Goelzer et al. (2012) paint a portentous picture of what is likely to happen to the global sea level over the next 1000 years. This worrying assessment is based on our current best understanding of how the world's giant ice sheets of Greenland and Antarctica, as well as a quarter of a million smaller glacial ice masses, and the ocean collectively respond to ongoing climate change. Theirs is a state-of-the-science study that integrates these key contributors of sea-level change based on the latest models and current understanding, and an integrated Earth-systems’ modelling approach termed LOVECLIM. As they point out in their study, only a handful of global climate models to date – i.e. models that are used to make predictions of future climate change – incorporate dynamically (fully) coupled ice-sheet models.

According to the scenarios presented by Goelzer et al. (2012), we could see between 2.1 and 6.8 m of global sea-level rise by 3000 AD, compared with "just" 1.1 m if the atmosphere is stabilized at 2000 CO2 levels. Much, up to some 4 m, of this contribution comes from increased melting and mass loss of the Greenland ice sheet, which is several times more sensitive than the Antarctic ice sheet to warming temperatures in these simulations. Interestingly, dynamical ice-mass losses through iceberg calving become increasingly less significant for Greenland as the ice sheet retreats further inland during the 1000 year runs (Sole et al. 2008). The latest modelling studies show that around a half, perhaps more, of the recent Greenland mass losses (Barletta et al. 2012, Rignot et al. 2011) are already through increased melt and runoff (Hanna et al. 2008, 2012, van den Broeke et al. 2009); note also the recent (summer 2012) record surface melting of the Greenland ice sheet (Nghiem et al. 2012) caused by atmospheric forcing (Overland et al. 2012) and the potential of such events to impact on ice flow (Bartholomew et al. 2011). By contrast, the greatest sea-level rise reported for Antarctica by AD 3000 is no more than 94 cm; Antarctica remains relatively insensitive for future sea-level rise given a temperature increase of no more than 5–6 °C (quite a lot) above present levels. Oceanic thermal expansion and, especially, glacier melt seem very much second-order effects, compared with the Greenland sea-level contribution, for the next millennium.

As expected, there are considerable differences between the outcomes of the model experiments depending on the time and level at which greenhouse-gas emissions are stabilized. I am not quite sure why they "prefer" the model version, which reaches stabilization at 2000 greenhouse-gas levels, since those levels have since been significantly exceeded and show no signs of tailing off yet – quite the reverse. According to the famous Keeling et al. dataset from Mauna Loa in Hawaii, atmospheric CO2 levels at about 369 parts per million by volume of the global atmosphere in 2000 have since risen to about 392 ppmv in 2012, and this increase shows no signs of abating. Realistically, it is going to be at least another decade or two (or longer) before we can effectively even begin to stabilize atmospheric greenhouse-gas levels, assuming the political will is there, which at the moment it is not. Of course this does not commit us to the other three more extreme experimental results (from greenhouse-gas stabilization at 2100) reported in the study but we are heading dangerously in that direction. In effect the simulations are sensitivity studies, which may be largely unrealistic but are still useful as a kind of guide to what might happen under future climate change.

Naturally, many uncertainties remain, especially concerning how ice-sheet motion ("dynamics") is represented in the models (e.g. the absence of so-called "higher-order physics" including longitudinal (push–pull) stresses, which can rapidly transfer peripheral ice-velocity perturbations inland (Price et al. 2011)). Furthermore, the atmospheric model used in LOVECLIM is very coarse at 5.625° latitude/longitude resolution. There appears to be a cancelling out of errors in LOVECLIM, where its climate sensitivity seems quite low (in comparison with other models) but the simulated enhanced high-latitude warming – often termed Arctic amplification and evident in observed climate data for the last 30 years – is quite high. It would be good to include precipitation as well as temperature changes when modelling the future response of glaciers, even though the former is likely to be less important. I do not agree that uncertainties in climate sensitivity can be adequately accounted for by varying boundary and initial conditions in ensembles of models, as all of the model simulations may be systematically biased due to some physical effect that is improperly considered – or unrepresented – by all of the models, but this is a widely used technique and probably the best that can be done here.

Despite these caveats, Goelzer et al.'s (2012) results will undoubtedly prove useful for the Intergovernmental Panel on Climate Change (IPCC)'s upcoming Fifth Assessment Report due to be released in 2014. The key challenge remains to further improve the individual components of the Earth system model, especially those concerning ice-sheet dynamics.

Acknowledgments

EH thanks Ben Brock, Amy Jowett and Andrew Sole for useful editorial suggestions to the text.

References

Barletta V R, Sørensen L S and Forsberg R 2012 Variability of mass changes at basin scale for Greenland and Antarctica Cryosp. Discuss. 6 3397–446

Bartholomew I, Nienow P, Sole A, Mair D, Cowton T and King M A 2011 Seasonal variations in Greenland ice sheet motion: inland extent and behaviour at higher elevations Earth Planet. Sci. Lett. 307 271–8

Goelzer H, Huybrechts P, Raper S C B, Loutre M-F, Goosse H and Fichefet T 2012 Millennial total sea-level commitments projected with the Earth system model of intermediate complexity LOVECLIM Environ. Res. Lett. 7 045401

Hanna E, Huybrechts P, Steffen K, Cappelen J, Huff R, Shuman C, Irvine-Fynn T, Wise S and Griffiths M 2008 Increased runoff from melt from the Greenland ice sheet: a response to global warming J. Clim. 21 331–41

Hanna E, Jones J M, Cappelen J, Mernild S H, Wood L, Steffen K and Huybrechts P 2012 The influence of North Atlantic atmospheric and oceanic forcing effects on 1900–2010 Greenland summer climate and ice melt/runoff Int. J. Climatol. at press (doi:10.1002/joc.3475)

Nghiem S V, Hall D K, Mote T L, Tedesco M, Albert M R, Keegan K, Shuman C A A, DiGirolamo N E and Neumann G 2012 The extreme melt across the Greenland ice sheet in 2012 Geophys. Res. Lett. at press (doi:10.1029/2012GL053611)

Overland J E, Francis J, Hanna E and Wang M 2012 The recent shift in early summer Arctic atmospheric circulation Geophys. Res. Lett. 39 L19804

Price S F, Payne A J, Howat I M and Smith B E 2011 Committed sea-level rise for the next century from Greenland ice sheet dynamics during the past decade Proc. Natl Acad. Sci. 108 8978–83

Rignot E, Velicogna I, van den Broeke M R, Monaghan A and Lenaerts J 2011 Acceleration of the contribution of the Greenland and Antarctic ice sheets to sea level rise Geophys. Res. Lett. 38 L05503

Sole A, Payne T, Bamber J, Nienow P and Krabill W 2008 Testing hypotheses of the cause of peripheral thinning of the Greenland ice sheet: is land-terminating ice thinning at anomalously high rates? Cryosphere 2 205–18

van den Broeke M, Bamber J, Ettema J, Rignot E, Schrama E, van de Berg W, van Meijgaard E, Velicogna I and Wouters B 2009 Partitioning recent Greenland mass loss Science 326 984–6