Clearing the air: Air quality modelling for policy support

The origin of particulate matter (PM) concentrations in the Netherlands is established using the LOTOS-EUROS model with a source attribution module. Emissions from the ten main economic sectors (SNAP1) were tracked, separating Dutch and foreign sources. Of the modelled PM10 in the Netherlands, about 25% comes from natural sources and 75% is of anthropogenic origin. Dutch sources contribute one third to the anthropogenic part, with the highest contributions from transport and agriculture. During episodes, the contribution of foreign sources becomes more important. This source attribution is based on model results only, and LOTOS-EUROS captures only about 60% of the measured concentrations. This is mainly caused by a strong underestimation of the concentration of organic carbon particles. Reducing the missing mass is an important step to improve the source attribution of PM. Source attribution of PM has also been performed for Flanders, with a focus on agriculture. Flanders has a large agricultural sector and therefore considerable ammonia emissions, which contribute to secondary inorganic aerosol formation. Manure transport data was used to estimate the temporal variability of ammonia emission from manure application in the LOTOS-EUROS model. This approach strongly improves the model performance for ammonia, but does not affect secondary inorganic aerosol. A scenario study showed that restricting manure spreading shortly before and during a period with high particulate matter concentrations is not an effective way to reduce particulate matter concentrations during these episodes. Two papers in this thesis investigate possible consequences of energy transitions for air quality. During the energy transition from fossil fuels to renewable alternatives there is likely a phase in which fossil fuel energy production will be mainly used to provide back-up capacity. This causes a change in the temporal emission variability from power plants. The effect of this projected shift in emission timing on air pollutant concentrations is assessed. For all PM components considered, air concentrations in a model experiment accounting for the change in emission timing were higher than expected based on the emission change. As back-up capacity, power plants are more likely to operate during stagnant weather conditions and during nighttime in this transition phase, conditions in which air pollution is transport and dilution is ineffective. This work illustrates that emission timing is important to look into when assessing the impacts of system changes on air quality. The second scenario study focuses on ozone. When bioenergy use in increased, more production area is needed causing a change in land use. In this work the effects of land use change as well as changes in anthropogenic emissions are taken into account using realistic energy and air quality scenarios for Europe. The increase in ozone damage caused by increased production of biomass is only marginal compared to the reduction in ozone damage because of reduced ozone precursor emissions in other sectors. The total effect is a significant reduction in health damage due to ozone, especially in southern Europe. However, taking into account the effects of projected climate change, ozone damage towards 2050 might actually increase