Snow in Antarctica, and other Science Fiction in Science Advances
There’s a recent paper in Science Advances that argues that the West Antarctic Ice Sheet is disintegrating, and researches a potential action to stop it. Gizmodo has a more readable summary, if Science is a little heavy. I thought that the article made some rather absurd claims, and in this post I’ll use Fermi estimation to explain exactly how absurd those claims are.
Credit: Alfred Wegener Institute/Stefan Hendricks
Stabilizing the West Antarctic Ice Sheet by surface mass deposition
There is evidence that a self-sustaining ice discharge from the West Antarctic Ice Sheet (WAIS) has started, potentially leading to its disintegration. The associated sea level rise of more than 3m would pose a serious challenge to highly populated areas including metropolises such as Calcutta, Shanghai, New York City, and Tokyo. Here, we show that the WAIS may be stabilized through mass deposition in coastal regions around Pine Island and Thwaites glaciers. In our numerical simulations, a minimum of 7400 Gt of additional snowfall stabilizes the flow if applied over a short period of 10 years onto the region (−2 mm year−1 sea level equivalent). Mass deposition at a lower rate increases the intervention time and the required total amount of snow. We find that the precise conditions of such an operation are crucial, and potential benefits need to be weighed against environmental hazards, future risks, and enormous technical challenges.
As Gizmodo summarizes the article, “Researchers used a climate model to simulate the ice sheet’s continued meltdown and then cranked up the snowfall on it, which in turn would become ice as it gets compacted down. The findings show that adding up to 32 feet (10 meters) of ice per year over the course of 10 years would be enough to protect the West Antarctic glaciers from completely destabilizing if our currently warmed-up climate stayed constant.”
Now, I am not a climate scientist or glacier simulation expert. I do believe in climate change, I do believe it is human caused, and I believe that something must be done, and lots of things attempted because we do not have the luxury of being efficiently slow or wrong on the matter. But I can do math, and that 7400 gigatonnes of snow over ten years struck me as an excessively large number. Fermi estimation is ‘back of the envelop’ calculations intended to check that your units make sense, and get an idea if something is impossible, or merely improbable.
The Gizmodo article quotes the amazing Dr. Jane Flegal on the impossibility of the plan because she’s an actual expert on geoengineering science and policy, but I worked out the math, which I’ll walk you through.
7400 Gigatonnes of ice is 7.4e15 kilograms of water. A kilogram of water takes up a volume of 1 liter (thanks, metric system!). A cubic meter of water is 1000 liters. 1e9 cubic meters is one cubic kilometer.
7400e9 Gt * 1000 kg/t = 7.4e15 kg
7.4e15 kg * 1 L/kg = 7.4e15 L
7.4e15 L/ 1e3 L/m^3 = 7.4e12 m^37.4e12 m^3/ 1e9 m^3/km^3 = 7400 km^3 of water
7400 cubic kilometers of water is a lot. That’s about as much water as is in Lake Michigan (4920 km³) and Lake Huron ( 3,540 km³) combined.
And of course, we’d have to turn that water into snow. The latent heat of fusion for water is 333 Joules per gram, so that’s how much energy we have to pull out of a gram of water at 0 degrees to freeze it. The article mentions that 12,000 windmills might be necessary to pump water from the Southern Ocean to the ice shelf, but let’s ignore that and just figure out how much energy it’d take to freeze all that water, assuming we had a perfect refrigerator.
7.4e15 kg * 1000 g/kg = 7.4e18 g7.4e18 kg * 333 J/g = 2.46e21 JWe have ten years to make freeze this water, so divide by ten to get 2.46e20 J per year.
China’s electricity production in 2017 was 2.25e20 Joules. I’m sure that China wouldn’t mind turning out the lights for a decade.
Of course, Antarctica is itself cold, which can help, but that heat of fusion and any waste heat from refrigeration plants would equalize with the local atmosphere, which might have negative impacts on the ice shelf. Did the model account for that?
Finally, we need to talk about the material scale of the project. 7400 gigatonnes is a lot of anything. Assuming it’s one tonne per cubic meter (ice is 92% as dense as water, so this is close enough) that’s 7.4e12 cubic meters of snow to spread out over Antarctica. The Panama canal involved excavating about 500 million cubic meters of earth (by Fermi estimating, summing up the Culebra cut and major locks). This project would involve moving material equivalent to 1,000,000 Panama canals.
Of course, we’re not moving material, we’re making snow. Different business entirely. Big Bear Mountain, CA, has snow making equipment that uses 5000 gallons of water a minute, or about 20 tonnes of water a minute. It’s hard to run a ski resort outside Los Angeles, and Big Bear likely has the most extensive artificial snow equipment in the world.
7.4e12 tonnes / 20 tonnes/minute = 3.7e11 minutes3.7e11 minutes / 365*24*60 minutes/year = 700,000 years.
We’d have to run it full blast for 700,000 years to get the required amount of snow, or use snowmaking equipment about 100,000 times better than Big Bear’s for a decade. I don’t have exact numbers, but “multi-million dollar investment” is what I see for how much Big Bear’s equipment cost.
So in summary, we’d need to freeze a couple of Great Lakes, using as much electricity as a major country, after investing hundreds of billion of dollars in snow making infrastructure, all in some of the remotest and most inhospitable terrain in the world.
I’m sure we’ll get right on that.
Does Science Have to Be Possible?
I can’t speak to the correctness of the paper’s application of the The Potsdam Parallel Ice Sheet Model. It is possible that solving this problem advances the state of the art in some fundamental way, which is not apparent to a reader outside of glaciology. Pure science is its own reward.
But I’d hazard that the goal of this paper is not pure research. Gizmodo quotes one of the authors.
“I am not proposing to do what we have shown to be possible but that is a decision for society,” Levermann said. “People have to understand that if we do not want to lose our cities we have to do something. We just laid out one possibility.”
This plan is exactly as possible as my plan, which involves a virgin pure of heart befriending a wish-granting unicorn, and then wishing that atmospheric CO2 be 350 ppm.
Certainly, we cannot foreclose any option prematurely. But Fermi estimation is a basic tool and best practice. If the orders of magnitude are off, it doesn’t matter how advanced and sophisticated your model is. An option that is impossible is in some ways worse than silence.
No physical resources will ever be diverted to this plan, but it’s possible that political and rhetorical resources will be, that blue ribbon commissions will write reports about ‘re-icing the Antarctic’, rather than engaging with the hard realities of climate refugees and energy system transitions.
Peer review is supposed to establish a baseline of credible scholarly statements. Science Advances’ mission statement is to publish “high-quality, original research and reviews in all disciplines of science.” If articles in Science Advances are in no longer bound by strict adherence to the laws of physics, then perhaps they should change their name to Science-Fiction Advances.
