The dynamics of the Lambert-Amery glacial system and its response to climatic variations

Antarctica's current and future contribution to sea level rise is uncertain, with changes in ice dynamics along the coast leading to mass loss while increasing precipitation in the interior is leading to mass gain. The Lambert-Amery glacial system drains a large region of East Antarctica, with the two largest glaciers within the glacial system, the Lambert and Mellor glaciers, having a substantial volume of ice grounded below sea level, suggesting a risk of marine ice sheet instability. The velocities of Lambert-Amery glacial system have been observed to be stable between 1968 and 1999, albeit with limited sampling. Recent mass balance and gravimetry studies also suggest a system in near balance. Here, visible spectrum satellite images between 2004 to 2012 have been used to compute surface ice velocities using a feature tracking approach. No significant changes in velocity were observed over the study region that included the Amery Ice Shelf adjacent to the grounding line and its three main tributary glaciers, the Lambert, Mellor and Fisher Glaciers. The stability of the Lambert-Amery glacial system allows for the initialisation of an ice sheet model by minimising the misfit between the simulate and observed system. A regional domain of the Lambert-Amery glacial system is simulated with the Parallel Ice Sheet Model. The control solution of the regional model is initialised by minimising the misfit to observations through an optimisation process. We investigate the importance of a primary boundary condition, geothermal heat flux to ice flow. Existing broad scale geothermal heat flux datasets fail to capture small scale localised variations in geothermal heat flux, such as estimates of geothermal heat flux in Prydz Bay suggesting that radiogenic crustal heat production can locally elevate geothermal heat flux by at least 100% compared to the background field. We insert high heat flow regions into a broad scale background geothermal heat flux field, and find that the presence of a high heat flow region can change the flow behaviour in regions from slow sheet flow to stream-like flow, while making no difference to regions of fast flow. This mechanism may contribute to the long term organisation of ice flow. Additionally, we use a range of different geothermal heat flux datasets, and compare simulation using them in place of our control geothermal heat flux. The simulations which use a relatively high GHF compared to the control solution increase the volume and area of temperate ice, which causes higher surface velocities at higher elevations, which leads to the advance of the grounding line. The grounding line advance leads to changes in the local flow configuration, which dominates the changes within the glacial system. To investigate the difference in spatial patterns within the geothermal datasets, they were scaled to have the same median value as the control dataset. These scaled geothermal heat flux simulations showed that the ice flow was most sensitive to the spatial variation in the underlying geothermal heat flux near the ice divides and on the edges of the ice streams. The Lambert-Amery glacial system is evidenced to change significantly during glacial cycles, with the grounding line advancing and retreating up to 700 km. This contrasts with the current stability of the glacial system. The Antarctic Ice Sheet responds to climate through several factors, including the temperature at the surface of the ice, the accumulation on the surface, the oceanic forcing at the base of the floating ice, and sea level change. We test the response of the Lambert-Amery glacial system to climatic variations by simulating the effects of a global air temperature change of ± 3°C. Each climate variable is simulated in isolation to test the sensitivity to each climatic variation, before a combined simulation. We find that the Lambert-Amery glacial system responds most rapidly to accumulation on the surface and the oceanic forcing at the base of the floating ice, while surface temperature eventually lead to the largest change, but on time scales longer than the recent glacial-inter-glacial cycles. The advance of the grounding line moves rapidly between negative sloping beds, where it then stabilises on a positive bed slope, with the ice sheet growing until a threshold is reached and the grounding line advances again. The model simulations are unable to recreate an advance simulation which was similar to the last glacial maximum, with the grounding line either not advancing to the continental shelf, or the ice sheet growing rapidly when the grounding line does advance. The contribution of Lambert-Amery glacial system to future sea level change is investigated through a range of future scenarios. Within our simulations, we find that under a range of plausible and extreme scenarios, the grounding line is unlikely to become unstable and retreat into the deep marine basins. This causes increases in precipitation to exceed mass loss through ice discharge within our simulations as long as a minimal ice shelf remains. This suggests that the Lambert-Amery glacial system has the potential to gain mass and mitigate the severity of sea level rise from Antarctica for the next 500 years.