Unprecedented spring flooding from Iowa to Illinois has forced thousands from their homes, led to millions in damage and doesn't show signs of stopping for days--if not weeks. After storms in Missouri already sent the Mississippi River's crest to deadly levels earlier this year, Congress and scientists alike are racing for answers. Some have compared this year's Midwestern wipeout to the catastrophic and deadly Great Flood of 1993, a supposedly once-in-a-lifetime event. What gives?
"There's a big problem with the calculation of what is a 100- and 500-year flood plain: it's based on the original shape of the Mississippi Basin," says Timothy Kusky, an earth scientist at St. Louis University. He believes that a triple whammy of factors is contributing to this year's flooding. The first, he says, is increased rainfall due to global warming, but the second two reasons are based on relatively recent physical changes in the region.
Bottomlands that used to absorb floodwaters have been overdeveloped. What's worse, the miles of new levees that were built to protect the malls, industrial parks and homes in those areas have hurt more than they've helped by constricting rivers into narrower channels. "Once we reduce capacity, things that used to be beyond the flood plains are at risk, because floods become higher and more frequent." Models from the Intergovernmental Panel on Climate Change predict 20 percent more rainfall in the region, which Kusky says could mean about 50 percent more water in the river systems.
Klusky can't stop the rain, but his digital models of possible flood outcomes might just help prevent developers from building in the most vulnerable spots around the country's biggest and most dangerous river. On a modest $250,000 budget, Kusky and his team at St. Louis University's Center for Environmental Sciences (SLUCES) are employing a powerful open-source system called to create super-detailed, three-dimensional models of the Upper Mississippi Basin--its mountains, valleys and riverbeds, as well as its parking lots, levees, and office towers.
To better understand how potential disasters--natural and unnatural--might affect the region, Geowall employs two digital light processing (DLP) projectors to generate two distinct, offset images that create a stereoscopic effect on screen. Researchers then don glasses similar to the cardboard specs used in the increasingly prevalent 3D screens at home and in theaters--but with much more complex readouts on the other side. "We'll be feeding in both hard data and mathematical models that involve river levels, rainfall, climate change and more," says SLUCES geospatial analyst Abduwasit Ghulam. , "and seeing the results play out in real time and space."
Rigging the high-end projectors to off-the-shelf graphics cards, Kusky's researchers have begun to recreate the St. Louis metropolitan area, recalculating the sample model from every angle with hopes of expanding their 3D blanket across every structure built at the confluence of the Mississippi and Missouri rivers. "We can make it so you'll feel like you're flying through," he says, "like you're zipping through the valleys in an airplane."
The data that the scientists feed to the system comes from a variety of sources, including U.S. Geological Survey aerial photographs and several NASA projects: the satellite-strapped image program, and the spacecraft cameras of the (SRTM). Each pixel in a given image provides information to the program about the represented land area, including elevation and permeability, which becomes data for Geowall's 3D projection.
Researchers also can look at images taken over several years to track shifts in land usage--from open bottomland to strip mall, cropland to forest or river-edge to levee--and how these changes affect flooding.
"One of the really valuable things is that once we have the topographic data, and the true heights and dimensions of the levees, we can run the whole range of climate models through the system," Kusky says. "Then, If Mr. Conservative X says this will happen, and Mr. Liberal Y says that will happen, we can show what may really happen between this range of values, and it will probably be somewhere in the middle."
As lawmakers debate spending nearly $2 billion in emergency funding for the ravaged Midwest, Klusky hopes that his DLP rig will have implications for the next great wave: deciding where new building should occur, showing how to protect already-developed areas and helping with plans for conserving wildlife--let alone human life.