Nioghalvfjerdsfjorden is a major outlet glacier in Northeast-Greenland. Although earlier studies showed that the floating part near the grounding line thinned by 30% between 1999 and 2014, the temporal ice loss evolution, its relation to external forcing and the implications for the grounded ice sheet remain largely unclear. By combining observations of surface features, ice thickness and bedrock data, we find that the ice shelf mass balance has been out of equilibrium since 2001, with large variations of the thinning rates on annual/multiannual time scales. Changes in ice flux and surface ablation are too small to produce this variability. An increased ocean heat flux is the most plausible cause of the observed thinning. For sustained environmental conditions, the ice shelf will lose large parts of its area within a few decades and ice modeling shows a significant, but locally restricted thinning upstream of the grounding line in response.

The high-Alpine ice-core drilling site Colle Gnifetti (CG), Monte Rosa, Swiss/Italian Alps, provides climate records over the last millennium and beyond. However, the full exploitation of the oldest part of the existing ice cores requires complementary knowledge of the intricate glacio-meteorological settings, including glacier dynamics. Here, we present new ice-flow modeling studies of CG, focused on characterizing the flow at two neighboring drill sites in the eastern part of the glacier. The3-D full Stokes ice-flow model is thermo-mechanically coupled and includes firn rheology, firn densification and enthalpy transport, and is implemented using the finite element software Elmer/Ice. Measurements of surface velocities, accumulation, borehole inclination, density and englacial temperatures are used to validate the model output. We calculate backward trajectories and map the catchment areas. This constrains, for the first time at this site, the so-called upstream effects for the stable water isotope time series of the two ice cores drilled in 2005 and 2013. The model also provides a 3-D age field of the glacier and independent ice-core chronologies for five ice-core sites. Model results are a valuable addition to the existing glaciological and ice-core datasets. This especially concerns the quantitative estimate of upstream conditions affecting the interpretation of the deep ice-core layers.

It can easily be expected that debris-covered glaciers show a different response on external forcing compared to clean-surface glaciers. The supra-glacial debris cover acts as an additional transfer layer for the energy exchange between atmosphere and ice. The related glacier reaction is the integral of local effects, which changes strongly between enhanced melt for thin debris layers and considerably reduced melt for thicker debris. Therefore, a realistic feedback study can only be performed, if both the ice flow and the debris-influenced melt is treated with a high degree of detail. We couple a full Stokes representation of ice dynamics and the most complete description of energy transfer through the debris layer, in order to describe the long-term glacier reaction in the coupled system. With this setup, we can show that steady-state conditions are highly unlikely for glaciers, in case debris is not unloaded from the surface. For continuous and complete debris removal from the lowermost glacier tongue, however, a balance of the debris budget and the glacier conditions are possible. Depending on displacement and removal processes, our results demonstrate that debris-covered glaciers have an inherent tendency to switch to an oscillating state. Then, glacier mass balance and debris balance are out of phase, such that glacier advance periods end with the separation of the heavily debris-loaded lowermost glacier tongue, at time scales of decades to centuries. As these oscillations are inherent and happen without any variations in climatic forcing, it is difficult to interpret modern observations on the fluctuation of debris-covered glaciers on the basis of a changing climate only.

<p>Debris-covered glaciers accumulate supra-glacial debris on the glacier surface in the ablation zone. As long as this debris layer is not at least partly removed, it can be expected that glaciers continue to grow in length, because the thickening debris layer continuously reduces surface melt rates. Removal of the debris layer, on the other hand, is a complicated process, which depends on a number of parameters, like surface slope, debris thickness, grain size distribution and water content to name just a few. However, the way how supra-glacial debris is removed might strongly influence the dynamic reaction of the glacier itself.</p><p>A realistic study of these interactions can only be performed, if the ice flow and the debris-influenced melt is treated with a high degree of detail. In our study, we coupled a 2-D full Stokes ice dynamic and surface debris transport model with a sophisticated description of energy transfer through the debris layer. This approach ensures that ice flow and surface melt rates are simulated at high detail, including the enhanced melt rates for very thin debris cover just below the equilibrium line. We restricted our experiments to rather simple initial conditions, in order investigate the fundamental feedback mechanisms between melt rates and glacier dynamics. Therefore, we introduced rather simple, but realistic formulations of debris unloading at the glacier front. The coupled experiments show that steady-state conditions are highly unlikely for glaciers with the debris layer remaining on the glacier. However, a balance of the debris budget and the glacier mass flux is possible, when introducing debris removal from the glacier tongue. We focussed on an as realistic as possible representation of the snout geometry, in order to allow a physically sensible debris discharge. The results show that for some removal processes debris-covered glaciers have an inherent tendency to enter an oscillating state, where glacier mass balance and debris balance are out of phase. In specific experiments glacier advance periods end with the separation of the heavily debris-loaded lowermost glacier tongue, at time scales of decades to centuries, followed by an advance of the remaining clean glacier. In such cases we assume that long-term “mean-steady-state” conditions modulated by oscillations in glacier length exist and are independent from climatic variations. This makes it difficult to interpret short-term geometry observations of debris-covered glaciers in the context of climate impact.</p>

No abstract