How Water, Ice, and Sediment Deform the Earth: Novel Developments and Applications of Models of Glacial Isostatic Adjustment

Abstract

Sea-level change in response to the growth and melt of ice sheets and glaciers is a process called glacial isostatic adjustment (GIA). This includes deformation of the surface of the Earth itself in response to the extreme mass exchanges between the oceans and continents, as well as changes to the gravitational potential that describe the sea surface in response to the redistribution of surface mass as well as mass within the Earth. This thesis describes four research projects I've conducted in the field of GIA modelling. Most GIA models represent the lithosphere, the outermost layer of the Earth, as functionally elastic. However, there is a large temperature gradient within the lithosphere that would lead to a reduction in viscosity with depth. Therefore, in Chapter 2, I developed and incorporated more realistic lithosphere structure into the GIA model, and demonstrate that this added structure results in a time-dependence to the response of the lithosphere. While the usual inputs to a GIA model are the ice load and Earth description, there are regions where other processes need to be accounted for. In the Mississippi Delta region, processes associated with the deposition of sediment carried by the Mississippi River are strong drivers of local sea-level change, and include isostatic adjustment as well as compaction of the sediment layers over time. Therefore, in Chapter 3, I incorporated a treatment of sediment isostatic adjustment into the GIA model and applied it to the Mississippi Delta region. Our results indicate that the sediment isostatic adjustment signal is important in the vicinity of the delta, but small otherwise. By comparing model projections to GPS measurements, we demonstrate that most subsidence in the region is due to non-isostatic processes (such as sediment compaction). Data used to constrain GIA models are generally sensitive to both ice and Earth structure. Therefore data parametrizations that are insensitive to one input or the other are valuable constraints. One such commonly used parametrization is the postglacial decay time. Previous research has shown that the decay times are relatively insensitive to the ice history, and therefore provide a more robust constraint on Earth structure. In Chapter 4 I tested the extent of the ice insensitivity of decay times by considering a suite of ice reconstructions. I found that decay times are sensitive to ice history, and that the sensitivity depends on the location of the data relative to the geometry of the ice sheet. In particular, my results suggest that James Bay (in Hudson Bay) is a location that should not be used in a decay time analysis. The GIA model applied in the projects described above is a 1-D, spherically symmetric model. However, it is known that the Earth's viscous structure is likely to feature significant lateral variation. This is evident in the differences in viscosities found in this thesis between what satisfies the RSL data in Hudson Bay (in Chapter 4) and the Gulf coast of the US (Chapter 3), as well as various previous studies. Therefore, in Chapter 5, I applied a 3-D model with lateral viscous structure determined by seismic shear wave velocity models, to determine whether incorporating this more realistic structure could resolve this apparent discrepancy. I demonstrated that the fit to relative sea level data on the Atlantic and Gulf coasts of the US can be significantly improved by incorporating lateral viscous structure, but also that there is significant uncertainty associated with the more complex viscous structure.