Examining the consistency between subglacial hydrology and basal friction inversion modeling for Slessor Glacier, East Antarctica.

Abstract

In light of rising global surface temperatures and sea-level rise, it is more important now than ever to understand what role the cryosphere will play in the Earth's evolution. The Antarctic ice sheet contains enough ice to raise the global sea-level by approximately 58 meters and for this reason alone, it is essential for scientists to be able to predict how the ice masses within will behave in the future. One way to study the future behaviour of glaciers and ice sheets is by applying geospatial data to mathematical models based on the relevant physics. In this thesis, the Glacier Drainage System model (GlaDS) and the Ice-sheet and Sea-level System Model (ISSM) are used to model subglacial hydrology and ice dynamics, respectively, for Slessor Glacier, East Antarctica. First, an in-depth description of the necessary physics and numerical framework governing the two models is presented. With the framework in place, a sensitivity test comprised of 48 transient runs is performed with the GlaDS model, and 13 inversion simulations are performed with ISSM. The sensitivity test consists of altering several poorly constrained parameters to understand their impact on the modeled hydrological network beneath Slessor Glacier. The results from the sensitivity test are then used to determine which model configuration is most appropriate based on the current understanding of subglacial networks beneath the Antarctic Ice Sheet, however, this is a limited method of model validation in the absence of observed data such as specularity content (data derived from geophysical radar surveys to determine locations of distributed subglacial water). To mediate this issue, the model outputs from the inversion simulations, observed ice sheet melt rates, and a hydrostatic inversion of high resolution surface data are used to validate the model outputs. This is followed by a suggested workflow that can be adopted by modelers to use inverse methods to validate subglacial hydrology model outputs. The model outputs from this study suggest an active subglacial hydrological network beneath Slessor Glacier and the surrounding area. There is good agreement between areas of fast observed ice flow and areas where the model predicts deep water, low effective pressure and an efficient drainage network. These results are consistent with areas of inferred low basal friction coefficient from the ice dynamics model, which also recovers observed velocities from a stress balance simulation. The results of this thesis demonstrate some control of basal hydrology on ice dynamics in the Slessor Glacier study area. Furthermore, the methods used here provide subglacial hydrology modelers an additional means of model validation, which is valuable where observed data is sparse or not available.