%This view is supported both by 
%fault plane solutions of intermediate depth earthquakes ($\ge$70 km) from 
%southern Tibet that show the similar east-west extension and normal faulting 
%(\cite{seism},\cite{seism1},\cite{seism2})
%as the earthquakes in the upper crust and by the correlation between the
%estimated strain within the crust and the observed shear wave splitting
%that is believed to reflect the strain in the mantle (\cite{strain_split},\cite{strain_split1}).
%At the same time, other evidence suggests that the middle crust of Tibet, at least 
%in some regions, is hot, if not partially molten, and hence very 
%weak (\cite{INDEPTH1996},\cite{Mak1999}).  
%Thus, crustal thinning might be dominated by, if not 
%confined to, a flow in the mid to lower crust 
%(\cite{Bird1991},\cite{ch_flow1},\cite{ch_flow3},\cite{ch_flow4}).
%As we discuss below, the crustal thinning can manifest itself by preferential 
%reorientation of mica crystals into the horizontal plane, which will result in radial
%anisotropy with horizontally polarized shear waves propagating faster than those vertically 
%polarized.