Under the Midnight SunAn Oregon State-led team and the search for ancient ice.

It was Dec. 23, our eighth night in a row of working the night shift in the drill tent. We were trying to finish drilling the longest ice core ever drilled in the Allan Hills of Antarctica. Radar data indicated that the bottom was somewhere around 300 meters down, and we had passed that mark yesterday. The previous night, I had felt so confident that I started playing Europe’s “The Final Countdown” on the sound system while we got set up. But now I was losing hope that we would reach our goal of bedrock by Christmas.

I was an ice core handler on the COLDEX 2025-26 Allan Hills field team. We had been living in tents on the ice sheet for just over a month, drilling cores in minus 4 degrees Fahrenheit weather. NSF COLDEX, the National Science Foundation Center for Oldest Ice Exploration, is dedicated to finding and analyzing the oldest ice possible. Specifically, the center is interested in the content of the air bubbles trapped within — tiny time capsules that provide important information about the climate millions of years ago.

COLDEX is part of an international effort to extend the ice core record back to more than 1 million years ago. When the ice core community conceived the Oldest Ice Challenge in 2004, the oldest ice core ever drilled was only about 800,000 years old. Taking that beyond 1 million years would uncover data from a time when climate cycles were shorter, with ice ages about once every 40,000 years, compared to the more recent 100,000-year cycle. Even more importantly, during interglacial periods earlier than 1 million years ago, our planet was very warm — even warmer than today. Greenhouse gas data from back then will help us understand how our climate could change in the near future, which could lead to new strategies for improving human resilience.

Several teams worldwide took up the challenge. Founded in 2021 as a multi-institutional NSF Science and Technology Center headquartered at Oregon State University, COLDEX is the United States’ entry. 

Unlike the other international teams, COLDEX focuses on blue ice areas. These cover less than 1% of Antarctica and are found mainly at the margins of the East Antarctic ice sheet. Here, ice flow to the sea is interrupted by mountains, especially the Transantarctic Mountain Range. At these barriers, ancient ice once buried deep under the ice sheet resurfaces, and harsh katabatic winds scour away the younger ice on top. What’s left appears blue due to millions of years of compression.

Recent COLDEX field seasons to blue ice areas have yielded ice far older than other oldest ice projects less than 200 meters below the surface. Unlike areas where deep ice cores are drilled, the Allan Hills have large amounts of ice from the Pliocene epoch (2.5 to 5.3 million years ago) and even before that. During the 2024-25 field season, COLDEX set a record, drilling a core with ice more than 6 million years old at its bottom.

But there are disadvantages to working in a blue ice area. The same ice flow patterns that result in the retention of such old ice make the ice very difficult to work with. It has traveled hundreds of kilometers and has folded and thinned unevenly. John Higgins, a professor at Princeton University and the science lead for this COLDEX project, likes to compare the ice core records from deep continuous cores to “pages in a book, because you can read them in order and they tell a continuous story.” Data from Allan Hills ice cores, however, are more like snapshots jumbled up in a loose box of photos, with some of photos missing altogether.

Because of this, we can’t assume that deeper ice is older than the shallower ice. COLDEX scientists determine the age back in the lab by analyzing argon gas isotopes trapped in bubbles in the cores. They’ve discovered that in one area of the Allan Hills, there’s a layer of much younger ice sandwiched in between two 3-million-year- old layers. Careful handling in the field — and detailed analysis in the laboratory — is required to accurately interpret data from these cores.

Greenhouse gas data from back then will help us understand how our climate could change in the near future.

---

Each time the drill pulls up another meter of ice, drill fluid drains out of the core barrel onto a sloped drip tray, which funnels it into a bucket. The fluid can be filtered to remove ice chips and then reused. The core handlers — wearing nitrile gloves over our wool glove liners to keep our hands dry, but not quite warm enough — vacuum the surface of the core to remove as much of the Estisol fluid from the surface as possible. The handlers then measure and log the core; label, photograph and bag it; and carefully place it in a large, insulated box, packed with snow, for storage and transport. These boxes can hold nine of the meter-long cores and, when full, weigh up to 200 pounds. In between drilling, the handlers try to warm up with thick gloves, hand warmers and a lot of dancing around.

Whenever Estisol splashes onto any of our regular clothing, the smell follows us around for days. It has a vaguely fruit-like or floral scent — not nearly as harsh as other solvents like acetone, but still bothersome.

At the Allan Hills, our camp didn’t have washing machines or bathing facilities, which meant that we wore the same layers of wool and fleece under our heavy overalls and coats for days or weeks on end. Fortunately, we had protective drill suits to cover our clothes, and boots that would not degrade when they came into contact with the drill fluid.

CLOCKWISE FROM TOP LEFT: This is as low as the sun ever gets when you’re at latitude 77 degrees south of the Earth’s equatorial plane. “Sunset” arrived at 1:30 a.m. on Dec. 17, 2025, in the Allan Hills of Antarctica. University of Washington graduate student An Li draws an orientation line on the second-to-last ice core section. The triumphant COLDEX team after loading boxes of ice cores onto the plane at the end of a successful field season. From left to right: An Li, Romilly Harris Stuart, Ivo Strawson, Danielle Whittaker (the author) and Martin Froger Silva. A slice of blue ice showing bubbles trapped within. These bubbles are like tiny time capsules, containing samples of ancient air that can be analyzed in the lab. Photos by Danielle Whittaker and Martin Froger Silva

This field mission was actually the first time we used the Shallow Wet Drill and the Estisol it requires. Previously, cores in the Allan Hills were dry drilled, a relatively easy and fuss-free method with minimal setup required. But compressed air in the bubbles in the ice can crack and badly fracture the cores while they’re being cut, contaminating the contents with modern air. COLDEX proposed to test whether fluid drilling would result in better ice core quality and commissioned this new drill from the NSF Ice Drilling Program for this season. This year’s goal wasn’t to find older ice than before, but to bring back better ice.

The improved core quality made all the extra effort worth it. We started drilling this particular borehole dry, waiting to add fluid until core quality started to degrade. One day in mid-December, at about 115 meters deep, the last core of the day came up as complete rubble, unusable for scientific analysis. The next day we filled the borehole about halfway with Estisol and immediately began drilling beautiful, perfect cores, proving that wet drilling would work in a blue ice area.

Still, despite drilling up to 20 meters a day, we were running out of time. In addition to drilling ice, our field team was also responsible for collecting data on the depth and movement of the ice sheet itself. We couldn’t take measurements until we were finished; otherwise the electricity from the generator-powered drill would interfere with radar measurements. We were scheduled to return to McMurdo Station on Jan. 11. We needed to speed things up and finish by Christmas in order to leave enough time for the other projects.

On Dec. 16,we had split the team into two shifts, with the goal of drilling about 16 hours a day until we reached bedrock. I volunteered for the night shift, figuring that bedtime was arbitrary when you had 24 hours of sunlight a day. An Li, a graduate student from the University of Washington who led the geophysics work on our field team, was the other core handler on the night shift. Andrew Haala and Dusty Brunner, engineers from the NSF Ice Drilling Program at the University of Wisconsin, rounded out our team.

The first shift included OSU postdocs Romilly Harris Stuart and Ivo Strawson, Martin Froger Silva from the University of Minnesota, and drillers Elizabeth Morton and Jay Johnson. Each day, they headed out to the drill tent to get started by 9 a.m. Second shift slept later, checking in around 1 p.m. to either swap out with a member of the day shift, or to work on tasks at camp like labeling sample bottles and bags or preparing equipment.

Dinner was the only time during this period that the whole team was together. We gathered in a large communal heated tent with 12 metal folding chairs set up in a circle around the perimeter. Our camp manager prepared surprisingly good meals for us, given the conditions: chickpea curry over quinoa, pasta with pesto and shrimp and grilled cheese sandwiches were some of my favorites. Second shift began right after dinner and lasted until after midnight. Once we were done drilling for the night, we spent an hour or so cleaning up, shutting down the drill and the generator and transporting the ice core boxes to a storage trench dug into the ice. At last, we headed back to camp, usually getting to bed around 3 a.m. under the still-bright sun.

The novelty of the situation had been energizing at first, and the two shifts enjoyed some friendly competition and cheered each other on. First shift left thermoses of hot water and encouraging notes for us to find when we returned. The second shift was sassier, writing “Second shift rules, first shift drools!” on the whiteboard in the community tent.

But the anticipation of the end and the failure to reach it day after day wore on us. By now, we had endured multiple days of winds between 30 and 40 knots (35 to 46 miles per hour), creating a wind chill as cold as minus 36 degrees Fahrenheit. Life in high winds is difficult: visibility is obscured, walking outside requires determined effort, and you must yell to be heard inside the flapping tents.

“You’re going to do it tonight, I am sure of it,” said Ivo at dinner on Dec. 23 when the night shift was suiting up. The first shift had stopped drilling just shy of 324 meters — the equivalent of an 80-story building, over three Statues of Liberty long. “You are going to go in there, drill 3 meters and hit bedrock!” I hoped he was right.

Feeling less optimistic than the night before, we set up for another drilling shift. The drill had some trouble cutting through the ice at first, retrieving only 11 centimeters of ice. Andrew and Dusty changed the cutters on the drill and tried again. This time, the core successfully pulled up a full meter, after a long struggle caused by cutting through a rock. The core had some very deep micro cracks, so thin that you might miss them if you didn’t know what you were looking for.

Looking into the borehole, you can see just how blue the ice really is. The blue color is the result of millions of years of compression. Photo by Martin Froger Silva

Such cracks cause the air bubbles to leak and gases within the core to mix together, rendering them useless as pristine air samples from specific moments in time. The appearance of micro cracks usually indicated that we needed to add more fluid to the borehole to counteract the pressure at deeper depths. The fluid column was now about 110 meters deep.

Andrew frowned at the cracks and asked, “How valuable is the ice at the bottom of this core?” An and I looked at each other. “Well … ,” she said, “the ice at the bottom is usually the oldest, so pretty important.”

Andrew gamely went out into the howling winds to get another 55-gallon barrel of Estisol and pump the fluid into the borehole.

It was after 11 p.m. when the next core came up, and it was only half a meter. The work was feeling especially slow and discouraging. Each drill run took half an hour or more — first to lower the drill all the way down, and then to slowly and carefully cut through the ice. “What do you think? Are you up for trying for one more core?” Andrew asked. Sure, why not, we shrugged.

While the drill made its way down deep into the ice, I asked Andrew, “How will we know when we’ve hit bottom?” I knew that when a dry drill reached bedrock, it would bring up bits of rock from the bottom, but I didn’t know what it would look like in wet drilling.

“Well,” Andrew said, “we won’t know for sure at first. The drill won’t go down any further and we’ll see sediment from the bottom. But most importantly, the cutters will be worn down really badly.” The three sharp cutters that cut into the ice were built for ice, not rock.

About 20 minutes later, Andrew pulled up the core. Again, only half a meter. But when Dusty pulled the core barrel out onto the table, An and I practically screamed. The ice chips that poured out from the barrel were sludgy and very brown, full of sediment. Dusty inspected the cutters. They were badly chewed up, despite being replaced just a couple of hours earlier.

I looked at Andrew. ”We’re going back down, right?” There was no question. Andrew and Dusty changed out the cutters and lowered the drill down the borehole once again. When it came back up, the barrel was empty and the cutters were once again very dulled. It was just after midnight. “Who wants to make the call?” asked Andrew, pointing at the handheld two-way radio.

A few team members had taken radios to bed with them the last night or two so they could be woken up with the news in case we reached bedrock. I grabbed the radio and pushed the talk button.

“Attention, Allan Hills. This is Shallow Wet Second Shift with a very important announcement. We have reached bedrock at approximately 327 meters. I repeat, we have hit bedrock. Have a good night.”

Our teammates back at camp had gone to bed hours before, so I didn’t expect an immediate response. But seconds later, cheers and whoops filled the airwaves. We had completed the deepest ice core ever drilled in blue ice.

Danielle J. Whittaker is a science writer and the managing director of COLDEX.