Nooksack flood risk to 'increase drastically' by 2080

📷 City of Abbotsford

In June of 2021, Canadian policymakers were not exactly waiting with baited breath for Evan Paul to finish his master’s thesis.

Two years of data analysis, writing, research, and cataclysmic flooding later, Paul’s findings have billion-dollar implications for the future of the Fraser Valley. Paul’s work suggests the Nooksack River will become much more likely to flood in the years to come, and that those floods will become more severe over the coming decades.

Paul’s thesis was overseen by a Western Washington University professor who is the authority on the Nooksack River, and its findings should shape how highways are built, how dikes are constructed, where pump stations are added, how emergency management plans are written, and how vast sums of money are spent. But first, policymakers, politicians, residents, and businesspeople have to actually read the research—and know it exists.

A pending study

As Sumas Prairie sat under feet of water, I called Western Washington University professor Robert Mitchell to learn how climate change was changing the behaviour of the Nooksack and its propensity to flood.

The 2021 flood was hardly unprecedented: the river had spilled its banks many times before, including in 1990, when it flooded a western portion of Sumas Prairie, shut down Highway 1, and sparked the creation of a crossborder task force. (That accomplished next to nothing—you can read more about that here). But scientists have suggested warmer temperatures and more extreme weather would broadly increase the frequency and scale of such events across western North America. And although any single flood can’t be blamed on climate change, there were clear signs that the Nooksack’s propensity to breach its banks would increase as the weather warmed. *

Mitchell said as much when I talked to him in 2021. (You can read the resulting National Newspaper Award-finalist story, The Life and Times of the Nooksack River, here.) But it’s one thing to know the existence of a broad trend, and another to understand how it will impact a specific river and the communities around it. Three years ago, existing research didn’t have a full grasp on the impact of climate change specifically on the Nooksack River. Because previous studies used daily, rather than hourly precipitation data, they weren’t able to accurately model the impact of heavy rain events. Which is a problem when you have governments deciding whether—and when—to spend billions of dollars to prevent future catastrophes.

In November of 2021, with Sumas Prairie still full of water, Paul’s impending thesis seemed likely to have particular value to American and Canadian flood experts as they tried to prevent future disasters.

Now his study is in, and it suggests climate change will have an even larger impact on the Nooksack than was already thought.

You can read the study here.

*While many Nooksack River locals blamed sediment loads, rather than climate change, for increasing flood risk, the reality is that both factors have a significant impact.

The conclusion

Previously, Mitchell had suggested that the magnitude of the Nooksack River’s floods could increase by at least 25% by the middle of the next century.

But Paul’s new study (for which Mitchell served as advisor) suggests planners might have to prepare for even larger floods in the years to come.

Paul found that the downstream flood risk will “increase drastically,” with peak flow magnitudes rising by between 34% and 60% “across all flow durations and return periods.” (A ‘peak flow’ is a high-water event of a certain rarity. Paul measured and examined both annual peak flow data—the highest a river got in any one year—and peak flows associated with floods of various sizes and expected return periods.)

That means that the infrequent-but-big flood events will be bigger and that more-common-but-smaller peak flows will also be significantly more severe. Essentially, the baseline for a “normal” high water events will rise and as that happens, the likelihood and frequency of larger events happening will also rise.

There are two specific and simple ways that human-caused climate change will increase the size and frequency of Nooksack floods.

First, atmospheric rivers are expected to be more frequent and more intense by the end of the century. Those findings come from an array of studies and climatic models, and studies on other Washington watersheds have suggested similar impacts there.

Previous modeling underestimated the impact of atmospheric rivers on the Nooksack because the research relied on daily, rather than hourly, weather data that obscured short-term downpours.

The second reason is even simpler. Progressively warmer temperatures will expose more of Mt. Baker’s slopes to precipitation falling as rain, rather than snow. That will reduce snowpacks, and funnel water much more quickly from the clouds into the Nooksack.

When precipitation falls as snow, its entry into a river like the Nooksack is delayed. So when there is a large rain event near sea level, the impact on the Nooksack’s water level is heavily dependent on the snow level. Precipitation that falls as snow will usually be released later in spring or early summer, rather than in winter, when the Nooksack poses the largest flood threat. Raise temperatures, and far more water immediately ends up in the river.

Snow can also reduce the flood risk of future atmospheric rivers. A pre-existing snowpack can soak up rain from later, warmer events. But as November and December snow becomes less common, and spring kicks in earlier, the snow line is expected to get higher, leaving less of the mountain able to absorb and store water that falls during warm atmospheric river storms.

Paul writes that the average April 15 snow line is projected to rise “from 800 metres in the 1900s to 1,200 metres in the 2080s.” And that will reduce the ability of Baker’s slopes to soak up heavy rain during large storms.

“In the future, we're expecting to see a smaller snowpack and therefore a larger area that is exposed to runoff, and therefore, that water is getting to the river quicker and therefore we see higher peak flows,” Paul told The Current.

Single peak

Paul’s thesis also found the timing of high-water events will shift significantly as temperatures rise.

Historically, the Nooksack has seen two “peak” seasons—one in the winter, and another in the spring as snow on Mt. Baker melts. The months of November to January saw the most frequent peak flows, with December the busiest time of year for high-water events.

Over the coming decades, Paul’s research suggests that will change, with high-water events expected to span a six-month horizon between October and March, with the busiest month becoming November. Meanwhile, water levels between May and September are expected to plunge, with much less snow—or glacial ice—to melt (though those months weren’t a focus of the study).

“By the end of the 21st century, spring peaks are nearly nonexistent in the hydrographs for each of the subbasins and the North Cedarville gauge,” Paul writes. “These changes to the timing and magnitude of streamflow indicate that the entire upper Nooksack basin will shift from a transient rain-snow basin to a rain-dominant basin by the end of the 21st century.”

(The North Cedarville gauge is located just upstream of Everson, Wash., where the Nooksack splits in two during large flood events. You can read more about the importance of the North Cedarville gauge here.)

In layman’s terms, that means that snow will stop being a prominent driver of the Nooksack’s rhythms. And that’s a huge change for a river that drains a mountain with a claim to being the snowiest place on earth.

The Nooksack’s cousin

Paul’s findings closely echo those of an earlier and separate study examining peak flow on the Fraser River. That study by a foursome of BC climate experts found that more rain and less snow would increase peak flow events in November and December while reducing the frequency of melt-related high-water events. (The Current reported on the study in January.)

The two studies each found that as the weather warms, the frequency and severity of winter floods will increase.

But the studies don’t just support the conclusions their researchers reached independently. When taken as a whole, the two studies suggest that a repeat of the calamitous 2021 Sumas Prairie flood could be exponentially more likely over the next 60 years.

The 2021 Sumas flood was particularly damaging because, as the Nooksack was flooding into Sumas Prairie, the Fraser River’s water level remained even higher and prevented the opening of Barrowtown Pump Station’s floodgates. Those floodgates would normally allow water to drain into the Fraser during high water events on the prairie. But with the Fraser even higher than the incoming Nooksack water, the floodgates had to remain closed. That caused the prairie to fill up like a bathtub.

Previously it had been expected that in the event of a Nooksack winter flood, the floodgates could be used to drain water into the Fraser. But 2021 proved that assumption wouldn’t hold if and when the Fraser rose concurrently with the Nooksack.

The two studies suggest that the winter flood threat on both the Fraser and the Nooksack will significantly increase. And that means there will be a dramatically higher chance that both rivers rise at the exact same time, as occurred in 2021.

If you were rolling a pair of (imaginary) 30-sided dice, the 2021 flood might represent snake eyes. To get the same result in 2080, you could roll a pair of ones, a pair of twos, or a combination of the two. By doubling the chances of each event, you quadruple the likelihood that the two events occur simultaneously.

The odds are different for each flood event, but the same statistical principle holds: when the likelihood of two independent variables each rise, the chance that they occur at the same time increases even more.

In this case, the fact that the Nooksack and Fraser floods aren’t independent from one another, only increases the chance that they will occur at the same time. After all, a single atmospheric river can be responsible for high-water events across southern British Columbia and Northern Washington, and there is a greater-than-average chance the Fraser and Nooksack will rise in tandem.

That suggests that a proposal to build a pump station to mechanically move water from the Sumas basin into the Fraser—a facility that may have been relatively redundant 50 years ago—may be increasingly attractive as the climate warms. The 2021 flood demonstrated that a winter atmospheric river could cause the Fraser River to rise to heights usually seen only in the spring. When the water gets that high, pumps—not gravity—are needed to drain Sumas Prairie.

The sediment wildcard

There is little reason to doubt the Nooksack River will have to accommodate more water more often in the coming decades. But the actual size, location, and frequency of floods themselves will depend on a couple factors.

The first wildcard relates to a common gripe about the Nooksack: the sediment on its bottom

Rivers possess incredible—and incredibly inconsistent—energy. With more water comes more energy, and a river can chew through banks or its own river bottom, moving sand, pebbles, rocks and sometimes boulders downstream, until the speed slows and its sediment-carrying energy drops.

The energy causes rivers to change, shift, and move. It can take years. Or days. The Current has written about how the Nooksack once carried its water north to the Fraser and then, possibly in a single day, its main course shifted and turned to the west, running toward the Pacific Ocean.

Often, these processes are less spectacular. But they still influence when and how the river floods. Locals have blamed higher sediment loads in the Nooksack for its increasing flood risk. But the river’s sediment isn’t consistently distributed.

Paul writes that the river channel near Everson “showed a one to three foot increase in average bed elevation from 2006 to 2015.” That reduced the capacity of the channel and has made it more likely to flood.

But south of Everson near the North Cedarville gauge, one local told Paul that during the 2021 flood, the water carved away at least eight feet of sediment.

Across the breadth of the Nooksack, sedimentation levels can vary widely and have a huge impact on the scale of water needed to breach the Nooksack’s banks near Everson and route water north toward Canada. The river’s shifting bed makes it difficult to know just what percentage of the Nooksack’s water, at what water level, ends up flooding north toward Sumas Prairie and Canada.

For that reason, Paul’s research focuses on the amount of water in the Nooksack just south of Everson, Wash. at the North Cedarville water gauge. South of that point, the destination for the river’s water becomes harder to predict.

“I don’t think anybody has really attempted to quantify how much water is going out of the Nooksack into the Sumas, because it’s difficult because the sediment level in the river is constantly changing,” Paul said.

There’s a second wildcard that when it comes to predicting the future: what steps people take to mitigate flood events or change the amount of water going west, rather than east.

More water flowing to Canada means less flooding downstream within the United States. That relationship has led the Americans to avoid work that could substantially restrict flooding that impacts Canada because doing so would increase flooding within Whatcom County.

Paul, probably wisely, leaves political analysis out of his thesis.

But his work suggests that unless political calculations change, the Nooksack’s large floods will get even bigger. And inevitably, that will send more water, more often north towards Canada and Sumas Prairie. The question is: will Canada be ready?