HR: 0830h
AN: S31E-0811 [PDF]
TI: Short Period Surface-wave Tomography Beneath the Central Tibetan Plateau
AU: * Arias, E
EM: arias@psu.edu
AB:
Data from the PASSCAL experiments, the Tibetan Plateau Passive Source Seismic Experiment in 1991-1992 (TIPLT), INDEPTHIII
experiment in 1997-1999 and BUTHAN and stations LSA and LZH, provide an opportunity to map the subsurface geology and advance
our understanding of the Plateau's dynamics and tectonic evolution. I present the results of the effort to construct
least-squares tomographic maps of short-period Love and Rayleigh wave group velocities across the central and eastern
Plateau. I utilize measurements made using stations and events within the Plateau to localize sensitivity to Plateau
structures. The inversion is based on observations using a set of 190 shallow events recorded on the 29 broadband stations. I
performed a least-square inversion using the conjugate gradient method by Paige and Saunders (1982) at the 6-to 45 s period
range. The ray path coverage and checkerboard tests obtained in this study show evidence of reliable tomographic images from
Rayleigh and Love wave group velocities across central and east of the Tibetan Plateau. The resolved region of the Qiangtang
terrane is one of the slowest features in the model for all periods for both Love and Rayleigh waves. The Lhasa terrane is
relatively fast until perhaps 40-seconds, where slow regions emerge in Rayleigh wave group velocities, particularly close to
the terrane's southern boundary. An interesting feature is the low-velocity region along the southern terrane boundary. The
resolution of such a small feature could be questioned, but it may indicate a region of abnormally low velocities along the
suture, caused perhaps by melt or structural anisotropy. Qualitatively, these observations suggest that the upper crust of
the Qiangtang terrane is slower than the Lhasa terrane to the south, and that the difference may decrease below a few tens of
kilometers, in the middle crust. There is a suggestion of a low-velocity zone at a depth of 20-30 km beneath the Lhasa
terrane, but the conclusion is not strongly supported by the present study. The recent characterizations of the upper-middle
crust of the Lhasa terrane as containing broad regions of partial melt is neither consistent with the regional and global
tomography results nor those of the local-regional tomography of this study. In constrast, the images produced here show the
upper crust beneath the Lhasa terrane to be faster on average than that to the north beneath the Qiantang terrane. Deeper
structures are beyond the reach of the short-period surface waves employed here, but the existence of wide-spread
low-velocity regions as shallow as 12 km is not consistent with the observed group velocities. The use of short-path
observations in this tomographic analysis has illuminated several problems. Foremost is the sensitivity of the observed group
velocities to source location and origin time. By comparing the Eurasian model mean dispersion values for Central Tibet of
Ritzwoller and Levshin (1998) and the mean values from this study the agreement is superb in the range where both data sets
are most reliable and the results deviate slightly at the shortest and longest periods. The use of a large number of data has
allowed us to overcome this limitation in this study as indicated by the consistent mean results after comparing with the
Ritzwoller and Levshin (1998) model, and the correlation of the tomographic images with the terrane geology. Future wok
involves an iterative tomographic and epicenter re-location procedure using surface-wave dispersion values to improve the
event epicenters, and therefore to precisely define the geological structures in the upper crust of the Plateau.
DE: 1734 Seismology
SC: Seismology [S]
MN: 2003 Fall Meeting