Search

InSAR is a useful technique to detect large-scale surface deformation from space. To place constraints on the rheological structure of the lithosphere in the Tibetan Plateau, two strike-slip earthquakes have been investigated. One is the Mw 7.6 Manyi earthquake, which occurred in the north-central Tibetan Plateau. The other is the Mw 6.5 Jiuzhaigou earthquake, which happened that on the eastern part of the Tibetan Plateau. My InSAR data cover 12 years following the Manyi earthquake, much longer than previous researchers' dataset. I test three viscoelastic models (Maxwell, Standard linear solids, and Burgers body) and one afterslip model. The viscoelastic models cannot match the observed temporal-spatial deformation patterns. The distributions of deformation in the viscoelastic models extend into the far field and the residuals tend to increase, which are inconsistent with the data. The afterslip model has the lowest misfit and explains the temporal and spatial pattern of the observed deformation with decent result. A combined model that considers the effects of both afterslip and viscoelastic relaxation has also been tested. In this combined model, the viscoelastic relaxation that occurs with an elastic layer of thickness of 30 km over a half-space place, produce an estimate for viscosity of 5 × 1019 Pa s for this area. Therefore, either the afterslip model or the combined model can be used to explain the 12 years postseismic deformation of Manyi earthquake. The long time series of the Manyi earthquake enable us to distinguish between afterslip and viscoelastic relaxation. The seismogenic fault of the Jiuzhaigou earthquake was previously unidentified and no surface rupture is found after the earthquake. I first determined the fault geometry and calculated coseismic slip model. The slip model indicates a left-lateral strike-slip pattern, which is consistent with focal mechanisms were determined by different agencies. There is no visible postseismic deformation signal of the fault, which means the surface deformation generated by fault creeping is smaller than the noise of our observation method over that period. Therefore, I try to find the lower bound of the viscosity for this area. My preferred minimum possible viscosity of the underlying half-space is ∼6 × 1017 Pa s. Together with previous geodetic studies, the viscosities obtained from central Tibet show at least one order of magnitude difference with the viscosities obtained from the eastern Tibet. The heterogeneity indicates the rheology has a relatively large spatial change through the whole Plateau. The viscoelastic model always been proposed to explain long-term postseismic deformation and afterslip is used to explain the short-term deformation or localised deformation. Sometimes, the viscoelastic deformation signal is invisible in the moderate earthquakes as the stress is not large enough to generate observable deformation.