Can the Hyporheic Zone be Restored? Designing and Constructing a Bold Experiment
Cutaway view of a plunge-pool structure designed to greatly enhance hyporheic flow. The water is flowing left to right. Zigzag arrows indicate warm water (red) and cooler water (blue).
Growing up in urban Seattle, I didn’t really realize how damaged the creek that I played in was. To me, it was a cool, leafy, watery oasis, where flowing water excited my sense of wonder. The deepened, straightened, rock-lined channel was just the way things were. It wasn’t until years later that I realized that the familiar smell of its water held definite hints of overabundant algae with a touch of raw sewage.
I also could not have imagined that I would one day become a key player in an effort to restore a portion of this very creek, using a bold, experimental design focusing on the hyporheic zone, a term which I don’t even think had been invented back then. I’m talking about Thornton Creek, which has a watershed occupying a big chunk of Northeast Seattle in Washington State, USA. The restoration would happen at two sites: the Kingfisher site (also known as Knickerbocker) in the Victory Heights neighborhood just down the hill from the first house of my uncle and aunt; and the Forks Confluence site, adjacent to the baseball field of my high school, Nathan Hale. What I would like to do here is describe this restoration, focusing mainly on the Kingfisher site, and then talk about the results of this bold experiment.
The South Fork of Thornton Creek originates in one of the most heavily modified landscapes of North Seattle, in the neighborhood, if you want to call it that, where North Seattle Community College is separated from the Northgate Shopping Center by Interstate 5. The creek emerges from a culvert just east of the shopping center, and then immediately enters a woodsy ravine. About a mile and a half (2 kilometers) further downstream, the ravine widens out, providing room for a small neighborhood with a few houses near creek level. This is the Kingfisher site. From here, the creek once again plunges into a steep, artificially rock-lined ravine to cross underneath the four lanes of Lake City Way and then head for the wide bowl -shaped Meadowbrook valley where it flows next to the buildings of Nathan Hale high school. Just downstream from here, is the confluence with the North Fork, and the second of the restoration sites.
Originally, the creek at the Kingfisher site was dredged, straightened, lined with boulders, and pushed against the hillside to make room for houses that were built so close to it that they were frequently being flooded during winter rains. The hillside against which the creek was forced to flow was a natural landslide area, and on several occasions, landslides forced the adjacent house off its foundations.
Because the creek had been straightened and rock lined, its water was forced to flow in an artificially narrow, deepened channel. During rain storms, this water would develop incredible velocities for a small creek; I once measured 8 feet per second (5.5 miles per hour, or 8.8 kilometers per hour) in the channel which was only 43 inches (1.1 meters) deep and 7 feet (2.1 meters) wide. The flow was so swift that it was bending the steel rod of the velocity meter into an arc. With frequent exposure to this kind of force, the streambed had become a thin veneer, about 6 inches (15 centimeters) thick, of sand and gravel, armored on top with well-cemented baseball-sized rocks. Beneath this thin veneer of streambed was the dense, compacted “ice-age” (Pleistocene) clay that makes up the core of all the hills on which Seattle is built.
The veneer of streambed was so embedded with fine sand that water could not really percolate through it, and the dense clay below was nearly a total barrier to any water flow. So there was essentially no hyporheic zone at Kingfisher. The useful ecological function of the hyporheic zone was all but lost. Thornton creek was like patient waiting for a liver transplant.
Fast-forward to the year 2014. By now the houses that had impinged on Thornton Creek were long gone, and the construction work began. A new floodplain was constructed, widening and deepening the valley make room for it, and make room for the creek to swing northward into a broad meander away from the hillside. The new stream channel was built much wider than the old one, and was lined with a thick layer of clean gravel. Dozens of logs, many with their roots still intact, were brought in to build specially designed structures that force the water to turn, to constrict, and to plunge over abrupt steps into deep pools.
These structures help to form the framework of the new stream channel, stabilizing it during the coming winter storms. They also create the kind of complex habitat that fish prefer, with a diverse mixture of swift and not so swift water, shallow and deep areas in close proximity, turbulent curtains of splashing bubbles to hide beneath, and overhanging protecting ledges. And finally, some of these structures were specifically designed to create enhanced hyporheic flow, and a hyporheic zone much greater in volume and depth than anything Thornton Creek has had in recent memory. Bigger, in fact, than the hyporheic zone of a natural stream of this size.
Plunge-pool log structure, in November, 2014. The hyporheic zone beneath this structure extends down 8 feet (2.1 meters) beneath the streambed.
The gravel making up the new stream channel was a minimum of 3 feet (90 centimeters) thick, which is six times greater than the previously existing streambed. Moreover, the new gravel was clean and permeable to water, containing very little of the fine sand and silt which was clogging the existing streambed and preventing hyporheic flow.
The log structures were placed within this thick gravel streambed, such that there was always gravel underneath the logs. The difference in elevation from upstream to downstream creates a pressure, causing water to enter the streambed on the upstream side, flow beneath the logs, and then re-emerge in the pool downstream. There were three main kinds of log structures built to enhance the hyporheic flow: pocket-water logs, plunge pool structures and hyporheic logs.
The pocket-water log structures are made by taking two logs, and embedding them into the streambed so that they cross each other, and force the water plunge over first one, and in the other. Any time water is forced to flow over a log, it will change direction, tending to plunge over the log in a direction perpendicular to it. And by plunging over the log, it will tend to scour a small pool, a “pocket pool,“ on the downstream edge. And because there is a small drop in elevation from the upstream to downstream side, there is pressure generated to force the water into the streambed, forcing it to flow underneath the log. This is hyporheic flow! There are three pocket-water log structures at the Kingfisher site.
Pocket-water log structure. Hyporheic zone beneath this structure extends down 3 feet (90 centimeters).
The plunge-pool structures were the most elaborate of the three types of structures installed, and the most effective at increasing hyporheic flow and volume. Construction of each of the six plunge-pool structures at the Kingfisher site began by digging a hole as much as 8 feet (2.4 meters) deep. The hole was partially backfilled with gravel, and then two logs, one on top of the other, spanning the whole channel, are put in to form the sill over which the water will plunge. For a distance extending from these sill logs to about 4 feet (1.2 meters) upstream, a layer of compacted soil was placed on top of the gravel and then more gravel placed on top of that, to a level flush with the sill logs. This compacted soil layer forms a barrier to water movement through the streambed, causing water that percolates into the streambed to be forced to circulate very deeply, all the way to the bottom of the 8 foot hole, before moving upward and reentering the surface water at the downstream end of the plunge pool. The temperature down there at the bottom of the hole is only about 52 degrees F (11 degrees C), so when the water reemerges, it is cooler than the stream water, which would typically exceed 70 degree F (61 degrees C) on a summer afternoon.
Hyporheic logs are the simplest of the three structure types. These are nothing more than a log, perpendicular to the channel, placed at the bottom of the excavation before the streambed gravel was poured in. Although these logs are not visible on the surface, they are performing an important function by forcing the hyporheic water percolating through the streambed to flow right or left, up or down, to go around the log. This increases the time that the hyporheic water spends moving through the streambed before it re-emerges, which is called its “residence time.” The longer the residence time, the greater the change in temperature and chemical composition will happen to the hyporheic water. So these logs amplify the cooling and purification that is going on.
Hyporheic water emerging from exposed streambed gravel downstream of a pocket-log structure, March, 2015. Water is emerging at lower left, and flowing towards the standing person. Burlap fabric was used to temporarily shield streambanks from erosion until vegetation regrowth.
At the Kingfisher site, there is a lot of groundwater emerging along the base of the valley, along the west and south sides. Because this water has been flowing underground, it is naturally cool. Most of this water flows into low spots built into the floodplain, forming wetlands. Wetlands are very important habitat for many things, including birds and amphibians. But at one place along the Southwest side of the valley, a channel was built and filled with gravel to capture some of this emerging groundwater and funnel it directly into the stream channel, just like a French drain. This “subsurface drain structure” forms another cool spot in the otherwise warm stream channel during the summer.
All of this construction work happened in the summer of 2014. There are a lot of details, and a lot of people, that I have left out of the story in order to keep it short. Like the long, sometimes tedious process that I led while employed by the U.S. Fish and Wildlife Service, collecting measurements of the stream channel, streambed, flowing water, and the groundwater in the years before the construction began. Or the collaboration between myself and Civil Engineer Michael Hrachovec, “Rocky,” of Natural Systems Design, to first fine-tune conceptual ideas for hyporheic zone restoration, and then for Rocky to turn these concepts into practical design drawings. Or the adept piloting of this project through the shoals and reefs of the City of Seattle bureaucracy by Aquatic Ecologist Katherine Lynch, of Seattle Public Utilities, tenaciously keeping it moving forward and keeping it funded, through numerous crises and changes in management.
The construction itself was not easy either. Unseasonably heavy rains in late July, and small flash floods initiated by work on a temporary reservoir upstream caused the site to be flooded several times, creating delays and more work. Construction of a pedestrian footbridge, by a different team of people, created time conflicts and further delays. Two concerns were also on our minds. First, what if the hyporheic zone was so successful that all of the water went into it, leaving the stream channel dry, and the fish stranded? And conversely, what if the first winter storms after the project was done caused the new streambed to get plugged up again with silt and fine sand, making all this work for nothing?
Finally, in October 2014, the construction was finished, the pumps turned off, and the creek allowed to flow through the new stream channel on its own. To our great relief, there was continuous surface flow throughout the site. The concern that the water would disappear turned out to be unfounded. We all walked around, slowly, looking, taking pictures, talking very little, each person allowing the magnitude of our success to settle in. We had done it! That morning, at least, all was well in the world.
And yet, my work was now just beginning. I was to lead a key portion of the scientific study to learn how the new hyporheic zone was working, and whether it would persist. It is to this three-year task that I will turn next.
References:
Paul D. Bakke, Michael Hrachovec and Katherine D. Lynch, 2020. Hyporheic Process Restoration: Design and Performance of an Engineered Streambed. Water 2020, 12, 425; doi:10.3390/w12020425, available at https://www.mdpi.com/2073-4441/12/2/425