The Life and Times of the Nooksack River

How the Nooksack River changed course in a day, and how it continues to shape the region—with help from humans.

This is the first story in a two-part series on how the Nooksack River has evolved, and how humans have shaped where and how it floods. The second part can be read here.

In the late 1990s, a dozen or so American politicians and community members made a fateful decision that would change the lives of Abbotsford and Yarrow residents two decades later.

They decided to do nothing.

The Americans had convened to figure out what to do with the Nooksack River, a waterway that was constantly making trouble for the denizens of northern Washington—and occasionally the Fraser Valley.

The Nooksack traces a broad arc from the foothills of Mt. Baker to the Pacific Ocean north of Bellingham, never leaving the United States. Like many rivers, it has a tendency to flood. But the Nooksack’s floods are incredibly complex, both geographically and politically, thanks to its natural and human history. Within that history includes one particularly important event: a single day, probably less than 1,000 years ago, when the entire Nooksack shifted course, abandoning its connection to the Fraser River for a more direct route to the Pacific Ocean. It is the legacy of that event, combined with ongoing hydrographic change and human activity, that has left the Nooksack so delicately balanced and internationally destructive.

Today, the river remains so close to the Fraser River’s basin that when the river rises too high, millions of gallons of water are detoured away from western Whatcom County and north toward Canada, Abbotsford, and the residents of Sumas Prairie.

This is partly due to nature and history, but also by design. The northern overflow route reduces the scale of flooding of the western Whatcom County communities downstream, and so officials have repeatedly declined to build a dike near Everson, Wash., that could help prevent or reduce flood damage in Canada and communities like Sumas, Wash.

In 1999, as they prepared a “comprehensive” flood plan for the Nooksack, officials voted to not even study whether to stop the flooding toward Canada, documents show. Political “considerations” played a key role in that decision, those same documents reveal.

Many of the discussions that influenced that decision featured questions about when and how humans should intervene in the natural course of a river. Today, officials are asking many of the same questions. But those thought exercises and conversations are again influenced by politics and frequently overlook an even-more fundamental question: just what, exactly, is the “natural state” of the Nooksack River?

This story, research on which began last spring, before The Current was even launched, is our best shot at answering that. A second story will follow next week on the complex politics behind how the Nooksack’s vital overflow is managed.

The evolution of a river

📷 Tyler Olsen

Geological time moves slowly. Until it doesn’t.

When we think of how nature changes itself over time, we usually think of long, slow geologic time scales: the rise and fall of mountains, the crawl of glaciers, the invisible erosion that transforms the earth over thousands of years.

Rivers move too. Each year, they chew into their banks, shifting course this way and that. As its current accelerates and slows, a river picks up and deposits rocks and sediment in another location. That sediment, then, redirects the current, which in turn affects erosion and sediment deposition. And so on. Over years, decades, centuries and millennia, a river wanders. Its masters are gravity and the other laws of physics, but its course and future have other influences too: wetlands; forests; plate tectonics; logjams; the weather.

Sometimes a river changes slowly or in ways that aren’t easily visible to the human eye: the Fraser River’s deepest channel, known as the thalweg, is known to have shifted recently between Mission and Abbotsford. But sometimes, the changes take place extremely quickly: satellite images show the Chilliwack River’s channel east of the Vedder Bridge (below) looks nothing like it did just 20 years ago.

In just 35 years, the Chilliwack River has changed course repeatedly as it approaches Chilliwack. 📷 Google/Tyler Olsen

Humans, inevitably, change rivers too by building dikes, dredging sediment, constructing bridges, and laying riprap. Indeed, it’s impossible to interact with a river and not change it a little.

Since European settlement, North Americans historically sought to constrain rivers, and slow their tendency to change. Maybe for that reason, we underestimate how quickly a river can still alter its geography. Last year’s atmospheric rivers, for instance, saw the Coquihalla, Coldwater, and Nicola rivers chart new courses straight through major highways, backyards, and entire communities.

Nearly every river carries that potential over the course of time. But almost none have changed course so dramatically, and so recently, as the Nooksack.

Today, the river charts a broad semi-circle before emptying north of Bellingham. But there was a time when the Nooksack went a different way. It flowed north toward Canada, Sumas Lake, and the Fraser River, until one day (and, yes, it may have happened that quickly) it didn’t. That moment has significant implications to this day.

The River That Was

Our story today is about the Nooksack River. Which means it is also about the Sumas River.

The Nooksack River is a child of Mt. Baker. The river originates on the slopes and within the valleys of the great volcano. It flows down those valleys in three separate “forks,” gaining energy and water. The three forks converge near the western edge of Baker’s foothills, merging together in a short, two-kilometre stretch of river. Just to the west of that confluence, Baker spits the river—full of silt and sediment from the mountain’s volcanic slopes—onto a large, flat plain. From there, the Nooksack arcs across northern Washington before dumping its water into the Pacific Ocean, just north of Bellingham. No other Washington River deposits so much sediment into the ocean.

Today’s Sumas River, on the other hand, isn’t much of a river. For much of its course, the Sumas is more of a glorified (or not-all-that-glorified) drainage canal. It dribbles north from the west-facing hillsides near Everson and weaves its way across the Canadian border before flowing, ever so slowly, northeast toward the Fraser River. Sometimes it doesn’t flow at all. And only the existence of floodgates at Barrowtown Pump Station stops the Sumas from flowing backwards when the Fraser River rises above the level of its meek tributary.

The Sumas River's headwaters begin just a few hundred yards from the Nooksack River.

Because of last year’s flood, people across the continent are now aware that the river once ran into Sumas Lake, before that waterbody was drained in 1921 and the valley’s watercourses dramatically re-arranged. But even before that, Sumas River would have been a lazy creek, its water largely sourced from groundwater and a short stretch of mountainside just below the 49th parallel.

It wasn’t always that way.

The Sumas River and the Nooksack were once the same river. And not just millennia ago.

Scientists have long believed the Nooksack once flowed north. But throughout the 20th Century, most research and writing on the Nooksack was premised on the theory that the change in course—known as an “avulsion” in scientific terms—occurred thousands of years ago.

Increasingly, however, those who study the river closest believe the change took place much more recently—potentially in the last 1,000 years and maybe as recent as three centuries ago.

Paul Pittman, a hydrologist who has studied the upper Nooksack for two decades, suggests that the Nooksack changed course recently—perhaps while Europeans were building cities on the continent’s east coast.

The evidence is perhaps best seen from space, he told The Current last spring.

The evidence from above

From a satellite image, you can see the Sumas River meander out of a narrow prairie squished between the Nooksack River and the foothills of Mt. Baker. Here, just southeast of Everson, the Sumas begins mostly as a series of drainage canals only a couple hundred yards from the banks of the much-larger Nooksack.

Both the Sumas and the Nooksack flow north from this cramped area, but near Everson, where the Nooksack turns left, the Sumas, still just a few hundred metres away, continues toward the Canadian border. It swerves this way and that, tracing large arcs.

Pittman looked at the broad C-shapes of those arcs, especially near Everson. And he looked at the piddly Sumas River, and they didn’t add up. The Sumas didn’t carve those meanders. The Nooksack did.

South of Everson, the Nooksack and Sumas River inhabit similar channel forms, but only one is a sizable river of any sort. 🗺 Bing/Tyler Olsen

Over the phone, Pittman spoke about a stretch of the Sumas, where the river has cut shortcuts between what were once broad meanders. Such shortcuts render old river curves redundant, often turning them into “oxbows,” wetland areas that, like a ghost town bypassed by a large highway, are a marker of how the area used to look.

“If you zoom into those oxbows, you see the tiny little Sumas River is occupying them, and it’s barely even a ditch at that point,” Pittman told The Current. “It didn’t make those channel forms. Those kinds of forms were made by a stream of a similar size to the Nooksack.”

A river might not be beholden to a nation’s laws, but it always must follow one set of rules: those of physics.

“River channels have a really distinct geometry and they always follow that law,” Pittman said. “The Sumas River can’t make an oxbow that size. The geometry and physics don’t work that way. So those oxbows were created by a river that was coincidentally the same channel width as the Nooksack River.”

The channels continue all the way into Canada.

These river relics aren’t just evidence that the Nooksack once flowed north. They are evidence that it flowed toward Canada over a broad enough time horizon to create a shallow valley. The quantity of the arcs is an indication of a river that had thousands of years to wander across that valley as it cut new channels and old ones fell into disuse. The converse can be seen in its sister river: the area surrounding the Nooksack River west of Everson has a distinct lack of oxbows and abandoned riverbeds. Its valley is also indistinct. Those are signs that it is a relative newcomer to the area.

The combined evidence, Pittman said, suggests that the Nooksack spent far more time flowing north than flowing west since the glaciers receded.

These maps, first published 19 years ago by in a presentation by Paul Pittman and Michael Maudlin, show how the region's watercourses may have looked before and after an avulsion of the Nooksack River. , show two

The case grows

Over the last two decades, more evidence has been collected in support of Pittman’s hypothesis.

In 2004, a Western Washington University student named Rich Hutchings published a paper that placed Pittman and company’s arguments in the context of what was known about human settlement in the region. Hutchings was particularly interested in archeological remains found close to the town of Ferndale, near what is now the Nooksack’s delta.

Large piles of processed marine shellfish, dated 4,000 to 5,000 years ago, had been found there, about 10km inland from the Pacific Ocean. The shell processing site, known as a “midden” would naturally have been located near the sea and close to where they had been originally harvested.

Garland Grabert, the archaeologist who found the site, had the job of trying to explain how so many shells were found so far from the ocean. He speculated that changing sea levels, along with terracing factors, could have eventually left the sites so far from the Pacific.

But using Pittman’s research, Hutchings suggested the midden’s location would make sense if it preceded the Nooksack River’s change in direction.

Before the Nooksack changed course, another shorter and tamer river would have met the ocean north of what is now Bellingham. Because that river was smaller, slower and didn’t originate in Mt. Baker’s valleys, it would have carried relatively little sediment and, compared to the river that would replace it, would have had a minor influence on the surrounding land.

Once the Nooksack turned west, it would have begun dropping huge amounts of sediment as it entered the Pacific and slowed. That would have created deltas and new land, moving the coastline west to its present location and away from the established midden area. That process wouldn’t have occurred overnight, but the sediment-heavy Nooksack would have created delta and new land far quicker than its predecessor.

The pattern would not be unique in the region, Hutchings pointed out. More than 100 years ago, Charles Hill-Tout noted the discovery of the Marpole midden site in southern Vancouver. The site (which is now a National Historic Site) is full of clams and mussels, but is located a substantial distance from the nearest clam beds. Other midden sites had been found by that time even further up the Fraser.

“I found it difficult to believe …these great piles had been laboriously brought up against the stream in canoes,’” Hill-Tout wrote. “It was too contrary to the genius of the people to suppose this.” Instead, Hill-Tout came to the same conclusion as Hutchings: as the meeting point between river and ocean moved, so too did the shell beds themselves and the people who created them.

There is other circumstantial evidence of connections, including closer linguistic similarities between Stó:lō people and the Nooksack First Nation—which is based in Mt. Baker’s foothills—than between the Nooksack and the Lummi First Nation, near the Nooksack’s delta.

“Rivers were like the I-5 of the day… The fact that the Lummi and Nooksack don’t have the same language and yet today they share a road is kind of odd,” Pittman said.

For Hutchings, the evidence suggests the mouth of the Nooksack only started moving sometime within the last 5,000 years.

But Pittman thinks it occurred much more recently than that.

“When we looked at the delta growth rate, our best guess is between 500 and 1,000 years ago,” he told The Current last year. That is a huge change from previous academic research, which had suggested that the Nooksack delta began forming more than 9,000 years ago. Pittman thinks it could have occurred even more recently, but whenever exactly the avulsion occurred, the idea that the Nooksack shift is relatively new is now being accepted by local governments and officials as they put together plans for how to manage the river’s unique flood risk.

Whether or not they fully grasp its implications, though, is another matter. Because the speed at which the Nooksack River changed directions speaks to the same issues and challenges underlying current discussions about its future.

One day that changed a region

It could have (mostly) happened in a single day.

For thousands of years, the Nooksack would have been a sizable tributary to an unconstrained, and unpenned Fraser River.

Exactly what it looked like at that time is hard to say.

The river systems of the vast Fraser Valley floodplain would have been constantly evolving and wandering, their paths and routes to the great river continuously in flux. Sumas Lake would also have changed, not just over decades, but seasonally too, rising and falling depending on the weather. Some of those changes would be natural and progressive. Others would be the result of flooding, when the river rose particularly high. The rich soil of the valley bottom is the result of that history of flooding and the existence of Sumas Lake.

Even with the context of this shifting landscape, the Nooksack’s course-change would have been epic and dramatic.

For thousands of years, as the Nooksack left Baker’s foothills, entered the flood plain and headed north, it meandered back and forth, gradually making its way downhill toward the Fraser and Sumas Lake. Unencumbered by modern-day human river “management,” the Nooksack would have cut new channels for itself. (One of the first presentations on the avulsion by Pittman, which you can view here, suggests a pair of parallel rivers over a broad plain may have fed the Sumas Lake.)

But because the area is so flat, and because the river was so close to the divide between two different drainage basins, the proto-Nooksack’s existence was unstable. And at some point, Pittman speculates, one of those channels sprung a leak to the west. That channel split; a tiny side channel entered an old glacial trough and flowed west to the Pacific Ocean. For an unknown amount of time, water from the Nooksack would have flowed both north into Sumas Lake and west into the Pacific Ocean.

A map by Pittman/Maudlin, adapted from a previous study, shows the divergence of the Nooksack River into two separate valleys with similar sediment and rock formations.

But rivers change. Over time, even as much of the river continued to go north, that westward opening migrated upriver. The opening kept moving until it arrived at a spot where much more of the Nooksack’s waters were able to enter this deeper trough. As the energy of this channel grew, it would have dug itself a larger channel, allowing even more water to head west. That water would have turned a smaller creek into a raging river. Anyone located further north, near Sumas Lake, would have seen the river suddenly shrink.

“I think it happened in a day. One day, it picked that other route and off it went,” Pittman said.

How long that process took, and exactly when it happened, is unclear, though Pittman says it wouldn’t be hard to find out. One would just need to radiocarbon date material in the now-mostly-dry oxbows. (And the funding for the exercise.)

Whatever the precise dating, over the last millennium, the Nooksack has worked on embedding itself in its current course. Humans, meanwhile, managed the river, building up its banks and occasionally removing sediment to more consistently channel the water toward the Pacific Ocean.

But the legacy of the avulsion and the circumstances that created it remain: although the river “picked” a side, it has yet to do so in a way that has created a valley large enough to carry all of its water during flood season.

And the arrival of settlers intent on taming rivers dramatically altered the Nooksack’s fate and ability to determine its own future. By building houses, roads, and dikes, humans have constrained the river’s valley and floodplain and reduced its ability to wander and carve new channels. It still flows, though, and it still deposits sediment, and so it still changes. But unable to grow and move laterally, it does so vertically. Now, when it drops sediment and gravel, its waters have less room to chart a new course around that material. And so, when the material isn’t removed, the riverbed rises, as do water levels—and the frequency of flooding.

(The sediment discussion is just one of the many political and philosophical debates surrounding the Nooksack. Although sediment removal may decrease the chances of moderate flooding, many worry about the impact on fish. They say it also doesn’t substantially reduce the risk of major floods.)

A bridge at Everson crosses the Nooksack River. The construction of bridges, roads, and dikes near the river has influenced how the river behaves and when it floods. 📷 Tyler Olsen

Enter human settlers

Given enough time, and a lack of deliberate human interventions, it’s possible the Nooksack would have developed enough of a valley and downstream flood storage to not spill its banks to the north.

But not necessarily.

Because it’s also possible for the river to revert to its old path and fully return to its old northward course.

Officials have occasionally warned that another avulsion is possible. The effect of sending all the Nooksack’s water—not just its overflow discharge—north toward Canada would be devastating. A previous report suggested there is a very slim but real chance that the Nooksack will change course once again and choose to go north, rather than west. Pittman says such a change is unlikely given the state of its current riverbed. The chance of another avulsion was previously estimated as having about a 0.2% annual chance.

But unlikely things happen. That is about the same chance of a major Fraser River flood, and the number also doesn’t account for climate change, or ongoing changes in the banks of the river. On the other hand, the odds also don’t factor in the likelihood that humans would intervene before an avulsion occurs. (In the ’90s, emergency steps were taken to reduce the avulsion risk.)

Another avulsion in our lifetime is unlikely. But it is guaranteed that the Nooksack will continue to change its riverbed and the surrounding landscape.

The Nooksack River flows past Everson this spring. 📷 Tyler Olsen


Which brings us to the present, and the future.

At the moment, the Nooksack’s banks have trouble containing all its water when rain pours down on Mt. Baker, especially near Everson. Officials now refer to the Nooksack’s tendency to split into two at Everson during peak flood levels as the river’s “flow split.” This split is the legacy of that previous avulsion, and the cause of so much heartbreak on Sumas Prairie last year.

Stopping, or significantly reducing, flooding toward Canada would not be particularly difficult or costly. One reason the river floods at Everson is because, unlike most other parts of the Nooksack near communities, there is no dike stopping water from escaping the river’s eastern bank near the town of Everson. That gap in flood defences is relatively short, but deliberate: the result of a calculated decision to not plug it and send more water downstream.

The consequences are real: in addition to the impacts on homes and businesses, downstream flooding also has the potential to close Interstate 5, shutting off the main travel corridor between the Western United States and British Columbia. And a 2004 study suggested that cutting off the overflow to Canada would increase downstream flows by around 8% during floods.

But the Americans have been reluctant to actually study whether sealing the overflow corridor off would still be worth the downstream consequences. The prime decision that has been presented to politicians south of the border is whether to reduce (but not eliminate) the overflow or to “maintain” it.

In documents and recent meetings, the split is frequently referred to as the natural state of the river: the result of gravity and natural forces.

But the flow split, and just how much water floods in which direction, is anything but constant. Indeed, it has never been so. It’s also not entirely natural. At this moment, scientists believe the river’s propensity to exceed its banks is increasing because of erosion, sediment deposition, and climate change (we wrote about how warmer temperatures increase the risk of the Nooksack flooding because precipitation falls as rain rather than snow)—two factors that are also both influenced by humans. The erection of bridges and addition of dikes has also altered the Nooksack’s behaviour and, in some cases, increased the amount of water flowing north toward Canada.

Twenty years ago, officials wrestled with these complex questions and delivered a plan that would prove to be insufficient to deal with the threat posed by the Nooksack. But in crafting that plan, a brutal truth was disclosed. The authors of that plan wrote that no amount of expert planning can answer what to do about the river. The answers of what to do about the Nooksack, they wrote, would depend as much on politics as any technical analysis.

That was the case 20 years ago. It may still be true.

In the second part, we’ll turn from the history of the river to look at how local and international politics could conspire to derail any plans to stop the Nooksack’s flood danger to Canada.

Want to read more of our unparalleled in-depth local coverage of the 2021 landslides and floods? Find all our previous stories here.


or to participate.