Why is the content of this map important?
Particulate matter air pollution affects human morbidity and mortality. Changes in natural emissions of desert dust and in dispersion conditions potentially affect the way particulate matter is emitted or transported. It is therefore important to understand the effect of a +2°C warming on ozone concentrations, and how related changes compare with those obtained from emission abatement policies.
Which sectors are affected by this result?
Particulate matter affects human health, visibility and the climate system itself. A higher PM2.5 concentration leads to a higher mortality risk and loss of life expectancy. Some compounds of particulate matter such as soot induce a warming while others induce an albedo effect and a climate cooling (sulphates).
What is shown on the maps?
Maps show the simulated PM2.5 concentration values for (i) the historical period and (ii) the future climate period when a +2°C warming is reached. PM2.5 is the volumic mass of particles smaller than 2.5 micron. Ozone concentrations are in microgram per cubic meter. Maps are calculated from statistics obtained using 4 model suites (CHIMERE, EMEP, MATCH, MOCAGE). For future periods changes in air pollutant emissions is also taken into account, so changes are due both to emission and climate change. By contrast, maps showing the climate change signal for a 2°C wamring only show the effect of climate change, using twin experiments where air pollutant emissions are fixed but climate period is changed.
Details and further information:
Using four suites of global, regional climate, air quality and health impact assessment models we have found that a 2°C climate change modifies the near-surface atmospheric composition of air pollutants in Europe. This is due to several reasons such as changes in weather variables (temperature, precipitation, water vapour, atmospheric flow, boundary layer turbulence) and to biogenic emissions (dust, sea salt, biogenic volatile organic compounds). We have identified areas where changes are robust, where 3 of the 4 models give a change with the same sign. However the changes remain small.
Robert Vautard
Centre national de la recherche scientifique (CNRS-IPSL), France