Description: The Recovery Glacier is draining about 8% of the East Antarctic ice sheet and feeds into the Filchner Ice Shelf. There were suggestions that the dynamics of the glacier is driven by large subglacial lakes that initiate the ice stream flow (Bell et al., 2008). As the Recovery Glacier is one of the least surveyed ice streams due to its remote location, this hypothesis could not be tested rigorously so far. In austral summer 2013/14 AWI carried out a survey of the Recovery Glacier including radio echo sounding, gravimetry, magnetics, and laser scanner. In total more than 22000 km survey lines were flown. Here we present an ice thickness map of the main trunk of the Recovery Glacier, as well as its tributaries Blackwall and Ramp glaciers. The ice thickness varies between 70 m in the vicinity of the Shackleton Range and nearly 3800 m close to the Bell lakes. Using different DEMs including one CryoSat-2 DEM (Helm et al., 2014), we determine the basal topography and the hydraulic head. We estimate the basal reflection coefficient and assess by this locations with potentially wet ice base. The distribution shows that few of the formerly proposed lakes show indeed a wet base, while others are missing clear lake like basal reflections.

Description: The occurrence of liquid water at the base of an ice sheet is believed to be a crucial component in its dynamic evolution. If temperatures at the base locally reach the pressure melting point, basal melt water lubricates the base and thus supports basal sliding. Faster basal sliding in turn reduces internal deformation and thus the internal heat production due to strain heating. If this loss of strain heating is not counterbalanced by frictional heat due to sliding or the advection of warm ice, the base of the ice will freeze to the bedrock again. Thus the presence of liquid water can lead to a cooling and a subsequent stagnation of fast ice flow, posing a negative feedback cycle. In addition, strain heating within a temperate ice layer generates a liquid water fraction in the ice, leading to a softer material and enhanced deformation. If the horizontal or vertical advection of cold ice to the base is weak, this will lead to a positive feedback. These feedback cycles are studied along numerical simulations of the present day ice flow in the area of the western Dronning Maud Land, Antarctica, including the adjacent Brunt and Riiser- Larsen ice shelves. To investigate the influence of basal water on basal sliding and ice rheology we use the three-dimensional thermo-coupled full-Stokes model TIM-FD3 on a 2.5 km horizontal grid. We use the enthalpy gradient method to compute the thermal evolution, including the microscopic water content, in temperate ice areas. Three different flux routing algorithms for the subglacial melt water and a modified Weertman-type sliding relation are implemented in the model to account for higher sliding velocities under wet basal conditions. We present our analysis of the involved feedback mechanisms between sliding, ice deformation, temperature and rheology, which are related to the occurrence of basal water.

Description: As ice flows over a subglacial lake, the drop in bed resistance leads to an increase in ice velocity and a subsequent lowering of the ice surface in the vicinity of the upstream lake edge. Conversely, at the downstream end of the lake a small hump is observed as the ice velocity decreases near the point of contact with land. There are two contributions arising from the ice/lake interaction: (1) changes in the thermal regime that propagate downwards with the advection of ice and (2) the increase in flow speeds caused by basal sliding over the lake surface. Sediment transport from upstream areas into subglacial lakes changes their size, thus reducing the area of the ice/lake interface. Here, we aim to study the effect that this reduction in size has on the flow dynamics and the surface elevation of an artificial ice stream and the temporal evolution of this effect. To this end, we use a full-Stokes, polythermal ice flow model, implemented into the commercial finite element software COMSOL Multiphysics. An enthalpy gradient method is used in order to account for the evolution of temperature and water content within the ice. This conceptual model uses prescribed boundary velocity and temperature profiles and a Weertman-type sliding law with a fixed parameter combination. In order to separate the effect of the slow thermal contribution from the fast mechanical one, we will present sensitivity tests that additionally involve a thermally-constant flow.

Description: Pine Island Glacier is a fast moving outlet glacier in the West Antarctic Ice Sheet. Several tributaries feeding the central ice stream characterise the flow field structure of this glacier. In the past decades the glacier has shown acceleration, thinning and a significant grounding line retreat. These ongoing processes are coinciding with a concentrated mass loss in the area around Pine Island Glacier, the Amundsen Sea Embayment. The area is of additional interest due to its retrograde bed slope. The postulated instability of the setting turns the glacier into an even more suitable object for modelling studies. One major challenge encountered when modelling the flow field of Pine Island Glacier is to reproduce the locally varying flow pattern, with its many tributaries. Commonly this difficulty is overcome by inversion for parameters controlling basal sliding. Our study is aimed at connecting basal sliding again to physical parameters. To achieve this we conduct experiments of Pine Island Glacier with the diagnostic 3D full-Stokes model COMice. The model is thermo-mechanically coupled and implemented with the commercial finite-element package COMSOL Multiphysics©. We use remotely sensed surface velocity data to validate our results. In a first step, the model is used to identify dominant local mechanisms that drive the flow of the different tributaries. We identify connections between the basal topography, the basal temperature, the driving stress and the basal roughness distribution. The thus gained information is used to confine basal sliding. Areas with similar qualitative characteristics are identified, and constant-sliding assumptions made for those. Additionally, the basal roughness distribution is matched onto a basal sliding parameter. This way the sliding law is again brought closer to its original meaning. Our results are important for prognostic model experiments, as we connect basal sliding to locally varying basal properties, which might lead to different responses of the tributaries to altered external forcing.

Description: The stability of ice shelves depends on the existence of embayments and is largely influenced by ice rises and ice rumples, which act as “pinning-points” for ice shelf movement. Of additional critical importance are interactions between ice shelves and the water masses underlying them in ice shelf cavities, particularly melting and refreezing processes. The present study aims to elucidate the role of ice rises and ice rumples in the context of climate change impacts on Antarctic ice shelves. However, due to their smaller spatial extent, ice rumples react more sensitively to climate change than ice rises. Different forcings are at work and need to be considered separately as well as synergistically. In order to address these issues, we have decided to deal with the following three issues explicitly: oceanographic-, cryospheric and general topics. In so doing, we paid particular attention to possible interrelationships and feedbacks in a coupled ice-shelf-ocean system. With regard to oceanographic issues, we have applied the ocean circulation model ROMBAX to ocean water masses adjacent to and underneath a number of idealized ice shelf configurations: wide and narrow as well as laterally restrained and unrestrained ice shelves. Simulations were performed with and without small ice rises located close to the calving front. For larger configurations, the impact of the ice rises on melt rates at the ice shelf base is negligible, while for smaller configurations net melting rates at the ice-shelf base differ by a factor of up to eight depending on whether ice rises are considered or not. We employed the thermo-coupled ice flow model TIM-FD3 to simulate the effects of several ice rises and one ice rumple on the dynamics of ice shelf flow. We considered the complete un-grounding of the ice shelf in order to investigate the effect of pinning points of different characteristics (interior or near calving front, small and medium sized) on the resulting flow and stress fields, focusing on the floating ice parts of the Brunt and Riiser-Larsen ice shelves. The major response of the ice is observed instantaneously and is caused by the time independent nature of the Stokes equations and the used Glen-type rheology. The influence of ice temperatures and therefore the time-dependent effect on the flow-rate are small, given a 100 year time frame and applying a fixed-geometry setting.. A particularly important result of the current project lies in the fact that we have numerically simulated the three-dimensional stress fields in an ice shelf. Common numerical models that utilize a vertically integrated Shallow Shelf Approximation (SSA-models), do not provide that information. Due to the detailed horizontal resolution of 1km in our models, we were able to also model the observed heavily fractured areas in the vicinity of McDonald Ice Rise, a region that is characterized by simulated tensile stresses reaching maximum vertical extension in the ice column.

Description: Numerical simulations moved in the recent year(s) from describing the cold-temperate transition surface (CTS) towards an enthalpy description, which allows avoiding incorporating a singular surface inside the model (As- chwanden et al., 2012). In Enthalpy methods the CTS is represented as a level set of the enthalpy state variable. This method has several numerical and practical advantages (e.g. representation of the full energy by one scalar field, no restriction to topology and shape of the CTS). The proposed method is rather new in glaciology and to our knowledge not verified and validated against analytical solutions. Unfortunately we are still lacking analytical solutions for sufficiently complex thermo-mechanically coupled polythermal ice flow. However, we present two experiments to test the implementation of the enthalpy equation and corresponding boundary conditions. The first experiment tests particularly the functionality of the boundary condition scheme and the corresponding basal melt rate calculation. Dependent on the different thermal situations that occur at the base, the numerical code may have to switch to another boundary type (from Neuman to Dirichlet or vice versa). The main idea of this set-up is to test the reversibility during transients. A former cold ice body that run through a warmer period with an associated built up of a liquid water layer at the base must be able to return to its initial steady state. Since we impose several assumptions on the experiment design analytical solutions can be formulated for different quantities during distinct stages of the simulation. The second experiment tests the positioning of the internal CTS in a parallel-sided poly- thermal slab. We compare our simulation results to the analytical solution proposed by Greve and Blatter (2009). Results from three different ice flow-models (COMIce, ISSM, TIMFD3) are presented.

Description: The benchmark experiment MISMIP3D (Pattyn et al., 2013) investigated the response of a artificial ice stream-ice shelf system to a sliding perturbation. We continued this experiment by applying cycles of pertubations at different time scales in order to see the long term response of the grounding line positions to changes in basal sliding. For this purpose we applied the finite-difference full-Stokes model TIM-FD3 on 2.5km and 1.25km using three different initial geometries. We found that our steady-state geometry shows a strong dependency of the grounding line position on the horizontal grid size and the chosen initial geometry. Not all experiments show a neutral equilibrium in subsequent basal sliding perturbation simulations.

Description: The evolution of the Antarctic ice sheet is highly affected by small scale topographic features that arise at the ice sheet margin where the ice flow towards the ocean ice channelized into ice streams and outlet glaciers. In general the ice sheet margin is characterized by pronounced basal relief and velocity gradients that require the consideration of both longitudinal and transverse stress gradients, as well as bridging effects in the force balance equation (Stokes equation). Furthermore, the widely temperate base of the ice sheet demands a reasonable law for basal sliding consequently also transport of basal melt water. We present numerical simulations of the present day ice flow using the three-dimensional thermo-coupled full- Stokes model TIM-FD3 on a 2.5 km horizontal grid in the area of the western Dronning Maud Land, Antarctica, including the three ice streams Stancomb-Wills, Veststraumen and Plogbreen and the adjacent Brunt and Riiser- Larsen ice shelves. We estimate the distribution of subglacial water based on flux routing methods. Since very little is known about the basal conditions in that area, we study the effect of different sliding laws for wet and dry conditions at the ice base on the overall flow field.

Description: Predictions of marine ice-sheet behaviour require models able to simulate grounding-line migration. We present results of an intercomparison experiment for plan-view marine ice-sheet models. Verification is effected by comparison with approximate analytical solutions for flux across the grounding line using simplified geometrical configurations (no lateral variations, no buttressing effects from lateral drag). Perturbation experiments specifying spatial variation in basal sliding parameters permitted the evolution of curved grounding lines, generating buttressing effects. The experiments showed regions of compression and extensional flow across the grounding line, thereby invalidating the boundary layer theory. Steady-state grounding-line positions were found to be dependent on the level of physical model approximation. Resolving grounding lines requires inclusion of membrane stresses, a sufficiently small grid size (〈500m), or subgrid interpolation of the grounding line. The latter still requires nominal grid sizes of 〈5 km. For larger grid spacings, appropriate parameterizations for ice flux may be imposed at the grounding line, but the short-time transient behaviour is then incorrect and different from models that do not incorporate grounding-line parameterizations. The numerical error associated with predicting grounding-line motion can be reduced significantly below the errors associated with parameter ignorance and uncertainties in future scenarios.