The sensitivity of marine-terminating glaciers to model parameters and geometry
MetadataShow full item record
- Geophysical Institute 
Greenland outlet glaciers are among the largest contributors to global sea level rise. With high velocities and calving rates, they discharge large amounts of glacial ice into the ocean. During the last two decades, the mass loss of these glaciers has increased dramatically. Jakobshavn Isbræ recently experienced dramatic acceleration to peak velocities of 17 km yr-1; in contrast to other fast Greenland glaciers, its high velocities have persisted. Many studies have explained the observed acceleration with increased ocean water temperatures, increased surface runoff and reduced buttressing by sea ice. However, it is still not completely understood how external factors, such as changes by the atmosphere and ocean, impact marine-terminating glaciers. Here, the impact of ice temperature, basal sliding, crevasse water depth, submarine melt rate, and buttressing by sea ice on glacier properties is studied with a numerical flowband model. A sensitivity study is conducted on an idealized marine-terminating glacier and on Jakobshavn Isbræ. Changes to the driving as well as internal parameters of the ice flow model have a great impact on the idealized glacier. Whilst a change in crevasse water depth, buttressing by sea ice, and submarine melt impact the thickness and length proportionally, basal sliding and ice rheology in uence rather the shape of the glacier. The ice temperature is represented by the rate factor, a complex parameter, found to influence the glacier in opposing ways through its control on the viscosity and lateral resistance. The study of Jakobshavn Isbræ shows that stabilization at pinning points dominates the impact of parameter uncertainties. The grounding line position can therefore be stable for hundreds of years, while the thickness of the glacier continues adjusting to previous perturbations. This adjustment may eventually lead to a dramatic change of the grounding line position. It is therefore crucial for ice sheet models to involve centennial to millennial time-scales.