Ice-sheet mass and sea level rise

As the popular climate change debate makes the transition from “is it happening?” to “what will the impacts be?”, one of the more controversial issues has been that of sea level. The dramatic collapse of parts of the Antarctic ice-shelf and the speed-up of Greenland’s glaciers have led to suggestions that the last IPCC report’s estimate (which didn’t include a contribution from ice-sheet dynamics) of a 0.18m to 0.6m rise in global mean sea level by 2100, was far too conservative. James Hansen, who heads the NASA Goddard Institute for Space Studies and is a proponent of strong action on climate change, has suggested that a rise of several metres by 2100 is possible, although work by Pfeffer et al. in 2008 concluded that “increases in excess of 2 meters are physically untenable”, and that 0.8m was a more plausible upper bound.

Ice-sheets are difficult to measure, and difficult to model. Ice-sheet dynamics are non-linear, and important processes happen on a very wide range of spatial and temporal scales, from days to millennia. A large part of the problem is that we haven’t had good enough (or long enough) observations of the important features of the big ice-sheets to be able to say precisely enough what they’re doing at the moment, let alone how they will evolve in the future. Rignot et al. (GRL, 2011) have made a big step forward this week, combining high-resolution modelling and satellite data to provide estimates of ice-sheet mass changes on Greenland and Antarctica. They use two separate methods – a budget of net surface accumulation minus melt at the edge, and measurements of the gravitational fields of the ice-sheets – which agree, within the error estimates. This provides some confidence that we’re getting a handle on at least the gross features of ice-sheet mass balance, and that we can distinguish the important long-term trends from the short-term variability. They estimate that, in 2006, the rate of ice loss was 475+-158 Gt/yr, a 1.3+-0.4 mm/yr contribution to global mean sea level, which is independently estimated to be going up at around 3.2+-0.4 mm/yr.

Rignot et al. also estimate that the rate of ice loss from both Greenland and Antarctica is increasing, and argue that the reasons for the speed-up are unlikely to change soon. They thus extrapolate their rates of change into the future, concluding that “if this trend continues, ice sheets will become the dominant contribution to sea level rise in the next decades, well in advance of model forecasts”, and even tentatively suggest that ice-sheet loss alone could contribute up to 56cm of sea level by 2100. But here we meet the major challenge in projecting the future of the ice-sheets. For all that we can fit curves to past and current changes of ice-sheet mass, it’s difficult to have any confidence in projecting any of these curves into the future without a detailed, physically based model – and we just can’t simulate the ice-sheets well enough for that yet. The much-criticised IPCC stance on the ice-sheet contribution to future sea level rise was not so much an under-estimate, as a non-estimate – they decided that, given the current level of knowledge, it was impossible to usefully put any numbers to the problem, concluding that “a basis in published literature is lacking”, and that “understanding of these effects is too limited to assess their likelihood or to provide a best estimate or an upper bound for sea-level rise”.

So, the arguments about ice-sheets and future sea level rise are probably going to rumble on for some time yet. It’s an interesting part of the climate change debate, not least because the relatively slow rate of change involved brings the question of adaptation to the fore – is climate change a problem if it happens slowly enough for society to get used to it? How slow is “slowly enough”? The massive thermal inertia of the oceans also means that some sea level rise is inevitable, and it’s very resistant to geo-engineering “solutions” to climate change. Work on better ice-sheet models is, of course, ongoing (including in this department) – but we’re probably not going to have definitive answers for a long time.

Leave a Reply

Your email address will not be published. Required fields are marked *