HIGH-WATER MARK

People living along the middle Mississippi River—from the Missouri River confluence north of St. Louis down to the Ohio River confluence at Cairo, Ill.—can expect more severe and frequent flooding in the future. That's the word from SIUC geologist Nicholas Pinter after delving into decades of water-stage data going back to the 1860s. 

"The size of a flood can be measured by quantity or by how high the water gets," Pinter says. "Many engineers focus on just the quantity of water, but if you’re a property owner, the stage—the height of the water—is the most important concern."

By that measure, his research shows that property owners may have a lot to worry about. Over the course of the last century, he says, "the same quantity of water has systematically caused larger and larger floods." That has happened, he says, because navigational and flood control structures built over the years have slowed river flow during floods. 

Pinter’s conclusions about river engineering causing an increased flood threat in the middle Mississippi are not new. Two studies published after record-breaking flooding in 1973 made some similar arguments, but they were slammed for faulty science. 

"The previous conclusions were right, but sometimes for the wrong reasons," Pinter says. "Floods have gotten worse on this stretch of the Mississippi, and now I, and others, think that for the first time we have the evidence to prove it."

Pinter, geology master's student Russell Thomas, and U.S. Geological Survey hydrologist Joseph Wlosinski analyzed daily water stage and discharge (volume) measurements taken at St. Louis and downstream at Chester and Thebes, Ill. The St. Louis records date to 1861; those at Chester and Thebes go back to the early 1940s. 

The team’s method, called the specific-gage technique, hadn’t been used before on the middle Mississippi. It allows scientists to track changes in water stage over years or decades for the same quantity of water. "We would pick, say, 500,000 cubic feet per second—a fixed discharge—and look at how the height of the water associated with that quantity changes over time," Pinter says.

The team did that for several fixed quantities, ranging from low-flow to flood conditions. After graphing the various stages reached by those discharges over the years, they found a pronounced upward trend in water stage at all three sites for all but low-flow conditions. And that means worse floods.

How pronounced a trend? In 1861, a flood measured at 700,000 cubic feet per second at St. Louis caused a flood stage of about 31 feet above normal. "If the same quantity of water came through today, it would push the stage to over 41 feet along that stretch," Pinter says. "The same rates of change are occurring up and down the middle Mississippi River."

Focusing on the relationship between discharge and stage allowed Pinter to isolate the effects of river engineering on flood behavior. "The technique excludes upstream variables like climate shifts and land clearing," he says.

The stage that a given discharge of water will reach depends on how fast the water flows—which is determined almost entirely by the channel and nearby structures, such as levees. So the team correlated stage/discharge trends with historical data on the river channel, such as cross-section measurements and water velocity.

"If you can push a given quantity of water through faster, it makes the stage of the water go down," says Pinter. "In contrast, if you do anything that slows the water down, the same quantity of water kicks the stages higher than they would otherwise be.

"It turns out that flood flows have been dramatically slowed down through the middle Mississippi since the 1930s and 1940s. That same time has been a period when a lot of construction for navigation improvement was being done on this stretch."

Levees are partly responsible for higher flood stages because they constrict the water instead of letting it spill out onto floodplains. But Pinter says that wing dams, also called navigational dikes, are an even bigger contributor to higher floods in the middle Mississippi. 

These dikes are jetties of rocks extending from the bank perpendicularly into the river. Hundreds of such structures exist on the middle Mississippi River. Their purpose is to funnel water into the navigation channel when the river is low so that barge traffic can continue. But when the river overtops the wing dams, they slow down the water, pushing flood stages higher than they would otherwise be, Pinter says. 

"It’s an alarming conclusion," he adds. "It means that the river navigation industry gains a benefit from wing dams, but property owners are paying for that in terms of increased flood risk and increased damage when a large flood occurs."

What fingers wing dams as a prime culprit? The team found that over the decades, water velocity has slowed down in the middle Mississippi not just at flood stages but also at normal and moderately low stages. Levees would not come into play under those conditions.

The team went on to translate its findings into flood hazard probabilities for the middle Mississippi. "There’s no denying the river has changed—but the flood probabilities haven’t been updated since the 1970s," Pinter says. 

Their most striking conclusion is that the 1993 flood, commonly called a 500-year flood for St. Louis, was really more like a 100-year flood, perhaps less. Pinter also calculates that the stage at St. Louis for a 100-year flood should be estimated at a minimum of 51 feet. In 1993 the river reached 49.5 feet at St. Louis, where the flood wall is only 52 feet high. 

The Army Corps of Engineers, which manages the river and has built the wing dams and many of the levees along its banks, also is recalculating flood hazard probabilities for the middle and upper Mississippi, with findings due out sometime in 2002. The Corps’ method considers only water volume, a technique called discharge-based flood frequency analysis. 

Pinter wants to test that against his stage-based technique for estimating flood frequency and severity, which he terms "a radically different approach." As he notes, much is at stake: millions of people will be affected by the Corps’ findings. "I think they’ll come up with good numbers," he says, "but our technique allows us to do it a lot more cheaply and quickly."

Pinter is a vocal critic of what some have termed the levee "arms race." "Right now up and down the Mississippi and Missouri Rivers they are talking about raising and strengthening the levees," he says. "At Jefferson City on the Missouri [the Corps] is considering a proposal for a 1,000-year levee, the highest protection level anywhere in the country.

"What others have speculated, and what I think our results have shown, is that all that’s going to do is boost future flood levels at that spot in the river and at other spots up and down the river—basically forcing those other spots to raise their own levees.

"What we saw in the 1927 Mississippi flood, the 1973 flood, the 1993 flood is that the stretch of floodplain with the weakest defenses loses. It’s a history of temporary local solutions driving a long-term regional problem."

The first step toward a better solution is to refrain from building in the floodplain, he adds—and then to reduce the infrastructure that’s at risk.

The team’s findings were published this past July in EOS, a journal of the American Geophysical Union. The research was funded by the Charles A. Lindbergh Foundation and by an SIUC internal grant. Doctoral student Rubin Heine, who holds a fellowship from the Association of State Floodplain Managers, will help Pinter continue exploring the link between engineering activities and flooding. 

Comments: Perspectives Webmaster
Copyright © 2001, Board of Trustees, Southern Illinois University | Privacy Policy