Changes in Iowa's weather patterns, landscape, cities and farms have rendered some of the state's most trusted flood prevention safeguards outmoded and inadequate, a review by The Des Moines Register shows.
That includes the state's system of dams - including Saylorville upstream from Des Moines - which were designed to meet climate conditions and a lay of the land that some scientists say haven't existed for decades.
That leaves Iowans, their homes and their businesses increasingly at risk for the sort of devastating floods that ravaged the state in 1993, 2008 and again last year, causing damage in the billions of dollars.
• Iowa's rivers and streams are running harder and faster. While the flow varies, the increases in some places are staggering. The Des Moines river is running 137 percent harder at Fort Dodge now than in 1951, according to a Des Moines Water Works analysis. Flow along the North Raccoon is up 60 percent. The projected increases in rainfall by 2040 would boost streamflows by 50 percent, ISU found.To find out why that's happening and what might be done to better predict and prevent future floods, the Register examined climate records, stream data, flood forecasts, weather modeling and dam operations and interviewed scientists wrestling to answer those same questions.
Among the Register's key findings:
• Iowa's rainfall has increased gradually over the past century, and has come more frequently in deluges. According to one analysis, rainfall has increased since 1951 along the Des Moines and Raccoon rivers by 15 to 20 percent. Iowa State University climate researchers expect the trend to continue, forecasting another 21 percent jump in annual rainfall by 2040, a function of global warming.
• Man-made changes to Iowa's landscape also contribute to increased flows and higher flood risk. In rural areas, improved drainage systems divert more water to rivers and streams, and much of the wheat and oats that once slowed runoff are gone. In urban areas, sprawl and large expanses of pavement produce the same effect.
• Computer models used to predict floods and control Iowa's system of dams do not account for the changes. That makes threats such as the double-cresting flood that surprised Cedar Rapids in 2008 and devastated its downtown easier to miss.
• The state's four reservoir dams - which protect the state's largest population centers - were not designed in anticipation of the changing conditions, documents show. Consider the Saylorville Dam near Des Moines. Assumptions in place during initial design work in 1946 led to the dam's design.
Engineers calculated a 14 percent chance that water would top Saylorville's emergency spillway four or more times during the dam's first 50 years. That's already happened six times in the 33 years since the reservoir opened, most recently in July.
Ron Fournier, a spokesman for the U.S. Army Corps of Engineers, which operates Saylorville and the state's other reservoir dams, said the dam was built to withstand what modelers consider to be the maximum possible flood. While the emergency spillway has been topped more often than expected, he said it is highly unlikely the dam itself would ever be topped.
Still, the Corps of Engineers is seeking money to pay for a new study to determine whether the four reservoir dams should alter operations.
David Miller, Iowa's homeland security director, is among those who wonder whether Iowa's system of 3,800 dams, some of which are substandard, are up to the task of handling the pressures posed by a wetter climate and angrier rivers.
"Some of the old models - no pun intended - don't hold water anymore," he said. "Is it time - because of the changing environment, knowing that the old science was used to build the dams - to give them another look? We think it probably is.
"More rain, intensity create new set of risks
Army Corps hydrologists are not convinced that climate change - or anything else - has permanently changed the natural variation in rainfall patterns, said Fournier, the agency's spokesman.
But Gene Takle, an atmospheric scientist at Iowa State University who models climate change, disagrees. While rainfall can vary widely from year to year, the trend in Iowa toward wetter weather is clear, he said.
Not only is the state getting more precipitation than it did just 60 years ago, according to annual rainfall records studied by Takle, it is getting more intense rains.
In Cedar Rapids, for example, there's been a five-fold increase in the number of times that more than 1.25 inches of precipitation have fallen eight or more days annually. That's important, because such rains are not absorbed by Iowa soil and result in significant runoff to rivers and streams.
Perhaps more troubling for flood-fighters: Most of the rain is falling in the first six months of the year, when spring floods loom.
In eastern Iowa, where a dam breach last year drained Lake Delhi, the average annual rainfall has risen at a greater clip than the rest of the state.
Takle said ISU researchers believe increased precipitation forecasted to occur by the 2040s could result in 50 percent more flow in the Upper Mississippi River system.
"The average flows are way bigger than what they have been, and everyone knows that," said Chris Jones, a Des Moines Water Works researcher who has analyzed river flow trends.
Iowa's rivers and streams are also carrying more water later into the summer, extending the potential for flood risk well beyond the usual wet spring months.
The Des Moines River, for example, was measured at 20 times its normal flow in August, when dry conditions have historically prevailed. The Des Moines tripled its previous record flow on Aug. 9. The Raccoon River at Fleur Drive also set a record that day, seven times the record set in 1998.
The rivers remained in their banks, but with far less cushion against floods than would be expected so late in the year.
Changing land uses speed water runoff
Scientists believe Iowa's swollen rivers and streams are not simply the result of more rain.
"We know it's raining more. We know stream flows are up. We know Mother Nature is doing something," said state geologist Bob Libra. "Does that account for what we are seeing? The increase is beyond what it would be for just precipitation."
Part of the answer lies in Iowa's changing landscape, according to Libra and others.
Prairie grasses, oats and hay that might have temporarily held back charging water from streams have largely disappeared.
When the Saylorville Dam was completed in 1977, Iowa farmers grew 1.4 million acres of oats. Last year, they grew 95,000. In the same period, wheat acres fell from 109,000 to 22,000.
Taking their places are more row crops such as corn and soybeans, which are not nearly as efficient at slowing runoff.
Also, grasses and woodlands that once were part of the Conservation Reserve Program have been turned into cornfields as the demand for corn-based ethanol has risen. Iowa has lost enough natural lands over the past 30 years to cover an 8-mile-wide strip stretching from Council Bluffs to Davenport.
Another factor: the drainage systems, known as tiling, that drain excess water from farm fields to streams and rivers. No one keeps data on field tiles, but Dan Lemke of the Iowa Department of Agriculture and Land Stewardship estimates that about 40 percent of the state's 23 million acres of cropland has subsurface drainage.
"We don't believe there are more acres being drained in the past 30 years," Lemke said. "There have been some improvements that moved water slightly faster."
Lemke's department wants to expand the systems to move water even faster, and use wetlands to help slow the flow into rivers and streams. Environmental groups are opposed, and contend increasing the drainage will only worsen floods downstream.
Takle, the ISU climate researcher, believes drainage tiles are likely to be an even bigger deal as rainfall increases by mid-century. A 24 percent increase in precipitation would increase tile flow by 35 percent, and a 32 percent jump would send 80 percent more water through the drainage systems.
The expanding roof lines and parking lots sprouting across Iowa's cities and suburbs also share blame for increased stream flows and flood risk.
In Des Moines alone, the area covered by pavement and other hard surfaces that promote runoff grew 17 percent between 1996 and this year, records show.
Acres of asphalt at such places as Jordan Creek Town Center in West Des Moines can make a difference to a nearby creek's flood habits, but Witold Krajewski, director of the Iowa Flood Center, says the impact is barely a blip in regional flood control models.
Outdated models do poor job of forecasting
Krajewski says the computer models used by the National Weather Service to predict floods are outdated and fail to account for changes in the state's climate and land use.
"Those models are based on fairly crude concepts of how water travels," said Krajewski, who has studied Iowa rainfall and floods for 23 years at the University of Iowa.
The result, he said, is increasingly unreliable assessments of flood risks.
A case in point was the models' failure to pick up the two-crested flood that swamped Cedar Rapids in 2008 and resulted in $5.7 billion in recovery expenses.
"Cedar Rapids was very poorly predicted," he said.
Krajewski also questions the effectiveness of a federal flood control system in which one agency - the U.S. Geological Survey - measures how fast rivers run, another - the National Weather Service - forecasts floods, and a third - the Corps of Engineers - operates dams. They often don't talk to each other enough, he contends.
Krajewski has company in searching for better forecast techniques.
"Better models are needed," said Libra, the state geologist. "The folks at the flood center are working on that. Something bad has to happen before resources are freed up to do something about it."
The National Weather Service, the Iowa Flood Center and others are working to improve and update the models.
It's important work. Beyond predicting floods, the models affect how dams are designed and operated, where homes and businesses can be built and when storm sewers are likely to overflow.
Krajewski and his colleagues at the U of I and ISU are trying to develop models that take into account runoff from big box stores and the loss of grasslands.
"The National Weather Service models don't have those sensitivities," Krajewski said. The Iowa Flood Center has been awarded an $8 million federal grant for pilot projects to study what happens on various landscapes when, say, buffer strips are planted.
"We want to figure out if we can somehow change the land use practices to mitigate this damage from flooding," Krajewski said. "We need to actually do something on the landscape to do this."
Jeff Zogg, hydrologist with the National Weather Service, said Iowa asked for funding to try new flood-forecasting techniques in 10 to 15 cities, but the Federal Emergency Management Agency turned down a grant request.
Zogg said part of the challenge is the intricate system of waterways lying underground in Iowa. It's hard for modelers to know what's down there.
Still, researchers hope new technologies will help tell their computers far more about what is on - and below - Iowa's landscape.
One of those is Lidar, a laser-based imaging method far more detailed than aerial photographs. Another is heliborne electromagnetic mapping, which reveals groundwater systems much as an MRI takes a look at a patient in a hospital.
Krajewski hopes these tools will help better protect Iowa against future floods.
"I'm optimistic. This is an experiment, but an experiment on a national scale. After the '93 flood, there were no special efforts at the state or federal levels. Now, more people are engaged."
The real beginning of the end of slavery?
