Why are some supraglacial lakes able to drain in merely hours?

Local stresses may open a pathway to the base of an ice sheet.

Shalini Saxena – Jun 11, 2015 1:50 PM | 18

The remains of a lake that has drained into the Greenland ice sheet. Credit: NASA

The Greenland Ice Sheet is the second largest body of ice in the world, covering roughly 80 percent of the island. As the melt season begins, surface meltwater reaches the bed underneath the glacier via conduits on the interior of the ice sheet. This process is often accompanied by the formation and sometimes sudden drainage of supraglacial lakes, bodies of water that form on top of glaciers.

While most supraglacial lakes drain relatively slowly, over many days or months, roughly 13 percent of lakes drain in less than one day, sometimes in a matter of hours. This less common type of lake drainage occurs when a crack forms at the base of the ice sheet and propagates to the surface, resulting in a crack that runs the entire depth of the glacier. These cracks are then opened through hydro-fractures that form directly beneath the lake basin and typically close up after the water has drained, though they can stay open for the melt season if there is continued stream flow.

Currently, scientists do not know what triggers the formation of kilometer length hydro-fractures, though they can only form when the supraglacial lake contains sufficient water volume to continuously fill the fracture as it is propagating from the surface to the bed. However, many lakes that contain the same volumes of water do not exhibit draining, even over multiple summers.

One area that researchers are studying is called North Lake, a supraglacial lake located south of the Jakobshavn-Isbræ catchment. Previous studies have been limited by insufficiently dense observations of motion at the surface of the ice, which are required to constrain the mechanism and location of hydro-fracture initiation. In order to address these limitations, a team examined data from a spatially dense array of 16 GPS stations positioned around North Lake between 2011 and 2013. This array was able to capture the dynamic response of the ice sheet during rapid lake drainage events, which enabled the team to infer information about the hydro-fracture geometry and spatial distribution of the meltwater at the ice-sheet bed before, during, and after drainage.

Ars Video

The GPS data indicates that there is a period of ice motion that occurs when there is a sufficient volume of water reaching the bed. This occurs only hours before each year’s local hydro-fracture initiation and rapid lake drainage. The scientists selected three time points for each drainage, designating them the start of the precursor, hydro-fracture initiation, and the maximum hydro-fracture opening.

Observations reveal that hydrologically induced displacement of the direction of the ice motion, also known as the flow-line, occurs before hydro-fracture events. These alterations occur over 6-12 hour periods in different areas under the lake. After this event, the scientists saw evidence that the main hydro-fracture (4 km long) opened, leading to a subsequent lake drainage over roughly three hours. They also saw that the hydro-fracture opening phase is accompanied by a large uplift at the ice's surface (>20 cm) that is enhanced along-flow-line motion.

The scientists used an algorithm to model the GPS time series of drainage-related surface motion by assessing the deformation caused by the hydro-fracture opening, opening of a cavity located in the glacial basin, and the extra motion of the ice at the base of the glacier. Results from these studies reveal that each drainage event is preceded by a period of enhanced basal slip and/or uplift; the team thinks this can be attributed to the injection of meltwater at the bed via neighboring hydro-fractures and Moulins, vertical well-like shafts that run through glaciers.

The observation of this precursor before rapid lake drainage strongly suggests that the period of motion plays an important role in triggering hydro-fractures. In particular, the team thinks that this motion could induce local stresses that overcome the background levels of compression found closer to the base of the ice sheet.

The scientists decided to test this hypothesis. They compared the background viscous stresses in the lake basin with the elastic stress changes that were caused by the precursory basal slip and cavity opening. Their calculations confirm that the drainage precursors can generate tensile stresses perpendicular to the crack near the surface that are large enough to temporarily overcome the compressive weight of the ice and water, promoting hydro-fracture initiation.

These results, as well as reinterpretations of previous studies, indicate that injection of surface meltwater into the bed is required to trigger hydro-fracture initiation and subsequent rapid lake drainage. Interestingly, in all three years explored in the study, North Lake contained approximately five times the critical volume of water necessary to keep a 4-km long crack open to the bed before hydro-fracturing occurred. This makes it highly unlikely that the volume of water is the exclusive threshold for triggering rapid supraglacial lake drainage.

doi:10.1038/nature14480