Saving coastlines from flooding is an uncertain science

Swings and roundabouts: sea level rises are hard to predict. Danny Lawson/PA

Coastal and river flooding struck Britain again this week with huge waves hitting southern and western coasts and around 100 flood warnings still in place by Wednesday evening. Disturbing but, sadly, not unfamiliar scenes were accompanied by an impressive catalogue of disruption and bitter tales of avoidable personal tragedy.

But was this weather or climate? Must we expect more of these kinds of events and if so what can we do about it? As with all things climatic the answers are far from simple, but we can start with three basic facts.

First, a warmer atmosphere can hold more water, so at the simplest level, the global hydrological cycle is likely to intensify. This suggests it’s likely to get wetter somewhere, but not necessarily everywhere. Second, the ocean will expand as it warms, leading to rising average sea levels in the coming decades – this is one of the most confident predictions of climate-change science. Third, any melting of the continental ice sheets on Antarctica and Greenland would raise global sea levels further.

Of these three factors, only more intense storms would affect river flooding. Changes in storms are extremely difficult to predict, and will vary between regions, while river flooding is also affected by highly local factors related to soils and land use. If river flooding is changing already then the signal is not yet clear, but the evidence from the past is that relatively modest climate fluctuations since the last glacial period have been accompanied by substantial changes in flooding regimes.

How much bigger will Blackpool’s waves get? John Giles/PA

But coastal flooding is another story. Although individual events are highly dependent on winds and tides, and hence subject to the same uncertainties as inland storm behaviour, the inexorable rise of sea level in response to ocean warming shifts the probability of coastal flooding steadily upwards.

The fact that the dynamics of ice sheets are hard to predict just means we can’t be sure how fast sea levels will rise, but the “likely” range from the IPCC is 26-82cm by 2100, so the uncertainty is as big as the signal itself. Indeed even predictions of the warming component of sea level rise depend on poorly understood processes such as the rate of transport of warm water to the deep ocean and typically vary by a factor of two.

Protection pays, but how much?

As rising sea levels will mean more coastal flooding it might seem the answer is clear – invest in sea defences now to avoid damaging floods before they happen. But how much to invest, where and when are much harder questions to answer. We’ve already seen that the rate of sea level rise is highly uncertain, but this is only one factor in the equation of cost versus benefit of sea defences.

Just as important is the value of land to be protected, which depends on how it is used. If land is built on, the degree of coastal development invokes public planning policy as well as private and commercial development decisions. And if land at risk is used for producing marketable agricultural or commercial output, then the value depends on the price of those goods on international markets.

What value the Aberystwyth foreshore? Antony Stone/PA

Science or policy first?

Any decision of where and how much to invest in coastal protection is thoroughly mired in uncertainty. The rate of sea level rise is highly uncertain, and the statistics of individual flooding events even more so. The value at risk depends on the propensity for coastal versus inland development; and the fact that coastal economies are embedded in a globalised world means macroeconomic forces, including international trade, affect the cost-benefit calculation.

So far, the instinct of scientists and politicians has been to take a “science-first” approach, working forwards from the most complete possible scientific knowledge to the best possible decisions. But the science will never be complete and time is short. Much better, some have argued, to start from the policy end – prioritising the science that is necessary to choose between the policy options that are actually available, and using models as tools to explore the most relevant options.

Science first or policy first, we need better ways to making decisions under conditions of uncertainty. The ERMITAGE project has taken important steps in this direction by combining models of climate, economy and agricultural production. Initial results confirm earlier conclusions: money put into coastal protection pays for itself handsomely.