Ice stream sticky spots: A review of their identification and influence beneath contemporary and palaeo-ice streams

Abstract

Rapidly-flowing ice streams are the arterial drainage routes in continental ice sheets and exert a major influence on ice sheet mass balance. Recent observations have revealed that ice stream flow exhibits considerable variability, with relatively rapid changes taking place in speed and direction. This spatial and temporal variability is intimately linked to the conditions at the base of the ice streams and the distribution of localised patches of basal friction, known as ‘sticky spots’. In this paper, we provide a detailed review of sticky spot observations from both contemporary and palaeo-ice stream beds in order to better understand their nature and influence. Observations and theoretical considerations reveal four primary causes of ‘stickiness’: (i), bedrock bumps; (ii), till-free areas; (iii), areas of ‘strong’ (well drained) till; and (iv), freeze-on of subglacial meltwater. These may act together in one location, or in isolation; and a progressive increase in their distribution could lead to ice stream shut-down. Bedrock bumps are influential under active ice streams, where they provide form drag and can create thinner ice which increases the likelihood of basal freeze-on. Increased bed roughness may prevent the lateral migration of some ice streams but bedrock bumps are unlikely to cause ice stream shut-down because, over long time-scales, ice stream erosion might be expected to reduce their amplitude. The influence of till-free areas beneath an ice stream will depend critically on the amount of water that might be drawn out of the surrounding till to lubricate such areas. They are likely to be most important in ice stream onset zones but their identification has proved difficult beneath active ice streams. If an ice stream operates solely by till deformation, it is conceivable that a progressive increase in the exposure of till-free areas could lead to shut-down through a process of sediment exhaustion. Areas of strong, well drained till have been identified beneath both active and ancient ice streams and are most likely to result from the reorganisation of subglacial meltwater. The collapse of an inefficient ‘cannalised’ system to a more efficient ‘channelised’ system can occur rapidly and this mechanism has been hypothesised as a candidate for ice stream shut-down in both contemporary and palaeo-settings. Basal freeze-on has also been observed and inferred from beneath modern and palaeo-ice streams, and a reduction in basal meltwater supply coupled with ice stream drawdown and the advection of cold ice increases the likelihood of switching off an ice stream. A paucity of data from ice stream sticky spots limits a better understanding of their nature, distribution and evolution beneath ice streams. Future technological advances are likely to improve the resolution of the data collected from the beds of modern ice streams but well-preserved palaeo-ice stream beds also hold potential for investigating their influence on ice stream flow and we present simple landsystems models to aid their identification. Such data will considerably enhance the basal boundary condition in ice stream models which will, ultimately, refine our predictions of the response of contemporary ice sheets to future changes in climate.