studying crustal anisotropy using the azimuthal variations of receiver functions
can be applied to regions with complex anisotropic feature, from which we can constrain the
lateral and depth distribution of anisotropy, as well as other important parameters. In this paper,
we first introduced the process of computing the response in anisotropic medium with some important
equations, then applied this method to the crustal anisotropy of TMR in North China, and
finally gave the seismic evidence of anisotropy existing within the crust.
1 Reverberations in anisotropic medium
1.1 Transverse isotropy
The symmetry of elastic media plays an important role in studying crystal anisotropy. In
crystallography, the anisotropic medium can be categorized into many kinds of symmetry systems
according to its symmetry plane and axis. Different anisotropic systems have different amount of
elastic coefficients. The higher the symmetry system, the fewer the elastic coefficients are required
to describe the anisotropy. Twenty one elastic coefficients are required for the lowest symmetry
system (triclinic systems) comparing with only two elastic coefficients for the highest (is
