Ancestral East Antarctic Ice Sheet - Anatomy of a Meltwater Drainage System Beneath it

Australia: The Land Where Time Began

Ancestral East Antarctic Ice Sheet - Anatomy of a Meltwater Drainage System Beneath it

Active meltwater drainage beneath contemporary ice sheets is not often accessible to direct observations, but subglacial hydrology is critical to understanding the behaviour of ice sheets. Simkins et al. identified a palaeosubglacial hydrological system that had been active beneath an area that had previously been covered by the East Antarctic Ice Sheet, by the use of geophysical and sedimentological data from the deglaciated western Ross Sea. They discovered a long channel network that had delivered meltwater at repeated times to the grounding line of an ice stream and was a persistent pathway for meltwater drainage events that were episodic. Embayments within the landforms of the grounding line coincide with the location of subglacial channels which marked reduced degrees of sedimentation and restricted growth of the landform. As a consequence, the degree to which these landforms could provide stability feedbacks to the ice streams was influenced by channelized drainage at the grounding line. The network of channels was connected to subglacial lakes upstream in an area of geologically recent rifting and volcanism, where sufficient basal melting would have produced elevated heat flux to fill the lakes over periods of decades to several centuries; this timescale was found to be consistent with the estimates by Simkins et al. of the frequency of the drainage events at the grounding line that was retreating. Simkins et al. hypothesised, based on these data that in this region the ice stream dynamics were sensitive to the underlying hydrological system.

The behaviour of these ice sheets and their grounding lines were influenced by subglacial processes, at the location of the most downstream ice sheets they are in contact with the underlying bed. Meltwater beneath the ice sheets, in particular, is associated with the onset of fast-flowing ice streams (Peters et al., 2007), shear margins that separate fast ice flow from slow ice flow (Perol et al., 2007), and enhanced deformation of subglacial sediments (Alley et al., 1986; Engelhardt & Kamb, 1997). Meltwater that is stored within subglacial lakes (Palmer et al., 2013; Howat et al., 2015; wright & stegert, 2012) can drain over periods of months to several years (Gray et al., 2005; Wingham et al., 2006; Scambos, Bertheir & Shuman, 2011; Fricker et al., 2016) as a result of changes in the hydrological gradient that are probably triggered by thinning of the ice and the retreat of the grounding line (Scambos, Bertheir & Shuman, 2011; Fricker et al., 2016). Periods of fluctuating and accelerated ice flow downstream of draining subglacial lakes have been interpreted to result from distributed water flow (Scambos, Bertheir & Shuman, 2011; Bell al., 2007; Stearns, Smith & Hamilton, 2008; Stegfried et al., 2016). Decelerated ice flow, in contrast, has been attributed to water pressure that has been lowered, as a result of the drainage of channelized meltwater (Bartholomew et al., 2010; Cowton et al., 2013; Andrews et al., 2014). Therefore, the distribution and movement of meltwater must be well understood in order to assess ice flow changes, though the model and theory have outpaced the acquisition of knowledge of subglacial hydrology that is based on direct observations.

The influence of subglacial meltwater drainage on the dynamics of the grounding line is a key question. Beneath the contemporary Whillans Ice Stream the termination of a subglacial channel coincides with an embayment of the grounding line, at a point where the grounding line is located several kilometres further inland from the adjacent grounding line and erosion of sediment within the channel and the mixing of water alter the behaviour of the grounding line (Horgan et al., 2013; Horgan, Christianson, Jacobel, Anandakrishnan & Alley, 2013). It is not clear if there is a causal relationship between channelised drainage and embayments in the grounding line; there is the possibility, however, that channelised drainage influences the position of the grounding line and the accumulation of sediments, which can reduce the thickness of the ice needed to remain in contact with the bed and even facilitate the advance of the ice (Alley et al., 2007, 2007; Christianson et al., 2016), and according to Simkins this should be explored. Also, subglacial channels that drain at the grounding line can release meltwater plumes that are buoyant which erode channels into the base of the ice shelves thermally (Le Brocq et al., 2013; Alley, Scambos, Stegfried & Fricker, 2016). It is suggested by 1 such example that basal melt rates of more than 1.5 m/year within an ice shelf that is actively forming a channel that is connected to a subglacial channel that has been hypothesised at the grounding line (Marsh et al., 2016). Though it is demonstrated by these observations that the subglacial channels drain at the grounding lines and can be the cause of melting of the ice shelf, what their impact is on the dynamics of the grounding line is still tenuous.

Broader spatial temporal perspectives on subglacial hydrology can be provided by the geological record. On the Antarctic continental shelf there are many palaeosubglacial channels that have been incised into bedrock and are now exposed on the Antarctic continental shelf (Lowe & Anderson, 2003; Nitsche et al., 2013; Anderson & fretwell, 2008; Domack et al., 2006; Campo et al., 2017), though the timing of their incision and occupation by melt water is not well constrained. Surficial subglacial channels on the Antarctic continental shelf that that are sediment-based are temporarily constrained (Wellner, Heroy & Anderson, 2006; Greenwood et al., 2013), but they have not yet been linked to former grounding lines. Simkins et al. provided the first evidence of a subglacial hydrological system that was active during the deglaciation of the LGM and connected specifically to the positions of the grounding line of a former ice stream in the western Ross Sea, by the use of geophysical and sedimentological data.

Conclusion

During deglaciation of the Ross Sea following the LGM an extensive subglacial channel network, that was sediment-based, was reactivated numerous times. Landform growth of the grounding line was restricted locally by channelised meltwater drainage, which consequently contributed to instability of the grounding line on a local scale. Grounding line landforms bisect and buried the channelised segments, which suggests meltwater drainage events occurred episodically, and at periodicities of 10s to several hundreds of years. Subglacial lakes in an area of geologically recent rifting, active volcanism, and elevated geothermal heat flow upstream fed the channel network. It appears that meltwater drainage configuration persisted through various phases of grounding line retreat, shifts in the direction of ice flow, and a circuitous pattern of retreat suggests that a degree of influence was exerted on the retreating ice stream by the stable location of the source lakes and ample supply of basal melting. During construction of the grounding line landform probable recurrence of drainage events suggests that an individual drainage event is not capable of dislodging a stable grounding line. Repeated drainage through embayments does, however, remain a possibility, and grounding line stability may be undermined by the development of pronounced sinuosity of the grounding line and feedbacks with ice melting that are plume-driven.

Sources & Further reading