Under the icefield and inside the mountain

Go deep below the Columbia Icefield in Banff National Park to “the most inaccessible place in Canada” with the 2020 Castleguard Cave Expedition

  • Apr 15, 2021
  • 2,144 words
  • 9 minutes
Expand Image

Castleguard. For a caver, the name invokes imagination and emotion. As Canada’s longest cave, it has enchanted expedition cavers from around the world who have contributed to its exploration over the last 50 years. A few unexplored leads still beckon, and armed with clear scientific and exploration objectives, a research permit from Parks Canada grants access for experienced teams.  

I was first there in 2009, fulfilling a dream of my own. And by 2012 I was the leader of a team at Castleguard supporting British cave diver Martin Groves. He was planning dry cave exploration beyond Boon’s Sump, a submerged tunnel nearly one kilometre in length. That year, an unprecedented ice blockage in a passage called the Ice Crawls scuttled the trip. Since then, a couple of teams have attempted to continue exploration, but with unpredictable ice blockages it is hard to commit to the logistics required to explore Castleguard.

Years passed until a seed was planted with a few visiting cavers who were in Canada decompressing after the massive 2018 cave rescue in Thailand. Australians Richard Harris and Craig Challen, along with another caver in our network, New Zealander Tom Crisp, were enchanted by the prospect of continuing where Martin left off in 2012. Our 2020 expedition was organized by Katie Graham of the Alberta Speleological Society. 

The massive effort would require transporting loads of scuba diving and caving equipment within the cave to Boon’s Sump. A dozen volunteers were needed on the front and back end of the expedition, along with an exploration group that would continue on pursuing underwater and dry cave exploration.  Reaching out beyond the caving community to the ACC and ACMG meant a few folks who were not primarily cavers were able to experience Castleguard, a rare opportunity.

Claustrophobia level: ice crawl

On the evening of March 8th, 2020, we received a desperate message from our support team who had arrived at the cave in advance. The ice crawls were tight. Not totally blocked, but not everyone could squeeze through them. Our solution to save the expedition? Chainsaw a channel through the ice restriction. So, the next morning while the main contingent started skiing the 20-kilometre route to the cave, I stayed back with our two Société québécoise de spéléologie cavers to scour Canmore hardware stores for 18-volt chainsaw batteries. 

Expand Image
Kathleen Graham at Boon’s Sump, Castleguard Cave. The start of an 845 meter
underwater tunnel. (Photo: Matt Kennedy)

The ski to the toe of the Saskatchewan Glacier was familiar, and bluebird skies made for a pleasant outing if not for the crisp wind coming from the Columbia Icefield. The most treacherous part of the route is ascending the moraine to join Castleguard Meadows. The steep, windswept slope contained no trace of the previous ski track and we progressed slowly upwards with eighty-pound pulks — albatrosses constantly pulling us down. The last kilometre descending towards the Castleguard River Valley was lit by headlamp and characterized by lots of swearing. Tree wells had exponentially increased the albatross factor of our pulks. Salvation was at Entrance Camp just inside the cave.

The next morning my “work shift” was with Tyler Neiss, our lone American caver. Beyond the imposing metal gate, the passage meanders until you arrive at an 8-metre drop.  We rappelled down to the base of the pitch and ducked into the start of the ice crawls. Laying flat out on the ice we slid along and eventually came to the blockage. Roughly 20-centimetres of airspace was between the ice floor and limestone ceiling. Tyler started first. The process was to cut some grooves with the chainsaw and then hammer out the ice with a hatchet and chisel to cut a trough a few centimetres deeper. Laying flat on the confined ice surface while damp and being blasted by the cold 25 km/h draft necessitated short work shifts. I was anxious to get chiseling.

Taking over from Tyler I slid forward into the trough with my head tilted to the side and my torso squeezed between the ice and rock. It didn’t take long for me to notice my heart pounding and the intense sensation of breathlessness that was overtaking me: panic. Completely overcome with the desire to extricate myself I hurriedly backed out of the trough and caught my breath. After a few minutes I made a second attempt. Trying again didn’t help, and I was left to watch Tyler take another crack at chiseling before sulking back to Entrance Camp.

Katie confronted me that night. If we could solve the ice crawl problem, we were going to do a six-day journey deep into the cave and if I couldn’t pass the ice crawls then I would have to stay at Entrance Camp. I felt like it would be more acceptable to have a broken leg, like somehow a physical injury would have been a better excuse. The thought of not being able to go deep with my team was devastating. My personal Sagarmatha was to reach Castleguard’s famous ice plug, where the cave ends underneath the Columbia Icefield. Our objectives in the farthest reaches of the cave would bring us there, and I was going to miss the opportunity.

Expand Image
Deep caving team at the Ice Plug, underneath the Columbia Icefield. (Photo: Jeremy Bruns)

The next day everyone mobilized to get the diving equipment to the sump and I kitted up with the team to try again. News that larger team members had fit through the trough was comforting. When my turn came, I felt my way forward with one arm stretched out in front pushing the cave pack ahead of me and the other arm to my side. The trough meandered until it opened up to where I could turn my head and see where I was going, and finally to where I could crawl on hands and knees. I had made it through.

The haul past the ice crawls consisted of a kilometre of crawling over sharp rocks. The team effort had moved all the rebreathers, dive cylinders, lead weights, and other equipment to the sump by nightfall.  Tom, Richard, and Craig would spend the next few days alone at Entrance Camp executing the dive. Our support team would leave the next morning, and Katie, Tyler, Jeremy Bruns, Christian Chenier, Catherine Tardy-Laporte and myself would spend the next six days doing exploration projects based from camps deeper in the cave.

Progress at the deep camps

Obstacles characterize the passages carving further into the mountain. After the ice crawls the next obstacle is the pools, a series of ice-cold waist deep ponds. Further along is a 24-metre rappel, and then The Subway, a near perfect example of a phreatic tube that goes arrow-straight for half a kilometre.  After this is The First Fissure: a vadose canyon sometimes ten metres deep. Cavers must bridge the passages that climb and descend within the rift. A day’s worth of hard travel finally brought us to Camp I, where flat ground with nearby pools of water in ancient drip cups were an oasis amongst the rifts and rubble.

From Camp I our team split up to tackle our exploration objectives. Katie, Jeremy, and Catherine tackled F7, an inhospitable passage yet to be fully explored. My team went to The Next Scene, a tight passage described both as a “terrific” lead by a British team and “terrible” by the last Canadians to see it. We brought a chisel and folding shovel to make The Next Scene’s lead bigger if necessary. The lead was too tight and we tried digging, but left convinced that it would take weeks to make meaningful progress. We returned to Camp I to find Katie’s team had been skunked too; the F7 passage being dreadfully wet, drafty and tight. Dejected, we were hopeful for better results deeper in the cave.

Rousing from a cold, damp sleep to put on cold, wet cave suits, our team was excited to get moving for the next phase. I was the only member of our team who had previously been as far as Camp I, so unfamiliar ground lay ahead for all of us. Beyond Camp I, Holes-in-the-Floor provides a series of 19 pits to traverse. Then comes Second Fissure. Within the start of this new rift is the most sporting passage in the entire cave. The crux is fissure-eroded by a small waterfall requiring long legs and teamwork to cross. 

Finally, we reached Camp II, a bizarre anomaly within the Fissure where a 50-metre section of the rift has an actual floor. 

With time left in the day we continued to our objective, the Munchkin Peon Trail. This passage was partially explored from the north in 1980 and then from the south in 1984, but the second expedition’s connection to the 1980 survey did not line up. This survey error cast doubt into the accuracy of documented passages in the headward complex of the cave.  

Climbing into the Munchkin Peon Trail from Second Fissure involved wedging seven metres up a side rift. Catherine and I started the resurvey at the top of the climb. Christian Chenier and Tyler went through the passage to where it nearly connects to the main passage and surveyed back towards my team. We were able to resurvey approximately half of the passage, a mixture of walking, crawling, and flat-out munchkining before returning to Camp II. Progress was good, and we excitedly planned to reach the ice plug the next day. 

The ice plug

There is something about being so incredibly remote and to finally find yourself in a place of your dreams. The juxtaposition of being within Banff National Park, but also being incredibly isolated, takes time to fully comprehend and appreciate. There was a time when the ice plug was the furthest one could get from a cave entrance anywhere in the world. 234 metres of rock and ice lay between us and the surface of the Columbia Icefield above. Simultaneously inside a mountain and underneath the glacier, for a moment we were in what has been described as the most inaccessible place in Canada. The team was smiling all around. Katie and Catherine became the third and fourth females to reach the ice plug.

At Camp III Katie and Jeremy left for a scientific sample mission at Dessert Glacé. The Swiss team that first explored that section of cave in 1987 had found peculiar organic matter in the passage and a new collection of this material was needed to facilitate further research. Christian, Tyler, Catherine, and myself resumed the resurvey of Munchkin Peon Trail. We connected the surveys and gleefully returned to Camp II for the night. Reviewing the results in camp by the glow of our PDA screens, we knew we had fixed the survey error that had persisted for 36 years. It was a great success, but the statistics would have to wait until I could crunch the data on my computer back home.

Returning to the entrance took two days. Emerging from darkness I was struck by the scent and the magnificent colors of the world, of pine forest and Watchman’s Peak across the valley. While we were away Craig and Tom had dived Boon’s Sump, and surfaced in what became a fissure-like rift. With the high risk of puncturing their dry suits on sharp projections within the passage, they safely surveyed what they could, and returned for the dive home.

Our support team arrived with the freshness required to move the equipment from the Sump back to the entrance, and over two more days all was ready for our departure. 

Unlike other caving trips, returning after 11 days from the underground to the regular world had a different flavour this time around: despite the highlights of the news that our support team had provided, nothing could have prepared us for a world changed by COVID-19. We emerged from under the Earth into a new reality, one which still persists a year later.

Moving the needle

Post-expedition cartography revealed the outcome of our trip’s work; a kind of delayed gratification that often comes with cave exploration. Between the new section beyond Boon’s Sump, the revised Munchkin Peon data, and our other new “dry” passages explored, the needle was moved in Canada’s longest cave. Castleguard is now 21,311 metres long and 388 metres deep. Unfulfilled objectives persist, and some of what lies within will be unlocked in our next attempt, scheduled for the winter of 2022.


Are you passionate about Canadian geography?

You can support Canadian Geographic in 3 ways:

Related Content

A mountain half covered in snow


Why mountains matter in Canada

They sustain us, enrich our lives and inspire us

  • 1287 words
  • 6 minutes
Mount Logan


Re-measuring Canada’s highest mountain

A team of climbers and scientists plan to summit the Yukon’s Mount Logan next spring in the name of climate change — and to re-evaluate its height with modern GPS technology

  • 1671 words
  • 7 minutes


Secret season in Jasper: 3 days of ice walking, ice skating and other winter adventures

Avoid the crowds and discover frozen canyons, scenic lakes and hearty cuisine in this epic Rocky Mountain town

  • 2010 words
  • 9 minutes


Endangered spaces and species in Canada’s mountains

Glaciers play a vital role in sculpting the landscape, and their retreat can bring about drastic consequences

  • 1218 words
  • 5 minutes