The magazine of the european research area European Commission Copenhagen, a missed chance?

15

SPECIAL REPORT

 

CLIM

A

T

E

Temperature and rainfall data from the past 

and present are used, for example, to test the 

model in regard to climate history using a pro-

cedure known as ‘hindcasting’. “We want to 

know how effective the model is and evaluate 

this compared with climate observations such 

as historical measurements of greenhouse gases,” 

explains Paul van der Linden. If the test shows 

an excessive departure from historical data, the 

model is deemed to be weak. As it takes an 

enormous amount of time to produce a global 

model, only certain elements of them can be 

tested and over a reduced period.  

Thousands of simulations

A single prediction from a model is insuffi-

cient. To improve precision, climatologists run 

the same model thousands of times with dif-

ferent data and are also able to run several 

models on repeated occasions using the same 

data. This method, known as the ‘ensembles 

model’, generates a more reliable result 

because the averages of several models pro-

vide a more precise picture than the result of 

one alone. One of the reasons why ENSEMBLES 

differs from previous projects is due to this very 

effect of scale.  

“It is the biggest project of its kind,” says Paul 

van der Linden. The team of forecasters, who 

are located across Europe, has designed a vast 

multi-model that combines seven European 

ensemble models. Its size enables it to provide 

a degree of detail superior to that of previous 

attempts. It also differs from its predecessors 

because it produces an ensemble on the basis 

of 15 regional models incorporated in the sev-

en global models. Researchers studied the 

potential impact of climate change on 14 sites 

in Europe. They are also able to simulate the 

effects of an average temperature rise of 2°C in 

Europe on agriculture, health, energy, water 

resources and insurance. “We have more of 

a  transverse than a top-down approach,” 

explains Paul van der Linden.  

Another innovation is the development of 

a new scenario, known as E1, which was tested 

by an ensemble of global climate models. This 

supposes that emission reduction policies are 

a success and that emission targets are achieved. 

This is an opposite approach to the IPCC sce-

narios in the sense that results are shared in 

terms of temperatures in order to then 

of the previous predictions that were based on 

data provided by global climate models.  

The median results of an ENSEMBLES pre-

diction, based on the A1B emissions scenario 

(in which our economies achieve a balance 

between the use of fossil fuels and other ener-

gy sources, including renewable energy), indi-

cate, for example, a temperature rise of 6°C 

and a 50 % fall in rainfall in summer in south-

west France between 2080 and 2099 compared 

with the 1961-1990 reference period. The 

degree of detail achieved by these new mod-

els is one of the project’s major contributions 

to climate change modelling. “We are achiev-

ing a much higher resolution than anything 

obtained previously. This marks major 

progress,” declares Paul van der Linden, the 

ENSEMBLES director who is based at the Had-

ley Centre for Climate Change in Exeter (UK). 

These results are the culmination of five years 

of hard work at the heart of the obscure sci-

ence of climate modelling.  

Passing the history test

Six IPCC scenarios (there are 40 in all) are 

currently being used in the models. These 

describe a series of future emission levels and 

are formulated on the basis of socio-economic 

hypotheses and conjectures on how we man-

age the climate problem. Some are based 

on ‘business as usual’ and are therefore neg-

ative; others are based on more active energy 

policies. They also include estimations for 

the sun’s rays and aerosols and their scope 

varies depending on whether they include 

data relating to 1, 2 or 10 to 15 pollutants, for 

example.  

The models were developed over several 

decades at 12 different centres – the seven 

centres located in Europe were included 

in  ENSEMBLES, which also includes non-

European  institutes among its 66 partners. 

The researchers start by scrutinizing the land 

surface, usually with a resolution of roughly 

three to four degrees. The models are subse-

quently launched on a global scale and the 

values of each bloc or cell reconstituted. Dif-

ferent models are used depending on the 

aim  of the research: they can be uniquely 

atmospheric or oceanic or both. In addition 

to the global characteristics of climate change, 

scientists can also model their impacts.  

models. They can include as many as 30 param-

eters, such as air speed, humidity, ground 

humidity, temperature or dew point. A model 

can consist of as many as a million lines of com-

puter code and take several months to devel-

op, in addition to the even longer analysis time. 

This is why there are only around 25 global 

models in existence.  

In its most optimistic scenario (low green-

house gas emissions), the IPCC predicts, at best, 

a temperature increase of 1.8°C in 2090-2100 

compared with 1990 (the temperature has 

increased by 0.7°C since the beginning of the 

industrial revolution), with a rise in sea levels 

from 18 to 38 cm. The most pessimistic scenar-

io (high emissions) predicts, on the basis of our 

present knowledge, a temperature rise of 4°C 

and a rise in sea level of between 26 and 59 cm. 

Whatever scenario proves to be true, the IPCC 

predicts increased damage by floods and 

storms. A 3°C temperature rise would thus 

cause a 30 % reduction in coastal wetlands. But 

there is no scenario depicting a drowned world 

in 2055…  

A European mega model

Climate models of this kind are improving 

all the time and serve both to refute alarmist 

scenarios and to combat denial, thereby ensur-

ing that the public are informed correctly. 

Measuring and modelling techniques have 

developed progressively over the past two cen-

turies. The very oldest data come from a series 

of temperature readings obtained in central 

England in the 17

th

 century. In the 19

th

 century 

meteorological observations became wide-

spread. In the 1920s balloons were sent up into 

the air full of measuring instruments. Thirty 

years later, aircraft were used to measure the 

atmosphere and weather stations were built at 

the North and South Poles. Today, data are col-

lected by satellite.  

The ENSEMBLES project, allocated funding 

of EUR 15 million under the Sixth Framework 

Programme (FP6), covers a whole series of 

major new climate models. Its predictions have 

a greater degree of certitude, not so much due 

to the precision of the observations as to the 

quality and the depth of the modelling. Clima-

tologists claim to have produced a picture of 

Europe’s climate at the end of this century that 

is clearer and more comprehensive than any