The distribution of glacier surge behaviour in Svalbard and implications for understanding unstable ice flow

Glacier surges are periods of significantly increased ice flow due to ice-dynamic feedbacks, in contrast to more conventional advances or other responses due to changes in mass balance. In the Arctic, a ring of surging glacier clusters can be found extending from Alaska-Yukon to Novaya Zemlya. The ‘Arctic ring’ encapsulates Svalbard, an archipelago with a long history of glaciological observations and consequently measurements of glacier surges. However, estimates of the number of surge-type glaciers across the archipelago range between 10% and 90% depending on the classification technique used. To better understand the causes, drivers and impacts of glacier surges in Svalbard, improved monitoring is required and new techniques developed to extend the observational record of active surge dynamics. In this contribution, we review the benefits and limitations of different approaches for monitoring and detecting glacier surges in Svalbard. We use this to compile a new database of surge-type glaciers in Svalbard, which also contains data on surge characteristics e.g. terminus change and velocity. We find that 36% of glaciers in Svalbard have displayed surge-type behaviour throughout our observational and landform record, rising to 51% when removing glaciers smaller than 1 km². Of all the glaciers in Svalbard, only 9% have been directly observed to surge in Svalbard. Since the 2000s, satellite monitoring has enabled detection of most surges of glaciers with large catchments, and the launch of the Copernicus Sentinels in 2014 has further enhanced our monitoring capabilities. Current surge detection is based upon tracking the speed of glaciers over time, elevation changes, terminus advances particularly in historical data sets, and more recently automatically detecting surface changes related to a surge such as increased crevassing. Geophysical sensors are critical for observing subglacial conditions and further work is required to improve deployment strategies on heavily crevassed glaciers. Past surge behaviour can be inferred by employing a landsystems approach and using historical archives such as maps, photographs and field notes. Improvements in our ability to detect surges has started to reveal more complex surge dynamics that suggests the binary classification of a glacier as surge-type or not breaks down. This has implications for how we understand the mechanisms through which glaciers build up energy during quiescence which enables ice flow acceleration during a surge.