Submerged buildings near Lake Providence, Louisiana, in May 2011. Floodwater released from a key Mississippi River spillway surged through the Louisiana bayou, and levees protecting the state's two biggest cities held as river flows neared their peak.Eric Thayer / Reuters

The levee is a technology fundamental to human civilization.  Artificial embankments were designed for the earliest cities, along with the first known draining systems and wells. In the ruins of great Bronze Age civilizations, lost now for thousands of years, the imprints of advanced networks of raised earth can still be traced.

Artificial levees in America predate the founding of the United States itself. Before European colonization, Native Americans made raised-earth structures along the banks of the Mississippi and Ohio rivers.

Levees can be made of mere mud and sand, yet still bedevil today’s engineers. They have been in use for millennia, yet still they fail. And in many ways, the story of the levee’s design and failure is a parable about the eternal battle between technology and nature.

Two of the best-remembered levee disasters in the United States both decimated New Orleans, once after the Great Mississippi Flood of 1927 and again after Hurricane Katrina in 2005. (The former was immortalized in a 1929 blues tune, “When the Levee Breaks,” that was covered and popularized by Led Zeppelin four decades later.) Local history, though, reveals a much longer record of levee failures; the region has always known anxiety about the integrity of manmade embankments.

Newspaper archives are filled with accounts of surging waters, submerged plantations, quicksand, and close calls. In the Great October Storm of 1893, women were saved from drowning, the story goes, only when their long hair wrapped around tree limbs. “Down the bayou we are used to dealing with sudden adversity,” the poet Martha Serpas wrote in 2010. “We calibrate history by big hurricanes.”

The word “levee” comes from the region, too—from the French lever, to rise—but the technology itself comes from nature.

* * *

Natural levees form gradually in response to floods. When high waters recede, they leave sediment on the banks. Those mud and sand deposits gradually build up into a buffer against subsequent floods. The idea behind manmade levees works the same way: They provide extra protection in areas prone to flood.

Today’s levees are designed to meet meticulous engineering standards. Inspection guidelines by the Army Corps of Engineers specify that urban levees should be built high enough to handle a 10-year flood, that is, a flood with a 10 percent probability of occurring in a given year.

But height, while critically important, is just one dimension of a technology that hasn’t ultimately changed that much over the millennia. Levee inspectors also look for signs of erosion, rutting or other pits that might indicate drainage problems. They look for cracking, slope stability, and signs of unwanted vegetation or animal burrowing. “Levee systems are complex because their successful performance relies on many different aspects working in conjunction with each other," said Tammy L. Conforti, the Levee Safety program manager for the U.S. Army Corps of Engineers, in a statement provided by a spokesman.

New Orleans now claims to have the best flood protection of any coastal community in the United States. “Today’s levee system is a far cry from the flawed structures that failed during Hurricane Katrina, devastating much of the region and killing hundreds of people,” the Times-Picayune wrote in 2013. “The new system was designed using better engineering, more advanced computer modeling and better construction materials. Just as important, it was designed to be a true system, and not just a system in name only as its predecessor.”

That may be true. It’s also been claimed before. “We have spent hundreds of millions of dollars to protect ourselves from water,” Louisiana Governor John McKeithen said in 1965, in the months before Hurricane Betsy devastated New Orleans, according to Craig Colten in his book, Transforming New Orleans and Its Environs. “We have built levees up and down the Mississippi,” McKeithen said. “We feel like now we are almost completely protected.”

They weren’t.

And the city still wasn’t protected half a century later. When dozens of levees catastrophically failed throughout New Orleans in 2005, the strength of the storm surge was only partly to blame. It wasn’t just that water flowed over the tops of levees; the levees themselves disintegrated and in some cases shifted suddenly in huge sections. “Think of a layer cake,” Thomas Zimmie, a civil engineer at Rensselaer Polytechnic Institute, told the Times-Picayune in the weeks after the storm. “In the middle I’ve got my icing. All of a sudden, I push on the top of my piece of cake, and what it's moving on is this weak, slick icing. The whole thing moves.”

Katrina left some 80 percent of the city flooded, with sections of New Orleans under 20 feet of water or more. The American Society of Civil Engineers called it “the worst engineering catastrophe in U.S. history.”

An independent report led by engineers at the University of California at Berkeley found several serious design flaws. Some levees were constructed atop too-weak soil, others contained too much sand and other “highly erodable material” that washed away in the storm surge.

“[It] should never have been used in a levee,” Raymond Seed, the author of the report, said in 2006. “A lot of that material is still in the levee, and if it didn't fail during Katrina, it will fail next time.”

* * *

Great floods, like so many calamities, have a way of returning once they've been forgotten.

In the late 1960s, archaeologists began to worry about their ability to continue excavations at the site of the Indus Valley civilization, in what is now western India and Pakistan, where the world’s oldest known levees had been found. A shift in the Indus River had created a rising table of underground water. Which meant that the remains of an ancient civilization known for its sophistication in handling water—not only advanced embankments but also waste-water drainage systems, sanitation, and wells—would soon be wiped out by it. Groundwater was seeping up from below, carrying with it concentrated minerals that ate away at ancient structures, salty earth turning artifacts to powder almost instantly on contact. In another irony, some scientists now believe it was drought—not flooding, for which the Indus appeared to have been prepared—that ultimately destroyed the early civilization. An earlier theory was that slowly rising waters weakened the ancient city's levees until they failed, and the people drowned in mud.

At the same time as archaeologists raced to preserve Indus Valley ruins, engineers in Louisiana gathered to discuss the threat to their own city. In 1965, Hurricane Betsy had slammed into the Gulf Coast, killing 81 people and destroying much of New Orleans. “I am here,” President Lyndon Johnson said from a storm-battered New Orleans tarmac, according to an account in Paul Martin Lester’s 2010 book, On Floods and Photo Ops, “because I wanted to see with my own eyes what the unhappy alliance of wind and water have done to this land and its people.”

What the president saw with his own eyes was utter desolation. Betsy prompted Congress to authorize a new ring of levees around the city. But half a century later, after Katrina, it became clear that those levees were not enough.

Today, along with structural fixes to existing embankments and new computer modeling to better understand and address vulnerabilities in New Orleans, some engineers believe more levees don’t always offer better protection. In some areas, opting for natural levees instead of manmade levees would actually decrease a storm surge, according to a 2013 study by researchers at Notre Dame and published in the Journal of Waterway, Port, Coastal and Ocean Engineering.

“Historically, the design of Southeast Louisiana’s hurricane-flood-risk reduction system has hinged on raising and adding levees in response to river or hurricane events that impact the region,” wrote William G. Gilroy, a university spokesman, of the research when it was published. “Now, it may be time to think and build smarter.”

The nagging question about any levee is always the same: Will it be enough? The hope is that the question never has to be answered, that the true test of such technology remains theoretical.

But in Louisiana, people know—from survival, from a history calibrated by big hurricanes—they will again be tested.

Or, in the case of New Orleans, as John M. Barry, the author of the book Rising Tide, put it in The New York Times: “The question is whether the ocean will engulf the city—whether the city can continue to exist.”

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.