results

Change in mean annual temperature and precipitation by the end of the century

About these maps

Climate data for PESETA originate from the EU funded research project PRUDENCE.
The early results from the PESETA project presented here are based on projections about temperature changes and precipitation changes. These two climate variables are represented in the maps for the 2071-2100 period, relative to the 1961-1990 period.
The change in mean annual temperature (Map 1) shows a general warming of several degrees by the end of the century. In most places the temperature change would exceed the average global warming expected for the same scenario (3.1 degrees). The Southern European heating occurs mostly during summer, whereas the large heating in the northeast generally occurs during winter.
Mean annual precipitation (Map 2) is expected to increase to the north and to decrease to the south. In winter most of Europe has an increase, largest in the north; during summer most of Europe experiences less precipitation, again with the largest drying in the south.

Map 1. Change in mean annual temperature by the end of this century
(figure 1 in the Green Paper about Adaptation)
Absolute change in mean annual temperature between control period 1961-1990 and 2071-2100, under the IPCC SRES scenario A2. Data from EC-funded project Prudence (HadCM3 global circulation model, and HIRHAM regional climate model in 12km resolution), map elaboration by EC JRC/IES.

 

How these variables computed and what are the sources of uncertainties

There are many uncertainties in climate model results. There is the uncertainty in the choice of emission scenario, the uncertainty coming from the differences between global climate models as well as between regional climate models, and finally there is the fact that the natural variability of weather does not permit exact determination of climatological means through simulations of a limited time period. On the other hand, many qualitative and quantitative climate change results are robust towards the choice of models and emission scenario, and so can be largely trusted due to this.

In order to get the detailed maps shown here, output of a global simulation including atmosphere, ocean and sea ice has been downscaled. This means that a regional atmospheric model in high spatial resolution has been set up, covering an area only slightly larger than the area of the maps. Sea temperatures inside the area as well as wind, temperature and pressure on the edges of this limited area have been prescribed by the global model.

This simulation is only one out of many possible projections of the future climate. Details of the results change when different models are used, and of course the general magnitude of climate change depends on the emission scenario in use. The simulations shown here employ the rather strong IPCC scenario SRES A2. However, the qualitative features of European climate change outlined above are seen in most climate projections.

Global and regional models

A regional climate model can describe the climate in much higher geographical detail than a global climate model can. On the other hand, it only covers part of the globe, and must therefore have a global model as parent, which prescribes the weather on the boundaries of the regional simulation area. In general, large-scale circulation is determined by the global model, and local weather phenomena by the regional model.

The effect of using different regional and global models is shown in the next figures for respectively winter and summer. More results can be found on the Prudence web page.

This figure shows temperature change in winter from four different simulations. The lower-left map shows the expected temperature changes in the short term (2011-2040). The other three maps show longer-term changes (2071-2100). The upper row uses two different regional models that downscale the same global model (i.e. with the same sea surface temperature and lateral boundaries), whereas the right column uses the same regional climate model, but different global models (thus different boundaries).

The regional model has a rather small influence in this case, due to the fact that temperature during winter is generally determined by large-scale weather systems and thus generally prescribed by the global model. The pattern that the largest warming is seen in continental climate and in the north is seen in all simulations. The short-term projection exhibits a spatial pattern very similar to the longer-term projections.

Change in winter temperature (°C) for different global and regional models for two periods (2011-2040 and 2071-2100) Source : PRUDENCE project

HIRHAM/HadAM3H 2071-2100

RCAO/HadAM3H 2071-2100

 

RCA3/ECHAM5 2011-2040

RCAO/ECHAM4 2071-2100

 

In the same comparison for summer, we also see a large effect of the boundary conditions, but the regional model differences are much clearer. Nonetheless all models agree that the largest warming will take place in Southern Europe. This warming is significantly larger than the global average warming.

Change in summer temperature (°C) for different global and regional models for two periods (2011-2040 and 2071-2100) Source : PRUDENCE project

HIRHAM/HadAM3H 2071-2100

RCAO/HadAM3H 2071-2100

RCA3/ECHAM5 2011-2040

RCAO/ECHAM4 2071-2100

 

This inter-model variability is important to keep in mind when interpreting results about the various impacts of climate change. The preliminary PESETA results presented at this stage are derived from single models and emission scenarios.
The project actually considers a combination of global and regional models coupled with different emission scenarios (see http://peseta.jrc.es/scenarios.html). This will be reflected in the final results.