We present detailed records of lake discharge, ice motion and passive seismicity capturing the behaviour and processes preceding, during and following the rapid drainage of a 4 km2 supraglacial lake through 1.1-km-thick ice on the western margin of the Greenland Ice Sheet. Peak discharge of 3300 m3 s−1 coincident with maximal rates of vertical uplift indicates that surface water accessed the ice–bed interface causing widespread hydraulic separation and enhanced basal motion. The differential motion of four global positioning system (GPS) receivers located around the lake record the opening and closure of the fractures through which the lake drained. We hypothesise that the majority of discharge occurred through a 3-km-long fracture with a peak width averaged across its wetted length of 0.4 m. We argue that the fracture's kilometre-scale length allowed rapid discharge to be achieved by combining reasonable water velocities with sub-metre fracture widths. These observations add to the currently limited knowledge of in situ supraglacial lake drainage events, which rapidly deliver large volumes of water to the ice–bed interface.
8 February 2024 in Nature Reviews Earth & Environment
Subglacial hydrological processes at tidewater glaciers remain poorly understood due to the difficulty in obtaining direct measurements and lack of empirical verification for modelling approaches. Here, we investigate the subglacial hydrology of Kronebreen, a fast-flowing tidewater glacier in Svalbard during the 2014 melt season. We combine observations of borehole water pressure, supraglacial lake drainage, surface velocities and plume activity with modelled run-off and water routing to develop a conceptual model that thoroughly encapsulates subglacial drainage at a tidewater glacier. Simultaneous measurements suggest that an early-season episode of subglacial flushing took place during our observation period, and a stable efficient drainage system effectively transported subglacial water through the northern region of the glacier tongue. Drainage pathways through the central and southern regions of the glacier tongue were disrupted throughout the following melt season. Periodic plume activity at the terminus appears to be a signal for modulated subglacial pulsing, i.e. an internally driven storage and release of subglacial meltwater that operates independently of marine influences. This storage is a key control on ice flow in the 2014 melt season. Evidence from this work and previous studies strongly suggests that long-term changes in ice flow at Kronebreen are controlled by the location of efficient/inefficient drainage and the position of regions where water is stored and released.
Concurrent ice sheet surface runoff and proglacial discharge monitoring are essential for understanding Greenland ice sheet meltwater release. We use an updated, well-constrained river discharge time series from the Watson River in southwest Greenland, with an accurate, observation-based ice sheet surface mass balance model of the ∼ 12 000 km2 ice sheet area feeding the river. For the 2006–2015 decade, we find a large range of a factor of 3 in interannual variability in discharge. The amount of discharge is amplified ∼ 56 % by the ice sheet's hypsometry, i.e., area increase with elevation. A good match between river discharge and ice sheet surface meltwater production is found after introducing elevation-dependent transit delays that moderate diurnal variability in meltwater release by a factor of 10–20. The routing lag time increases with ice sheet elevation and attains values in excess of 1 week for the upper reaches of the runoff area at ∼ 1800 m above sea level. These multi-day routing delays ensure that the highest proglacial discharge levels and thus overbank flooding events are more likely to occur after multi-day melt episodes. Finally, for the Watson River ice sheet catchment, we find no evidence of meltwater storage in or release from the en- and subglacial environments in quantities exceeding our methodological uncertainty, based on the good match between ice sheet runoff and proglacial discharge.
Meltwater drainage across the surface of the Greenland ice sheet (GrIS) is well constrained by measurements and modeling, yet despite its critical role, knowledge of its transit through the subglacial environment remains limited. Here we present a subglacial hydrological analysis of a land‐terminating sector of the GrIS at unprecedented resolution that predicts the routing of surface‐derived meltwater once it has entered the basal drainage system. Our analysis indicates the probable existence of small subglacial lakes that remain undetectable by methods using surface elevation change or radar techniques. Furthermore, the analysis suggests transient behavior with rapid switching of subglacial drainage between competing catchments driven by seasonal changes in the basal water pressure. Our findings provide a cautionary note that should be considered in studies that attempt to relate and infer future response from surface temperature, melt, and runoff from point measurements and/or modeling with measurements of proglacial discharge and ice dynamics.
14 May 2015 in Journal of Geophysical Research: Earth Surface
The rapid drainage of supraglacial lakes injects substantial volumes of water to the bed of the Greenland ice sheet over short timescales. The effect of these water pulses on the development of basal hydrological systems is largely unknown. To address this, we develop a lake drainage model incorporating both (1) a subglacial radial flux element driven by elastic hydraulic jacking and (2) downstream drainage through a linked channelized and distributed system. Here we present the model and examine whether substantial, efficient subglacial channels can form during or following lake drainage events and their effect on the water pressure in the surrounding distributed system. We force the model with field data from a lake drainage site, 70 km from the terminus of Russell Glacier in West Greenland. The model outputs suggest that efficient subglacial channels do not readily form in the vicinity of the lake during rapid drainage and instead water is evacuated primarily by a transient turbulent sheet and the distributed system. Following lake drainage, channels grow but are not large enough to reduce the water pressure in the surrounding distributed system, unless preexisting channels are present throughout the domain. Our results have implications for the analysis of subglacial hydrological systems in regions where rapid lake drainage provides the primary mechanism for surface-to-bed connections.