Paul Detjens is driving us from his Martinez office to a
restoration site near the mouth of Walnut Creek on Suisun Bay, a project he
spearheads as an engineer for the Contra Costa County Flood Control District.
These lower reaches of the creek — straightened, widened, and leveed by the
U.S. Army Corps of Engineers — have been a sluggish, silt-filled problem for
more than half a century. Detjens has worked to find a solution for the last 17
Now that the district has taken the unusual approach of
parting ways with the Corps in favor of local control, a fix is finally in
sight. Goals include protecting people from floods, restoring habitat, reconnecting
the creek with its historical floodplains, and offering public access so people
can enjoy the wonders of newly restored marshes. “We want to work with
natural processes, not against them,” Detjens says.
The closer we get to our destination, however, the more industrial the landscape becomes, and the more improbable it seems that we’re in the right place to realize this vision. We’re heading east from I-680, just before the bridge to Benicia, on Waterfront Road. That sounds picturesque. But first we pass Copart’s vast parking lot, crammed with thousands of junker cars, and next comes the immense bulge of the Acme Landfill. Though the road parallels Suisun Bay, you can’t see that far. Instead drivers have a close-up view of the Union Pacific Railroad tracks and an array of elevated petroleum pipelines. The massive tanks and towers of the Marathon Martinez Refinery lie straight ahead.
Then Detjens takes a right on a narrow levee road and it all
begins to make sense. Here, Walnut Creek makes its way down a slender strip of
land — squeezed between the landfill and the refinery — as it traverses the
final few miles to the bay. The project Detjens heads will restore this lowest
stretch. Confined by levees and hemmed in by industry, the creek still holds
the tremendous power of water and thus the potential to reshape land according
to the laws of nature.
This winter, the last parcel needed for the restoration fell
into place. “I’ve been here since the start in 2003 and have pretty much
worked on it since then,” Detjens says. “This is an emotional,
exciting time to finally be so close to implementation.”
Bounded by the 1,400-foot Briones Hills to the west and
3,800-foot Mount Diablo to the east, Walnut Creek drains Contra Costa County’s
largest watershed at nearly 150 square miles. The creek’s woes began in the
early 1800s, when Spanish ranchers grazed cattle in the watershed’s plentiful
grasslands. They also introduced shallow-rooted non-native grasses. “The
land became unstable and erosive,” Detjens explains. The post-World War II
development boom made matters even worse. The watershed’s population jumped
15-fold from 1940 to the mid-1960s, and each new house, business, and road
added to the creek’s troubles.
“When you pave over a watershed, you disrupt the
natural function where water soaks into the ground,” Detjens says. With
nowhere else to go, stormwater rushed into the creek. “The banks were
eroding and falling in, and there were lots of floods,” he adds. The town
of Walnut Creek was inundated repeatedly, suffering four major floods in the
Back then, people were at the mercy of floods. “Nothing
was regulated the way it is now,” Detjens says. Citizens demanded action,
resulting in the formation of the flood control district where he works. The
district then sought outside help for taming the waters of Walnut Creek.
“We couldn’t handle it on our own, it was too expensive,” he says.
“Like many others, we turned to the federal government.” That ended
up causing yet another problem for the creek.
In the early 1960s, the district teamed up with the U.S.
Army Corps of Engineers, which designed, built, and largely paid for a
flood-control project extending 22 miles upstream from the creek’s mouth on
Suisun Bay. The district’s responsibilities were to provide the land, and to
own and to maintain the project. The latter proved to be impossible for the
lowest four miles of Walnut Creek, near the mouth.
Here the land flattens, water slows, and silt settles. And
it turned out that there was a lot more sediment coming off the watershed than
the Corps had calculated. “They were off by a factor of five and the project
immediately filled up,” Detjens says. “It never worked right, it was
completely flawed – it just turned into a big sediment trap.”
That put the district in a no-win situation. Bulrushes,
cattails, and other water-loving plants soon sprouted in the trapped sediment,
creating habitat for protected species like the salt marsh harvest mouse,
Ridgway’s rail, and the California black rail. The district then faced
conflicting directives. “One branch of the Corps was saying, ‘You must
dredge out the sediment,’ and the other branch was saying, ‘No way will we give
you a permit for that,'” Detjens says. “It was untenable — it was
like the two-headed monster Muppet in Sesame Street that couldn’t agree
with itself.” Wildlife and water-quality permitting agencies concurred
that dredging was not an option.
Being out of compliance with the maintenance agreement had
major repercussions for the district. To be eligible for disaster relief from
the Corps, flood-control projects must pass annual inspections and the impasse
over the lower four miles put the entire 22-mile Walnut Creek system in danger
At first, Detjens worked with the Corps to find a solution.
In 2004 the district and the Corps embarked on a joint study to reengineer
lower Walnut Creek. “Everyone was like, ‘We’re going to fix this,'”
he says. But a decade and $3.6 million later, the study was only on step three
of the Corps’ ten-step planning process. In addition, the Corps had run out of
planning money so the district would have to shoulder that entire cost.
“We had already put in our half — $1.8 million — for the study,”
Detjens says. “We really didn’t have an appetite for more Corps
So the district took matters into its own hands. “We
ended up doing something pretty unknown,” Detjens says. The solution
hinged on the fact that lower Walnut Creek is fundamentally distinct from the
rest of the project. “The lower four miles are tidal, so they have
different habitat and operate differently from the upper 18 miles,” he
says. Importantly, sediment accumulation in the lowest part of the creek does
not cause floods in the upper part, where the cities are. The district decided
to take back local control of the lower reaches of Walnut Creek while retaining
its partnership with the Corps for the upper reaches.
Called selective deauthorization, the process requires an
Act of Congress. Representative Mike Thompson sponsored the legislation and in
2014 President Obama signed the deauthorization of lower Walnut Creek into law.
The upper 18 miles comply with the maintenance agreement and so remain eligible
for disaster relief, while the lower four are under local control. “It
keeps the Corps’ management intact where it works and gets them out of the way
where it doesn’t,” Detjens says.
By the time selective deauthorization of lower Walnut Creek
went through, Detjens had already laid the groundwork for moving forward
independently. “We had gone through a community visioning process and had
come up with a more compelling vision of what we wanted,” he says, stressing
that the plan reflects local expertise and values. His outreach is ongoing and
includes tours of the restoration site, community meetings, and a stakeholder
advisory group. The local plan offers far more than the Corps’ proposal —
including habitat restoration and public access in addition to flood control —
and will cost far less at about $19 million instead of $50 million.
A Vision for Reconnecting
Detjens stands on the levee the Corps built along lower
Walnut Creek, just off Waterfront Road. A constant low rumble from the Marathon
refinery fills the air, but there are also high notes from birds and that
fresh, clean smell after a rain. In the distance, the peaks of Mount Diablo
disappear into low, gray clouds. Detjens points out the creek side of the levee,
which is crowded with tall marsh plants rippling in wind. You can’t even see
the water. On the other side of the levee lies a remnant of the creek’s former
floodplain. Starved of sediment, the peaty soil has subsided and is visibly
lower than the marsh that thrives between the creek levees.
The land around lower Walnut Creek is so altered that it
offers few clues of what was once here. To find out, historical ecologist Sean
Baumgarten and colleagues at the San Francisco Estuary Institute sifted through
archival records like maps, photos, and written documents. “Drawing on
multiple, independent sources paints a relatively complete picture of the
pre-European landscape,” Baumgarten says.
Baumgarten found that lower Walnut Creek originally
meandered through extensive wetlands, including a broad swath of tidal marsh
between present-day Highway 4 and Suisun Bay. Much of that is now gone. Also
missing is the connection between the creek and the lands around it.
“Understanding what’s been lost helps identify key ecological functions
that you want to restore,” he says.
Next came setting priorities for restoration. A multiagency
team called Flood Control 2.0 brainstormed where and how to bring back
ecological function given the limitations of today. “Flood Control 2.0 is
more than a report on a shelf,” Detjens says. “Their work was
foundational to our project planning.”
The Corps’ levees disconnected the creek from the marsh,
diminishing the tides that scour sediment out of the channel. Moreover,
disconnecting the creek from its floodplains concentrates stormwater in the
channel, resulting in flooding. To undo this harm, the team recommended setting
levees back as well as breaching them in strategic spots. Reconnecting lower
Walnut Creek with its remaining floodplains will redirect stormwater to places
where it can spread out and percolate into the land. Reconnecting the creek
with tidal wetlands will let sediment feed new marshes and enhance their
resilience to sea-level rise.
Habitat for All
A surprising amount of tidal marsh can be brought back. Even
the narrow stretch along the landfill offers room to work with. The creek
channel is more than 450 feet across here, and the floodplain remnants outside
it are wide enough to set the levee back another 350 feet. Altogether, freeing
this part of Walnut Creek to meander and overflow will restore about 55 acres
of tidal marsh.
Far more tidal marsh will be restored just before the
creek’s mouth on Suisun Bay. Here the land opens up, so while the creek is
still channelized, its future is not as tightly constrained by industry. In
2004, the Contra Costa County Flood Control District bought a 122-acre parcel
adjacent to this part of the creek, in partnership with the East Bay Regional
Park District and the John Muir Land Trust. They got it for a song. “The
property was in default, and we purchased it for about $700,000 in back
taxes,” Detjens says.
Called Pacheco Marsh in a nod to the historical landscape,
today the parcel looks nothing like its name. The land is high and dry,
dominated by short non-native grass. “It was drained and filled,”
Detjens says, as he leads the way towards the Bay. Yet this heavily degraded
land still offers the promise of renewal. Frogs call nearby, and salt marsh
harvest mice and nesting black rails live here. In addition, protected tidal
marsh lies on both sides: the 275-acre Peyton Hill Marsh to the west and the
760-acre Point Edith Marsh to east. Ultimately, restoring Pacheco Marsh and the
creek’s upstream floodplains will give wildlife an uninterrupted corridor of
The 122-acre parcel purchased in 2004 stops just short of
Suisun Bay. In between sits a 19-acre parcel, dubbed the donut hole because it
is completely surrounded by protected lands. Late last year, the donut hole
finally joined the lower Walnut Creek restoration project. The Marathon
Martinez Refinery bought this last piece of the project for $4 million in
December and will donate it to the John Muir Land Trust. The transaction was
shepherded by Tim Fitzpatrick, an engineer at the refinery and longtime member
of the restoration project’s stakeholder advisory group. “The term
‘win-win’ is often overused,” Fitzpatrick says. “In this case,
however, it is not.”
Detjens couldn’t agree more. “It’s absolutely
strategically placed — right between our parcel and Suisun Bay,” he says.
“It will allow full tidal marsh restoration.”
The Marathon parcel sits high above the Bay, overlooking a fringe of tidal marsh lush with cattails and tules. The restoration project will add a new tidal channel that ties into an existing one with a mouth on the Bay. The restoration will also take advantage of Pacheco Marsh’s unnatural height, creating a gradient from wetlands to uplands. The latter will provide high-tide refuges for wildlife, and give tidal marsh room to migrate inland as sea level rises.
Michelle Orr of Environmental Science Associates, which
developed the restoration plan, calls the project groundbreaking. “It
remakes an old flood-control channel into something that’s more in line with
our current thinking,” she says. “It’s a new way of doing flood
protection and a new approach to baylands restoration that builds in a lot of
The project will also offer public access. The John Muir
Land Trust plans to build 2.5 miles of trails, boardwalks, and bridges in
Pacheco Marsh. “You’re a world away once you’re out there,” says
Linus Eukel, who directs the land trust. “It’s an opportunity to allow
people to really connect with shoreline habitat. It will be a sanctuary for
birds, habitat, and people.”
In 2019, the project was awarded $7.9 million for
construction from Measure AA, the $12 annual parcel tax that supports
restoration projects in the San Francisco Bay. Rosalie Howarth, a Walnut Creek
homeowner and avid birder, can hardly wait. “For a dollar a month — one
less biscotti with your latte — we’ll forever protect this land for everybody,
people, wildlife, and plants alike.”
The funding is in place, the permit applications have been
filed, and Detjens is more than ready to begin restoring lower Walnut Creek.
“Local control isn’t the answer to everything but in this case, it was
definitely the answer,” he says. “It worked so much better to be
nimble, flexible, and inclusive instead of being confined to the rigid federal
process ― we have the local expertise to know how the creek actually works and
to develop plans that are sustainable for natural processes and the
Corte Madera Creek, which drains the small watershed between
Mount Tamalpais and San Pablo Bay, is barely four and a half miles long. But
this little creek is an outsized problem for people in Ross and other towns
built right up to its banks. Nearby streets can flood to depths of three feet
during the rainy season, running like streams and leaving a shocking amount of
mud in their wake.
“Our peaceful creek turns into a rushing torrent in
winter,” says Chris Martin, who grew up in Ross. One of his neighbors kept
a canoe outside of his house for rescuing those stranded by floods, only to be
stranded himself at age 100 years. “He called 911 and they said to handle
it himself,” says Martin, a former Ross mayor who ran for office to work
towards a better solution.
Finding a fix has been contentious since 1971, when the U.S.
Army Corps of Engineers put a mile-long concrete flood control channel through
Ross. “It never worked,” says Sandra Guldman, board president of
Friends of Corte Madera Creek. “It disconnected the creek from the
floodplains, wrecked the coho salmon fishery, and didn’t accomplish its purpose
― it’s a failure any way you
look at it.”
The local community has repeatedly rebuffed the Corps’ attempts
at redoing the project. The most recent effort, developed under a 2014
agreement between Marin County and the Corps, was shot down in 2018. The plan
received so many public comments that responding to them would have taken more
staff time than the County was legally able to provide under its agreement with
the Corps. “We were working under a strict 50/50 cost share that did not
allow us to exceed the funding currently allocated to the project,”
explains Liz Lewis, who oversees the project as water resources manager for the
Marin County Department of Public Works.
So the County and the Corps terminated the 2014 agreement.
“I think everyone heaved a huge sigh of relief,” says Guldman. This
short-term shift to local control keeps the project under Corps authorization,
leaving open the possibility of partnering with the Corps on future flood
control efforts further up Corte Madera Creek.
Local control will speed reworking the flood control channel
in Ross. “The Corps has a lot on its plate ― projects can take a long time,” says Martin, a longtime
member of the region’s flood control advisory board. Another benefit of local
control is that it will facilitate aligning the project with community
priorities. “It will be a comprehensive flood control program,” he
continues. “It will be as natural as possible and will provide public
access where possible.”
Key elements of the local plan include removing one or both
sides of the concrete channel where feasible, and then restoring floodplains
and riparian corridors. “Widening the creek will increase its capacity to
accommodate flood flow,” Guldman says. Ultimately, the project could also
help steelhead trout and coho. “We’re working to improve passage for
salmon through the concrete channel until they reach a natural stream bed in
Ross,” says Lewis.
The project, estimated at $13.5 million, is on track to be
constructed by 2023. For the thousands of people who live, work and go to
school along this part of Corte Madera Creek, finally getting a respite from
flooding will be the biggest relief of all. RM
Robin Meadows is an independent science journalist in the San Francisco Bay Area. She covers water and climate change adaptation for Estuary News, is the water reporter for the Bay Area Monitor, and contributes to Bay Nature, Frontiers in Ecology and the Environment, PLOS Research News and Water Deeply. Robin also enjoys hiking and photography.