Investigating marine-terminating glacier behaviour in Greenland

Abstract

Recent changes at marine-terminating glaciers in Greenland have suggested a dynamic and sensitive response to climate via atmospheric and oceanic warming. Ice loss from marine-terminating glaciers has already contributed significantly to global sea level rise and this is likely to continue with future warming. However, understanding the processes connecting this warming to ice loss from marine-terminating glaciers remains confounded. This is in part due to the complex and dynamic interactions between various potential forcing mechanisms and also due to limited observations of marine-terminating glacier behaviour at sufficiently high spatial and temporal resolutions. This thesis combines remote sensing techniques and field-based observations to investigate ice loss at marine-terminating glaciers at a range of spatial and temporal scales. Here, the overarching aim is to better elucidate the dominant controls on ice loss at marine-terminating glaciers and thus the role they play in the sensitivity of the Greenland Ice Sheet to future environmental change. This thesis first aimed to address one of the primary gaps in observations regarding how key regions of the ice sheet have responded to environmental change, by deriving a detailed record of marine-terminating glacier terminus positions in southeast and northwest Greenland between 2000 and 2015. Through analysing these two key regions of ice loss alongside potential controls on retreat, the ambition was to better constrain the mechanisms driving region-wide change. The results revealed that irrespective of individual spatial or temporal variations in glacier terminus behaviour within or between these two regions, there was an overriding signal of marine-terminating glacier retreat (and thus ice loss) during the 21st Century. Overall, 97% of glaciers experienced terminus retreat, with mean annual retreat rates of -90 m a−1 and -70 m a−1 in the northwest and south-east respectively. This work also showed that changes in bed and fjord geometry were key controls on retreat, which prior to this work had only been indicated at a local scale. Whilst this first results chapter highlighted a clear and dominant signal of retreat and gave an indication to the important controls on this retreat, it was not possible to fully isolate key mechanisms of ice loss by analysing marine-terminating glacier behaviour on large spatial and temporal scales. To further improve understanding of the complex processes driving ice loss, this research next turned to an analysis of marine-terminating glacier behaviour at considerably smaller, and thus more detailed, temporal and spatial scales. Time-lapse imagery was used to generate a very high temporal resolution record of iceberg calving activity at Kangiata Nunaata Sermia (KNS), a large marine-terminating glacier in southwest Greenland. This was combined with meteorological data and modelled estimates of subglacial discharge to investigate the timing and magnitude of ice loss from the glacier terminus over the course of one melt season. The key finding from this work was that the seasonal evolution of subglacial hydrology exerted a critical control on the spatial pattern of iceberg calving across the KNS ice face and ultimately, critically impacted the rate and magnitude of ice loss. More specifically, the subglacial hydrological system was an important control on meltwater plume-focused submarine melting across the KNS ice face. At the start of the melt season, subglacial hydrology was in a relatively distributed state, whereby meltwater emerged through multiple, small portals at the grounding line. This promoted iceberg calving across the entire ice face, likely as a result of instabilities in the calving margin driven by submarine melt-enhanced terminus undercutting. Later in the melt season, the subglacial hydrological system evolved to a relatively more channelised configuration, with a focused discharge of meltwater through fewer portals at the grounding line. This led to a focusing of plume-enhanced submarine melting across the ice face, visible here by changes in terminus planform shape, with the formation of embayments and 'crenelated' terminus geometries. During this time there was also a substantial reduction in terminus-wide iceberg calving, coincident with a ∼50% decrease (0.014 km3 d −1 to 0.005 km3 d −1 ) in frontal ablation flux. Despite the importance of thorough analyses, it is important to recognise that the acquisition of detailed data required to conduct this level of investigation is often not possible. However, the time-lapse work at KNS revealed a potential solution to this problem. An important finding was the changes in the planform geometry of the KNS calving margin, which were strongly controlled by meltwater plume-focused submarine melting. Furthermore, emerging studies have shown morphology to be an important indicator of controls on ice loss, such as bed and fjord geometry. If these and further additional processes of ice loss can be attributed to characteristic morphometric changes observed across glacier calving margins, it may be possible to supplement existing regional scale observations with new types of data. In light of this, the final results chapter utilised a new method for analysing calving front morphology at marine-terminating glaciers. The aim of this work was to establish the potential of using calving front morphology to predict marine-terminating glacier retreat behaviour. However, despite a detailed analysis of the changes in morphology at the terminus of Narsap Sermia in southwest Greenland, during a period of sustained retreat, it was not possible to link specific morphometric change to patterns of ice loss. Whether this reflects the lack of direct process link between terminus planform morphology and calving, or the fact that morphometric changes were a result of interlinked processes and thus could not be isolated to individual controls, remains unclear. The findings of this thesis report on an exploration of marine-terminating glacier behaviour over different spatial and temporal scales. Each investigation has developed new insights advancing our understanding of the processes driving ice loss from marine-terminating glaciers. However, the work has also demonstrated that whilst both large-scale approaches and very detailed analyses are able to elucidate drivers of change, they cannot uniquely isolate key controls. This thesis therefore supports the emerging view that the individual processes of ice loss are too complex and interlinked to isolate effectively and thus, more generalised, large-scale parameterisations of the processes of ice loss may be the best way forward to advance future projections of Greenlandic marine-terminating glacier behaviour. Going forward, regional and ice-sheet wide observations of patterns of marine-terminating glacier behaviour will be essential to our knowledge of the key processes of ice loss, and an improved understanding of the future sensitivity and response of the Greenland Ice Sheet to environmental change.