Is there a pub bore in the country who has not been shaking his head over his pint and opining, "Well, if you will build on a flood plain…"? These armchair prophets were merely parroting phrases they heard from Westminster. "Gordon Brown has to accept…" said shadow communities secretary Eric Pickles, "that if you build houses on flood plains, it increases the likelihood that people will be flooded." But David Orr, chief executive of the National Housing Federation, countered that, "There's simply no way we could [say]… that we can't build any more new homes because of concerns about flood plains… much of the country is a flood plain."
What exactly is a flood plain? The answer might start with a reminder that rivers, unlike canals, are not compelled to respect fixed boundaries. They are, in fact, not things at all, but processes. Surface water flow from rain or snow melt, erosion, and sediment transport combine to produce a river channel that constantly shifts, redefining its landscape. The meanders gradually push back the surrounding hill slopes and smooth out a broad, flat valley floor, thick with fertile sediment: the perfect setting for agrarian settlements, or so it seems. But when the waters rise above the banks, there is nothing to stop them washing across the plain. Levees may try, but struggle against the fact that a river's curves are always moving: the Mississippi shifts its tracks by up to 20 metres a year. A fundamental problem for construction near rivers is that buildings stay put, but rivers don't.
What's the solution? To judge from recent events, it hasn't changed much in a hundred years: pile up sandbags. But some precautions are still little heeded: replacing soil with concrete exacerbates the dangers by increasing run-off, and the inadequacies of Britain's Victorian drainage system are no secret. Flood defence is not sophisticated: it is largely a matter of installing physical barriers. But permanent walls can conflict with access and amenity—no one would want a three-foot wall all along the Thames. And some areas are impossible to protect this way. So there is no real call for new technology, only a need to recognise that flood threats now have to be considered routine, not once-in-a-lifetime risks.
A lot wetter in Asia
The British floods were the worst for 60 years, and claimed at least nine lives. But a soggy summer in Gloucester is put in perspective by the recent situation in parts of Asia. In China, heavy rainfall in the north brought flooding to the Yangtze, and storms affected one tenth of the population. Meanwhile, a heatwave in the usually moist southern provinces left more than 1m people short of drinking water. And an unusually intense monsoon devastated parts of India and Bangladesh, killing more than 2,000, and displacing hundreds of thousands from their homes. A map of the flooded areas of Bangladesh was almost surreal, showing over half the country "under water."
But there's little new in this. Low-lying Bangladesh floods to some degree most years. The Yellow river, deadlier even than the Yangtze, is known as "China's sorrow" for having brought catastrophe to the country's great plain well over a thousand times despite Herculean efforts to contain its flow. A flood in 1887 created a lake the size of Lake Ontario, directly killing an estimated 1m.
Perhaps surprisingly, some Chinese have been more ready than westerners to attribute recent events to global warming. Dong Wenjie, director-general of the Beijing Climate Centre, claims that the frequency of extreme weather events is increasing, and that this "is closely associated with global warming." Well, maybe. The most we can say is that it is in line with what global warming models predict, as the hydrological cycle that moves water between the seas and skies intensifies—although that does not imply a uniform increase in rain across the globe. In fact, global variation in rainfall trends was central to scientists' recent claim to have detected evidence of global warming in 20th-century weather: models predict that this influence has a geographical fingerprint that has now been identified in the data. This suggests that the models—which predict more extreme weather in future—can be trusted.
One question so far given little consideration is what this implies for the major hydraulic engineering projects under way in Asia. Ten years ago, specialists in water resource management thought that the problems of existing projects, such as the Aswan dam on the Nile, might curtail the era of mega-dams. Now that looks unlikely: China's Three Gorges dam is more or less complete, and both China and India appear set on ambitious schemes to transfer waters between major rivers. The "south-to-north water diversion" project in China is scheduled to deliver water to Beijing for the Olympics from over 1,000km away, while the "interlinking rivers" project in India would convert the entire country into a grid of waterways controlled by dams, with the aim of alleviating floods and drought.
These schemes are already fraught with economic, environmental, social and scientific questions. The prospect of more extremes of rainfall will only make the issues more uncertain, and prompts us to shed the illusion that we understand what rivers can and will do.
