The most successful flood control system in American history has activated more than two thousand times since 1922. Total flood damage in the areas it protects over the past century: zero dollars. Most people who live behind it have never heard of it.
That last fact is the whole story, if you think about it long enough.
On March 25, 1913, the Great Miami River and its tributaries buried downtown Dayton, Ohio under twenty feet of water. The flood killed more than 400 people, with at least 428 by some accounts depending on how indirect deaths are counted. Fires broke out in flooded buildings and burned unchecked because the fire department couldn't reach them. Property damage ran past $100 million, which in 1913 was real money. Before the flood, Dayton's engineers had considered a flood control plan designed for a hypothetical flow of 90,000 cubic feet per second. The actual flood delivered nearly three times that. The plan that would have been inadequate hadn't even been built.
Five weeks later, a committee of Dayton civic leaders led by industrialist Edward Deeds selected a man named Arthur Ernest Morgan to figure out what to do about it.
Morgan had three years of high school and six weeks at the University of Colorado. Before he was an engineer he'd been a logger, a surveyor, a ranch hand, a miner, a typesetter, and a beekeeper. He learned water control working alongside his father and recognized it as a field undeveloped enough that a man without credentials could establish himself in it. By twenty-six he'd volunteered to draft statewide drainage standards for Minnesota. The state engineering society adopted his proposals and wrote them into law, then offered him the position of State Engineer. He declined.
By 1913 Morgan was running his own firm out of Memphis, specializing in drainage and flood control. He was forty-five, obscure, and lacked the credentials to impress the engineering establishment of his day. His practical experience with water ran deep, though, and he carried a habit of mind that would prove more valuable than any diploma: he did not trust his own assumptions.
Morgan hired nearly fifty engineers and surveyors and sent them across the 4,000-square-mile watershed. He sent a separate team to Washington to undertake what appears to have been the first comprehensive rainfall analysis ever attempted in the United States, gathering half a million data points and mapping the 160 greatest storms in the country's recorded history.
On October 3, he delivered his report. It contained eight different flood control plans.
The specific contents of seven of those eight plans have not survived in accessible records, though Morgan's October 1913 report and the Miami Conservancy District's historic bulletins may hold them. What is documented is what Morgan concluded and what Dayton chose.
Morgan's analysis ruled out channel widening alone as insufficient. He had studied European flood control and found his model in the Loire Valley, where retarding basins had been used to manage river flooding for generations. The principle was simple and, to conventional engineering thinking, backwards: you build space to let water in. On your terms, at locations you choose, releasing it at a rate the downstream channels can handle.
The plan called for five earthen dams across the Great Miami watershed, each creating a retarding basin. The basins would hold floodwater and release it through concrete conduits at controlled rates. Between floods, the land behind the dams would be farmed. The dams would have no moving parts. No gates, no valves, no operator standing by to make the right call at the worst possible moment. When water rose above channel capacity upstream, it would simply back up behind the dam, and the conduits would meter it out at whatever rate the downstream channel could safely carry.
Morgan did something else worth understanding.
Having gathered his half-million rainfall data points, having mapped the 160 worst storms, having calculated the worst case the historical record could show him, he added forty percent.
He designed the system not for the 1913 flood, but for a flood forty percent larger than the 1913 flood. Nine to eleven inches of rain across the entire watershed in three days, plus another forty percent on top of that. A storm nobody had seen and that might not come for a century, or might come next spring.
Morgan called his method "conclusive engineering analysis," by which all options were studied in detail until they were ruled out. The name is modest almost to the point of misdirection for a method whose central act was to take the worst thing the data could show you and then assume it wasn't bad enough.
The five dams went up between 1918 and 1922, built by more than 2,000 workers. Germantown, Englewood, Lockington, Taylorsville, and Huffman. Dam lengths ranged from 1,210 to 6,400 feet, heights from 65 to 110 feet. Englewood Dam, the largest, contained as much earth as the Great Pyramid of Giza. Forty-three miles of levees were built through the downstream cities. The combined retarding basins covered 35,650 acres.
At the time the largest public works project in the world, the system cost over $30 million (~$649 million today), funded almost entirely by local tax initiatives. No federal money. No Army Corps of Engineers.
It was not painless. The village of Osborn, Ohio, eight miles east of Dayton, had suffered little damage from the flood itself. But it sat in the area designated to become part of the Huffman retarding basin. Engineers calculated that if a flood matching 1913 returned after the dam was built, the tallest church steeple in Osborn would be nine feet underwater.
Osborn's residents fought the relocation to the Supreme Court and lost. Beginning in 1922, nearly 200 houses were physically moved to a new site, along with businesses and shade trees. The move took two years. A newspaper in Troy had claimed the enabling legislation would "bring Ruin, Death or Starvation to Miami County." Upstream landowners challenged the constitutionality of the eminent domain provisions in legal battles that ran from 1915 to 1919.
Morgan had designed a system that worked with flood dynamics rather than against them. The people whose land and homes sat in the path of that design experienced it as sacrifice compelled by law for the benefit of a larger city downstream. The archive preserves the engineering triumph in rich detail. What the families of Osborn thought about being moved so Dayton could stay dry is harder to find. Wright State University holds the Osborn Removal Company Records in its special collections. Even the name of the archive tells you whose perspective organized it.
In January 1937, twelve days of rainfall equaled the total that had fallen in the four-day 1913 catastrophe. All five dams and all the levees performed exactly as designed. Dayton stayed dry. Nobody drowned. Nobody needed to.
By the end of 2021, the retarding basins had collectively stored floodwaters 2,085 times. The system has never come close to reaching capacity. The highest river stage since 1913 occurred in January 1959. Today the dams and levees protect roughly a million people and $3.2 billion in property.
In 1933, Franklin Roosevelt appointed Morgan as the first chairman of the Tennessee Valley Authority. Morgan brought the Miami Valley approach directly into TVA's early design: the principle that flood control meant treating an entire watershed as a single system rather than fortifying individual points along a river, and the labor-camp practices that had produced low accident rates and high worker morale in Ohio. His tenure ended badly. He was rigid, paternalistic, unable to tolerate opposition, and FDR removed him in 1938. But the watershed-as-system philosophy outlasted the man who carried it there.
As Mark Bernstein wrote in Ohio Magazine in 1987:
"The Miami Conservancy District project was so successful that most in the Gem City today are unaware of it."
The dams are still there. The farmland between them is still farmed. The river still rises, and the system still catches it. No moving parts, no human decision required at the moment of crisis. Just earth and concrete and the physics of water, arranged by a man who gathered every piece of data he could find, then built for forty percent worse than the worst of it.
Two thousand and eighty-five times the water rose and the dams caught it and the conduits metered it out and nobody downstream had to know.