A river always floods through it: The combined force of dollars and engineering cannot subdue the Mississippi, says Sandy Balfour

THE ferocity of the now receding Mississippi/Missouri floods caught everyone by surprise. But one has only to look at the seemingly endless record of 'flood years' to understand that in the Mississippi Valley - a rich, wide, fertile and flat alluvial floodplain - there is no such thing as a surprise flood.

The timing of this year's devastation may be a little odd (normally high water comes with the spring thaw; by July the waters are subsiding), but that the Mississippi should flood is as natural as sunshine in Florida. More usually, of course, the great floods occur in the lower river, in the last 1,000 miles below Cairo, Illinois. This is where the plain flattens out (the river drops less than 400 feet in the distance) and where the Ohio - and therefore the Tennessee and Wabash - flow into the Mississippi.

Of the water that flows past Memphis, on the lower river, only about 38 per cent comes from the Missouri/ Mississippi network. The bulk comes from the Ohio and Tennessee watershed, from the lush Appalachians, rather than the dry mid-West. 'We don't care too much about the Missouri,' says Donna Willett, speaking for the US Army Corps of Engineers in Vicksburg, Mississippi. 'Hell, it can rain there for weeks, and we wouldn't mind. We can handle three times the water coming down in those floods. But the Ohio, well, that's another story. When that starts rising, we start watching.'

Watching is the responsibility of the US Army Corps of Engineers. Watching, waiting and preparing for the floods, which always come. The corps was put in charge of the river after the war of 1812. Its mission was to keep out the French and the British, but when those armies never returned they focused on keeping out the water. In 1879 the Mississippi River Commission - four parts military, three parts civilian - was formed.

Charged by Congress to 'correct, permanently locate and deepen the channel and protect the banks of the Mississippi River, improve and give safety and ease to the navigation thereof, and prevent destructive floods', the Commission began, or rather, continued, a policy of 'hold by levees'. Gradually the entire lower river and its tributaries came to be lined by levees, and with each successive flood, the levees and costs grew. A typical levee near, say, Memphis would have had an elevation of 9ft feet in 1882, 15 in 1896, 27 in 1928 and 35 or more today. The base would correspondingly have grown from 53 feet wide to more than 300 feet.

The levees are sophisticated earth dams. Often set back quite far from the river, they simply provide artificial canyons through which the river is made to flow. But the system has its problems. One is that during floods, the water level inside the levees is higher than the surrounding land. This is not natural. Nor is the confinement of water to a single channel. Louisiana and parts of Mississippi exist in their current form precisely because the river did not stay in one channel, but would suddenly and radically change course over an arc of 300 miles. In the 5,000 years since the last ice age, the river has made five major shifts in direction. Were this not so, Louisiana would be a long peninsula sticking out into the Gulf of Mexico.

The natural process is that as the mouth extends southwards, the current slows, sediment builds up, and the river spills to one side. The main channel of 3,000 years ago is now the quiet backwater of the Bayou Teche. Donna Willett explains: 'The river used to wander all over its floodplain. When the waters rose, people would move their stuff, and that was that. But not now. You can't move Vicksburg.'

In the last 300 years the United States has come into being, and it cannot afford nature's perambulations. Driven by the twin engines of oil and agriculture, it has sent the US Corps of Engineers in to do battle with the river and has created a hugely complex water system that could threaten, rather than protect, millions of people, livestock and acres.

Each year, usually in July, Congress passes a Flood Control Act, with the appropriate funding for new engineering works on the Mississippi. You can chart the floods by looking at the funding provided by Congress for the subsequent year. The great flood of 1927, when a river normally a mile wide became more than 90 miles wide, led to the Flood Control Act of 1928. The basis of policy then survives today: hold by levees, but also by distributaries, spillways and fallbacks. As the current floods illustrate, there is only so much levees can do.

The most spectacular piece of engineering, and one that demonstrates more than any the determination of the corps to impose its will on the river, is at the Old River Control Structure. By the Thirties it was becoming clear that the river was gearing up for its sixth major change in direction, and that the Atchafalaya, its only distributary left after all the levees were built, was threatening to 'capture' the flow of the Mississippi, sending it due south to Morgan City, leaving Baton Rouge and New Orleans stranded to become ghost towns alongside a derelict tidal creek.

The volume of flow down the Atchafalaya had been increasing year on year to the point, in 1950, where 30 per cent of the Mississippi waters went that way. And there, Congress deemed, it should remain. The 1954 Flood Control Act authorised the river commission and the Corps of Engineers to build a dam and a control structure that would, forever, preserve this ratio of water: 30 per cent down the Atchafalaya; 70 per cent on past Baton Rouge. The dam was built - but the 1973 flood nearly destroyed it. So it was rebuilt.

There is doubt as to what 'forever' means. It could mean until the next glacial age in 5,000 years. Or it could mean until the river decides to go where it likes. A 1980 editorial in the Washington Post put it neatly: 'Who will win as this slow-motion confrontation between humankind and nature goes on? No one really knows. But after watching Mount St Helens and listening to the guesses about its performance, if we had to bet, we would bet on the river.'

In the aftermath of such a disaster, it is tempting to apportion blame. The corps should have known. Their worst case scenarios should have allowed for this. And so on. This is not so. Any attempt to control nature is a balancing act. There is always some risk too expensive to guard against. But in the lower river, where the water is so much higher than the land, we can be confident that, sooner or later, it will flood, and do so with a vengeance.

(Photograph and graphic omitted)