The summer of 2003 was one which broke all records, in every respect. Central and southern Europe suffocated under exceptionally high temperatures over a period of several months. "From a statistical point of view such a summer should never have happened," says Christoph Schär, Professor at the Institute for Atmospheric and Climate Science at ETH Zurich (1). But what is exceptional today could become the norm within a measurable space of time. This at least is what models indicate, published by Schär and his team this week in "Nature" (2).

Beyond statistics

Within the national research project "Climate" (3) researchers have compared temperatures of summer 2003 with data from the past 150 years. "Last summer is way beyond anything that has ever been registered. It even surpassed 1947, which until now counted as a year of exceptional records," explains Schär. Proceeding from the available data, Schär has calculated that, assuming a constant climate, such a heatwave should only occur every 10,000 years – at the most. "Of course, this is a fictitious number," he relativises, "but it illustrates how exceptional the situation was."

The researcher is convinced that last summer is a clear sign that the climate is changing. "Since 1975 we have observed a pronounced warming of the climate in the northern hemisphere. When the average temperatures rise, the probability of a very hot summer grows over proportionally." This interpretation is confirmed by simulations that his team have carried out with a high resolution regional climate model. "If we assume a not-too-optimistic scenario, based on a doubling of the current level of the concentration of greenhouse gases in the Earth's atmosphere, average summer temperatures in central Europe will be around 4.6°C higher at the end of the century than they were during the period of reference between 1961 and 1990", says Schär.

Increase in variability

Apart from the rise in temperature, a further – almost more important – factor must be taken into account. "According to our results the standard variation of summer temperatures in central Europe will increase by up to 100 per cent. This means that towards the end of the century the climate in summer will be extremely different from year to year," explains Schär. Just how determinant the factor of variability is can be demonstrated with an example. If the standard variability increases by 50 per cent, then the probability of another summer like that of 2003 increases by a factor of 150. "So the frequency of extremely hot summers depends on the variability of the climate as a whole," says Schär. "And variability has probably already increased in recent years in any case."

The dried up river-bed of the Töss on the 28th August 2003. (Picture C. Schär) large

Climate change scenarios for European summer temperatures for the period 2071-2100 (compared with 1961-1990) based on a high-resolution regional climate model. The panels show that the effect of mean warming will dominate in the Mediterranean (left-hand panel). At the same time a stronger increase in year-to-year variability is likely in central and eastern Europe (right-hand panel). Both factors increase the likelihood of extremely warm summers like that of 2003. large

Very variable summers

Indeed, existing indications confirm that this assessment was not pulled out of a hat. As Schär points out, the past two summers were very different. While drought was very widespread last year, floods grabbed the headlines in the summer of 2002. Moreover, the conclusions reached by the research group in Zurich are confirmed by other scientists working on the European project "Prudence" (4) in their climate models.

Results also show that the variability will not increase evenly all over Europe. "In Spain we expect to see a strong rise in average temperatures. But the mean deviation will change only slightly. This means that there the climate will remain about as stable as it is today, despite warming."

Ground humidity as a key factor

The assumption that the variability in central Europe will strongly increase can be substantiated by physics. "The key factor in this system is the water content of the ground," explains Schär. "In our latitudes the greater part of the radiation energy is transformed into vaporisation during the summer months. However, if the subsoil is dry, vaporisation is repressed and the air heats up. This is precisely what happened in central Europe last summer. This makes an increase in variability in Spain, where the earth is nearly always dried-out in summer, very unlikely."

The kind of uncertain climate development that Schär predicts would cause major problems for central Europe. "In principle one can become accustomed to a warmer climate if it's more or less constant," he says. "Farmers, for example, can cultivate other plants that do well in a warmer climate. But if mean temperatures become erratic and change greatly from one year to the next it can be far more difficult to adjust to the unpredictable conditions."