Accurately forecasting the contribution of the Greenland Ice Sheet to global sea-level requires precise observations to constrain present-day processes and incorporate them into models. However, the spatial and temporal resolution of satellite imagery and representativeness of in situ measurements often precludes or obscures our understanding of mass loss processes. This thesis investigates whether imagery from unmanned aerial vehicles (UAVs) have the potential to 1) bridge the scale gap between in situ and satellite observations and, 2) resolve processes of mass loss which are beyond the resolution of satellite imagery. It is found that the footprints of ground-based pyranometers are insufficient to capture the spatial heterogeneity of the ice surface as it progressively ablates and darkens. Point-to-pixel albedo comparisons are therefore often invalid, meaning that satellite-derived albedo measurements may be more accurate than previously thought. A 25 km transect intersecting the dark zone reveals that distributed impurities, not cryoconite nor surface water, govern spatial albedo patterns and may have implications for the future expansion of the dark zone. Repeat surveys over Store Glacier show that UAVs can be used to quantify calving rates and surface velocities of tidewater glaciers. The surveys indicate that large calving events cause short-term terminus velocity accelerations and can explain the seasonal pattern of acceleration and retreat. Any process which accelerates calving, such as removal of the ice m ́elange, therefore has important implications for the glaciers future behaviour.

Contemporary observations show an increase in the melting of the Greenland Ice Sheet (GrIS) since the early 21st century. Located near the critical sites of oceanic deep convection and deep water formation, the melting of the GrIS has the potential to directly impact the Atlantic Meridional Overturning Circulation (AMOC) by freshening ocean surface waters in these regions. The majority of the Coupled Model Intercomparison Project Phase 5 (CMIP5) models project a decline in AMOC strength by 10-50% during the 21st century, in response to the increase in atmospheric greenhouse gas (GHG) concentrations. However, due to the simple treatment of polar ice sheets and the lack of a dynamical ice sheet component in these models, these projections likely underestimated the impacts of the GrIS melt, leading to uncertainty in projecting future AMOC evolution and climate change around Greenland. To better understand the impact of the GrIS melt on the AMOC, we perform a series of 21st century projection runs with a state-of-the-art Earth System Model-GFDL ESM2Mb. We consider a medium and a high Representative Concentration Pathway (RCP) scenario (RCP4.5 and RCP8.5, respectively). Unlike the CMIP5-standard RCP runs which included only radiative forcing, the new model experiments are also forced with additional and potentially more realistic meltwater discharge from the GrIS. This meltwater discharge is estimated based on a model-based relationship between the GrIS surface melt and the 500hPa atmospheric temperature anomalies over Greenland. The model simulations indicate that compared to the RCP4.5-only and RCP8.5-only projections, the additional melt water from the GrIS can further weaken the AMOC, but with a relatively small magnitude. The reason is that radiative forcing already weakens the deep convection and deep water formation in the North Atlantic, therefore limiting the magnitude of further weakening of AMOC due to the additional meltwater. The modeling results suggest that the AMOC's sensitivity to freshwater forcing due to the GrIS melt is highly dependent on the location and strength of oceanic deep convection sites in ESM2Mb as well as the pathways of the meltwater towards these regions. The additional meltwater contributes to the minimum surface warming (so-called "warming hole") south of Greenland. These simulations with ESM2Mb contribute to the Atlantic Meridional Overturning Circulation Model Intercomparison Project (AMOCMIP), a community effort between international modeling centers to investigate the impacts of the melting of the GrIS on the AMOC and quantify the associated uncertainty.

This work aims to study the behavior of the western margin of the Greenland Ice Sheet during a period of pronounced ice retreat roughly 10,000 years ago, after the end of the last glacial period. It explores the efficiency of subglacial erosion, the spatial dynamics of ice retreat, and the rates of ice retreat. To address these questions, I use the radionuclides 10Be and 26Al, which form in rocks due to the bombardment of cosmic rays, only after the rocks have been exposed from underneath retreating ice. These nuclides can be used as a geologic dating technique to explore exposure history. Before applying this dating technique to address geological questions, it was critical to first perform methodological development. My work in the University of Vermont‘s new Cosmogenic Nuclide Laboratory served to improve the precision and efficiency of the pre-existing laboratory methods. New methodological advances ensured that samples from Greenland, which contained only low concentrations of 10Be and 26Al, could be used to yield meaningful results about ice behavior. Cosmogenic nuclide dating was applied at two sites along the ice sheet margin in central-western Greenland. At both of these sites, I collected paired bedrock and boulder samples in a transect normal to and outside of the present-day ice sheet margin. Samples were collected from a variety of elevations at numerous locations along the transects, thus providing three-dimensional coverage of the field area. After isolating the mineral quartz from the rocks, and isolating the elements Be and Al from the quartz, isotopic analysis was performed using accelerator mass spectrometry to quantify the relative abundances of the radionuclides against their respective stable isotopes. The southern study site, Ilulissat, is located on the western coast of Greenland at a latitude of 69N. Much previous work has been conducted here due to the presence of one of the largest ice streams in the northern hemisphere, Jakobshavn Isbræ. My work in Ilulissat demonstrated that subglacial erosion rates were high during previous glacial periods, efficiently sculpting and eroding the landscape. Ice retreat across the land surface began around 10,300 years ago, and the ice sheet retreated behind its present-day margin about 7,600 years ago. Ice retreat occurred at a rate of about 100 meters per year. My work in this area suggests that retreat in the large ice stream set the pace and timing for retreat of the neighboring ice sheet margin. The northern site, Upernavik, is located on the western coast of Greenland at a latitude of 73N. Little research has been conducted here in the past. Unlike in Ilulissat, my work here shows that the ice sheet did not efficiently erode the landscape, especially at high elevations, during previous glacial periods. This is likely because the ice was thinner, and therefore had a colder base, than the ice in Ilulissat. My work suggests that ice cover was lost from this area very rapidly, likely at rates of about 170 meters per year, in a single episode around 11,300 years ago. Comparison between the two study sites reveals that ice characteristics can vary appreciably over relatively small distances.

There is increasing recognition that the hydrology of the Greenland Ice Sheet plays an important role in the dynamics and therefore mass balance of the ice sheet. Understanding the hydrology of the ice sheet and being able to predict its future behaviour is therefore a key aspect of glaciological research. To date, the ice sheet’s hydrology has tended to be inferred from the analysis of surface velocity measurements, or modelled in a theoretical, idealised way. This study focuses on the development of a high spatial (100 m) and temporal (1 hour) resolution, physically based, time-dependent hydrological model which is applied to the ~2,300 km2 Paakitsoq region, West Greenland, and is driven, calibrated, and evaluated using measured data. The model consists of three components. First, net runoff is calculated across the ice sheet from a distributed, surface energy- balance melt model coupled to a subsurface model, which calculates changes in temperature, density and water content in the snow, firn and upper-ice layers, and hence refreezing. The model is calibrated by adjusting key parameter values to minimize the error between modelled output and surface height and albedo measurements from the three Greenland Climate Network (GC-Net) stations, JAR 1, JAR 2 and Swiss Camp. Model performance is evaluated in two ways by comparing: i) modelled snow and ice distribution with that derived from Landsat-7 ETM+ satellite imagery using Normalised Difference Snow Index (NDSI) classification and supervised image thresholding; and ii) modelled albedo with that retrieved from the Moderate- resolution Imaging Spectroradiometer (MODIS) sensor MOD10A1 product. Second, a surface routing / lake filling model takes the time-series of calculated net runoff over the ice sheet and calculates flow paths and water velocities over the snow / ice covered surface, routing the water into ‘open’ moulins or into topographic depressions which can fill to form supraglacial lakes. This model component is calibrated against field measurements of a filling lake in the study area made during June 2011. Supraglacial lakes are able to drain by a simulated hydrofracture mechanism if they reach a critical volume. Once water is at the ice / bed interface, discharge and hydraulic head within subglacial drainage pathways are modelled using the third model component. This consists of an adaptation of a component (EXTRAN) of the U.S. Environmental Protection Agency Storm Water Management Model (SWMM), modified to allow for enlargement and closure of ice-walled conduits. The model is used to identify how the subglacial hydrological system evolves in space and time in response to varying surface water inputs due to melt and lake drainage events, driven ultimately by climate data. A key output from the model is the spatially and temporally varying water pressures which are of interest in helping to explain patterns of surface velocity and uplift found by others, and will ultimately be of interest for driving ice dynamics models.

The dynamic response of a faster-flowing Greenland Ice Sheet to climate change is modulated by feedbacks between ice flow and surface meltwater delivery to the basal environment. While supraglacial melt processes have been thoroughly examined and are well constrained, the response of the englacial and subglacial environment to these seasonal perturbations still represent the least-studied, understood, and parameterised processes of glacier dynamics due to a paucity of direct observation. To better understand these processes in the wake of a changing climate, novel in-situ geophysical experiments were undertaken on Store Glacier in west Greenland to quantify rates of englacial deformation and basal melting. The records produced from these experiments yield new insights into the thermodynamic setting of a major outlet glacier, and the physical mechanisms underlying and resulting from fast glacier motion. The deployment of autonomous phase-sensitive radio-echo sounders (ApRES) $\SI{30}{\kilo\metre}$ from the calving terminus of Store Glacier between 2014 and 2016 revealed high rates of both englacial deformation and basal melting, driven primarily by the dynamic response of the basal hydrological system to seasonal surface meltwater influxes. Thermodynamic modelling of this process revealed a convergence of large-scale basal hydrological pathways that aggregated large amounts of water towards the field site. The warm, turbulent water routed from surface melt contained and dissipated enough energy at the ice-bed interface to explain the observed high melt rates. Simultaneously, changes in the local strain field, involving seasonal variations in the morphology of internal layers, were found to be the result of far-field perturbations in downstream ice flow which propagated tens of kilometres upglacier through longitudinal stress coupling. When observed in multiple dimensions, the layer structure revealed complex internal reflection geometries, demonstrating ApRES as not just a monitor of depth changes in ice thickness, but also as an imaging instrument capable of characterising the subsurface environment within and beneath ice sheets. Altogether, the observations and analyses comprising this thesis provide new and significant insight and understanding into the structural, thermal, and mechanical processes tied to Store Glacier and its fast, complex, and dynamic ice flow.

No region has felt the effects of global climate change more acutely than the cryosphere, which has changed at an unprecedented rate in the past two decades. The scientific consensus is that these changes are driven largely by increasing ocean heat content at high latitudes. In southeast Greenland, acceleration and retreat of the marine-terminating glaciers contributes significantly towards global sea level rise. Circulation in the fjords which accommodate these glaciers is thought to be driven both by freshwater input and by barrier wind-driven shelf exchange. Due to a scarcity of data, particularly from winter, the balance between these two mechanisms is not fully understood. In Svalbard, increasing water temperature has decimated sea ice cover in many of the fjords, and had substantial implications for the local ecosystem. While there is a relatively comprehensive literature on shelf exchange mechanisms in Svalbard fjords, questions remain over how the internal circulation interacts with exchange mechanisms. The region shares a similar underwater topography and oceanographic setting with southeast Greenland, with marine-terminating glaciers in close proximity to warm Atlantic waters, and results from Svalbard can hence be used to inform studies of high-latitude fjord-shelf exchange in a broader context. A realistic numerical model was constructed with the aim of better understanding the interaction between Kangerdlugssuaq Fjord and the adjacent continental shelf, and quantifying heat exchange during winter. The model was initially run in an idealised configuration with winter climatological forcing fields, incorporating a parameterisation for melting at the terminus, and used to test the impact of barrier wind events. The Earth's rotation played a crucial role in the nature of the circulation and exchange in the fjord, with inflow on the right (looking up-fjord) and outflow on the left. While the heat delivered into the fjord-mouth was smaller than that observed in summer, the background internal circulation was found to efficiently distribute waters through the fjord without external forcing, and the heat delivered to the glacier terminus was comparable to summer values. Barrier winds were found to excite coastally-trapped internal waves which propagated into the fjord along the right-hand side. The process was capable of doubling the heat delivery. The process also enhanced the background circulation, likely via Stokes' Drift. The model was then adapted to simulate winter 2007-08 under historical forcing conditions. Time series of glacial melt rate, as well as the heat flux through fjord cross-sections, were constructed and compared to the variability in wind forcing. Long periods of moderate wind stress were found to induce greatly enhanced heat flux towards the ice sheet, while short, strong gusts were found to have little influence, suggesting that the timescale over which the shelf wind field varies is a key parameter in dictating wintertime heat delivery from the ocean to the Greenland Ice Sheet. An underwater glider was deployed to Isfjorden, a large fjord system in Svalbard, to measure the temperature, salinity and depth-averaged currents over the course of November 2014. Like in Kangerdlugssuaq, the circulation in Isfjorden was found to be heavily influenced by the Earth's rotation and by wind activity both locally and on the shelf. The combination of hydrography and high-resolution velocity data provided new insights, suggesting that the approach will be useful for studying high-latitude fjords in the future.

The Greenland Ice Sheet is the largest ice sheet in the northern hemisphere. Ongoing melting of the ice sheet, resulting in increased mass loss relative to the longer term trend, has raised concerns about the stability of the ice sheet. Melt water generated at the surface is temporarily stored in supra-glacial lakes on the ice sheet. Connections between melt water generation, storage and ice sheet dynamics highlight the importance of the surface hydrological system. In this thesis different methods are used that improve our ability to observe the supra-glacial lake system on the west Greenland Ice Sheet. This region of the Greenland Ice Sheet has the most extensive supra-glacial hydrological system with a dense network of streams connecting lakes that can exceed several square kilometres in area. Synthetic Aperture Radar (SAR) and visible-near infrared (VNIR) images are used to explore the potential of different sensor systems for regular observations of the supra-glacial lakes. SAR imagery is found to be a useful complement to VNIR data. VNIR data from moderate resolution sensors are preferred as these provide high temporal resolution data, ameliorating problems with cloud cover. The dynamic nature of the lakes makes automated classification difficult and manual mapping has been widely used. Here a new method is proposed that improves on existing methods by automating the identification and classification of lakes, and by introducing a flexible system that can capture the full range of lake forms. Applying our new method we are better able to analyse the evolution of lakes over a number of melt seasons. We find that lakes initiate after approximately 40 positive degree days. Most lakes exist for less than 20 days before draining, or later in the season, and less often, freezing over. Using the automated method developed in this thesis lakes have been mapped in imagery from 2001–2010 at approximately five day intervals. / <p>At the time of the doctoral defense, the following papers were unpublished and had a status as follows: Paper 3: Manuscript. Paper 4: Manuscript. Paper 5: Manuscript.</p>

Atmospheric particulate matter pollution is a challenging environmental concern in both urban and remote locations worldwide. It is intrinsically difficult to control, given numerous anthropogenic and natural sources (e.g. fossil fuel combustion, biomass burning, dust, and seaspray) and atmospheric transport up to thousands of kilometers after production. In urban regions, fine particulate matter (particles with diameters under 2.5 m) is of special concern for its ability to penetrate the human respiratory system and threaten cardiopulmonary health. A second major impact area is climate, with particulate matter altering Earth s radiative balance through scattering and absorbing solar radiation, modifying cloud properties, and reducing surface reflectivity after deposition in snow-covered regions. While atmospheric particulate matter has been generally well-characterized in populated areas of developed countries, particulate pollution in developing nations and remote regions is relatively unexplored. This thesis characterizes atmospheric particulate matter in locations that represent the extreme ends of the spectrum in terms of air pollution the rapidly-developing and heavily populated Pearl River Delta Region of China, the pristine and climate-sensitive Greenland Ice Sheet, and a remote site in the Colorado Rocky Mountains. In China, fine particles were studied through a year-long field campaign at seven sites surrounding the Pearl River Delta. Fine particulate matter was analyzed for chemical composition, regional variation, and meteorological impacts. On the Greenland Ice Sheet and in the Colorado Rocky Mountains, the carbonaceous fraction (organic and elemental carbon) of particulate matter was studied in the atmosphere and snow pack. Analyses included quantifying particulate chemical and optical properties, assessing atmospheric transport, and evaluating post-depositional processing of carbonaceous species in snow.

The recent thinning, acceleration and retreat of tidewater glaciers around Greenland suggests that these systems are highly sensitive to a change in climate. Tidewater glacier dynamics have already had a significant impact on global sea level, and, given projected future climate warming, will likely continue to do so over the coming century. Understanding of the processes connecting climatic change to tidewater glacier response is, however, at an early stage. Current leading thinking links tidewater glacier change to ocean warming by submarine melting of glacier calving fronts, yet the process of submarine melting remains poorly understood. This thesis combines modelling and field data to investigate submarine melting at tidewater glaciers, ultimately seeking to constrain the sensitivity of the Greenland Ice Sheet to climate change. Submarine melting is thought to be enhanced where subglacial runoff enters the ocean and drives energetic ice-marginal plumes. In this thesis, two contrasting models are used to examine the dynamics of these plumes; the Massachusetts Institute of Technology general circulation model (MITgcm) and the simpler buoyant plume theory (BPT). The first result of this thesis, obtained with the MITgcm, is that the spatial distribution of subglacial runoff at the grounding line of a tidewater glacier is a key control on the rate and spatial distribution of submarine melting. Focussed subglacial runoff induces rapid but localised melting, while diffuse runoff induces slower but spatially homogeneous melting. Furthermore, for the same subglacial runoff, total ablation by submarine melting from diffuse runoff exceeds that from focussed runoff by at least a factor of five. BPT is then used to examine the relationship between plume-induced submarine melting and key physical parameters, such as plume geometry, fjord stratification, and the magnitude of subglacial runoff. It is shown that submarine melt rate is proportional to the magnitude of subglacial runoff raised to the exponent of 1/3, regardless of plume geometry, provided runoff lies below a critical threshold and the fjord is weakly stratified. Above the runoff threshold and for strongly stratified fjords, the exponent respectively decreases and increases. The obtained relationships are combined into a single parameterisation thereby providing a useful first-order estimate of submarine melt rate with potential for incorporation into predictive ice flow models. Having investigated many of the factors affecting submarine melt rate, this thesis turns to the effect of melting on tidewater glacier dynamics and calving processes. Specifically, feedbacks between submarine melting and calving front shape are evaluated by coupling BPT to a dynamic ice-ocean boundary which evolves according to modelled submarine melt rates. In agreement with observations, the model shows calving fronts becoming undercut by submarine melting, but hints at a critical role for subglacial channels in this process. The total ablation by submarine melting increases with the degree of undercutting due to increased ice-ocean surface area. It is suggested that the relative pace of undercutting versus ice velocity may define the dominant calving style at a tidewater glacier. Finally, comparison of plumes modelled in both MITgcm and BPT with those observed at Kangiata Nunata Sermia (KNS), a large tidewater glacier in south-west Greenland, suggests that subglacial runoff at KNS is often diffuse in nature. In addition to the above implications for submarine melting, diffuse drainage may enhance basal sliding during warmer summers, thereby providing a potential link between increasing atmospheric temperature and tidewater glacier acceleration which does not invoke the role of the ocean. This thesis provides a comprehensive investigation and quantification of the factors affecting submarine melting at tidewater glaciers, a complex process that is believed to be one of the key influences on the current and future stability of the Greenland Ice Sheet. Based on the magnitude of modelled melt rates, and their effect on calving front shape, the process of submarine melting is a likely driver of retreat at slower-flowing tidewater glaciers in Greenland. For melting to influence the largest and fastest-flowing glaciers requires invoking a sensitive coupling between melting and calving which is as yet obscure. It should however be noted that modelled melt rates depend critically on parameters which are poorly constrained. The results and parameterisations developed in this thesis should now be taken forward through testing against field observations - which are currently rare - and, from a modelling perspective, coupling with ice flow models to provide a more complete picture of the interaction of the Greenland Ice Sheet with the ocean.

The Greenland ice sheet is an area of great interest to the scientific community due to its role as an important bellwether for the global climate. Satellite-borne scatterometers are particularly well-suited to studying temporal changes in the Greenland ice sheet because of their high spatial coverage, frequent sampling, and sensitivity to the presence of liquid water. The dry snow zone is the largest component of the Greenland ice sheet and is identified as the region that experiences negligible annual melt. Due to the lack of melt in the dry snow zone, backscatter was previously assumed to be relatively constant over time in this region. However, this thesis shows that a small seasonal variation in backscatter is present in QuikSCAT data in the dry snow zone. Understanding the cause of this seasonal variability is important to verify the accuracy of QuikSCAT measurements, to better understand the ice sheet conditions, and to improve future scatterometer calibration efforts that may use ice sheets as calibration targets.This thesis provides a study of the temporal behavior of backscatter in the dry snow zone of the Greenland ice sheet focusing on seasonal variation. Spatial averaging of backscatter and the Karhunen-Lo`eve transform are used to identify and study the dominant patterns in annual backscatter behavior. Several QuikSCAT instrumental parameters are tested as possible causes of seasonal variation in backscatter in the dry snow zone to verify the accuracy of QuikSCAT products. None of the tested parameters are found to be related to seasonal variation. Further evidence is given that suggests that the cause of the seasonal variation is geophysical and several geophysical factors are tested. Temperature is found to be highly related to dry snow backscatter and therefore may be driving the seasonal variation in backscatter in the dry snow zone.