Using the past to inform the future: the role of palaeoglaciology

The study of glaciers and ice sheets has traditionally been split into three communities: those conducting observational research of the present day cryosphere; those developing theory and models to simulate cryospheric processes; and those who use the palaeo record to interpret glaciations of the past. Isolation of these three branches of glaciology can hinder both improved understanding of physical glacial processes and improved interpretations of the palaeo record. In a recently published paper (Greenwood et al., in press) my colleagues and I reviewed the current state of knowledge of one aspect of glaciology (ice sheet hydrology) from the perspectives of these three research communities. The paper further discussed key questions and uncertainties in understanding of the glacial hydrological system, leading to conclusions including: that the traditional bi-modal view of subglacial hydrology is challenged by both modern observations and palaeo interpretations; that we do not fully understand the extent to which different landforms represent different drainage modes or are different expressions of one drainage mode; and that the geomorphic imprint of the multitude of frequencies and magnitudes (figure 1) of drainage events remains poorly understood. These conclusions highlight only some of the many research challenges for the glaciological community which cross-disciplinary cooperation could help to tackle.


Figure 1. Peak discharge circles, scaled by area. The peak discharge of the Glacial Lake Missoula floods is 340000 times greater than the Adventure Trench drainage; like comparing the Amazon River with a small brook. From Greenwood et al. (in press).



Figure 2. Glacial lineations visible on the sea floor: A) rat tails c. 200 m long; B) drumlins c. 1-2 km long; C) mega scale glacial lineations c. 10-20 km long; D) lineations of cross-cutting directionality (pink and blue).

Our understanding of physical glaciological processes, and controls on the stability, dynamics and retreat of ice sheets must continue to improved if we are to better predict ice sheet sea level contribution in response to future climatic change. Accessing the proglacial and subglacial environments of ice sheets to observe processes first-hand is, however, very much restricted due to both logistical and safety constraints. Furthermore, despite the advent of remote sensing, observational records from present-day ice sheets are short (up to tens of years) in comparison to glacial cycles (thousands of years). The glacial landform record, a footprint of previous glaciations left behind on both present day land and the sea floor, provides a window into these environments, and a great deal of insight into the processes operating beneath past ice sheets. Mapping and analysis of glacial landforms can reveal a significant amount of information, including the direction and speed of ice flow, the topology of the subglacial hydrological system, the retreat pattern of the ice sheet margin, and the extent of iceberg calving activity. As an example figure 2 illustrates some of the multitude of forms of glacial lineations, from which both direction and relative speed of ice flow can be interpreted.


Figure 3. Flow path of the Bothnian Sea ice stream inferred from glacial lineations. Longer lineations depicted in green, shorter lineations in red. From Greenwood et al. (2015).

Marine-terminating catchments are the source of much of the mass loss from the Greenland and Antarctic ice sheets, sensitive to changes in both climate and in the ocean. It is thus particularly important to better understand controls on marine ice sheet instability if we are to predict the effect of future atmospheric and environmental change. Recent work in the Gulf of Bothnia region of the past Fennoscandian Ice Sheet gives a unique insight to the dynamics, retreat pattern, and processes operating under a marine-terminating ice stream during the last deglaciation. In addition to ongoing modelling work and data analyses, mapping of glacial lineations allowed us to identify the onset of fast ice flow from the north-east Swedish coast (Greenwood et al., 2015; figure 3). Relating bedform length to relative ice flow speed (Stokes & Clark, 2002), the ice stream width was constrained by a sharp transition between relatively short and relatively long lineations visible on both present day land and the sea floor. Aside from an improved understanding of ice dynamics and retreat in this region of the Fennoscandian Ice Sheet, this research offers a unique perspective on marine-terminating ice stream catchments. The broad, shallow, low salinity setting of the Gulf of Bothnia contrasts with the topographically-constrained continental shelf settings typical of studies of present-day marine outlet glaciers. Palaeoglaciological study can thus improve our understanding of the multitude of responses exhibited by ice sheets, in a range of geographical settings, to external forcings.

In concert with palaeo proxies for reconstructing past climates and environments, glacial landforms can provide a record of long-term response to changing conditions. The study of ice sheets of the past can thus offer important insights to ice sheet response in the future. There is a real danger of isolation between glaciological research groupings, leading to disconnects in understanding of processes important for predicting the future response of present day ice sheets to a warming climate. Fostering cross-disciplinary research is imperative if the true diversity of processes and controls exhibited by contemporary ice sheets are to be understood; using the past to inform the present, and vice versa.


Greenwood, S.L., Clason, C.C., Mikko, H., Nyberg, J., Peterson, G. & Smith, C.A., (2015), Integrated use of LiDAR and multibeam bathymetry reveals onset of ice streaming in the northern Bothnian Sea, GFF, DOI:10.1080/11035897.2015.1055513

Greenwood, S.L., Clason, C.C., Helanow, C. & Margold, M., (in press), Theoretical, contemporary observational and palaeo perspectives on ice sheet hydrology: processes and products, Earth-Science Reviews, doi:10.1016/j.earscirev.2016.01.010

Stokes, C.R. & Clark, C.D., (2002), Are long subglacial bedforms indicative of fast ice flow?, Boreas, 31 (3), 239–249, doi:10.1080/030094802760260355


Paper overview: Modelling the transfer of supraglacial meltwater to the bed of Leverett Glacier, Southwest Greenland

Figure 1. Moulin on Leverett Glacier, southwest Greenland

Figure 1. Moulin on Leverett Glacier, southwest Greenland

The delivery of surface-generated meltwater to the subglacial (basal) environment beneath glaciers and ice sheets is a driver of increased ice flow through raising water pressure in subglacial hydrological systems and lubricating the interface between the base of the ice and the rock or sediment beneath. The extent to which this surface-to-bed meltwater transfer influences the average annual ice velocities of outlet glaciers on the Greenland ice sheet is debated, but we know that at least in the short term, the delivery of meltwater into the subglacial system promotes localised acceleration of ice flow (e.g. Bartholomew et al., 2011a: Hoffman et al., 2011). Our new paper in The Cryosphere describes the application of a model for predicting the formation of moulins (figure 1) and the drainage of supraglacial lakes (surface meltwater ponds) to the bed of the Leverett Glacier catchment in southwest Greenland.

Figure 2. a) Daily average air temperatures at 457 m a.s.l. for 2009 and 2010, and moulin formation and lake drainage through the b) 2009 and c) 2010 melt seasons with elevation. Blue diamonds in panel b) represent observed drainage of lakes in events between two MODIS images, and red diamonds represent lakes which drained over a period of several MODIS images; after figure 2a of Bartholomew et al. (2011a).

Figure 2. a) Daily average air temperatures at 457 m a.s.l. for 2009 and 2010, and moulin formation and lake drainage through the b) 2009 and c) 2010 melt seasons with elevation. Blue diamonds in panel b) represent observed drainage of lakes in events between two MODIS images, and red diamonds represent lakes which drained over a period of several MODIS images; after figure 2a of Bartholomew et al. (2011a).

The model can be split into three main components: the generation of meltwater at the surface through melting of snow and ice; the routing of meltwater across the ice surface and storage in supraglacial lakes; and the propagation of fractures through the ice driven by water pressure from meltwater inflow. The outputs of modelling provide information on the timing and location of moulins penetrating from the surface to the bed of the ice, the timing and location of the drainage of lakes through hydrofracture, the storage of meltwater in the supraglacial and englacial (internal) hydrological systems, and the quantity of meltwater reaching the glacier bed.

The results of our modelling were compared with ice surface velocities measured with a series of 6 GPS stations (Bartholomew et al., 2011a) and observations of supraglacial lake drainage events (Bartholomew et al., 2011b) in the Leverett glacier catchment. The pattern of modelled lake drainages compared favourably with that observed from satellite imagery during 2009 (figure 2) . Modelling for the 2010 melt season produced a 44% increase in the number of lakes which drained, with lake drainages starting earlier in the season and occurring more frequently at higher elevations in comparison to the cooler 2009 melt season. Formation of moulins in both years occurred at increasing elevation with time into the melt season, reflecting retreat of the snowline and water availability for driving fractures through thicker ice inland of the margin.

Figure 3. Supraglacial meltwater delivered to the bed each day through modelled lake drainages, moulins, and for the control simulation within ice surface elevation bands of 250 m during 2009. Ice surface velocities from GPS sites 1 – 6 are plotted within their respective elevation bands (after Bartholomew et al., 2011b).

Figure 3. Supraglacial meltwater delivered to the bed each day through modelled lake drainages, moulins, and for the control simulation within ice surface elevation bands of 250 m during 2009. Ice surface velocities from GPS sites 1 – 6 are plotted within their respective elevation bands (after Bartholomew et al., 2011b).

To demonstrate the need to account for supraglacial storage and routing of meltwater, and the ability of that water to reach the ice sheet bed through hydrofracture, we conducted a control simulation where all surface-generated meltwater was transferred locally and instantaneously to the bed. The importance of accounting for this was particularly apparent above 750 m a.s.l. (figure 3), where discharge modelled including supraglacial storage, routing and hydrofracture greatly outperformed the control simulation when compared against measured ice surface velocities. Modelled surface-to-bed meltwater transfer characterises well the up-glacier progression of increased ice surface velocities, with the relative contribution of meltwater drained through lakes increasing with elevation.

Figure 4. Density of moulins and lake drainages predicted for the IPCC (2007) A1B mean June, July and August Arctic scenario for temperature change within 250 m ice surface elevation bands.

Figure 4. Density of moulins and lake drainages predicted for the IPCC (2007) A1B mean June, July and August Arctic scenario for temperature change within 250 m ice surface elevation bands.

With meltwater production rates responding to increases in air temperature, we tested our model with a future climate scenario (IPCC 2007, A1B June, July, August Arctic scenario) to investigate possible future changes in the occurrence of ice surface-to-bed meltwater transfer. The density of moulins and drained lakes at higher elevation in the Leverett catchment increased in response to this temperature forcing (figure 4), giving a 68 % increase in moulin numbers, 182 % increase in the drainage of supraglacial lakes, and, most importantly, a 14 % increase in the proportion of surface-generated meltwater reaching the glacier bed. At elevations above 750 m a.s.l. these increases result in more widespread delivery of meltwater to the bed through ice of large thickness, beginning earlier in the melt season.

While below the ELA (equilibrium line altitude – the average elevation on a glacier where snow accumulation equals ice melt) ice dynamic response to meltwater inputs may be negligible over annual time scales (e.g. Sole et al., 2013), above the ELA we may expect a positive relationship between summer temperatures (and melt rates) and ice surface velocities where efficient subglacial drainage of meltwater is hindered by large ice thicknesses and shallow ic surface slopes (e.g. Meierbachtol et al., 2013). Crucially, the implications of a warming climate and increased ice surface melt for subglacial drainage configuration and ice dynamics are difficult to assess fully without incorporating a model such as this into models of subglacial drainage and ice flow. This work thus contributes to efforts to couple physically based models of surface meltwater generation, subglacial hydrology and ice sheet dynamics, as are required to better understand past, current and future sensitivity of ice sheet mass balance and dynamics to climate change.

The full paper is available to read here.


Bartholomew, I. D., Nienow, P., Sole, A., Mair, D., Cowton, T., King, M. A., and Palmer, S., (2011a), Seasonal variations in Greenland Ice Sheet motion: Inland extent and behaviour at higher elevations, Earth and Planetary Science Letters, 307, 271–278

Bartholomew, I., Nienow, P., Sole, A., Mair, D., Cowton, T., Palmer, S., and Wadham, J., (2011b) Supraglacial forcing of subglacial drainage in the ablation zone of the Greenland ice sheet, Geophys. Res. Lett, 85, L08502

Hoffman, M. J., Catania, G. A., Neumann, T. A., Andrews, L. C., and Rumrill, J. A., (2011), Links between acceleration, melting, and supraglacial lake drainage of the western Greenland Ice Sheet, Journal of Geophysical Research: Earth Surface, 116, F04035

Meierbachtol, T., Harper, J., and Humphrey, N., (2013), Basal drainage system response to increasing surface melt on the Greenland ice sheet, Science, 341, 777–779

Sole, A., Nienow, P., Bartholomew, I., Mair, D., Cowton, T., Tedstone, A., and King, M. A., (2013), Winter motion mediates dynamic response of the Greenland Ice Sheet to warmer summers, Geophys. Res. Lett., 40, 3940–3944

New paper: Dye tracing for investigating flow and transport properties of hydrocarbon-polluted Rabots glaciär, Kebnekaise, Sweden

Our new paper about the application of dye tracing to investigate the flow of pollutants on Rabots glaciär has just been published online in Hydrology and Earth System Sciences Discussions, and you can find a link to it here.




Over 11 000 L of hydrocarbon pollution was deposited on the surface of Rabots glaciär on the Kebnekaise Massif, northern Sweden, following the crash of a Royal Norwegian Air Force aircraft in March 2012. An environmental monitoring programme was subsequently commissioned, including water, snow and ice sampling. The scientific programme further included a series of dye tracing experiments during the 2013 melt season, conducted to investigate flow pathways for pollutants through the glacier hydrological system, and to gain new insight to the internal hydrological system of Rabots glaciär. Results of dye tracing reveal a degree of homogeneity in the topology of the drainage system throughout July and August, with an increase in efficiency as the season progresses, as reflected by decreasing temporary storage and dispersivity. Early onset of melting likely led to formation of an efficient, discrete drainage system early in the melt season, subject to decreasing sinuosity and braiding as the season progressed. Analysis of turbidity-discharge hysteresis further supports the formation of discrete, efficient drainage, with clockwise diurnal hysteresis suggesting easy mobilisation of readily-available sediments in channels. Dye injection immediately downstream of the pollution source zone revealed prolonged storage of dye followed by fast, efficient release. Twinned with a low dye recovery, and supported by sporadic detection of hydrocarbons in the proglacial river, we suggest that meltwater, and thus pollutants in solution, may be released periodically from this zone of the glacier hydrological system. The here identified dynamics of dye storage, dispersion and breakthrough indicate that the ultimate fate and permanence of pollutants in the glacier system is likely to be governed by storage of pollutants in the firn layer and ice mass, or within the internal hydrological system, where it may refreeze. This shows that future studies on the fate of hydrocarbons in pristine, glaciated mountain environments should address the extent to which pollutants in solution act like water molecules or whether they are more susceptible to, for example, refreezing into the surrounding ice, becoming stuck in micro-fractures and pore spaces, or sorption onto subglacial sediments.



Evaluating the environmental impact of a plane crash on a melting glacier in Arctic Sweden

Figure 1. Clean-up operation in the accumulation zone of Rabots glaciär

On the 15th March 2012 a Royal Norwegian Air Force Lockheed Martin C-130J Super Hercules aircraft crashed into the western face of Kebnekaise (Sweden’s highest mountain) in Lapland, during a military exercise. In addition to five fatalities, the crash deposited some 11000 litres of kerosene jet fuel across the mountain wall, snow and ice, which was then further distributed following a large avalanche triggered by the crash. Debris from the crash has been found on Storglaciären, Björlings glaciär, and near the summit of Kebnekaise, but the majority of debris and fuel was deposited on Rabots glaciär, and the pollution was not subject to any immediate decontamination efforts. While much of the aircraft debris has now been removed during clean-up exercises by the Swedish Military (Figure 1), little was known about the fate of the hydrocarbon pollutants.

Figure 2. Rabots glaciär location and pollution source zone (from Clason et al., submitted)

Figure 2. Rabots glaciär location and pollution source zone (from Clason et al., submitted)

Rabots glaciär is subject to ongoing terminus retreat and thinning, and responds more quickly to climatic changes than neighbouring Storglaciären (Brugger, 2007). Although the Rabots glaciär catchment is in a very remote location, the glacier meltwater outlet (proglacial river) feeds rivers used for drinking water by numerous backpackers and the local Sami, who work with reindeer husbandry along the banks of the Kaitum river system. With a possibility for environmental, social and economic impacts arising from water pollution in this region, it is imperative to evaluate the evolution of the polluted area and mechanisms for hydrocarbon pollutant dispersal from the source zone (Figure 2) in the accumulation area of Rabots glaciär. To tackle this, a monitoring programme has been run by researchers at Tarfala Research Station and Stockholm University (led by Gunhild Rosqvist, myself, and Jerker Jarsjö), including repeat sampling in the snowpack, ice/firn, and the proglacial river for chemical analysis to detect jet fuel components. To investigate the pathways and transit times of pollutants from the source zone (Figure 2), a series of dye tracing tests were conducted as a proxy for flow of pollutants in solution through the glacier system (Clason et al., submitted). In these tests an instrument called a fluorometer detects the emergence of the dye in the proglacial river, following initial injection into flowing water above a moulin (Figure 3), a vertical shaft in the ice which provides entry for meltwater to the internal and subglacial drainage systems. Dye tracing experiments provide an indirect method for understanding what the hydrological system inside and underneath a glacier may look like, how fast and efficiently meltwater flows through a glacier, and the extent to which meltwater is stored during its journey through glacier.


Figure 3. Injection of the water tracing dye rhodamine into a supraglacial stream directly above a moulin on Rabots glaciär.

The results of dye tracing revealed likely storage of meltwater and pollutants near the source zone within or beneath the glacier, released in pulses when sufficient water flux from precipitation or melting flushes out the hydrological system. This was supported by infrequent, sporadic detection of pollutants in the proglacier river. The levels of pollutants detected are not thought to currently be a threat to the drinking water supply, but under conditions of higher melting in a warming climate, the frequency and levels of pollution released into the river system could increase. Based on the solubility of the hydrocarbon compounds of the fuel, the lifetime of these compounds in the source zone on Rabots glaciär was modelling with a coupled melt-mass flux modelling approach (Clason et al., in preparation). Assuming no change in summer melting conditions (based on 2013 meteorological data), and perfect contact between the meltwater and the fuel compounds, some of the lighter hydrocarbon compounds have likely already left the glacier. However, heavier components have the potential to persist for tens of thousands of years, far outliving the glacier itself.

Under warming climate scenarios we may see increased interaction of these jet fuel compounds with the surrounding pristine Arctic environment, with potential not only to affect drinking water supplies, but also the local ecosystem and even microbial life on the ice. With increasing use of snowmobiles and helicopters in this region, evaluation of both the long and short term impacts of hydrocarbon pollutants must continue.

Brugger, K. A., (2007), The non-synchronous response of Rabots Glaciär and Storglaciären, northern Sweden, to recent climate change: a comparative study, Annals of Glaciology, 46, 275-282.

Clason, C.C., Coch, C., Jarsjö, J., Brugger, K., Jansson, P. & Rosqvist, G., (submitted), Dye tracing for investigating flow and transport properties of hydrocarbon-polluted Rabots glacier, Kebnekaise, Sweden, Hydrology and Earth System Sciences

Clason, C.C., Jarsjö, J., Rosqvist, G. & LaBianca, A., (in preparation), Glacier melt modelling and mass flux of hydrocarbon pollutants on Rabots Glacier, Kebnekaise