HR: 17:45h
AN: S34B-08 [Abstracts]
TI: SAsia3D: A New Crustal and Upper Mantle P- and S- Velocity Model in Central and Southern Asia from Joint Body- and Surface-Wave Inversion
AU: * Reiter, D
EM: delaine@westongeophysical.com
AF: Weston Geophysical Corp., 181 Bedford St., Ste 1, Lexington, MA 02420, United States
AU: Rodi, W
EM: rodi@erl.mit.edu
AF: Massachusetts Institute of Technology, Dept. of Earth, Atmospheric, and Planetary
Sciences
77 Massachusetts Ave., Cambridge, MA 02139, United States
AB:
Accurate travel-time and amplitude predictions for regional seismic phases are essential for locating and
characterizing small seismic events with the accuracy needed for explosion monitoring decisions. Parameter
estimates calculated through 3D Earth models have the best chance of achieving acceptable prediction errors, if
the models are constrained by sufficient data. With this motivation, we have developed and applied a joint body-
wave/surface-wave inversion method to produce a new 3D P and S velocity model (SAsia3D) for the crust and
upper mantle to a depth of 400 km in the region of central and southern Asia between 10-50° N and 40-
110° E. The method uses Pn and Pg arrival times to determine the P velocity structure and Rayleigh-wave
group velocities in the period range 10-150 s to constrain the S velocity structure and depth to Moho. The body-
wave and surface-wave inverse problems are coupled through an assumed correlation coefficient between P and
S velocity perturbations and the imposition of bounds on the velocities and Poisson's ratio as a function of depth.
Both body-wave and surface-wave forward modeling are performed in 3D models with the aid of finite-difference
numerical raytracing to calculate body-wave raypaths and 2D raytracing to calculate non-great circle surface-wave
paths. Nonlinearity is addressed by iterating the inversion method with updated raypaths.
The regional P-wave arrival-time observations used to obtain SAsia3D were collected from the Engdahl, van der
Hilst and Buland (1998; EHB) bulletin, restricted to well-located earthquakes in the years 1988-2004. The group-
velocity measurements were provided by groups at the University of Colorado and Lawrence Livermore National
Laboratory. Our initial model for the inversion procedure was taken as a hybrid of the CRUST2.0 3D model
(Bassin et al., 2000) and the upper mantle portion of the global 1D AK135 model (Kennett et al., 1995). SAsia3D
was obtained with four iterations of our technique, achieving a fifty percent variance reduction for both the body-
wave and surface-wave data. Relative velocity (in particular the P velocity) variations with respect to the AK135
mantle model are in good agreement with previous studies and reflect the major tectonic features across
southern and central Asia. For example, at 250 km depth, P variations across new model range from -2.0 to
+2.3%, while S variations vary from -2.1% to +1.5%. We have also noted intriguing differences between the P
and S velocity models in regions of significant tectonic activity, such as the Tibetan Plateau and South Caspian
Basin. Some of these variations may prove useful in explaining the tectonic evolution of the Indo-Asian collision
zone.
We have also completed a number of validation exercises to demonstrate
the accuracy of SAsia3D in regional seismic event location. Most notably, SAsia3D performs well when both
regional P and S phase arrivals are included in the location. The regional P/S location obtained with SAsia3D are
frequently superior to the locations obtained with a large set of teleseismic and regional P arrivals and the AK135
reference model.
DE: 7219 Seismic monitoring and test-ban treaty verification
DE: 7230 Seismicity and tectonics (1207, 1217, 1240, 1242)
DE: 7270 Tomography (6982, 8180)
DE: 7290 Computational seismology
SC: Seismology [S]
MN: 2007 Fall Meeting