Plate Boundary Kinematics and Dynamics in New Zealand

I investigate deformation, forces, and material properties of the New Zealand plate boundary zone through a series of kinematic and dynamical models. I use an iterative method based on physical and statistical principles applied to thin-sheet models. I construct a kinematic model of Quaternary long-term deformation to find a velocity field that fits fault slip rate observations, has consistent off-fault strain-rate style, and is constrained by known plate motion. The kinematic model balances on-fault and off-fault deformation and provides improved estimates of fault slip rates and uncertainties in a statistically rigorous manner. I predict shortening rates of 45±8, 34±3, and 20±3 mm/yr across the northern, central, and southern portions of the Hikurangi subduction margin and these rates are better constrained by regional kinematics than estimates from on-fault data. Additionally, the model predicts large strains adjacent to the Alpine Fault, indicating ~9 mm/yr of plate motion in central South Island on faults with currently unknown slip rates or on unknown faults. Differences between my long-term velocity field and a contemporary velocity field arise mainly through interseismic locking on major faults, but in northern North Island <5 mm/yr of missing fault movement is identified near the North Island Dextral Fault Belt.

I build on the kinematic model by creating a smoothed velocity field from it that removes discontinuities at faults. The smoothed velocity field provides input to dynamical models that investigate forces and material properties at the New Zealand plate boundary. Stress magnitudes are estimated at 10–50 MPa and effective viscosities are 0.5–5×10^21 Pa s within actively deforming regions. Models that include only far-field forces and gravitational potential energy can fit observations in most of the plate boundary but basal tractions are required to fit extension in Havre Trough. For models that include non-topographic forces, I specify a rheology to acquire a unique solution. I test three rheologies: a power law with n=3, a power law with n=5, and an equal-stress pseudo-plastic rheology. I predict forces in Havre Trough equivalent to basal tractions of 7–10 MPa at 20 km depth. Models with n=3 and n=5 require antiparallel forces to drive deformation in South Island, implying a plate boundary that becomes more localised with depth, which is not supported by observations. The equal-stress pseudo-plastic model drives deformation with plate motion boundary conditions and localises it with variations in effective viscosity. The effective viscosity of South Island is most realistically modelled by an equal-stress pseudo-plastic rheology, i.e. one with a near constant yield stress.

I construct a further series of dynamical models that include faults. I assume a pseudo-plastic rheology and find a best fit to fault slip rate observations at 20 MPa. Modelled fault tractions are generally 10–20 MPa, similar to stress magnitudes adjacent to faults. Modelled stress orientations and dips suggest that many faults are not optimally oriented for their style of faulting. Notably small tractions are modelled for the central North Island Dextral Fault Belt, which are also modelled as being poorly oriented in relation to the regional stress field. Estimated long-term friction coefficients on faults are generally <0.2, indicating that faults are weaker during slip than undamaged rock surrounding them.