Scientist at work: Tracking melt water under the Greenland ice sheet

The crew of scientists prepare to put the drill stem into the Greenland ice sheet to probe water flows about a half of a mile below. Joel Harper, Author provided

During the past decade, I’ve spent nearly a year of my life living on the Greenland ice sheet to study how melt water impacts the movement of the ice.

What happens to the water that finds its way from the melting ice surface to the bottom of the ice sheet is a crucial question for glaciologists like me. Knowing this will help us ascertain how quickly Greenland’s ice sheet could contribute to global sea-level rise. But because doing this type of research requires studying the bottom side of a vast and thick ice sheet, my colleagues and I have developed relatively unique research techniques.

Our approach is to mimic the alpine style of mountaineering to do our polar research. That involves a small group of self-sufficient climbers who keep their loads light and depend on speed and efficiency to achieve their goals. It’s the opposite of expedition-style mountaineering, which relies on a large support crew and lots of heavy equipment to slowly advance a select few people to the summit.

We bring a small team of scientists who are committed to our fast and light field research style, with each person taking on multiple roles. We use mostly homemade equipment that is designed to produce novel results while being lightweight and efficient – the antithesis of “overdesigned.” The chances of scientific failure from this less conventional approach can be unnerving, but the benefits can be worth the risks. Indeed, we’ve already gained significant insights into the Greenland ice sheet’s underside.

Mysterious place

Our science team from the University of Montana and University of Wyoming sleeps in backpacking tents, the endless summer sunshine making shadows that rotate in circles around us. Ice-sheet camping is challenging. Your tent and sleeping pad insulate the ice as it melts, and soon your tent rises up into the relentless winds on an icy drooping pillar. Occasionally people’s tents slide off their pillars in the middle of the night.

But it’s not the melting on the surface that concerns us so much as what’s happening at the base of the Greenland ice sheet. Arctic warming has increased summer melting of this huge reservoir of ice, causing sea levels to rise. Before the melt water runs to the oceans, much of it finds its way to the bottom of the ice sheet.

The additional water can lubricate the base of the ice sheet in places where the ice can be 1,000 or more meters thick. This causes the ice to slide more quickly across the bedrock on which it sits. The result is that more ice is transported from the high center of the ice sheet, where snow accumulates, to the low elevation margins of the ice sheet, where it either calves into the sea or melts in the warmth of low elevations.

A system of pumps and heaters generates a high pressure jet of hot water that is used to melt a hole to the bottom of Greenland ice sheet.

One school of thought is that a feedback may be kicking in; the more water added, the faster the ice will move, and so ultimately the faster the ice will melt.

An alternative hypothesis is that adding more water to the bed will create large water flow pathways at the contact between the ice and bedrock. These channels are efficient at flushing the water quickly, which could limit the effects of increased melt water at the bed. In other words, by adding more water there is actually less lubrication – not more – because a drainage system develops that quickly moves the water away.

We know flowing water generates heat and melts open the channels in the ice. However, the enormous pressure at the base of the ice acts to squeeze the channels shut. Competing forces battle in a complicated dance.

We can represent these processes with equations, and simulate the opening and closing of the channels on a computer. But the meaningfulness of our results depends on whether we have properly accounted for all of the physical processes actually taking place. To test this, we need to look under the ice sheet.

The bottom of the ice sheet is a mysterious place we glaciologists spend a lot of time hypothesizing about. It’s not a place you can actually go and have a look around. So our team has drilled boreholes to the bed of the Greenland ice sheet to insert sensors and to conduct experiments designed to reveal the water flow and ice sliding conditions. They are essentially pinpricks that allow us to test and refine our models.

Homemade heat drill

Our approach to penetrating many hundreds of meters of cold ice (e.g., -18 degrees Celsius) is to run a light and nimble drilling campaign. We use alpine climbing tactics so that we can move quickly around the ice sheet to drill as many holes as we can in different places, to see if conditions vary from place to place. Our drill can be moved long distances in just a few helicopter loads, and we carry it ourselves for shorter hauls.

We don’t have devoted cooks or mechanics or engineers; we have a small group of faculty and carefully selected students who need to do it all. We rely on people who can fiddle with the electronics of homemade instruments while being unafraid of hard manual labor like moving fuel barrels and hooking up heavy pumps and hoses in the biting cold Greenland wind. Back in the lab, these same people must have outstanding skills to apply math and physics to data analysis and modeling.

The drill is moved long distances by helicopter, and shorter distances by hand-carrying over the ice. Our goal is to keep the drilling equipment as small and light as possible to permit easy transport. Joel Harper, Author provided

Our homemade drill uses hot water to melt a hole through the ice. We capture surface melt water flowing in streams, heat it to near boiling and then pump it at very high pressure through a hose to a nozzle that sprays a carefully designed jet of water.

Our drilling days are long, extending from morning to well into the night. When the hole is finished, that’s when our work really begins because we only have about two hours before the hole completely freezes shut again. We need to get the drill out of the hole and all experiments completed before that happens. Like astronauts who rehearse their spacewalks, we plan every step and try not to panic when something unexpected happens.

We conduct experiments by artificially adding slugs of water to the bed to measure how the drainage system can accommodate extra water. We send down a camera to take pictures of the bed, a suction tube to sample the sediment and homemade sensors to measure the temperature, pressure and movement of the water. We build the sensors ourselves because you just can’t buy sensors designed for the bottom of a 800-meter-deep hole through an ice sheet.

Joel Harper (Univ. of Montana) and Neil Humphrey (Univ. of Wyoming) operate the hot water drill. Joel Harper

I’ll admit our fast, light approach to drilling comes with risks. We don’t have redundant systems and we don’t carry lots of backup parts. Our lightweight drill makes a narrow hole, and the top of hole is freezing closed as drilling advances the bottom. We’ve had scary episodes where we’ve almost lost the drill.

A generator fails or a gear box blows, and now the hole is freezing shut around the 700 meters of hose and drill stem. If we can’t come up with a fix within minutes, the drill is lost and the project is over. We could take much less risk by scaling up logistics and reducing our goals. But that would mean doubling the crew and the pile of equipment, and adding another zero to our budget, only to drill one or two holes a year.

Our light-and-nimble approach has allowed us to drill holes quickly and to move large distances. We have drilled 36 boreholes spread along 45 kilometers (28 miles) of the ice sheet’s western side. The holes are up to 850 meters deep, or about a half of a mile, and have produced multi-year records of conditions under the ice.

Different physics than thought

Our instruments have discovered the water pressure under the ice is higher than portrayed by computer models. The melting power of flowing water is less effective than we thought, and so the enormous pressure under the thick ice has the upper hand – the squeezing inhibits large channels from opening.

This does not necessarily mean the ice will move faster due to enhanced lubrication as more melt water reaches the bed. This is because we have also discovered ways the water flows in smaller channels and sheets much more quickly than we expected. Now we are retrofitting our computer models to include these physics.

Our ultimate goal is to improve simulations of Greenland’s future contributions to sea level. Our discoveries are not relevant to tomorrow’s sea level or even next year’s, but nailing down these processes is important for knowing what will happen over upcoming decades to centuries. Sea level rise has big societal consequences, so we will continue our nimble approach to investigating water at Greenland’s bed.