Yousefi, Maryam2020-09-302020-09-302020-09-30http://hdl.handle.net/10393/41138http://dx.doi.org/10.20381/ruor-25362The Pacific Coast of Central North America is a geodynamically complex region which has been subject to various geophysical processes. The isostatic response of the Earth to the advance and retreat of ice sheets and glaciers, called glacial isostatic adjustment (GIA), is known to be important on millennial timescales in this area. The other dominant geodynamic process at play along this coastline is that associated with two active plate boundaries. The work presented in this thesis aims to better isolate the signals associated with different geodynamic processes that are observed simultaneously by a variety of data types in order to arrive at a better understanding of these processes. The principal aim of this thesis is to develop an improved GIA model for western North America. In Chapter 2, I quantify the significance of the GIA signal (and its associated uncertainty) by constraining its two fundamental components- the Earth rheology and history of past ice loading- using relative sea-level (RSL) observations. Due to the possible influence of non-GIA processes on the RSL observations, a careful approach was taken in order to confirm that parameter inferences are robust. The results obtained from a 1D (spherically-symmetric) Earth modelling analysis of two- and three-layer mantle viscosity models suggest that there are significant variations in the inferred viscosity structure from north to south. The data-model misfits (and also sensitivity to the ice model) are largest in northern parts of the study area which were once located close to the ice margin. The optimal parameter sets are then used to estimate the GIA contribution to present-day crustal movement and future RSL change. My results suggest that GIA contributes significantly to present-day vertical motion; mostly at those parts of the coastline experiencing forebulge subsidence. The data-model misfits evident in Chapter 2 motivated me to improve the applied Earth model by incorporating, in Chapter 3, lateral variations in viscosity structure., and this work is presented in Chapter 3. Three different approaches are used to construct 3D realizations of the Earth structure. For the first approach, lateral variations in sub-lithosphere viscosity are calculated using four different global seismic tomography models and are used with two models of lateral variations in lithosphere thickness. The other two approaches incorporate regional structure into the Earth model by, in one case, inserting a regional seismic model into two of the global seismic models and, in the other case, by explicitly incorporating the subducting slab, the overlying mantle wedge, and the plate boundary interface. Neither of these approaches reveal any clear preference over the 1D models. However, the two different approaches used to determine 3D Earth structure give us insight to the spread (and thus uncertainty) of modelled RSL values and are a necessary step towards trying other approaches. In Chapter 4, I apply late Holocene RSL observations (over the past 4 kyr) and GPS data (over the past few decades) to infer the associated vertical land motion (VLM) rates along the Pacific coastline of North America. I model and remove the signals associated with 20th-century glacier mass loss and ground water depletion, so that the residual VLM rates are expected to represent the signal due to the Cascadia earthquake cycle. The residual rates compare well with those predicted using a recent locking model of Cascadia subduction, which is encouraging given that the model was developed using only horizontal land motion observations. Performing a sensitivity test on locking model parameters suggest that the residual VLM rates are able to provide useful constraints on key subduction model parameters such as the (near-trench) locking state.enGeodynamicsNorth AmericaGlacial Isostatic AdjustmentTectonicsSea Level ChangeLand MotionCascadia subduction zoneMantle rheologyThesis