When Will Greenland's Ice Sheet Melt?

If Greenland's ice sheets melted global sea-levels could rise by several meters. But how hot will it need to get for that to happen?
28 August 2013

Interview with 

Anthony Payne, University of Bristol


Greenland is home to some of the largest ice sheets in the world, and if those melt, they could cause global sea-levels to rise by as much as several metres. For low-lying parts of the world, including Cambridge, that could be very bad news...

So, how much of a rise in global temperatures would be required for those ice sheets to melt? Anthony Payne, from the University of Bristol, has built computational simulations to find out, which he presented in the journal PNAS this month.

Anthony - Greenland represents one of the major sources of freshwater in the world and is the second greatest potential contributor to future sea levels with something like 7 or 8 meters worth of potential sea level rise. So that's why a lot of research is centred on understanding what's happening in Greenland over the last decade.

Dominic - You've looked at two mechanisms by which this melt can occur - both the melting itself and also, sliding of that ice. How do you separate those two effects?

Anthony - We did two separate sets of experiments. One, where just looked at the effects of melt by itself. We incorporated results from recent climate models on how air temperatures were warmer in Greenland and how that will affect the amounts of ice melt in Greenland. Then in the second set of the experiments, we allow that additional melt water to interact with the flow of ice sheets, with the idea being that if there's more melt water around, then potentially, the flow of the ice sheet could be lubricated and therefore, the ice sheet would flow faster.

Dominic - How do you go about doing those experiments though?

Anthony - Well, the experiments are performed using ice sheet models that run on supercomputers and typically take 2 or 3 days to run. You give the model information on the geometry of the ice sheets, how much ice there is there, or where it is on Greenland, you give information on things like future climates, how much warming is there like to be over Greenland, how much melt will there be over Greenland, and how much snowfall will change over Greenland. The most important part of our work has been to develop a mathematical relationship that depicts how increased amounts of melt water could affect lubrication and therefore, the flow of the ice sheet.

Dominic - I'm imagining that lubrication must be a very difficult thing to model because there must be all sorts of factors affecting how stable that ice is.

Anthony - What we have done is to take a shortcut and using empirical field data to parameterise the effect, which means that we treat the effect as a black box and we use field data to say, "There's  this amount of imput i.e. melt water. This is what the observation suggest the output would be." In particular, what we've used is information on the seasonal speedup of the flow of the ice sheet. So, they use GPS receivers scattered over the ice sheets, that record the velocity of ice flow, and then you can look at the records from those recorders and they'll show that the ice flow speeds up in the summer or most likely, the early spring, and then slows down in the winter. So, what we assume is that the winter flow is with no lubrication and the summer flow is lubricated. So, if we look at the ratio of those two, we get a handle on how important the melt water effect is.

Dominic - Are you validating those models in comparison with your observational data from the ground?

Anthony - So, the validation that we have attempted is to compare observations of ice flow against what the models predict. On the whole, they do a fairly good job.

Dominic - And obviously, you're also very dependent on these models of the future climate to know how this melt will proceed in the future. Where are you getting this from?

Anthony - For our particular study, we cooperated with Zavio Satweis who works in the University of the Age and does regional climate modelling of Greenland. But I think the important thing to say is that, when you compare his models to other models that have been used to predict what's going to happen in Greenland, his model falls pretty much in the middle of the group. The results we get are so strong that we think we could use any climate model and get similar set of results.

Dominic - Often, with these questions of climate change, you find yourself in what we call very nonlinear processes where suddenly a process reaches some critical point where it takes off. How do you know that as the climate changes, we're not going to move in to a very different regime which your models aren't covering?

Anthony - The modelling wasn't intended to be a prediction of the future. It was intended to assess whether a particular set of effects, those related to basal lubrication of melt water availability were going to be important. So, while there may be lots of other thresholds in the system that we're currently unaware of or weren't the subjects of the particular paper that we've written, we're after understanding what affect this particular effects and whether this particular effect is going to be important or not.

Dominic - And what's the conclusion? What kind of sea level rises might we expect to result from this?

Anthony - We found that there was very little sea level rise associated with this particular effect. Our ballpark figure is roughly 8 millimetres by 2200 which in comparison to other sources of sea level rise over the next century is fairly minimal.


Add a comment