IN 1543, GARCILASO DE LA VEGA, a member of Hernando de Soto’s expedition, was one of the first white men to see the Mississippi River. He recorded its power: “Then God, our Lord, hindered the work with a mighty flood of the great river, which…came down with an enormous increase of water, which in the beginning overflowed the wide level ground between the river and the cliffs”—meaning the river’s banks, which towered above the river at low water—“then little by little it rose to the top of the cliffs. Soon it began to flow over the fields in an immense flood, and as the land was level, without any hills, there was nothing to stop the inundation. On the 18th of March, 1543,…the river entered with ferocity through the gates of the town of Aminoya [an Indian village near the present site of Greenville, Mississippi]. It was a beautiful thing to look upon the sea that had been fields, for on each side of the river the water extended over twenty leagues”—nearly 60 miles—“of land, and [within] all of this area…nothing was seen but the tops of the tallest trees…. These floods occur every fourteen years, according to what an old Indian woman told us, which can be verified if the country is conquered, as I hope it will be.”
IN THE LATTER PART of August 1926, the sky darkened over much of the central United States and a heavy and persistent rain began to fall. Rain pelted first Nebraska, South Dakota, Kansas, and Oklahoma, then edged eastward into Iowa and Missouri, then into Illinois, Indiana, Kentucky, and Ohio. Great cracks of lightning seemed almost to break open the sky, booming thunder made buildings quiver, and the wind rattled windows and tore away anything loose. Rain poured from the sky in sheets. Even in the intervals when the rain ceased, the darkness did not lift. The sky was impenetrable and uniform, the clouds themselves lost in gray.
The great storm lasted for days. Finally it broke and the sun shone, but within forty-eight hours a second low-pressure system heavy with moisture moved up the Mississippi valley and poured precipitation over this same region. It was followed by still another storm system. In the last fourteen days of August 1926, an area covering several hundred thousand square miles received little relief from drenching rain.
The rain drowned crops and ruined harvests. Though it fell in the dry season, it saturated the soil and filled the riverbeds. The rivers rose. On September 1, water poured over the banks of dozens of streams and flooded towns from Carroll, Iowa, to Peoria, Illinois, 350 miles apart. All the land in between lay heavy, wet, and glistening, reflections of silver drowning the green of crops and grass.
Still more rain fell. On September 4 floods deluged much of Nebraska, Kansas, Iowa, Illinois, and Indiana, killing four people. The Mississippi River rose rapidly in the upper Midwest and washed out bridges and railroads. A few days later there was another storm. Flooding stretched from Terre Haute, Indiana, to Jacksonville, Illinois. Seven more people died.
The rains continued. Storms hung over the region. On September 13, the Neosho River roared through southeastern Kansas, killing five, causing millions of dollars in damages. In Illinois a flood rammed a tree through an oil pipeline, setting it ablaze, spreading fire on the water.
In northwestern Iowa, 15 inches of rain fell in three days over the Floyd River valley, the Sioux River valley, the Dry Creek valley. With the land already saturated and the streams bursting, this water exploded over the riverbanks, drowning ten, inundating 50,000 acres that included Sioux City, adding millions of dollars more in damages. On September 18, the national headquarters of the American Red Cross sent its special disaster team to Iowa, while the waters near Omaha rose to threatening levels.
When the rains began, they had first relieved the region of the summer’s heat. As they persisted, they had irritated, then depressed. Now they frightened. People could do nothing but watch their crops drown and their rivers rise and, reminded of their own impotence and of the power of God and nature, pray.
In many a church preachers spoke of the rain as a sign from God of man’s wickedness. Even without a preacher’s sermon, the good, decent folk of the prairie had to be thinking of the story of Noah, of the end of the world, of the coming of judgment.
All through the month of September and into early October it rained. Floods followed in Nebraska, South Dakota, Oklahoma. The Neosho River in Kansas and 600 miles to the east the Illinois River in southern Illinois rose to their highest levels in history. Flooding in those two states was the greatest and most disastrous ever—an extraordinary occurrence in October, when rivers normally run low.
The Mississippi itself grew fat and swollen, until it too overflowed above Cairo. Tens of thousands more acres went under water.
Along the 1,100 miles of the lower Mississippi, from Cairo to the Gulf, only levees were in place to contain the energy of the river. For the length of that levee line the great earthworks seemed an impregnable fortress, towering two and three stories above the flat delta land. The Mississippi River Commission had pride and confidence in them.
Indeed, that year, even while threatening clouds formed over much of the drainage basin of the Mississippi River, General Edgar Jadwin, new chief of the Army Corps of Engineers, had for the first time officially stated in his annual report that the levees were finally in condition “to prevent the destructive effect of floods.”
But the gauge readings were disturbing. The U.S. Weather Bureau noted that the average reading through the last three months of 1926 on every single river gauge on each of the three greatest rivers of North America, the Ohio, the Missouri, and the Mississippi itself—encompassing nearly 1 million square miles and stretching the width of the continent—was the highest ever known. The Weather Bureau later stated, “There was needed neither a prophetic vision nor a vivid imagination to picture a great flood in the lower Mississippi River the following spring.”
But that fall no one at the Weather Bureau or the Mississippi River Commission correlated or even compiled this information. The individuals who made the readings simply noted them and forwarded the information to Washington.
The gauge at Vicksburg, which lay at the foot of the Delta and roughly halfway between Cairo and the Gulf, was even more disturbing. In October the Vicksburg gauge usually hovered not far above zero, a low-water mark. Only six times in history had the river exceeded 30 feet on the Vicksburg gauge in October. Each time, the following spring saw a record or near-record high water.
Usually, records on gauges are broken by inches, rarely by more than a foot. No October reading had ever broken 31 feet. In October 1926, the gauge at Vicksburg exceeded 40 feet.
Late in October the rain ceased. Those watching the river relaxed.
SIX WEEKS LATER unusually violent storms carrying heavy precipitation began pelting the Mississippi valley again. On December 13, in South Dakota the temperature fell 66 degrees in 18 hours, followed by an intense snowstorm. Helena, Montana, received 29.42 inches of snow. In Minnesota snowdrifts of 10 feet were reported. As the storm swept south and east, 5.8 inches of rain fell on Little Rock in one day, with Memphis reporting 4.11 inches and Johnson City, Tennessee, near the Virginia line, 6.3 inches. By Christmas, 1926, heavy flooding had begun.
To the west, three children drowned in Arkansas as the continuing rains turned streams into torrents. To the east, the Big Sandy River, dividing West Virginia and Kentucky, overflowed. The Cumberland River rose to the highest level ever recorded and flooded Nashville. The Tennessee River rose to near record levels and flooded Chattanooga. At least sixteen people died in Tennessee, with thousands homeless over Christmas. The Yazoo River, running through the heart of the Mississippi Delta, overflowed and left hundreds more homeless. Goodman, Mississippi, had the highest water in thirty years. The Illinois Central, running north-south, and the Columbus & Greenville, running east-west, both suspended railroad traffic across Mississippi.
THE CHIEF, but by no means sole, determinant of how dangerous a flood will be is the height of its crest. This crest is not a wave but a gradual swelling; it is by definition simply the highest point to which a river rises. Flood height depends on several factors, with volume of water only the most obvious. Another is the speed with which a crest moves downriver. The slower it moves, the more dangerous it is: slower floods exert pressure on levees for a longer time, and slower floods carrying the same volume of water rise higher.
Common sense explains why. Floods are measured in cubic feet per second, also known as “cfs” or “second-feet,” a dynamic measure of both volume and force. (When considering issues like storage and irrigation, engineers instead use the static measure of “acre-feet”; 1 acre-foot of water covers 1 acre of land with water 1 foot deep. Although 1 second-foot of water flowing for one day almost exactly equals 2 acre-feet, the two terms represent different concepts and do not equate easily.)
The number of second-feet is obtained by multiplying the average speed of the current times the river’s “cross section.” A large river might have a width of 1,000 feet and an average depth of 10 feet; its cross section would measure 10,000 square feet. If the current is moving at 10 feet each second—almost 7 miles per hour, a current fast enough that a person would have to break into a run to keep pace with driftwood floating downstream—then the river would be carrying 100,000 cubic feet per second. If the current slowed to 5 feet per second, the cross section of the river would have to double to pass the same 100,000 cfs. The river would have to either spread wider or rise higher, or both. Similarly, if the current accelerated to 15 feet per second, the river could accommodate the same 100,000 cfs with a cross section one-third smaller. So the slower the current, the bigger the cross section—and the higher the flood height—need be. The faster the current, the smaller the cross section—and the lower the flood height—need be.
Current velocity depends on the slope of the river toward sea level and on conditions in a particular stretch of river. In some reaches the river flows in a straight line and moves faster; in others it constantly collides with bends or rough spots on the bottom and slows down. Friction—with wind, the riverbank, the riverbed, sediment pushed along the bottom or carried in suspension—can influence current velocity. So can tides, whose influences on the Mississippi reach north of Baton Rouge, and other factors.
Even in a particular locality, the average speed of a river current is just that, an average. In midstream, water contends with less friction than near the banks and generally moves faster; water 20 feet deep faces less friction than on the surface, and so on. On the Mississippi violent differences in currents can create undertows that pull 100 feet straight down, or whirlpools as large as 800 feet long and 200 feet across, large enough to swallow trees, flotsam, or boats. As Ellet observed, “It is no unusual thing to find a swift current and corresponding fall on one shore towards the south, and on the opposite shore a visible current and an appreciable slope towards the north.”
Floods increase the height of the river—in some sections of the Mississippi the average high-water mark may be 50 feet higher than low water—and therefore increase the slope and the speed of the current. If the river is low, and a sudden surge of water is poured into it, the current speeds up. But if the river is already high when more water enters, the river can act like a dam, forcing the additional water to pile up and slow down. Backwater flooding occurs when the main river is so high a tributary cannot empty into it; water from the main river can actually push water upstream into the tributary.
A classic study reveals how flood crests can move at very different speeds. It compared two different flood crests flowing down the same 307 miles of river, from Cairo, Illinois, to Helena, Arkansas, in 1922.
In one case the flood crest poured into the river at Cairo when the Mississippi was low and swept downriver at a speed almost double the average current velocity of the river. The crest, in effect, was a separate layer of water that skidded down the top of the river, traveling the 307 miles in three days.
In the second case the crest entered the lower Mississippi when the river was already high and flooding, and the river dammed up the new flood crest. The new crest moved only one-third as fast as the average speed of the whole river, and took eight days to go the same 307 miles. This crest was, in effect, a layer of water that had to wait for the river channel to empty out before it could flow south. In the meantime, it rose higher than it otherwise would have.
There is no standard speed for a flood. Engineers have observed a maximum sustained current velocity in the Mississippi of 13 feet per second, roughly 9 miles an hour. The power of a mass several miles wide and 100 or more feet deep moving at 9 miles per hour is, literally, awful. And a Corps study concludes that, in a large lower Mississippi flood, “stage transmission” averages 419 miles a day. (The stage is the height of the river surface as measured by gauges up and down the river.) This does not mean that a flood crest covers 419 miles in a day, but some of the force of an approaching crest—some indication of it—travels downriver at almost 18 miles an hour.
The most dangerous floods are those that contain several flood crests. The first crest fills the storage capacity of the river, causing later ones to rise higher than they otherwise would. Meanwhile, the river’s pressure on the levees intensifies. In 1927 the U.S. Weather Bureau station at Cairo, Illinois, noted ten distinct flood crests moving down the Mississippi.