How New Orleans Drained Its Way Below Sea Level

In 1853, yellow fever killed 7,849 people in New Orleans. In 1878 it killed another 4,046 and cost the city what the Louisiana State Board of Health put at $10.5 million, roughly $350 million today. Smaller outbreaks came back in 1897, 1898, and 1899, killing far fewer people but, as Mayor Martin Behrman later wrote, depressing business confidence and pushing residents and investors out. Between the Mississippi's natural levee and the Metairie-Gentilly Ridge lay the back swamp: low, saturated ground that held standing water for mosquitoes, flooded the streets after every rain, and produced what the city's own engineers called unsanitary conditions and "obnoxious gases."

So the city drained it, and by the standards of what anyone knew in the 1890s that was among the most defensible infrastructure commitments in American urban history. The complication is what the swamp was sitting on, which was peat, and what peat does when you take the water out of it.

The soil nobody was thinking about

The shallow subsurface under much of low-lying New Orleans is organic clay, peat, and soft deposits laid down over millennia in swamp and floodplain. A 2024 field study that analyzed 88 soil samples from 72 boreholes across the city found peat averaging 39% organic matter, with individual samples reaching 67%. Under Lakeview the peat layer ran about three meters thick.

Saturated peat holds its shape. It is spongy, soft, nothing anyone would stand on and call firm, but it occupies space and it maintains elevation. Water keeps it submerged, and submersion keeps it intact for the same reason a log on the bottom of a lake survives for centuries while an identical log left on the bank is gone in a decade. Water blocks oxygen from reaching the organic material, and without oxygen decomposition effectively stalls.

Lower the water table and two things start at once. The soil skeleton loses the buoyant support the pore water was providing, so it compresses under its own weight and under whatever has been built on top of it. And the organic material, now in contact with air, begins to oxidize — microorganisms that could not operate underwater start converting peat into carbon dioxide and water vapor. Soil volume shrinks. The ground surface comes down. Neither process reverses when the pumps stop, because the material that was holding the elevation has not been displaced to somewhere else; it has left as gas.

None of this appears in the 1895 engineering report, in the public meetings, or in any documented objection from the period.

The decision to drain

The timeline runs earlier than most tellings suggest, and the sequence matters. The drainage campaign preceded the city's celebrated 1905 victory over yellow fever by nearly a decade.

Ordinance 7170, approved February 6, 1893, authorized the topographical survey a comprehensive drainage plan would require. A second ordinance created the Advisory Board that would judge the proposals. The board held public meetings through late 1893 and early 1894, threw out the submitted plans on the grounds that the survey data behind them was not reliable enough, and in April 1895 approved a system designed by City Engineer L. W. Brown.

The 1895 report rewards reading for what it contains and for what it leaves out. It identifies the problems precisely: streets flooding after rain, groundwater saturating the soil, stagnant canals, and large tracts of undeveloped land the city needed if it was going to grow. It reprints an 1857 recommendation from City Surveyor Louis Pilié, who had argued that draining the swamps would improve health, keep residents in town during the "sickly months," allow building on lakeside and swamp land, increase commerce and population, and turn nominally valueless property into taxable real estate. Pilié's proposal had been sitting there for nearly forty years. The city had grown around the swamp without ever being able to grow through it. By 1895 the board was designing a system whose stated purpose was to make a "large and now mostly uninhabited territory" permanently dry and open to improvement — blocks close to the business district that were still, in the board's own words, "practically swamps."

The pressure behind the commitment was not abstract. Behrman, writing in 1914, described the 1897–1899 recurrences as threats to the city's viability as a place to live or invest. The mortality was small by comparison with earlier decades — 298 dead in 1897, 57 in 1898, 23 in 1899 — and the damage to outside confidence was wildly out of proportion to it. What was being measured in those years was not bodies. Civic organizations pushed modern sewerage, water, and drainage as the price of remaining a commercial center at all, and the swamp stood in the way of every part of that.

The Louisiana Legislature created a Drainage Commission in 1896. Construction began in 1897, made possible by the Wood screw pump, a low-lift, high-volume machine that could move water at the scale the back swamp demanded. Earlier proposals, Pilié's included, had no technology capable of dewatering territory this large and this low; the Wood pump is what turned the 1895 plan into something executable. Substantial portions of the system were running by 1900. In 1899 property taxpayers approved a special improvement tax of two mills annually for 43 years, and the legislature constituted the Sewerage and Water Board, which absorbed the Drainage Commission in 1903.

The 1905 epidemic killed 437 people and was fought mainly with mosquito control aimed at Aedes aegypti breeding in domestic cisterns and containers. It gets remembered as the event that launched the drainage campaign. The infrastructure was already in the ground. What 1905 supplied was confirmation — proof that coordinated public health intervention worked, and institutional confidence behind a system that had been operating for five years.

What the engineers understood and what they couldn't

The 1895 planners understood the elevation problem perfectly well. The city is a shallow bowl: the Mississippi's natural levee forms the high southern rim, Lake Pontchartrain sits to the north, and the Metairie-Gentilly Ridge — an old river distributary built of sandier, more stable material — runs roughly east-west between them. The back swamp filled the low ground on either side of that ridge. The report notes that a "large portion" of the basin already sat below mean Gulf level, which meant every drop of rainwater and groundwater would have to be lifted mechanically before it could be disposed of. Continuous pumping was in the original design. The report treats groundwater seepage as an "ordinary flow" to be collected and pumped even in dry weather; Behrman later put that constant seepage at about a million gallons per square mile per day.

They knew, in other words, that they were building something that could never be switched off.

What the report does not mention, model, or warn against is that removing the water would drop the ground. There is no discussion of soil oxidation, none of organic compaction, nothing describing the back swamp as performing any structural or water-storage function. In the report's frame the swamp is an obstacle — unsanitary, economically dead, in the way.

That is not a failure of intelligence. The mechanism by which peat oxidizes and compacts when dewatered was not part of the engineering knowledge available to that board. They solved the problem they could see with the tools they had. The problem they could not see was underground, running at millimeters per year, and would take decades to show up anywhere a person could read it.

Millimeters per year, for a century

It surfaced first in buildings. By 1913 cracks had opened in St. Louis Cathedral, and architects told the Times-Picayune that the drainage system's lowering of the groundwater had caused the settlement. Campanella's historical reconstruction dates the earliest architectural warnings to around 1910. In 1918 the Weather Bureau forecaster Isaac Cline wrote that pumping had caused "some subsidence or sinking," with adverse effects on older structures, the cathedral among them. On August 3, 1919, the Times-Picayune ran W. S. Callender's account under the headline "New Orleans Is Sinking Slowly but Steadily Down Toward China." By 1927 the Sewerage and Water Board had pulled soil samples at Napoleon Avenue and South Johnson Street and photographed how much they shrank after drying.

All of these observers could see settlement. What none of them had yet was the scale. The drainage system was not settling a few masonry buildings; it was lowering the surface of whole neighborhoods, year over year, at a rate no single season made obvious and every decade made permanent.

Before comprehensive drainage, natural subsidence in the area ran around a millimeter a year or less. After 1900 the human-induced rate reached 5 to 25 millimeters a year across much of the city, with some studies recording roughly 50. A millimeter a year is nothing — below the threshold at which anyone standing on the ground would notice. Twenty-five millimeters a year is an inch. Run that for fifty years and four feet of ground that used to be under the house is not there anymore.

The 2024 study measured the accumulation. In Lakeview, comparing a 1950s Army Corps cross-section against modern borehole data, the land surface had dropped roughly 2.2 meters in about 70 years, and the three-meter peat layer was down to about 0.7 meters. Most of that loss the authors attributed to oxidation of the shallow peat. One location south of the ridge went from roughly 1.5 meters above sea level in the 1950s to about a meter below — a loss of some 2.5 meters. City Park now sits about 1.7 meters below mean sea level. Areas north of the ridge run to nearly three meters below. One study borehole in the Lower Ninth Ward measured 1.5 meters below.

The ridge itself is only a few decimeters above sea level, and it is now the high ground, standing roughly two meters over the neighborhoods on either side.

The process is not finished. The 2024 study estimated up to another 40 centimeters of oxidation-related subsidence still available north of the ridge, wherever shallow peat remains above the groundwater.

Something on the order of 60% of the present urban area lies below mean sea level.

The pumps that can never stop

The system built to reclaim the swamp became, through the subsidence it produced, the system required to keep the reclaimed ground from flooding every time it rains. The 1895 planners had designed for perpetual pumping already. Subsidence deepened that obligation by an order of magnitude nobody had budgeted for. As the land came down, the hydraulic difference between the city's interior basins and the surrounding water grew. Lake Pontchartrain and the outfall canals, which were supposed to be where the city's water went, became sources of groundwater recharge flowing back toward the subsided neighborhoods. The system had to work harder against a problem it was making worse by working.

The Sewerage and Water Board now runs 24 drainage pumping stations holding 120 drainage and constant-duty pumps, tied to roughly 90 miles of open canals and another 90 miles of subsurface canals. Part of it still runs on dedicated 25-hertz power, an artifact of the original equipment that was never fully converted. Stations are staffed or monitored around the clock. When pump capacity drops, the water stays where it is.

An independent analysis of the August 5, 2017 flood found that about 45% of constructed capacity at one major station and 37% at another were unavailable because of maintenance conditions, while insufficient 25-hertz power tripped pumps that were otherwise operable. More working pumps and adequate power, the report concluded, would have lowered flood levels and shortened drainage time — it stopped short of saying a fully functional system would have prevented the flooding.

The authors of the 2024 subsidence study stated their counterfactual plainly: Katrina would likely have done less damage had earlier subsidence been smaller, because the inundation would have been shallower and the water would have drained off higher ground more readily. That is a modeled judgment rather than an observed comparison. The topography behind it is not in dispute. Water finds the low point, and the low point is where the peat used to be.

What the record can and cannot support

There is a version of this story in which the drainage engineers are villains who should have known better, and the documentary record will not carry it. The 1895 report contains no discussion of subsidence. The public meetings left no transcript of dissent on the question. The earliest awareness anyone has found in the accessible record is an 1836 newspaper article in the New Orleans Bee observing that "drainage cannot give substance to the spongy soil" — a remark that shows someone sensed dewatering would not produce firm ground, and that is a long way from a quantitative forecast, and it was written 57 years before the project it might have warned about was proposed.

The men who built the system were responding to a crisis that had killed thousands within living memory and had twice come close to shutting the city's economy down. They built something that worked exactly as designed. It drained the swamp, dried the ground, opened the territory, and contributed to the sanitary conditions that helped end yellow fever in New Orleans. It also did a second thing they had no framework for predicting, at a rate too slow to register in any given year and too fast to survive a lifetime of accumulation.

This is not the only celebrated adaptation that reorganized its own landscape on the way to succeeding — Dayton's flood-control districts and Johnstown's engineered channels each produced exposures that came with the fix rather than despite it. What distinguishes New Orleans is that the intervention altered the ground. The peat that oxidized is carbon dioxide now, distributed through the atmosphere, and the elevation it was holding cannot be put back by any means currently available. The city sitting in the resulting basin depends on 120 pumps running continuously, some of them on an electrical frequency the rest of the country abandoned generations ago, to keep rainwater out of neighborhoods that a century ago were swamp holding its own water without being asked.

North of the ridge, where shallow peat is still drying above the water table, the ground has another 40 centimeters to give.