Conclusion

Using the data visualizations one should be able to answer this project's research questions:

1. How did air quality vary in different cities in Europe in 2018?
In terms of temporal variation, most of the cities show similar patterns and trends (not magnitudes though), not only on annual scale but on a weekly and daily scale as well. However, in terms of spatial variation, there are immense differences. There seems to be a divide between Western and Eastern Europe. In Western Europe, efforts to reduce air pollution have led to an improvement of the situation over the last decades. Unlike in Western Europe, air pollution is a much bigger issue in Eastern Europe. Lower standards in industry, older vehicles as well as more common wood and coal heating lead to a higher air pollution in Eastern Europe. However, there are not only Eastern European regions suffering from bad air quality. For instance, due to its enclosed nature the Po Valley in Italy is prone to air pollution as well (look at Milan).

2. What conclusions can be drawn from the relationship between air pollution and weather data?
Numerous correlations between pollutant concentrations and weather characteristics can be found, indicating a strong connection between weather an air quality. For instance, ozone concentrations are usually higher when air temperature and sunshine duration are high. This matches findings of higher mortality due to ozone during heat waves (Filleul et al. 2006). As heat waves will probably occur more frequently in the future, ozone is likely to become a bigger issue for urban health.
Higher wind speeds seem to result in lower NO₂-concentrations but this is probably due to more mixing of the polluted local air with fresh, unpolluted air from higher up or rural areas.
Another correlation is between PM₁₀ and precipitation, as rainfall washes particles from the air. This effect could be observed in scientific studies as well (Ouyang et al. 2015). Consequently, if this washing effect is absent during droughts, PM₁₀-concentrations can rise to very high levels. This can lead to significant health issues for inhabitants of urban regions.

Further interesting findings

• Events with big fireworks are clearly visible on the map. For instance, have a look at New Year's Eve or France's national day (14. July), when PM₁₀-concentrations rise to extremely high levels.
• Comparing the NO₂-concentration to the week-day reveals the significantly lower amount of traffic during week-ends compared to working days. This shows that NO₂ is mostly emitted by traffic. Additionally, the PM₁₀-concentrations vary according to the week-day as well, which was already looked at in a case study in Switzerland (Barmet et al. 2009).
• Plotting the concentrations of O₃ and NO₂ against each other nicely shows the interdependency of the two pollutants. As ozone formation reduces the NO₂-concentration, it is basically impossible to have high concentrations for both pollutants at the same time. Usually NO₂ reaches its maximum concentration in the morning and during colder seasons, while O₃-concetration is very high in the afternoon and during summer.