Annual heat-related (left) and cold-related (right) mortality induced by climate change per 100 000 persons for a high-emissions scenario (IPCC A2) for the period 2071-2100 relative to the period 1961-1990, map elaboration by EC JRC/IES.
(Figures 9 in the Green Paper on Adapation)
Data from EC JRC project PESETA. All climate data for PESETA originate from the EU funded research project Prudence
How the results have been computed
Human health can be affected by climate conditions. In particular, high ambient temperatures can be associated with more mortality. On the other hand, cold ambient temperatures are also associated with mortality. These heat and cold related effects have been assessed in previous epidemiological studies (linking mortality with average temperature, at a resolution below country level), and these provide quantitative relationships (exposure-response functions). These have been combined with estimates of the future modelled climate, along with estimated changes in the population (size and age), to predict the likely consequences of climate change. The models show that climate change will increase average temperatures (compared to the historical climate of the 1960-1990 period), and so will lead to increased incidence of heat related mortality. However, these rises in temperature will also reduce winter excess deaths, i.e. they will lead to benefits as well as damages.
The mortality impacts will also be affected by acclimatisation and adaptation. Acclimatisation includes the physiological (and behavioural) change that is likely to take place in humans in adjusting to a future climate, especially over longer periods of time. Adaptation involves actions taken in a planned and proactive way to address climate change (e.g. through heat-alert systems). The following results do not include acclimatisation or adaptation.
Limitations of the approach
There are many sources of uncertainty that should be considered when interpreting the results. First of all, there are uncertainties in predicting the climate in future periods, and, although results based on only one model are shown here, the study has used several models to investigate this. Different models show large differences in predicted effects. There are also uncertainties related to the projections of future population size and age distribution. Larger uncertainties exist over the exact relationship between climate and mortality (both heat and cold) and how these vary with geographical conditions, and how they will change in the future with acclimatisation, socio-economics (e.g. the evolution of income), and adaptation.
The maps show only the results for one family of exposure-response functions. Alternative functions have been used in the underlying study and these show significantly different levels of impacts and different geographical distributions.
The maps above do not include the effects of acclimatisation or adaptation. Strictly speaking, it is incorrect to simply apply the temperature-mortality relationships from today’s climate to future climates, as this is likely to overestimate impacts. Analysis in the PESETA projects shows that acclimatisation could reduce effects significantly (for heat- and cold-related mortality).
There are also other health effects from climate change that should be considered alongside heat and cold related mortality. The effects of heat waves, which require looking at the daily extreme temperatures, are not included in the analysis. Other effects include: cold spells, heat and cold related morbidity (illness); food, water and vector borne disease; and changes in the risk of accidents and wider well being from extreme events (e.g. storms and floods).
How to interpret the maps
The maps show one set of results for the change in heat and cold temperature related mortality across Europe from climate change in the period 2071 – 2100, compared to the historic modelled climate (1961-1990). They are presented as the change in the number of deaths per 100 000 people.
They show the changes in heat related mortality (on the left in red) and cold related mortality (on the right in blue). The pattern of changing heat and cold related mortality largely matches the predicted change in temperature from the climate models. These show increased heat related mortality in Southern Europe (where the climate signal is strongest), and reduced cold related mortality in Northern and Eastern Europe.
Note that the increase in heat related mortality on the east coast of the Baltic (which appears as a dark red line) is thought to be due to a bias in the climate model used.
It is highlighted that an alternative analysis was undertaken in the PESETA study, using a different set of functions based on a limited number of individual country studies, each with a country-specific function and threshold. These show a greater rate of change in heat related mortality in Central Eastern Europe than shown here (though there is still a high rate of change in Southern Europe). Interestingly they show a very different pattern for cold related mortality, with greater reductions predicted in Southern Europe and less change in Northern Europe. While this may appear counter-intuitive at first, it can be explained by the existing climatic adaptation in Europe, i.e. those in the Southern countries are currently less adapted to the cold. These country studies bring more local specific aspects to the analysis, but the small number of them, and the fact that they are not part of a consistent wider study, makes it problematic to apply and interpret them at a European scale. An emerging EC funded study producing harmonised relationships is due to be published shortly (the PHEWE project).
Overall, the net balance of heat and cold related mortality is complex and is determined by the assumptions and approach used. Taking Europe as a whole, the benefits of reduced cold-related mortality projected for winter are at least as large, and in most analysis runs, larger, than the negative impacts of higher heat-related mortality projected for summer (though this is not always the case in all runs). It is clear that the net effects predicted and the spatial distribution across Europe varies according to the quantification method and the assumptions about acclimatisation. Because of this, the results presented here should only be considered as an initial, interim assessment until better information becomes available and some parts of the methods are elaborated in more detail.