The Hakai team launches an ocean glider into the Pacific off Calvert Island. The autonomous instrument will spend weeks crossing the continental shelf, measuring temperature, salinity, and oxygen through the full water column in conditions too rough for a research vessel to work in.
Every summer, hundreds of meters below the surface of the Pacific Ocean, a river begins to flow toward the British Columbia coast.
Nobody can see it. It carries no sediment, leaves no surface trace, appears on no weather map. It moves cold and oxygen-poor, pressed upward from the deep ocean by summer winds that draw warmer surface water seaward, and then it slides eastward across the continental shelf, filling a long submarine valley, inching toward one of the most productive coastlines on Earth.
For most of human history, this river ran largely unseen. Fishermen worked the waters above it; oceanographers, sampling from research vessels, caught occasional glimpses of it. But tracking it was hard—no one could follow where it went across a whole season, or watch how it changed from one year to the next.
What changed, in the last few years, was that scientists at the Hakai Institute found a way to follow it. As Wiley Evans, the Institute's lead oceanographer, puts it: “We're tracking ocean change with robots.”
Calvert Island sits in a geographic gap that is easy to overlook on a map. It lies roughly halfway between Vancouver and Alaska, in the open water between Haida Gwaii and the north end of Vancouver Island—neither fully sheltered inside the coastal channels nor fully exposed to the open Pacific. The Heiltsuk town of Bella Bella is 80 kilometers away by sea. The nearest road is much farther.
This position, which might seem like remoteness, is actually the point. The continental shelf off Calvert Island is the broadest in western Canada—roughly 120 kilometers wide, a feature unlike anything else along the coast, or anywhere else on the west coast of North America. The channels of the central coast converge here. If you want to understand what the Pacific is doing to the BC coast, this is where you would want to be.
And cutting across that broad shelf is a riverbed of sorts—a long, deep channel called Goose Trough, incised to about 300 meters in places, running from the open ocean almost to the doorstep of Hakai Pass. It works, Evans explains, something like the inverse of a river valley.
The red lines show the actual paths traced by Hakai's ocean glider across the continental shelf, from the open Pacific to Hakai Pass. The long submarine valley running through the center of the image is Goose Trough—the channel that guides dense, oxygen-poor water toward Hakai Pass each summer.
Fly through Goose Trough. This animation reveals the underwater valley that guides deep Pacific water toward Hakai Pass.
Instead of water flowing downhill to the sea, this is a channel along which dense ocean water can move inward and upward. “You could think of it as a river basin in reverse,” he says. “It's a conduit for deep, dense ocean water—because it's hard for water to go up and over things.” The trough is the path of least resistance. The invisible river follows it because water, like everything else, takes the easiest way through.
To follow the water—not just visit it, but track it as it moved—the scientists needed a different kind of tool.
The ocean glider looks less like a scientific instrument than like a large yellow torpedo. About two meters long, finned, sealed against pressure, it carries no propeller. It moves by changing its own buoyancy—pumping fluid between internal chambers to sink or rise, using small wings to translate that vertical motion into forward travel. The result is a slow, undulating path: down toward depth, back toward the surface, down again, tracing a zigzag through the water column as it goes. At each turn it surfaces briefly, transmits its position by satellite, takes new instructions, and dives again.
A cross-section of the 150-kilometer shelf crossing—one leg of the glider's longer run out to the open Pacific and back. The glider moves by changing its own buoyancy, tracing a slow zigzag from surface to seafloor. Each time it surfaces, it transmits its data by satellite and receives new instructions before diving again.
What makes the glider revolutionary is not any single measurement but its endurance. It can operate for months without human contact, crossing open water in conditions that would keep a research vessel in harbor. “The glider doesn't care about bad weather,” Evans says. “It's out there in conditions that would be really tough for us to work in, even from a large vessel.” A single six-week mission—running from Hakai Pass out across the shelf to the edge of the open Pacific and back, a round trip of some 300 to 400 kilometers—costs around $100,000. A research ship costs roughly that much every two days. And all along its path, the glider measures continuously: temperature, salinity, oxygen, chlorophyll, a profile of the entire water column from surface to seafloor, collected in weather no crew could work in and at a fraction of the price.
Launching the ocean glider is one of the few moments when the autonomous instrument depends entirely on people. The Hakai oceanography team coordinates each deployment, sending the glider on a weeks-long journey across the continental shelf through conditions that would often keep a research vessel in harbor.
Before gliders, oceanographers relied largely on moorings—instruments anchored in one spot, recording continuously but only ever from that single location. A mooring watches one place; a glider watches a journey. And following an invisible river is nothing if not a journey.
The glider program isn't Hakai's alone. It runs as part of C-PROOF—the Canadian Pacific Robotic Ocean Observing Facility—a partnership with the University of Victoria, Fisheries and Oceans Canada, and the University of British Columbia that puts autonomous instruments to work across Canada's Pacific waters.
Under certain conditions each summer, northwesterly winds pull surface water away from the coast, and cold, dense water from the depths rises to replace it. This process—upwelling—is fundamental to the productivity of eastern Pacific coastlines: it draws nutrients from the deep ocean into sunlit surface waters, fueling food webs that support everything from krill to salmon to humpback whales. Scientists expected to find upwelling along the BC coast. What the glider revealed, in far more detail than had been possible before, was what that upwelled water actually carries—and where it lingers.
The water is hypoxic: low in dissolved oxygen. In the deep ocean, far from sunlight, oxygen is consumed by microorganisms over long timescales, and the water that rises onto the shelf carries that deficit with it. Through the summer, glider data showed this water settling into Goose Trough—lying along the seafloor, persisting for months, growing steadily more oxygen-starved. And the scale of it was greater than anyone had anticipated. “The extent and intensity were more than we were expecting to see,” Evans says. Analysis by postdoctoral researcher Sam Stevens, who set the glider observations against decades of historical records, showed that these hypoxic events have been growing more frequent, more intense, and more widespread over time. They now appear essentially every summer.
Glider data from 2023 shows low-oxygen water—shown in orange—moving up the continental shelf through summer and persisting into the deep channel of Goose Trough before retreating by October. The scale and intensity of these seasonal hypoxic events has been growing over time.
Then came 2023, and a wrinkle. In earlier years, when autumn storms arrived and the dense upwelled water was pushed back offshore, oxygen levels on the inner shelf recovered along with it. In 2023 the recovery lagged. Oxygen stayed low even after the dense water was gone—this time held down not by upwelling but by something local. Phytoplankton had bloomed in the nearshore, their cells sinking to the seafloor and feeding a bloom of bacteria that consumed oxygen as they broke the organic matter down. The oxygen did come back, but on its own schedule, decoupled from the tides and winds that had always governed it before. The story was more complicated than a river arriving and departing on schedule.
For marine life, none of this is abstract.
When hypoxic water fills the depths and warm surface water—above about 19 °C—caps the shallows, mobile species like salmon and herring are squeezed into an ever-narrower band of livable water in between. Evans calls it habitat contraction. “Part of the water column is no longer favorable habitat,” he says. “That could be problematic for species trying to forage in a really confined area.”
For animals that cannot swim away—crab, and the slow-moving invertebrates of the seafloor—there is no in-between to retreat to. On the Oregon coast, comparable hypoxic events have produced mass die-offs, and crab have been seen migrating inshore en masse, fleeing the low-oxygen water until there is nowhere left to go.
But the glider revealed something else, too—something no one had fully expected.
At the same depths where Goose Trough holds its hypoxic water through the summer, the water on the other side of Hakai Pass, in the deep basin of Fitz Hugh Sound, stays comparatively rich in oxygen. The question, as Evans frames it, is simply: why?
If Goose Trough is the riverbed, Hakai Pass is the gate—and it may be doing something remarkable. The pass is narrow and turbulent; tides rip through its constriction. The leading hypothesis is that this violent mixing works oxygen into the water as it crosses the threshold, and that the newly oxygenated water then sinks to fill the basin beyond. “In early spring, the oxygen content in the deep part of Fitz Hugh Sound is highest, and the water there is also freshest,” Evans explains. “You can't mix fresh, oxygenated water straight down from the surface to 300 meters—not here, anyway. So it has to be getting oxygenated somewhere else. And that somewhere is Hakai Pass.” Work by Martin Williamson, a University of Victoria graduate student working with C-PROOF co-lead Jody Klymak, is pointing in exactly this direction.
The inside waterways, in other words, may be protected—not by distance, not by policy, but by geography. By the shape of the seafloor and the turbulence of the water forced through it.
It is not a guarantee, and Evans is careful with the word. The same intensification building on the shelf—hypoxia more frequent, more severe, more widespread—could eventually overwhelm the mixing that Hakai Pass provides. So could ocean acidification, as human carbon continues to accumulate in the sea. “I'm not sure that protection will be lasting,” he says. “Eventually, the benefit of mixing in Hakai Pass might be swamped.”
To learn how long the gate holds, the scientists need to watch it continuously—and the glider, for all its range, cannot. It passes through Hakai Pass only twice per mission: once on the way out, once six weeks later on the way back. It catches glimpses, not the full story of what happens hour by hour, tide by tide.
After weeks crossing the continental shelf, the ocean glider is recovered by the Hakai oceanography team. Autonomous beneath the surface, the instrument still depends on an experienced field crew to launch it, recover it after weeks at sea, and prepare it for its next mission.
So this coming spring, Hakai will put a different instrument into the pass. A C-PROOF Wirewalker is a profiling platform that rides up and down a moored cable using nothing but the energy of the waves, measuring oxygen, temperature, salinity, and current through the full depth of the water, around the clock, for months at a time. It will resolve the questions the glider can only raise: whether the strong spring tides drive more oxygenation than the weaker neaps, how storms modulate the renewal of oxygen in Fitz Hugh Sound, and how long the pass can go on protecting the waters behind it.
The glider mapped the approach. The Wirewalker will watch the gate.
The ocean keeps its workings hidden. For nearly all the time people have lived and fished along this coast, the water simply moved—upwelling, mixing, carrying its chemistry toward shore—with no way for anyone to watch it happen. Scientists could read the ocean's past in fossil shells and seafloor cores; what they had no way to do was follow the process live, across a whole season, as it unfolded. What has changed, in the last few years, is not the ocean. It is our ability to observe it.
Next summer the river will begin its journey again, as it always has. Now we can follow where it goes.
