The Soos Creek Trail is a long ribbon of asphalt that follows Big Soos Creek for about six miles until the trail culminates at a small town near Kent in Puget Sound. Like the creek, the trail wends through a variety of habitats — small family farms, residential neighborhoods, parks and playgrounds, miscellaneous scrub — but the creek itself is only occasionally visible through a dense thicket of trees and shrubs, at least for the first couple of miles. In the late summer it is slow and sluggish. You might be forgiven for mistaking it for a long skinny pond.

This changes once the creek’s Big and Little arms converge and the creek widens, its water picking up speed. Then, Soos Creek looks more like what it is: one of King County’s more significant minor water bodies, if you will.

All in all, nearly 60 miles of tributaries feed it, draining about 45,000 acres, or 70 square miles. Less than a mile up from where the creek joins the Green River, near Auburn, the state has run a hatchery since the early 1900s. Every year, biologists send thousands of salmonid smolts on their way to Puget Sound and beyond in the hopes that some of them will come back and spawn; Chinook, chum, coho, pink, and sockeye salmon, as well as winter steelhead and sea-run cutthroat trout all have a history of spawning in the creek. And it is on their behalf — especially the Chinook and coho, which are federally listed species — that several projects have converged here. Scientists hope that these studies will make it easier for managers to decide where to invest critical but limited resources to reduce one of Puget Sound’s biggest environmental problems: stormwater pollution.

Stormwater study

Almost every body of water in the Puget Sound region that passes through or near a city or town has been affected in some way by stormwater runoff and the pollution it carries. The sickening and sometimes lethal effects of stormwater on salmon are increasingly well documented. The problem is caused first by the impervious surfaces—roads, highways, streets, asphalt trails, and so on — that make an urban environment an urban environment. It is made worse by fateful timing.

Heavy fall rains are not only the source of the most potent stormwater, washing a summer’s worth of accumulated pollution into Puget Sound; they are also a signal to all the salmon that have been waiting in Puget Sound to begin their journeys upstream. The fish, in effect, are called to their own undoing.  

But while proven remedies for stormwater pollution exist, including everything from grassy swales along highways that filter out pollutants to more regular street sweeping, these remedies are not cheap. According to one estimate, to restore Puget Sound to a point that it might function in something approximating a natural state could theoretically cost a staggering $400 billion over three decades (admittedly an extreme amount according to the authors of that figure).

In any case, such a vast expenditure seems unlikely. The state's current outlay for its Stormwater Financial Assistance Program, for example, is about $23 million per year. That figure goes up substantially if you take into account additional expenditures for toxic cleanup and other financial incentives, but there continues to be an intense interest in determining the most effective ways to prioritize limited resources to restore polluted water bodies. This is where Soos Creek comes in. “Several years ago, the EPA and the Washington Department of Ecology had an idea to go out and do a pilot stormwater study,” says Teizeen Mohamedali, an environmental engineer at Ecology’s field office in Bellevue. They selected Soos Creek in part because stormwater was a known problem in its basin, and also because data collected over the years showed the watershed’s aquatic health was impaired.

The study’s goal was to go above and beyond the usual technical methods that hinge traditional metrics such as dissolved oxygen, pH, temperature, water flow rates and sediment loads. All of these physical and chemical measures are important to salmon, but are limited in one important respect: They don’t show whether a stream can actually support a healthy freshwater ecosystem’s suite of organisms, known as a stream’s “biological integrity.”  “The biological integrity of these water bodies can often get overlooked in monitoring,” Mohamedali says. “It’s just a lot easier to focus on physical or chemical measurements.”

It starts with bugs

Biological integrity in this case is part of a management-specific term related to life of a somewhat squirmy kind. The Benthic Index of Biotic Integrity, or B-IBI (pronounced B-I-B-I) is based on the community of benthic invertebrates — specifically mud-dwelling bugs — found in a given water body. In essence, a B–IBI  score is a measure of the insect community that occupies a particular point in a stream. “Some bugs are more tolerant of certain types of pollution than others,” says Cleo Neculae, the watershed cleanup lead with the Department of Ecology in Bellevue. “So you know that if you find more of those more tolerant bugs you probably have more pollution, versus the types of bugs you would find in much healthier environments.”

Different insect species also point to different stressors — some respond more to flashy flows, others to particular types of chemicals. Knowing which bugs are there tells you what contaminants to look for, and at Soos Creek, an earlier study identifying stressors to the bug population had shown that flashy flows during stormwater events, sediment loads, and habitat destruction were responsible for low B-IBI scores in the watershed.

What has been missing so far in the traditional framework was the integration of physical or chemical metrics with biological health measures. Relating the two will ultimately allow managers to undertake projects that can address both aquatic health and stormwater. Since 2012, Mohamedali, Neculae, and their colleagues at other government and tribal agencies have worked to simulate flow and sediment transport — part of traditional measures known as Total Maximum Daily Loads or TMDL — in Soos Creek and marry those conditions to B-IBI scores. They would thus figure out how to develop management targets that will increase those scores, and from that the likelihood that a waterbody is actually good for salmon and other species of conservation interest. 

“It’s a unique approach,” Mohamedali says, and one that was in part motivated by the Clean Water Act’s requirements that the state establish water quality standards that encompass not only the physical and chemical integrity of a water body, but its biological integrity. At present the analysis and modeling are ongoing; Mohamedali anticipates the project will be completed within the next couple of years. “We want to improve it all to level of confidence to the point of where we can start using it in scenarios,” she says. “Ultimately it will help us prioritize our action, point us to which parts of the watershed will benefit the most from mitigation efforts.”

Stormwater heatmap

If Mohamedali, Neculae, and their partners are at work in the physical spaces, then Emily Howe and her colleagues occupy a more digital realm. An aquatic ecologist with The Nature Conservancy, Howe has for the past four years been part of a team working to develop a program called the Stormwater Heatmap. The heatmap is a tool that will quantify pollution loads at fine scales to help managers identify pollution hotspots for a suite of contaminants. “Once it’s done, it will let people answer the ‘where and how much’ questions at the heart of effective intervention,” Howe says. All of this comes as the state is now requiring eighty-five jurisdictions to develop and implement stormwater management action plans; previously, only four were.

This may sound fairly straightforward at first — gather information on regional water bodies and pollution inputs, plot it on a map, identify the problem spots — but it requires a tremendous amount of information: land use, landcover, imperviousness, topography, soils, slope, runoff, stormwater monitoring data for a range of contaminants (copper, zinc, nitrates and nitrites, phosphorous, polycyclic aromatic hydrocarbons), rainfall scenarios and climate data, hydrologic modeling. “When we first ran the model, we had something like three hundred billion rows of output,” Howe says. “So that was a lot to sift through.”

With all this in place, the designers are incorporating the last and perhaps most crucial layer: the social-ecological element. “This will let us intersect stormwater conditions with its social and ecological impacts on people and nature,” Howe says. In the end, after all, the heatmap’s greatest power will come from its predictive capacities—it’s ability to show, for example, the coho spawning streams that are most likely to be harmed by a stormwater event, or the swimming beaches or shellfish beds most likely to close. Then, managers can identify the most important places to work.

“It’s taken a lot of work on the part of a lot of people,” Howe says, “but the hope is that all of this will lead to more rapid recovery of marine and freshwater ecosystems throughout Puget Sound.”