Deformation of Tibet: InSAR analysis and viscous flow models

The Tibetan plateau in central Asia is a prime example of the distributed deformation that occurs in the lithosphere as a result of continental collision. Large scale lithospheric deformation can be estimated using viscous continuum models that balance the vertical stress induced by lateral variations of potential energy, and horizontal stress induced by tectonic boundary forces. I find that the 2-dimensional Thin Viscous Sheet (TVS) model gives a good approximation to deformation during continental collision, providing that the indenter half-width is greater than the lithospheric thickness. However even when this ratio approaches one, reasonable correspondence exists when the strain-rate exponent (n) of the rheological constitutive law is ≤ 3. By applying the TVS to model the contemporary deformation of Asia, I find that the first order features of the geodetically-determined velocity field can be explained. Models which can best predict the observed velocity field have n between 2 and 5 Argand numbers of between 1 and 4, and the strength of the Tibetan plateau and Tien Shan is between 3 and 8 times weaker than the foreland regions. Models with these parameters give a value of FL = 7-15x1012 N m-1 for the vertically integrated horizontal driving force on the Himalayan arc. I describe the π-rate method for determining slow linear deformation rates from Interferometric Synthetic Aperture Radar (InSAR) observations and validate it using synthetic data. When using real data, the π-rate method out-performs the conventional stacking method. The RMS difference between the two methods and observed GPS measurements are 3.7 and 7.1 mm/yr for π-rate and stacking respectively. I used the ~rate method to determine the interseismic velocity field across the Tibetan plateau in an approximately north-south orientated, ~1000 km-long swath using ESA Envisat data spanning a period of 6.23 years. The resulting InSAR rate map indicates a factor of 2 variation in the magnitude of line-of-sight (LOS) velocity between the latitudes of 29-40oN. Significant localisation of deformation around mapped fault zones is not observed. A deviation of up to 8 mm/yr in LOS between the InSAR rate map and GPS-derived horizontal velocity field suggests either ~8 mm/yr of vertical uplift, an additional ~20 mm/yr of eastward motion, or a combination of horizontal and vertical motion that has not been measured using horizontal-component campaign GPS data. Comparison of InSAR and GPS observations with predictions of kinematic block and viscous continuum models suggests that the latter provides a more useful description for large-scale continental deformation.