Abstract
Societal adaptation to sea level rise is expected to be one of the defining challenges of the coming decades, potentially affecting hundreds of millions of people and significantly impacting the global economy. The primary driver of sea level rise in this time-frame is land ice loss, but there remains substantial uncertainty regarding the magnitude and pace of these contributions, even under specific climatic warming scenarios. Much of this uncertainty stems from our lack of understanding of glacier dynamics, particularly the processes that control the rate at which ice flows to lower elevations or towards ocean, where it is more vulnerable to melting or calving. As such, improving our understanding of the mechanics that regulate ice flow speeds, and especially those that can trigger sudden increases in ice velocities, is critical for improving predictions of future sea level rise.
This dissertation gains insights into the mechanisms that drive fast ice flow by investigating glacier surges, periodic multi-year increases in ice velocities by up to orders-of-magnitude. As such, surges are powerful examples of rapid change in glacier dynamics. Glacier surges have long been linked to changes in subglacial water volume and pressure that are driven by changes in the drainage efficiency of the subglacial hydrological system. The canonical and decades-old models describing glacier surging remain incomplete; and despite recent academic interest in glacier surging, a detailed image of the spatio-temporal evolution of the subglacial drainage system throughout the surge cycle remains unclear.
The findings in this dissertation stem from modeling and monitoring the pathways of liquid water flow during the surge cycle. Over the course of three chapters, we show through observations and modeling that glacier surging is facilitated by a gradual build-up of water near the glacier bed during the slow, or quiescent, phase of the cycle rather than by a complete disruption of basal drainage during the surge itself. We base this on field observations of efficient basal water flow during a surge of S\'it' Kus\'a, Alaska; along with observations of increasing volumes of water stored near the glacier bed of S\'it' Kus\'a during its subsequent quiescent phase. Our work also identifies multiple modes of oscillation in the evolution of the subglacial drainage system, that modulate ice velocities. Through numerical modeling, we show that simultaneous variability in the drainage system at multiple timescales can account for a wide range of ice velocity patterns. Furthermore, we observe strikingly similar multi-modal oscillations in basal water volume and drainage efficiency at S\'it' Kus\'a. Our observations also hint at spatial heterogeneity in the basal drainage system. Future research should aim at tracking spatially distributed changes in basal drainage and in the connectivity of the drainage system at a higher spatial resolution.Together, the work presented in this dissertation expands our knowledge base on hydrologically driven glacier surging. It shows that there is a viable explanation for unstable ice-flow as primarily driven by change in subglacial hydrology.