An estimated 5,800 homes and businesses have been flooded in England and Wales this winter, now officially the wettest ever.
By comparison, the total was 55,300 properties after the 2007 summer floods, and more than 10,000 following those in November 2000. So while rivers have been at their highest on record in some places this month, the flooding has been less widespread and less damaging than previously. This is little consolation to those currently underwater of course, but it leads to the obvious conclusion that this scale of flooding is not unprecedented. The fact is simply that we are not sufficiently resilient to floods.
The Environment Agency estimates almost 500,000 properties in England and Wales are insufficiently flood protected. This figure is based on those properties that will flood on average at least once in 75 years, but will include many that will flood much more regularly.
On a small island, our relatively large population is concentrated in lowlands and valley bottoms. Flat land next to rivers and coasts is attractive and cheap to develop, but most exposed. Planning controls have not been sufficiently restrictive, and as a wealthier society now than in the past we have more to lose. Risk is the combination of the probability of an event occurring and the loss it causes, and if the frequency of floods and the damage they cause rise, then flood risk will increase significantly.
One can question the 500,000 figure, but one does not need to be a statistician to establish that there are many thousands of properties that can expect to flood every few years, and many tens of thousands that can expect to flood every decade or so.
A muddy history
So why do we seem continually surprised that floods occur? Collective memory tends to be short: people live more transient lives, less rooted to a particular community. Just because no one can remember a recent flood doesn’t mean the risk has disappeared.
Our records of river flow and water levels are surprisingly short. Most monitoring equipment was installed in the 1960s and 1970s, and with only 30-40 years of data it is difficult to estimate the magnitude of extreme events that occur, on average, much less frequently.
It’s not surprising then that new records are set in every major incident – the combination of only relatively recent historical data and the fact that each flood tends to affect a different set of places makes this statistically inevitable. Worse still, the usefulness of these already short records is reduced by both changing land use, which alters the way water flows through river catchments and the storage capacity of river channels, and by a changing climate, which alters the location, timing and duration of extreme rainfall events.
We know natural variation brings us cycles of flood-poor and flood-rich decades. In the very few places where we have long records that stretch back 100 years or more there is clear evidence of this. The 1890s and 1920s were two examples, and over the last 50 years we have moved from the flood poor decades of the 1960-1980s, to the flood rich decades from the late 1990s to the present.
This high natural variability in extreme weather events including flooding is probably driven by changes in global atmospheric and ocean circulation patterns, although this is by no means definitively established. This means estimating flood frequency from historical records is difficult, as records from flood-poor years may not tell us much about what to expect from the flood-rich period we currently find ourselves in.
So floods are inevitable, and worse now than in the recent past, and we have made things worse for ourselves as a society in the way we have approached building and flood protection. Moving 500,000 homes is not a viable option. We can spend money on flood defences for vulnerable properties, or share the risk collectively through insurance. Both are elements of the solution, but the extent to which society as a whole should pick up the costs for the risky decisions of some needs to be widely debated.
Start small and local
A better solution may be to ensure vulnerable properties are made more flood-resilient so homeowners can recover more quickly and cheaply when they are flooded, rather than the enormous cost of trying to protect them entirely with massive flood defences. In this respect the government’s offer of £5,000 to fund alterations to each flooded property that would make it more resilient – moving electric circuits and plug sockets up above flood levels, fitting flood-proof doors, waterproofing masonry – is a step in the right direction.
We also have to recognise that some of the uses to which we put our land are unsustainable. The profits from carefree building on floodplains benefit the developer, but the cost of flooding is borne by the homeowner and, in the long run, by society through taxpayer funded schemes and insurance costs.
Intensive farming practices also exacerbate the problem, by compacting soil which prevents water being absorbed into the ground and encourages more surface run-off. This washes away sediments that clog drainage systems and rivers, and washes fertiliser into ditches and streams that stimulates excessive plant growth, forcing water out onto the floodplain more quickly.
Both man-made climate change and natural variability are changing the frequency of flooding over time, but we don’t yet know which is more important. Our view is that over the next 20 years natural variability will have a greater impact, but over the next 100 years there is much greater scope to see an increase in flooding that is unequivocally the result of climate change.
Change is certain; but the details of how, where, and why are still unclear. What is unarguable is that in the UK there are steps that can be taken now to reduce the impact floods have, and to ensure that our ability to cope and respond to floods is not compromised in the future – we would be wise to take them.
Paul Bates is on the Royal Society working group examining human resilience to climate change and disasters, whose report is due later this year.