Note: This article is from Conservation Magazine, the precursor to Anthropocene Magazine. The full 14-year Conservation Magazine archive is now available here.

Harnessing the Restoration Potential of Artifical Floods

July 29, 2002

By Ross Freeman

Illustration ©Ken Orvidas

In the spring of 1963, as teams of ironworkers rolled the last massive steel gate shut, Glen Canyon Dam in northern Arizona officially opened for business. Ten million tons of concrete formed a plug over 700 feet high, capable of detaining almost 2 year’s worth of the Colorado River’s flow. At the time, no one considered the potential impacts (including those to native fish or their habitat) of the dam on the Colorado River ecosystem in Grand Canyon National Park. The Endangered Species Act was a full ten years into the future, Environmental Impact Statements (EIS) didn’t exist, and commercial river rafting was in its infancy. Few of those witnessing the completion of the monolithic dam that day could ever have imagined that 33 years later the dam operators themselves would allow an intentional flood to course through its bypass tubes.

It happened one March day in 1996. A man was permitted to traverse the catwalk at the foot of the dam and operate a valve resulting in formidable and unprecedented consequences. He opened one of the four jet tubes that bypass the money-making hydropower plant, initiating an experimental flood of 45,000 cubic feet/second (cfs), several times the normal dam outflow volume but still only half the volume of the historic average annual spring floods. He later recalled a cloud of mist billowing hundreds of feet in the air and the roar of escaping water, both sensations confirming the river would surely rise as it had done historically in the springtimes of millennia past. That man was Bruce Babbitt, then Secretary of the Interior. He was not alone—on the same platform other key participants opened the three remaining bypass tubes: dam management and hydropower officials, river guides and conservationists, and finally, scientists whose data supported using a test flood as a restoration tool. Unlike the observers 33 years prior, this collection of stakeholders was a microcosm of the diverse and disparate interests that had struggled to guide the dam’s operation in line with society’s new interest in the health of the downstream environment. For the past three decades, flooding had been avoided at all costs. Scientists never considered using the dam as a tool for habitat restoration.

The Grand Canyon experimental flood lasted a full week, and although it now appears to have altered the local ecosystem for less than a year, it has changed the face of large river restoration in North America forever. This remarkable project demonstrated the powerful ecological influence of flow variability, the importance of considering unprecedented management options, and the need to keep scientific objectives afloat amidst the turmoil of competing administrative goals.

How Did We Get Here?

For most of the dam’s history, highly variable daily releases were governed mainly by the air-conditioned power needs of desert cities like Phoenix. Each year, the reservoir is gradually lowered in the winter to accommodate runoff from spring snowmelt. But in June 1983, unusually high runoffs forced the Bureau of Reclamation to send 92,500 cfs of water roaring through the dam and over the spillways. The hydraulic force of these releases even caused partial disintegration of the spillway tubes. To river runners and avid environmentalists, this was heaven. To dam operators, this was a potential nightmare.

At about the same time, in part due to rising public concern, the Bureau of Reclamation began its first program to assess the effects of dam operation on downstream environmental and recreational concerns. Dave Wegner, Glen Canyon Environmental Studies (GCES) program manager at the time, explains: “After the floods of 1983 and 1984, the consensus was that all floods were bad.” Nobody publicly considered using one as a tool for positive downstream management. Floods in general were to be avoided.

One of the many unusual elements of this case study is the degree of existing protection that this stretch of the Colorado River enjoys. The area has been a National Park for over 80 years, and in 1989, the Secretary of the Interior proclaimed that dam operation must also have an EIS1. By 1992, Congress had even passed the Grand Canyon Protection Act, which required dam operation that would protect resources within the Grand Canyon and mitigate adverse impacts. To focus on these two new directives, GCES orchestrated a huge collaborative effort among many state and federal agencies, private consultants, and university researchers to collect data on the biological, ecological, and archaeological characteristics of this highly regulated river. The data suggested that the canyon ecosystem was suffering both from the daily surges of hydropower releases, and the lack of a traditional and regenerative spring flood.

River runners weren’t surprised. They’d noticed for years that the river’s broad sandy beaches seemed to be shrinking. Trees on slopes above river level were dying from lack of spring floodwaters. Native fish populations, perfectly adapted to warm, silty, turbulent waters were declining, and some appeared to be retreating to the less altered tributaries. Historically, an average of 66 million tons of silt entered the Grand Canyon annually before the closure of the dam gates upstream—the river was considered “too thick to drink, too thin to plow.” This silt, often resting on the river bottom, was reworked annually by spring floods and deposited in terraces above normal water level, creating beaches and backwater pools believed to be critical to native fish habitat.

But, with the dam in place, water emerging from its base is crystal clear and perennially cold. It’s drawn from inlets hundreds of feet below the lake surface, where the sun’s warmth cannot reach; it also contains no sediment because that all settles out at the head of Lake Powell (the 250-square-mile reservoir behind the dam). The water thus enters the Grand Canyon “sediment-starved.” A 1996 summary study by the U.S. Geological Survey (USGS) concluded: “the release of clear water into a canyon that once carried extremely high sediment loads is a recipe for substantial environmental change.”

The GCES team began to concentrate on ways to improve the sediment budget within the Canyon as a means to address other related habitat issues. “What we learned after several years of study was that periodic disturbance was what a normal ecosystem commonly experienced,” notes Wegner. Until this point, scientists and conservationists had generally focused their research on specific adverse impacts of dam operation. No one thought that proactive and potentially beneficial dam management was possible. Indeed, according to one report published by the American Geophysical Union, most researchers focused on how “Glen Canyon Dam could be operated to minimize sandbar erosion,” rather than on how to remobilize critical beach-building sediments. But, according to Wegner, integrating the physical, biological, and recreational studies into one coherent package led to a critical change in thinking.

Intentional Flooding

Intentional flooding is not a concept that enjoys a great deal of public popularity, especially in the arid West, where it comes across as an incomprehensible waste of water that should be used for hydropower and irrigation. When GCES scientists and other stakeholders proposed artificial restorative floods as a habitat management tool, power and water interests were not at all enthused: because the flood would partially bypass the turbines, it was as if wads of dollars were going to shoot right out of the four jet tubes. Strangely, the U.S. Fish & Wildlife Service also initially opposed the idea because of several endangered and sensitive species that potentially stood to suffer. Larry Stevens, a pre-eminent ecologist associated with Grand Canyon research for 20 years, says it took reams of persuasive field data and numerous back-room negotiations over three years to get both parties to sign on to the flood idea.

By this point, the draft of the EIS was almost as massive as the dam it concerned: tens of thousands of public comments were received. After much negotiation, the “preferred alternative” option in the final 1996 EIS contained a recommendation for occasional high volume releases. Wegner remembers the extreme amount of patience, planning, and effort required to bring all parties to the table. However, he also feels the collaborative approach was far from faultless: “Too often, the meeting of an administrative process [tended] to take precedence over the environmental and scientific process.” Scientific goals bogged down by the endless need for visible consensus.

Other opposition hailed from Indian tribes concerned about archaeological damage, and trout-fishing enthusiasts anticipating the annihilation of the rather incongruous trophy fishery at the base of the dam. Sheer determination, and the inclusion of every possible stakeholder, allowed the agenda to trundle slowly forward. Nevertheless, some scientists allegedly continued to disagree with the idea over fears that a damaging precedent might be set if the flood failed to deliver benefits.

The largest fully controlled experimental flood in the history of North American river regulation eventually took place, notable in particular because exhaustive scientific studies were conducted before, during, and after the event to document its effect. Officially known as the 96 Beach Habitat Building Flow, its overarching intent was simply to mobilize river-bottom sand (delivered for years by free-flowing desert tributaries) and redeposit it on downstream beaches as the flood slowly receded, mimicking in some way a natural springtime flood event common to large rivers.

As stated in the final EIS, the flood’s specific scientific goals were to:

1. redeposit high elevation sand bars,

2. preserve and restore camping beaches,

3. flush non-native fish,

4. rejuvenate backwater habitat for native fish,

5. scour new high water zone invasive vegetation, and

6. provide water to established riparian vegetation at the old high water zone.

All this was to be done while protecting endangered species and cultural resources.

Although this project may have been precedent-setting in North America, other global venues had already explored and implemented prescribed floods at least 10 years prior. Dr. Rich Beilfuss, a wetland hydrologist with the International Crane Foundation, says several African nations—such as Cameroon, Nigeria, and Senegal—have been leading the way in experimental releases for restoration. Other controlled releases have occurred in Japan.

Political versus Ecological Success

After the flood, perhaps the most consistent point of agreement was that the power of floodwater is impressive, delivering results far sooner than researchers expected. Although the purpose of the experiment was quite clear, the results were less so. In retrospect, initial reports of the scientific success may have been a little overzealous.

Barry Gold, Director of the USGS’s Grand Canyon Monitoring and Research Center, now coordinates the ongoing studies. He says that politically the flood was a success, but few of the natural resource goals were realized on a sustained basis. This may be in part due to a by-product of the consensus-based process. Gold says the EIS suggested that test floods should occur in low-water years, when high-volume releases from the dam were not expected. This would prevent the immediate erosion of any gains enjoyed by the beaches. But there was a complication. Certain states were claiming that 1968 legislation outlawed releases bypassing the power plant except for safety puroses. Thus, in order to address the dispute, the official 1996 Record of Decision stipulated flooding between January and July in high-water years—times when the states could perhaps more comfortably concede to “wasting” water for such a purpose under the rubric of dam safety. As a result, says Gold, the big new beaches built following the 1996 flood were quickly lost to high flows through the Canyon.

Dave Wegner says the scientific goals (listed above) were met in the short term, but hopes for long-term establishment of beaches went unfulfilled. He’s also careful to add that another more politically driven goal went entirely unmet: “that the flood would solve the conflict between [dam] operations and the downstream environment.” One flood does not a natural river make.

However, Larry Stevens points out that the thorough monitoring and years of painstaking planning allowed huge amounts of useful data to be gathered, and he thinks it was good value for the money spent. Certainly, the flood cost more than most local or regional conservation projects could hope to raise: approximately US $2-2.5 million in scientific overhead with perhaps US$2 million in lost power revenue. But Stevens doesn’t know of “another single, small federal expenditure that has produced as many peer-reviewed papers, or opened so many doors to river management directions and issues.”

A year after the flood, a scientific paper touted the benefits of returning rivers, as much as possible, to their “natural flow regime.” Although this research was done before the Grand Canyon experiment, the paper did mention the ‘96 flood. The lead author, aquatic ecologist LeRoy Poff of Colorado State University, sees prescribed flooding as an increasingly valid option in river management. Although critics of the Grand Canyon experimental flood claim it was a “band-aid” solution to a much bigger ecosystem problem, the process was nonetheless a highly visible way of introducing adaptive management concepts to dam operators as well as to the interested public.

Barry Gold firmly maintains that the flood itself was just a single “event in a much richer story.” He emphasizes that it was an experiment from which scientists learned many valuable lessons for the future. This is, first and foremost, applied science that “may or may not need more basic research in order to understand what is going on.” Rather than focusing on the actual flood, he argues that “it was revolutionary to pick the preferred alternative, but implement it in an adaptive management process”—a process involving long-term monitoring, continued research, and learning from past experiments. Embracing such uncertainty and experimentation is not normally associated with dam operation.

Looking Downstream

A number of other experiments have been conducted based on the findings from the original ‘96 flood. In 1997, a 31,000-cfs habitat maintenance flow was tested to evaluate its ability to conserve sediments recently delivered by silty tributaries. In 2000, low steady summer flows were used to simulate the seasonal pattern of the natural hydrograph to benefit native fish. Gold notes that until very recently, the strong focus on adjusting dam operations has unfortunately diverted attention from the ecological effects of non-native fish—new experiments are planned.

Over the past few years, Gold has seen adaptive river management considered or used on numerous other rivers in the United States and internationally. He says these initiatives all fit coherently into a larger picture: “The way to do large natural resource management in the future is to use collaborative, science-based, adaptive management processes.”

The 1996 flood demonstrated that even an idea as outlandish as “squandering” water for research could be sold to the traditionally intransigent hydropower interests and supported by the Bureau of Reclamation, an agency not known for progressive environmental restoration. Moreover, the flood showed that a new, unconventional perspective can restore vitality to a stagnating restoration process. Admittedly, the ’96 flood was a single event, and cannot compensate for the continued dam presence. However, the use of a variety of managed releases to simulate historic flow events has now been firmly planted on the river management map.

What to Read Next