Recent and future variability of the ice-sheet catchment of Jakobshavn Isbræ, Greenland

Glacier catchments, refer to the upstream area of an ice sheet that is being drained by a single glacier. Delineated catchments are used to study the mass balance of individual glaciers or conglomerates of glaciers. Single glacier basin mass balance studies allow us to gain an understanding of the controlling processes and monitor changes of individual glaciers. Such studies, however, are highly sensitive to how the surface mass balance collection area is delineated for comparison with the grounding line ice discharge. Frequently, catchments are assumed to be temporally invariant, which can impact studies of basin wide mass balance on longer time scales. We have explored variations in present day delineated area, and potential temporal variability, in the catchment of Sermeq Kujalleq, or Jakobshavn Isbræ, Greenland. Five observation based, ‘present day’ catchment delineations are evaluated along with delineations based on modeled surface velocities from a sixteen-member ensemble of ice-sheet models within the Ice Sheet Model Intercomparison Project for CMIP6 (ISMIP6).  The ISMIP6 ensemble mean area was found to be ~5.4% larger than the mean of the observed catchments. Observed and modeled ensemble spreads were comparable, ±12.3% and ±15.4%. Hence, models are able to delineate the present day catchment with the same degree of uncertainty as observational methods. The ISMIP6 ensemble mean catchment area shows temporal variation, increasing ~4% from 2015-2100, primarily as the southern catchment boundary migrates southward. This is interpreted as Jakobshavn Isbræ exhibiting dynamic piracy, redirecting ice away from adjacent land terminating glaciers.  Repeat velocity observations in the catchment suggest a recent acceleration in ice flow ~100 km inland from the terminus of Jakobshavn Isbræ. This observed acceleration is not believed to be attributable to changes in meltwater lubrication nor driving stress changes. Instead it is theorized that changes in the basal temperate ice layer could be the cause of the acceleration. The trend in simulated ensemble velocity and azimuth values over the 2015-2100 period is evaluated against trends from the recent velocity observations. The ISMIP6 ensemble is underestimating the observed regional acceleration, not fully capturing the deep inland dynamic response of the ice sheet to the recent climatic changes.