Deadly water on the brain

On the streets of Uckfield, last month, nobody was safe. Most people stayed indoors, but one man, a 54-year-old Vernon Bishop, ventured outside, only to be knocked down outside his jewellery shop and carried off, against his wishes, to a remote spot more than a mile away. Throughout the small Sussex town, windows were smashed, shops emptied of stock. Homes were violated, with furniture, carpets and valuables destroyed. Police, if they could capture the lawbreaker responsible for this crime wave, would be certain to achieve a famous conviction. But it’s difficult to make an arrest when the culprit is water.

Thankfully, nobody was killed in Uckfield. But just two days earlier a couple of girls were dragged to their deaths in a swollen river in Yorkshire, and by the end of the same week 25 more people died in floods around the border of Switzerland and Italy.

Welcome to flood season. From now until summer, events like these are more or less to be expected. The first warning – perhaps 24 hours beforehand – will be a flood alert during BBC Weather reports, featuring new logos unveiled at the start of the season. After a long period of rainfall, land near your home will be saturated, and your local river will be swollen: a single shower will be enough to burst its banks.

As water starts rising in your street you’ll feel terribly anxious. Later, as it swills between the walls of your living room, you may find it hard to fall sleep. Finally, when the water subsides, you’ll experience a bizarre sense of alienation from your most cherished possessions – because they will stink of raw sewage, washed indoors with the flood. “If you have not been flooded,” says a scientist at the Centre for Ecology and Hydrology, in Wallingford, Oxfordshire, “you might imagine that it’s fresh water that comes into your house. But people who have been flooded know that shit comes inside too, and the psychological effect of that is terrible.”

Does this appall you? It should, because more than 1.3m homes across the country are at risk of flooding, and that number is rising. Many people believe this is because of climate change, and hydrologists – the scientists who study the behaviour of water – don’t entirely disagree. They too believe the numbers of people at risk will rise. But the principal cause, they argue, is not climate change. It’s the way we use our land.

Hydrologists have worked at Wallingford since 1949, when the government established a testing facility at the request of the civil engineers responsible for rebuilding the war-torn country. Just outside Wallingford the hydrologists found a site, previously used by the Canadian air force, with access to a plentiful supply of water from the Thames. Since then, the town has become celebrated, among hydrologists worldwide, as a hothouse of British expertise.

But to speak of Wallingford’s hydrologists as a single group would be a mistake. Back in 1968, they were split into two groups, of which one was privatised during the 1980s. Since then, HR Wallingford – it stands for Hydraulics Research – has shifted away from pure research. Nowadays, the company turns over many millions of pounds as a consultancy for water-related projects both here and abroad.

>From the outside, the main office resembles an urban swimming pool: it’s blue-green, with huge portholes. Standing directly opposite is the building in which scientists carry out their research on enormous models, fashioned in concrete and gravel, of real and proposed harbours, bridges and jetties. One, near the door, represents an exact replica of a British dam. With summers generally becoming drier, and winters more stormy, water authorities wish to increase the amount of water stored during winter. Before they do that, they must consider the effects. Will the dam be strong enough? And if it’s fuller than usual, how would the authorities deal with overflow in the event of an unusually heavy rainstorm?

A second model, much bigger than the first, shows the Thames from the Dartford tunnel to the North Sea. A pneumatic generator, containerised inside a huge metal box, simulates the inflow of seawater, while pumps running below the concrete supply a constant surge of river water. With the entire tidal process reduced to just an hour, hydrologists can run as many as eight tides in a single day. This particular geographical facsimile is designed to establish the consequences of new building works on the banks. What kinds of currents, and eddies, might this cause? And would the modified flow damage neighbouring installations? (Another common problem with estuaries involves power stations discharging warm water from their cooling systems: how do they avoid harming the local ecosystem, and can they avoid drawing that warm water back into the cooling system if the tide should push it back upstream?)

How do they measure the flow? “It’s simple,” says Richard Wooldridge, the civil engineer in charge of publicity, holding out a handful of tiny white balls. “We scatter these bits of polystyrene on the surface and follow them with cameras mounted on the gantry [overhead].” To measure currents below the surface, they dangle rods into the water with propellors on their ends, and measure the movements electronically. As he explains this, with a big grin, it becomes hard to avoid the feeling that hydrologists are like overgrown schoolboys – and most of them do seem to be male – mucking about with big toys.

Of course, the models are infinitely more complex than they initially appear. Building them, to precisely calibrated contours, takes weeks – and they can’t be made any smaller because then surface tension and viscosity would skew the results. Some experiments require a level of detail provided only by, say, running salt water (which is heavy) against the flow of freshwater, which weighs less. To achieve this effect is not simple – or cheap – because the scientists, unable to dump vast quantities of brine in the Thames, must ship it away in tankers. An alternative, suitable in some circumstances, is to run cold water, which is heavy, against hot.

Towards the side of the hangar-like space, a hydrologist and a technician carry out an altogether different experiment. They stand on step-ladders, leaning over scaffolding from which a machine noisily throws off drops of water. Below them stands a system of pipes with a U-bend, much like the ones in an ordinary lavatory. At the invitation of Yann Gasowski, whose shirt is by no means dry, I step onto the ladder to inspect what turns out to be a mechanism for draining rainfall from motorways. The machine pumps water with a whoosh into an aluminium gulley, from which it runs through a metal grate and into the U-bend, technically known as a gulley pot. The purpose of the exercise is to establish the capacity of the grate to take in water while also keeping out sediment. If this can be improved, water companies will be saved some of the expense of emptying sediment from gulley pots – which must currently be done two or three times a year. (The maximum capacity of the grate, Gasowski later informs me by phone, is 14 litres per second.)

That’s not the last of the extraordinary contraptions I find at HR Wallingford. In a long, glass-walled flume, Andy Steele simulates storm waves against a rubble mound. Fist-sized stones – some orange, some green, each one individually numbered and weighed – have been positioned precisely, enabling him to measure any displacement caused by the waves. And the last item, monstruously tall, comprises an elevated see-saw with something like a lengthy fish-tank mounted on top. This was recently used to assess the force – and quantity – of water required to shift debris on a mountain stream. “We can vary the angle,” says Wooldridge, “and then vary the discharge. And try it again with a range of sand, gravel and stone mixtures – then work out the algorythms to predict events.” In other words, if given details about the type of debris, and the gradients, in any given terrain, hydrologists will be able to state precisely how much rainfall would cause a landslide.

That kind of practical experiment ties in neatly with a groundbreaking piece of work recently completed by the other group of hydrologists at Wallingford, in the Centre for Ecology and Hydrology. Formerly known as the Institute of Hydrology, CEH is now part of the Environment Agency. As such it recently unveiled a Flood Estimation Handbook – actually a five-volume boxed set, worth £750 plus VAT, with additional CD-ROM – giving guidance on the likelihood of floods in any given part of the country. Armed with this handbook, developers, planners and insurers will be able to ask, and tentatively answer, questions such as: “How rare was that flood?”

In an airless office on the second floor, Adrian Bayliss, Rob Flavin and Neil Runnalls load the CD-ROM to give me a demonstration. At first glance, the onscreen image resembles a brain scan, with fibrous neurons trailing haphazardly towards the edges. Clicking on a button, Bayliss brings up tumour-shapes in grey. Then he clicks on another button. This one flashes up place names – making it immediately apparent that the grey areas actually represent urban developments, and the “neurons” are rivers and streams.

“Imagine you’re standing here ” – he says, pointing at a coordinate on the screen – “with a bucket of water at your feet. This programme will help you to work out the direction the water would flow if you knocked the bucket over. In the past, it would take hours to work that out” – in the event that heavy rain should fall in the catchment area of any particular river, such as the Uck, last month, in Sussex – “but now you can do it in seconds.”

To produce the map, the entire country was measured at 50 meter intervals, giving a total of 100 million grid points. “If you work out the flow,” explains Flavin, “and use the digital terrain you can produce massive areas of Britain at risk of flood. What homeowners – and insurers – want to know is whether, say, 3 Poplar Road is going to flood. But even with gridpoints at every 50 meters, the map can’t be entirely accurate – and the difference between No 3 and No 9 can be crucial, because roads dip, here and there… This is powerful information,” he adds gravely. “Nobody wants to hear that their home is on dangerous ground, because that devalues the property.”

The risk of flooding is not exclusively a matter of high and low ground. CEH’s digital map also takes account of details such as the prevailing type of soil, because that affects the capacity of the land to absorb rainfall: London clay, for instance, is absorbent, whereas impermeable rock in Scottish mountains, obviously, is not. Nor is the issue exclusively geographical: history, too, is important. To assess an area, it’s necessary to know what has happened there in the past. Moving away from the computer, Flavin lifts a pile of maps from a table to find one showing Oxford, and outlying villages, in black and white – with blue bands running way beyond the river boundaries to indicate areas at risk. “If you look at this,” he says, “you can see that people have historically resisted building on the flood plain… except for, er, here, with the Park and Ride. Also, New Botley and New Hinksey are both built on the flood plain. Of course they’re both protected, but I love the fact that both of those places are called “New”. Look at old Oxford – it’s entirely out of risk!”

Actually, that’s not entirely true, as Flavin immediately acknowleges. “We had a chap working here who was really into history. He found some information about Oxford, 800 years ago, which showed that one of the college chapels had been flooded under three feet of water.” That colleague, Frank Law, spent some years working in Australia, where he was surprised to find the collective memory of flooding is more detailed than in the UK. Since retiring, he has set up a web site, www.hydrology.org, which he hopes will be useful in gathering historical records. Of 6,000 accounts gathered so far, about half are about floods (others are about exceptional tides, or frozen rivers). The kinds of information that Law is looking for might include the name and location of a pub, where a wall has been marked to show the level of some long-forgotten inundation – or an entry in your ancestor’s diary.

Using historical information of that kind, hydrologists can calculate flood frequency curves (which relate the severity of floods to the frequency at which they occur). They can also determine the size of the “100-year flood” – a slightly confusing hydrological concept, meaning a flood that has a one in 100 chance of occuring in any given year. British cities on major rivers are typically protected up to the level of the 100-year flood, but that means they may be flooded when a 100-year flood eventually occurs. And that can happen – not just once in 100 years. Confused? Think of the Grand National: it’s entirely possible for a horse to win the National against odds of 100-1, and then for another with identical odds to win the following year. Thus, similarly, parts of Yorkshire recently experienced 100-year floods in two consecutive years. (Uckfield, before the big flood in October, was inundated as recently as May.)

Hydrologists at CEH argue that homebuyers should make the search for information about flood risk a matter of routine – just as we would expect conveyancing solicitors always to search for information about plans to build motorways at the end of the garden. But that’s not going to help the many people currently in possession of properties at risk. Why don’t the authorities build defences to cope with every conceivable flood? The answer is boringly simple: money. Even relatively modest schemes, such as one currently in progress at the river Nene in Northampton, cost millions of pounds. And for floods that occur no more than once in 800 years – like the one in the Oxford college chapel – building defences is out of the question. (This a matter of priorities. In Holland, building defences against 600-year risk is considered economically worthwhile.)

“When a river floods,” says Flavin, “that’s a totally natural event. Floods happen in the winter all the time. With the Severn, for example, it’s all over the fields, for miles. But that’s deliberate ” – because, since the 1800s, there have been mechanisms in place enabling the authorities to flood farmland in order to prevent water hitting towns; where this has been agreed, farmers are compensated for this inconvenience by the Environment Agency.

The greatest problem facing us is that those fields are rapidly being converted to other uses. By covering up farmland (which absorbs moisture) with tiled roofs and acres of tarmac (which don’t) we produce conditions in which flooding is inherently more likely than before. And fields are not the only spaces where open soil is being covered up. CEH’s Runnalls, who lives in Reading, reports with evident horror that developers are constructing homes where previously there were gardens. “Land use,” he says sternly, “is much more significant than climate change.”

What really bothers Runnalls, and many others like him, is the thought that changes in the population will lead to the construction of millions of new homes across the south east. “Four million new homes?” he asks, aghast. “The graph will go like this…” (Runnalls runs his hand upwards, not at 45 degrees like an aeroplane but directly vertical, like a rocket, to show how much more common floods may become.) To the residents of Uckfield, Runnalls does not bring good news: “The people down at the bottom of the catchment will be hit by it all at once.” And that, adds Flavin, is why hydrologists choose to live on higher ground.

Forecasts and warnings

At the Met Office, last month, they knew heavy rain was coming before the clouds even formed. Using calculations generated by supercomputers, forecasters warned, early on October 10, that downpours were on the way. And already, by then, enough rain had fallen to saturate the ground and swell rivers to dangerous levels.

Typically, in the UK, rain comes from the Atlantic, falling heavily on western hills: Exmoor, Dartmoor, the Cambrian mountains, the Lakes, and the West Highlands. That was the pattern in October. By 6.00am on the 11th, clouds had formed over southern Ireland and begun crossing the Irish Sea. At 9.44, the Met Office issued a new warning. Heavy rain was now more likely than not. Its destination: south-eastern England. Worse, the rain would not pass swiftly over, because high pressure, above Moscow, was causing weather systems over England to concertina, grinding virtually to a halt.

As the clouds gradually moved east – in a diagonal band, from the Isle of White to Norfolk – forecasters at the Met Office and the Environment Agency watched its progress on radar screens. At the centre of the band was a small area of extraordinarily heavy precipitation. “Each pixel,” explains Nigel Reed, a Met Office manager, “covers roughly 5km. If it’s dark blue, that indicates light rain. Then it becomes light blue, then green, yellow, orange, and red. Red means it’s absolutely heaving down, anything from 16mm an hour to 32mm. To give you some idea what that means: if you stepped outside and it was falling at 4mm an hour, you’d say ‘Yuergh! It’s really pouring down!’ or you might say something even stronger. So just imagine if there was eight times as much…”

At 5.44pm, the Met Office issued a final warning – just before forecasters left their Reading office to step into the worst of the rain. On the BBC, that evening, weather reports included an official Flood Warning, and a telephone helpline for anxious viewers.

Over the next few hours, it gradually became clear which rivers were filling most dangerously, and eventually the EA activated its warning system for homes and businesses at risk. OpenTalk, developed for the EA by Kingston Voiceware, delivers messages – pre-recorded to describe every conceivable scenario – at a rate of 1,700 calls an hour. Anybody who registers for the service – at no charge – is telephoned automatically.

For residents of Uckfield, on October 12, the warning came around 2.30am. Some, sitting up late to watch TV weather forecasts, will have seen the Flood Warning upgraded to a Severe Flood Warning. Most were woken by the telephone, and a pre-recorded message. They had less than three hours to carry valuables upstairs, turn off gas and electricity supplies, and put sandbags around the front door. That done, many then battled through the dark – and the rain – to do the same for shops and offices.

3138 words. © FT Magazine