The far reach of ice-shelf thinning in Antarctica

1Potsdam Institute for Climate Impact Research (PIK), Member of the Leibniz Association, PO Box 60 12 03, Potsdam, Germany. 2Institute of Physics and Astronomy, University of Potsdam, Potsdam, Germany. 3British Antarctic Survey, Cambridge, UK. 4Lamont–Doherty Earth Observatory, Columbia University, New York, NY, USA. *e-mail: [email protected] Floating ice shelves, which fringe most of Antarctica’s coastline, regulate ice flow into the Southern Ocean1–3. Their thinning4–7 or disintegration8,9 can cause upstream acceleration of grounded ice and raise global sea levels. So far the effect has not been quantified in a comprehensive and spatially explicit manner. Here, using a finite-element model, we diagnose the immediate, continent-wide flux response to different spatial patterns of ice-shelf mass loss. We show that highly localized ice-shelf thinning can reach across the entire shelf and accelerate ice flow in regions far from the initial perturbation. As an example, this ‘tele-buttressing’ enhances outflow from Bindschadler Ice Stream in response to thinning near Ross Island more than 900 km away. We further find that the integrated flux response across all grounding lines is highly dependent on the location of imposed changes: the strongest response is caused not only near ice streams and ice rises, but also by thinning, for instance, well-within the Filchner–Ronne and Ross Ice Shelves. The most critical regions in all major ice shelves are often located in regions easily accessible to the intrusion of warm ocean waters10–12, stressing Antarctica’s vulnerability to changes in its surrounding ocean. Owing to their ability to regulate upstream ice flow, Antarctic ice shelves play a key part in future sea-level rise in a warming world5,13,14. At the same time, as they are in direct contact with the ocean at their base and have low surface elevation, they are particularly vulnerable to a changing climate15–19. Ocean-induced thinning of ice shelves, which has been accelerating over the past decades4, has the potential to reduce the restraining stress to ice sheet flow provided by ice shelves1,2 and thereby enhance ice discharge across the grounding lines of the Antarctic Ice Sheet6–9,20. Although the importance of ice shelves in modulating ice flux from the interior of the Antarctic Ice Sheet across the grounding lines and into the ocean has been recognized for some time21–24, no study has set out to systematically map the dependency of ice flux across the grounding lines to different spatial patterns of ice-shelf thinning. By modifying the stress balance along the grounding line, ice shelves can ‘buttress’ the ice flow from the (grounded) ice sheet into the ocean. At any point along a grounding line ice-shelf buttressing can be quantified in terms of the impact that an ice shelf has on the state of stress compared to the state of stress in the hypothetical absence of the ice shelf25. Ice-shelf buttressing is primarily affected by how strongly the flow of an ice shelf is restricted both laterally and through local grounding (an unconfined ice shelf of uniform width provides no buttressing). However, it is also dependent on the geometry, thickness and rheological properties of the ice shelf3,26,27. Thinning in any part of a confined ice shelf affects the stress regime within the whole ice shelf, and therefore has the potential to impact the ice flux of all surrounding grounding lines. The impact of ice-shelf thinning on flow across grounding lines is expected to depend on a number of factors such as the location of thinning, how strongly the flow of the ice shelf is confined, ice flow properties upstream of the grounding lines, the ice softness and the shape of the grounding lines. Fully assessing the impact of ice-shelf thinning on the discharge from the Antarctic Ice Sheet therefore requires estimating the effect of thinning at any location on all grounding lines. In order to identify the most critical ice shelf regions, we provide a comprehensive, quantitative assessment of the impact of any local ice-shelf thickness change on the discharge across the grounding lines of present-day Antarctica. Based on input inferred from data assimilation of present-day ice thickness28 and velocities29, we conduct a series of Antarctic-wide simulations with the finite-element model Úa23, which solves for the ice flow in sheet and shelves simultaneously (using the shallow-shelf/shelfy-stream approximation of the momentum balance, see for example, refs 30,31). At the ice front, the boundary condition is given by the vertically integrated, horizontal static fluid pressure of the ambient ocean. The grounding line position is diagnosed in Úa using the floatation criterion. We express the effect of ice-shelf thinning as the ratio between the total changes in annual mass flux across all grounding lines to the magnitude of locally applied thinning (Fig. 1). In each diagnostic experiment, a 20 km × 20 km region is thinned by 1 m, and the immediate response in ice flux across all grounding lines determined. These perturbations mimic ice-shelf thinning patterns resulting from changes in ocean-induced melting. Idealized perturbations allow us to separate the contribution from different ice-shelf areas and therefore identify regions of particular importance for grounded ice loss. Subdividing the ice shelves into 20 km × 20 km sectors was found to be of sufficiently fine resolution to adequately resolve the spatial response pattern, and 1 m thinning sufficiently small to be within the linear response range (see Methods and Supplementary Figs. 1, 2). The impact of each local perturbation is then obtained by dividing the resulting increase in annual mass flux across all grounding lines R by the applied perturbation in ice shelf mass P. We refer to this non-dimensional ratio as the buttressing flux response number θB = R/P and express this number as a percentage. If, for example, θB = 100%, then the added annual mass flux across all grounding lines equals the mass-flux perturbation applied to the ice shelf, or 20 km × 20 km × 1 m × 910 kg m−3 = 0.364 Gt (assuming an ice density of 910 kg m−3). Where the perturbed sector is crossed by part of the grounding line, the removed ice mass decreases accordingly. To ensure that the recorded flux response is only caused by changes in ice-shelf buttressing, and to exclude any potential impacts related to changes in driving stress over grounded regions, ice-shelf thinning is only applied to those elements of the computational grid that share no nodes with any elements crossing grounding lines (see Methods). With the buttressing flux The far reach of ice-shelf thinning in Antarctica