Four months ago, they failed spectacularly, causing the costliest natural disaster in United States history. Today, New Orleans' levees are considered the linchpin of its revival. If the city can rebuild them, it can begin reconstructing itself in earnest.
There's just one problem. While engineers agree that the levees failed because of improper design, they're not yet sure how to fix them. It's a key question, which will determine whether the city's current hodgepodge of steel and earthen barriers can be upgraded and rearmored - or if they need a complete overhaul.
Such mysteries need to be solved quickly if New Orleans is to be ready for the start of the next hurricane season in June.
"As we go forward, the question we have and all the people living here have is: 'How good is the rest of the system?' " says Paul Mlakar, a senior research scientist for the US Army Corps of Engineers. "If it breached here, how do we know it's not susceptible somewhere else?"
As he talks, he watches engineers pull up eight sheet pilings from the 17th Street Canal levee, whose failure caused most of the city's flooding. These pilings, long pieces of steel driven into the soil, increasingly have become the focus of several teams of scientists - some working in cooperation, others independently - as they search for clues to the disaster.
After hurricane Katrina hit on Aug. 29, forecasters initially breathed a sigh of relief. The storm had weakened before hitting the coast (scientists now say it was a Category 3, not 4, hurricane) and the city missed the full fury of its winds. But what it did get was a storm surge of seawater, which battered the 475-mile levee system and began to attack the pilings.
The pilings support the floodwalls and they're supposed to be long enough and driven deep enough to keep water from seeping under them. But with this hurricane, water seepage weakened the soil and several floodwalls collapsed.
The culprit? The Army Corps of Engineers, says Team Louisiana, a group of six Louisiana State University professors and three independent engineers investigating the levee failures for the state transportation department. The team says its computer simulations showed that pilings driven 17.5 feet below ground weren't deep enough and that the corps was supposed to run the same kind of tests. Reports from two other investigating groups - the American Society of Civil Engineers and the National Science Foundation - say the pilings along the internal canals should have been sunk to at least 18.5 feet.
For its part, the corps admits that sheet pilings may not have been deep enough. But it has formed its own investigative team to study the question, called the Interagency Performance Evaluation Taskforce.
In addition to sheet pilings, scientists are studying whether the materials used in the pilings and floodwalls was appropriate; whether dredging of the canal to 18.5 feet stripped away supporting soil in some places; and whether large trees planted close to the levees weakened the already soft soil. "At this point, we are ruling nothing out," says the corps's Mr. Mlakar.
Most everyone agrees that answers are needed quickly, since thousands of residents are delaying rebuilding until they get assurances that the levees will be made stronger than before. Two weeks ago, President Bush made that promise when he asked Congress for an additional $1.5 billion in levee funding. Previously, he had only committed to bringing the levees back to their original condition.
Even that may not be enough, some experts say. "Now they want to build the levees a little higher and add some rocks on top of the canal," says Robert Bea, a professor of civil engineering at the University of California at Berkeley and part of National Science Foundation's investigative team. "It's still a patchwork-quilt approach, which has already failed miserably."
After hurricane Betsy flooded the city in 1965, the corps began building I-walls, jointed walls sunk into the soil that are the least expensive and least stable of the solutions available, experts say. Even more disturbing is that no one was responsible for the overall integrity of the system after the work was completed, Mr. Bea says.
"It's warm weather and boats are sailing on Lake Pontchartrain and some neighbors are reporting wet backyards. So they send out a Sewerage and Water Board crew, which tests the water and says that it's salty," he says. "But soon the tide drops and the seep goes away and people quit complaining."
The next time, though, city officials are not likely to be so laissez-faire.
There are problems that engineers can address now, says Thomas Wolff, associate dean of engineering at Michigan State University in East Lansing and a member of the American Society of Civil Engineers investigatory team. They could further flatten slopes along the levees, reconfigure drains to reduce subsurface water pressures, and add pumping facilities so the internal canal levees would no longer form the main line of protection.