Using sea-level and land motion data to develop an improved glacial isostatic adjustment model for the British Isles

Abstract

The Glacial Isostatic Adjustment (GIA) of the British Isles is of interest due to the constraints that can be provided on key model parameters such as regional ice sheet history, viscoelastic earth structure and the global meltwater (eustatic) signal. Many past studies have used as their primary observable the high quality relative sea level (RSL) data from this region. However, as indicated in these studies, the data are notoriously difficult to fit due to the highly complex non-monotonic nature of the observed sea level change. In addition, the model predictions show a strong sensitivity to both regional ice and earth model parameters as well as the global meltwater history, resulting in a high degree of non-uniqueness when seeking an optimum model solution.

The principle aim of this thesis is to address this inherent non-uniqueness in the British Isles GIA problem and reduce the viable solution parameter space by considering additional datasets that display distinct sensitivities to local and global components of the GIA signal. This was achieved in a number of stages, using a combination of relative sea level data from both near and far-field sites, Continuous GPS data and the most recent geomorphological field constraints.

GPS observations of present day vertical crustal motion at sites across Great Britain are employed to provide an independent constraint on the choice of regional earth model parameters (viscoelastic earth structure). It was shown that the data are relatively insensitive to plausible variations in both the regional and global ice model history and as such, the data are used to identify an optimum range of earth model parameters for the region.

Using a set of previously un-modelled far-field sea level data from China and Malay-Thailand, a revised model of eustatic sea level change over the Holocene was developed. Constraining the eustatic component of sea level change is useful since it provides a direct measure of continental ice volume that can be compared to results from oxygen isotope methods. Additionally, inference of the eustatic signal over the mid-late Holocene, which was the primary focus of one aspect of this thesis can provide information on both: (i) the rate and timing of major ice melting at the end of the last deglaciation and (ii) the magnitude of melting during the late Holocene, which is an important baseline that can be compared to estimates of global sea-level rise in the 20th century. This new global ice model is characterised by an initial slowdown in the rate of eustatic sea level rise at 7 kyr BP, followed by a continuing rise, until 2 kyr BP, driven by continued melting from the Antarctic Ice sheet.

A new British-Irish Ice sheet model over the most recent glacial cycle was produced. This model was constrained to fit to most recent geomorphological field constraints and includes an extensive two-stage glaciation across the North Sea basin, with a greatly thickened and extended ice sheet within the Irish Sea, out along the NE Atlantic margin and across Ireland. During deglaciation, the Irish ice sheet now undergoes a very rapid thinning and retreat.

These results have been combined to produce a final new GIA model for the region, which has been constrained, for the first time, using an extended RSL database which includes observations from both Great Britain and Ireland. This new model captures reasonably well the regional trends in the observed sea level, with previous unresolved misfits relating to the timing and height of the Holocene highstand now largely removed.