Nonprofit journalism dedicated to creating a Human Age we actually want to live in.


Glacial Elevation Operations

It was pretty audacious: suck enough water from the underside of the glacier for the whole block of ice to lose its water cushion and crash back down onto bedrock.

By Kim Stanley Robinson

The Climate Parables series engages the powerful imaginative forces of science fiction to explore what it’s like to live in a future in which humans have discovered creative ways to mitigate climate change—and live well. Learn more here >

I was back down there again, helping with the seawater pumping experiment, even though it was obvious to all of us that it was a crazy idea. Ten million wind power turbines? Thousands of pipelines? Not going to happen. It was a fantasy cure.

But someone had to try it. And the project had one more season of funding. So we got the pump intake back through the sea ice into the water. Then we followed the pipeline up the big white hill. It was laid right on the ground, because snow or ice was a better insulator than air, and warmer too. Still, a big part of the total energy budget was for heating the pipes to keep the water liquid on its way to its destination. The rest of the energy was simply to move the water uphill. And water is heavy, and Antarctica is high. So, whatever. An experiment or an exercise in futility, depending on your view.

There were people proposing to generate energy from ocean currents. The Antarctic Current runs around the continent like a belt, clockwise as seen from above, and of course it gets channelized through the Drake Passage; if electricity could be generated from that faster section of the current, great. But none of us thought it would work. The sea eats everything you put in it, and the size and number of turbines that could spin up enough electricity to do the job was off the charts.

Then there were those who still held the dream of space-based electricity. Russians for the most part. They had used their Molniya orbit for communications satellites for a long time—this being an almost polar orbit, in an elliptical shape that brings it close to Earth twice a day. So the Russians were putting up satellites with solar panels and microwave transmitters to send power down to Earth. Microwave collection stations were to be located by the Antarctic pumps and heating elements, and electricity thus beamed down from space to help power the warmed seawater uphill and inland, even during the long night of the Antarctic winter.

Maybe, we said. Although the truth is that solar power from space is not likely to work very well. Capture, transmission, reception, all problematic. Even if a sufficient power supply was found, people would still be required at the upper ends of the pipelines to oversee the water getting poured out up there. So this season we tried that part too, and it was a weird sight to see. Typical polar plateau scene, Ice Planet Zero, a sastrugied white plane to every horizon, domed by a dark blue sky very low overhead, stupendously awesome, you feel like the Little Prince and have to pinch yourself from time to time, also do some Pete Townshends to keep your hands warm.

When the water pours out of the end of this pipe, it steams madly in the dry polar air and then sploshes down onto the ice and runs away, just as we had planned it, having aimed the nozzle down a slight hill. But the tilt of this so-called hill was about two meters in every kilometer, as good as we could get in the region. So we got surprised by how fast the water froze. Maybe we shouldn’t have been, most of us had tried the old Antarctic trick of taking a pot of boiling water outside and tossing it into the air to watch it steam and crackle and freeze to ice bits before it hits the ground—it’s an experiment that never ceases to amaze. But pouring it out in quantity, as from a fire hose or a sewer outflow, we thought it would take longer.

We could deposit about a meter of water per year on any part of the polar plateau and still have it freeze successfully . . . At a meter thick, that would be about a third of Antarctica. Not going to happen, no way in the world.

Not so. In fact, newly frozen ice stacked just a few meters from the end of our pipe, creating a low dam, making downhill no longer down, so that the unfrozen water began to flow back toward the pipe outlet, and then past it in the other direction. Oh no!

We hustled down to the little ice dam and started trying to break it, which worked about as well as you might expect. In the midst of our fuss, Jordi shouted, Hey I’m stuck! Help!

He was standing in what had been ankle-deep water near the outlet, which was now ice that had him stuck in place, with more water flowing over his boots all the while. Help!

We laughed, we cursed, we tried to cut him out, nothing worked. He was not in imminent danger, but on the other hand we couldn’t free him. And in the race between rising ice and the newly emerged water sliding slickly over it, the ice was winning. Pull him out of his fucking boots, I said. Leave his boots.

It's time to upgrade not just our technology, but also our collective imagination.

Discover Anthropocene’s newest and most forward-looking project: Climate reporting from the future.

Learn More

So Jordi was saved, but the problem remained—water was going to freeze fast enough to create a problem in getting it to flow across the surface of the ice. The pipe outlet would have to be whipping back and forth like a hose on the driveway with too much pressure in it—which maybe could be arranged, but yikes, how to control that, it’s kind of nonlinear. A sharper gradient would help too, although on the polar plateau those are not easy to find.

So we closed down for a week, and rebuilt one outlet to emerge under pressure and snake back and forth like a windshield wiper, see how that went. And when we tried it again, water came out of the pipe and flowed downhill and froze along the way, and finally pooled pretty well away from the outlet, where it mounded and the new water shifted and flowed around it and slid down yet again. And we got better and better at mapping where it might ultimately pool, and staying out of those pools while it happened.

The estimate we came away with was that we could deposit about a meter of water per year on any part of the polar plateau and still have it freeze successfully. More than that and we would be exceeding the capacity of the air and ice in combination to chill the water. So we would need a wide spreading zone; at a meter thick, that would be about a third of Antarctica.

Not going to happen, no way in the world. We had firmly established a trifecta of impossibilities: not enough energy, not enough pipe, not enough land.

So the straight-up seawater pumping solution wasn’t going to work. It was a fantasy solution. The beaches of the world were fucked.

As we sat in our little habitats, like mobile homes half-filled with insulation, we would gather around the table, looking at maps and talking it over. World maps, I mean.

The endorheic basins of the world, meaning basins where water does not drain to the sea, were many in number. And many of them in the northern hemisphere were dry playas, where water had existed at the end of the last ice age but dried out since, partially or all the way. The Caspian Sea had been helped to dry down to its current level by people, the Aral Sea even more so. The Tarim Basin was completely dry all on its own, Utah’s Great Salt Lake was the remnant of a much bigger lake from the past—on and on it went, mostly in Asia and North America, and the Sahara. Of course there were people living in some of these places, but not many of them, given the problems of desertification, or disasterated shorelines in the case of the Caspian and Aral. If you added up their volume of empty available space, it was considerable. A lot of seawater could be relocated there, in theory. We ran the numbers; well, it would do for a meter or two of sea level rise. But then all those basins would be full, and you’d be back to the unworkabilities of Antarctica.

No. We needed to go back to the plan to pump water out from under the big glaciers, to drop them back on rock beds to slow them down. It was the only thing that was going to work. We had been following the money, taking it where we could get it, and doing what they asked us to do with it. The billionaires and oil companies and Russia—even the NSF has said, Pump seawater back onto Antarctica, cool idea! Do it! But we were the glaciologists, so we had to guide the process if it was going to have any chance of success.

So, the next Antarctic season we were back. This time we went to the Pine Island Glacier above the WAIS, the West Antarctic Ice Sheet. Pine was running narrow and fast into the sea, right next to Pine Island. It had been a study site for many years, so NSF had the logistics in place for getting a camp installed there.

Although in fact they weren’t experienced in getting as much stuff there as we needed. We needed almost as much as McMurdo itself, not really, but NSF had decided that if we were going to give this idea another test we had better make it a good test, or we wouldn’t know what our results meant.

So we borrowed the ships that resupplied McMurdo at the end of every summer, and the Russians sent a couple of their monstrous beasts south to help us bash an open-water lane to Pine Island, where we could land stuff and drag it over and up onto the glacier, using the same snow tractors that had been dragging fuel and gear from Mactown to Pole for many years.

It was like dragging a village over the ice, some kind of Baba Yaga thing, the monster tractors pulling trains of four or five huts in a row behind them from spot to spot. We started between the Hudson Mountains and Pine Island itself; this was a perfect pinch point in the glacier’s fall, a place where if we could thump it back down on its bedrock, it would slow for sure. So we circled the wagons and got to work.

The numbers were holding good. Amount of water lubricating the bottoms of Antarctica’s glaciers roughly sixty cubic kilometers . . . a lot of water for humans to pump. But not outside the zone of what we already pump every year.


The ice borers were still as simple as showerheads. Slow but effective. In the old days we burned a lot of fuel to get the showerheads’ water hot. Now solar panels helped to power the heaters. The meltwater under the showerheads gets suctioned and pumped out of the hole and reheated and used again, with the excess piped a distance away to freeze somewhere else. To both sides of the Pine Island Glacier are regions of ice moving slowly, so the water could be dumped there. In truth it was such a small volume that where we dumped it didn’t matter.

We also began to make use of the microwave energy beamed down to us from the Russian satellites in their almost polar orbits, to power the pumps and the showerheads when the sun went down. Test whether we could make that work, so we could keep the system going year-round.

That first experiment had at least taught us this, that you probably wanted to choose a homogeneous block of ice within the flow of the glacier, so that the movement of the glacier wouldn’t deform the hole as we drilled it. We had chosen our site partly for that reason—it was a single block that was 40 kilometers long and extended all the way across the glacier. Still 130 kilometers upstream from the ice shelf, and a couple hundred meters above sea level. Perfect.

We had done the calculations as to how many drill holes we needed, and how far apart they should be and so on. It was pretty audacious: suck enough water from the underside of the glacier for the whole block of ice that we were working on to lose its water cushion and crash back down onto bedrock, hopefully with a mighty squeal and maybe a crashing sound, as of tires braking on asphalt followed by car hitting wall. Even if that only happened in our heads, it did seem like it was going to be palpable when it happened.

The numbers were holding good. Amount of water lubricating the bottoms of Antarctica’s glaciers roughly sixty cubic kilometers. Not insignificant, a clear cube of ice about four kilometers on a side and the same high, so, half as tall as Everest—yes, a lot of water for humans to pump. But not outside the zone of what we already pump every year.

Still, a lot of water to pump, but that’s for all of Antarctica. Around the circle of the continent, seventy four glaciers dump the majority of the ice now rushing into the sea, with a few being the major contributors. So, sixty cubic kilometers sucked out from under seventy four glaciers. Okay, not so bad!

So we were melting and casing twenty boreholes into Pine Island Glacier. After that we’d pump up all the water that we could. It didn’t seem so bad! In the same realm as wells up in the world, draining fossil water for farms all over the Ogallala and other places set atop irreplaceable groundwater resources. Can be done! Solves all problems!

All right, it doesn’t solve all problems. But let’s not get picky. If sea level rises even a meter, all the beaches in the world are gone, and seaports and coastal infrastructures and salt marshes and you name it. And as Hansen and his team pointed out in their 2016 paper, if the rate of rise doubles every ten years, quickly you are fucked, all the coastal cities of the world devastated, damage in the quadrillions, if you think you can put a price on it. What’s the monetary value of human civilization? Trying to answer that question proves you are a moral and practical idiot. Well, economists make such calculations all the time, but that’s their job, and they think it makes sense. In this case, better just to throw up your hands and say civilization is effectively a fiscal infinity, a human infinity.

Greenland is easier, it’s basically a stone bathtub with narrow cracks in it. Tack down those suckers falling through the cracks, and you’ve got a somewhat stabilized sea level.

Tacking down the Antarctic glaciers won’t completely stop sea level rise, of course. But if we could get the big ones back to one-tenth of their current speed, like the good old days, that would help a lot. Best to do it in Greenland too—the ice there is going even faster, and even though it has only a tenth the ice of Antarctica, that’s still a seven-meter sea level rise, if all of it were to liquify. So fine, do them too; Greenland is easier, it’s basically a stone bathtub with narrow cracks in it. Tack down those suckers falling through the cracks, and you’ve got a somewhat stabilized sea level, rising at about a millimeter a year—meaning a thousand years before it rises a meter. Enough time to draw down enough CO2 to get back to 350 parts per million—hell, enough time to start another ice age if you want to!

Basically, the sea level rise problem gets solved. Beaches still in existence.

So, someone asked tonight in the mess tent, is what we’re doing down here geoengineering? Who the hell knows! What’s in a word? Call it Glacier Elevation Operations, Based on Estimates of Godawfulness Gobsmacking Interested Nations’ Goodness: GEO-BEGGING. Call it whatever you want, but don’t immediately clutch your pearls and declare we can’t predict the unintended consequences, we are sure to create backlash effects so bad they overthrow the good we intended, etc. There are some things man was not meant to know—my ass! We are meant to know everything we can find out. So get over that whole wimpy line of objection. And I’ll tell you what the unintended side effects of slowing down the glaciers of Antarctica will be: nothing. Nada. No side effects what-soever, and the beaches and coastal cities of the world will stay out of the drink.

So if this works, and it looks like it will, I think we’ll be doing it. Our team is relatively small. The project is expensive but not that expensive. Like drilling a few wells anywhere else, pumping water anywhere else, plus keeping everything warm. And getting the stuff and people here in the first place. That will be expensive, yes, and dangerous, but a quick calculation of the cost of this operation of ours multiplied by something like a hundred, or even a thousand—that gets us to ten billion dollars. Well, probably it will be more expensive than that, everything is when you actually get into it. But whatever—call it fifty billion dollars. This is such a bargain!


Today I made the rounds and found that every hole was working. Pumps pumping, heating elements keeping the water liquid. The lines from each pump feed into a bigger pipeline, like a piece of the Alaska pipeline lying right on the ice, with repeater heaters every kilometer, and joints where you can get into the line and run a pig if you need to melt an ice clog. The pipeline is running slightly uphill, which I think is an interesting decision, as you could actually just let it run downhill into the ocean, the addition to sea level rise would be so trivial, well below detection level. But people like to be neat.


Looks like tomorrow we’ll be able to run the final tests and declare the job done. Then in about a year we should be able to tell how it’s working in terms of slowing this beast down. Maybe five years to be safe. Although people will be in a hurry to declare one way or other. But if you’re going to be putting billions into it, and training a bunch of crews, you’d do better to be sure. Actually, one of the choke points in the supply chain, in terms of getting this done, if they decide to do it, is simply people. It takes a certain expertise. On the other hand if we’re shutting down all the oil operations, as we really ought to do, then that’s a lot of people out of work. And the work in question isn’t that different. Some of them might even regard this job as easier. Simple stuff, although colder. But if you’re working in Saudi Arabia, maybe you like the idea of cold. And if you’re working in Alaska, maybe it makes no difference. Yeah, that part will probably work out. We’ve only got fifty people here now, and you could do it with thirty—almost half of us are here to study the other half, or do science while we have the camp here, in the usual way. Again, scale that up and the number of people is still trivial. We’ll have a party tomorrow to celebrate.

I’m sorry to report that this is the last entry in this file of Dr. Griffen’s laptop. He took a last inspection run on February 6, and on the way back to camp he took a shortcut that left the snowmobile road, which is clearly flagged. Visibility was good, so no one is sure why he did that. Usually he stuck to the flagged road, just like the rest of us. So we are all mystified.

It was Jeff, one of our mountaineers, who followed a snowmobile track to an unobtrusive hole. He then came back and got Lance, the team’s other mountaineer, looking grim. They went out to near it and then stuck in some ice belays and roped up, and walked over to look in the hole. We watched from the dining hut, standing around silently. Lance belayed Jeff, who descended into the hole. He disappeared and was gone about, I don’t know, twenty minutes. It seemed longer. Finally he reappeared and climbed out and stood there, then walked back to Lance. They conferred; Lance put out an arm and held him by the shoulder. They hugged. They looked our way, saw us. Jeff shook his head. We understood it as clearly as if he had said it: Dr. G was dead.


Top Photo: The IceCube Neutrino Detector by Dr. Kathie L. Olsen, NSF

From: The Ministry for the Future: A Novel by Kim Stanley Robinson. Copyright ©2020 by Kim Stanley Robinson. Reprinted by permission of Hachette Book Group, Inc.

Anthropocene Magazine Logo

Get the latest sustainability science delivered to your inbox every week


You have successfully signed up

Share This

Share This Article