LETTERS PUBLISHED ONLINE: 11 JANUARY 2009 DOI: 10.1038/NGEO394 Large-scale changes in Greenland outlet glacier dynamics triggered at the terminus Faezeh M. Nick 1 * , Andreas Vieli 1 † , Ian M. Howat 2 and Ian Joughin 3 The recent marked retreat, thinning and acceleration of most of Greenland’s outlet glaciers south of 70 ◦ N has increased concerns over Greenland’s contribution to future sea level rise 1–5 . These dynamic changes seem to be parallel to the warming trend in Greenland, but the mechanisms that link climate and ice dynamics are poorly understood, and current numerical models of ice sheets do not simulate these changes realistically 6–8 . Uncertainties in the predictions of mass loss from the Greenland ice sheet have therefore been highlighted as one of the main limitations in forecasting future sea levels 9 . Here we present a numerical ice-flow model that reproduces the observed marked changes in Helheim Glacier, one of Greenland’s largest outlet glaciers. Our simulation shows that the ice acceleration, thinning and retreat begin at the calving terminus and then propagate upstream through dynamic coupling along the glacier. We find that these changes are unlikely to be caused by basal lubrication through surface melt propagating to the glacier bed. We conclude that tidewater outlet glaciers adjust extremely rapidly to changing boundary conditions at the calving terminus. Our results imply that the recent rates of mass loss in Greenland’s outlet glaciers are transient and should not be extrapolated into the future. Two main hypotheses have been advanced to explain the rapid dynamic changes of Greenland’s outlet glaciers. The first postulates that the dynamical changes result from processes that act at the terminus and trigger a retreat and reduce along-flow resistive stresses (backstress) 2,3,10 . This leads then to faster ice flow and thinning that propagates rapidly upstream and leads to further retreat. Several climate-related processes may initiate these near- terminus changes, such as surface-melt induced thinning and increased calving due to enhanced hydro-fracturing of water-filled crevasses from increased surface melt 11 . For Helheim Glacier, the sensitivity to such processes may be further enhanced by a basal overdeepening in the fjord 12 , as has been suggested for tidewater glaciers 13–15 . The second hypothesis is that warmer air temperatures increase the amount of surface meltwater reaching the glacier bed, increasing basal lubrication and the rate at which ice slides over its bed, leading to glacier acceleration, thinning and retreat 16,17 . To better understand the processes driving rapid outlet glacier change and assess their potential future impact, we developed a numerical flow model for Helheim Glacier that includes horizontal (along-flow and lateral) stress transfer and a dynamically determined adjustment of the grounded calving front (see the Methods section and Supplementary Information, Model). 1 Department of Geography, Durham University, South Road, Durham DH1 3LE, UK, 2 School of Earth Sciences, Byrd Polar Research Center, Ohio State University, 1090 Carmack Road, Columbus, Ohio 43210-1002, USA, 3 Polar Science Center, Applied Physics Lab, University of Washington, 1013 NE 40th Street, Seattle, Washington 98108, USA. *Present address: Geological Survey of Denmark and Greenland GEUS, Ostervolgade 10 DK-1350 Copenhagen, Denmark. † e-mail: andreas.vieli@durham.ac.uk. We test the above hypotheses and triggering mechanisms by carrying out a series of modelling experiments in which we perturb the boundary condition and then run the model forward in time and compare the output to the observations (Fig. 1a,b). First, we carry out a step increase in the longitudinal stress boundary condition at the calving front (‘front-stress perturbation’, see Supplementary Information, Model). Physically this can be interpreted as an along- flow rheological weakening of the ice at the terminus or a reduction in backstress. The modelled surface elevation, velocity and terminus position generally agree with the observed changes (Fig. 1c,d). An instantaneous velocity increase occurs through the transfer of longitudinal stresses and extends up to 20 km upstream of the terminus. This acceleration initiates thinning near the terminus, which steepens the surface, increases the driving stress and leads to further acceleration. This interaction between increased driving stress and flow acceleration causes thinning and acceleration to propagate upstream. As a result of the thinning, the ice near the calving front approaches flotation and causes the terminus to retreat (Fig. 2a). Within the first few months after the perturbation, rates of acceleration and retreat decrease (Figs 1c and 2a), which is mainly a result of the applied step change in perturbation. Applying an extra experiment with a gradual perturbation with time produced a continuous acceleration similar to that observed. When the terminus eventually retreats over the bedrock high into deeper water, ice speed and discharge begin to increase again leading to further thinning and retreat (Figs 1c,d,2a). This positive feedback between thinning and retreat results in an unstable retreat over the reversed bed slope and thinning of more than 100 m in two years. In our model, this feedback is solely the result of enhanced ice flux with increasing ice thickness, as hypothesized by the ‘marine ice sheets instability’ 18 . Other effects, such as a thinning-induced decrease in effective pressure near the terminus, may contribute to the instability 19 , but here we find they are not necessary to explain the observations. The model successfully reproduces both the acceleration to 12 km yr -1 near the front, as it retreats down the reversed bed slope into deeper water, and the subsequent deceleration once the bottom of the overdeepening is reached (Fig. 1d). Despite this deceleration and stabilization of the terminus, a wave of acceleration and thinning continues to diffuse upstream as observed. In our experiment, the perturbation imposed at the terminus has been removed when the terminus reaches the 2005 position, enhancing the deceleration. Without this removal, the calving front still decelerates, but retreats over another bedrock low before stabilizing 110 NATURE GEOSCIENCE | VOL 2 | FEBRUARY 2009 | www.nature.com/naturegeoscience © 2009 Macmillan Publishers Limited. All rights reserved.