Floods are the most destructive, most frequent and most costly natural disasters on earth. And they are getting worse. In recent weeks, 14 African nations have seen their worst floods in decades. More than a million people have been affected, over 200 drowned, and countless others made homeless across the continent.

At least some of this suffering was preventable.

The recent flood disaster in Ghana was reportedly greatly worsened when dam operators in Burkina Faso opened a floodgate to prevent the Bagre Dam overflowing after heavy rain.

That water raced downstream to Ghana, through the Black and White Volta Rivers, hitting riverbank dwellers hard. This is not the only time this has happened in Africa. Hundreds of people were killed and hundreds of thousands harmed when Nigerian dam operators opened their floodgates without warning in 1999 and 2001.

In these cases - and in many others the world over - dams supposed to help reduce floods only made them worse. Yet a number of governments in Africa and elsewhere are proposing more dams as the answer to floods.

Flood damages have soared internationally in recent decades, partly because global warming is leading to more intense storms, and partly because more people are living and working on floodplains. The UN estimates that by 2050 the number of people at risk of damaging floods will double to two billion. But one key factor behind the spiraling flood damages are the very flood control measures that are supposed to protect us.

Dams and embankments can never be fail-proof, and when they fail, they do so spectacularly and sometimes catastrophically. They provide a false sense of security that encourages risky development on vulnerable floodplains. Too often in the case of non-flood-control dams, downstream residents are put at risk by agencies more intent on

wringing the most electricity or irrigation water possible from a reservoir, rather than

keeping water levels low enough to absorb flood waters.

The limitations of conventional flood control will become ever more evident as global warming-induced super-storms test dams and embankments beyond their intended limits.

Author Jacques Leslie aptly describes dams as "loaded weapons aimed down rivers." Dams kill not only because of dam-operators’ negligence and failure to warn people downstream when they suddenly open their gates, but also because they collapse (as many as 230,000 people died from a chain of dam failures in central China in 1975). Several large dams have collapsed in Nigeria with deadly consequences, notably, the Bagauda Dam in Kano

State in 1988; the Cham Dam in Gombe State in 1991 and the Bagoma Dam in Kaduna State in 1994.

Conventional "hard path" flood control ignores the complex workings of rivers and coasts.

Dams, embankments and the straightening and dredging of rivers trigger profound changes

in the ways in which water and sediment flow through watersheds. Flood damages soar when engineering projects reduce the capacity of river channels, block natural drainage, increase the speed of floods, and cause the subsidence of deltas and coastal erosion. In addition, "hard path" flood control often ruins the ecological health of rivers and estuaries.

There is a better way to deal with floods - the "soft path" of flood risk management.

Flood risk management assumes that all anti-flood infrastructure can fail and that this failure must be planned for. The "soft path" is also based on an understanding that some flooding is essential for the health of riverine ecosystems.

Instead of spending billions of dollars vainly trying to eradicate floods completely, we need to recognize that floods will happen and learn to live with them as best we can.

This means taking measures to reduce their speed and size (for example, restoring meanders and wetlands) and duration (e.g. improving drainage). It means protecting our most valuable assets by raising houses on mounds or stilts, and defending built-up areas behind carefully planned and well-maintained embankments. It also means doing all we can to get out of floods’ destructive path with improved warning and evacuation measures.

Such practices are already in use in many parts of the world. In China, efforts are

underway to restore 20,000 square kilometers of Yangtze wetlands to act as flood

absorption areas. Artificial flood releases are being tested in Nigeria as a means to

revive wetland ecosystems downstream from the Tiga and Challawa Gorge dams. Alterations

in water management are being proposed for Mozambique’s Cahora Bassa Dam that could

reduce the impact of major floods and restore downstream ecosystems.

In the United States a 10-year project to reduce floods on the Napa River in California will restore tidal marshlands, remove some buildings from the flood zone and set back embankments to give the river room to spread. Communities along France’s longest river, the Loire, persuaded the government to scrap a planned "flood control" dam in favour of river restoration and a new flood warning system.

Despite a growing worldwide consensus that mitigation, not elimination, is the only realistic flood policy, there remain powerful factions devoted to outmoded "hard" flood control. An iron triangle of politicians, bureaucrats and dam builders continues to promise salvation through embankments and dams after floods strike (even when such floods have been worsened - or caused - by existing dams and embankments).

Instead of adopting a method of flood management that has failed in other parts of the world, Africa can learn from the mistakes of other nations and adopt a more flexible, effective and sophisticated set of techniques to cope with floods. Such an approach will provide greater protection, at less cost, than engineered flood control, and more flexibility for adapting to future climate changes.