Subglacial drainage and patterns of surge propagation in Svalbard

During glacier surges, regions of fast flow can propagate down, up, or across the glacier. In most models of surging, propagation of fast-flowing regions is attributed to an increase in the spatial extent of stored water at the bed, although the processes controlling spatial and temporal patterns of surge propagation are not well understood. Using a combination of historical observations, landform mapping, and direct exploration of glacial conduits, we show how a persistent subglacial drainage system has constrained surge propagation in two tributary fjord-terminating glaciers in Svalbard. Bakaninbreen, the right-hand tributary, surged during the 1980s and 1990s. The surge bulge propagated at least 7 km downglacier, but the lateral boundary of the surging ice closely followed the medial moraine separating Bakaninbreen from the left-hand tributary, Paulabreen, throughout the event. Paulabreen surged in 2005. Again, the surge did not propagate across the medial moraine, although the region of fast flow propagated ~ 5 km down Paulabreen.
Several lines of evidence show that the pattern of surge propagation is linked to the presence of a persistent subglacial conduit below the medial moraine. First, during retreat of the Bakaninbreen-Paulabreen system in the 20th C, upwelling water was observed in a deep embayment in the glacier margin in front of the medial moraine. Second, detailed bathymetric mapping has revealed an esker running the full length of the fjord, coincident with the recorded positions of the embayment. Third, at the confluence between Bakaninbreen and Paulabreen, summer meltwater drains into a moulin and rapidly reaches the bed.
We argue that this persistent conduit acts as ’firewall’ between the glacier beds during surges. If a distributed drainage system below one branch of the glacier system reaches the conduit, it can be readily evacuated, preventing its propagation into the other branch. Escape of water along the conduit will not cause surge termination if water recharge at the bed (due to strain heating) is high enough to maintain water pressures in the surging part. It is notable that the subglacial conduit persisted during a long period in which both glaciers were quiescent, when the bed was presumably below pressure-melting point. However, we have frequently encountered water in englacial and subglacial passages in cold ice in wintertime in Svalbard, and attribute its presence to solute enrichment during partial freezing.