Outdoor air pollution data shows how a city breathes and lives. As the day starts, cars start filtering in towards the city center as people start commuting to work. The public transportation system starts revving up to meet the trickle of passengers in the early hours of the morning which slowly ramps up as the city eases itself into another beautiful day.
The weather of each individual day partially dictates how people travel in certain cities. Like Copenhagen, for example, on a rainy day, we might see slightly reduced bike traffic as commuters switch to the public transportation and carpooling. In addition to these environmental dynamics of precipitation, wind, temperature, and humidity, all of these citizen dynamics also have a high influence on a city’s pollution profile.
Based on data from multiple cities where our CPHSense devices are installed, we can see how different cities breathe and how the pollution varies over the day. Let’s try to look at an example!
The graphs above show the distribution of air quality levels in two different cities over individual hours of the day for the month of August 2017. The two cities show completely different conditions across the day. City B is on average less polluted than City A. City A shows pollution peaking in the morning and the evening while City B shows a consistent pollution profile throughout a day.
The differences we see in these two pollution profiles hints that the two cities have very different citizen dynamics. But is that the only difference portrayed in the graph? The sizes of the cities being compared are integral to the study. City B is actually smaller than City A. This difference in size is seen in the pollution levels in the city. To see how it translates into daily values let us take another example.
The graph below shows a comparison of how the pollution values are distributed across the morning half of a single day. In the mornings, there is a much wider spread of the pollution values registered in both cases. The afternoons are much calmer with pollution values quite dense with the spread well controlled. And as the heatmaps above suggest, City B does have much lower pollution levels compared to City A, but the daily trends are the same.
The above data, generated from a cumulative dataset combines together the pollution levels on weekdays and weekends. Different explanations can be given for the extent to which the range of values are registered over the morning and the afternoon. It gives us a general idea that the city wakes up at different times over the course of a week but goes back to the cave at around the same time irrespective of the weekday or weekend.
The question can now change to see whether there are differences in the mornings during weekdays, versus the mornings during weekends. Intuition suggests that there is bound to be a change and this should reflect the values we see in the dataset.
Our intuition suggests that the weekend should have lowered pollution levels relative to the weekdays. The data shows that the weekends do have lowered pollution levels compared to the weekdays. The cumulative effect of the data averaged all the values across the days reflecting a wider range of values registered.
Similar trend changes can be seen in both cities during the evenings. For City B the weekends have a similar profile as the weekdays during the evenings. The hypothesis is that the size of the city, being considerably smaller, lends itself to having a closer to normal working day trend as we approach weekends. How so? Citizens who travel closer to the city to do their shopping or other weekend purposes would contribute to those localized phenomena. This hypothesis, of course, needs proof, but it does provide a framework to test different things in the future.
The way a city breathes is fundamentally different from city to city. Culture, environment, and a lot of local factors contribute to the pollution metric assigned by regulations. As citizens, we should be aware of these metrics and trials to see how well these values change if something changes in the environment. With the dataset as it stands, without context, hypotheses can be generated. These hypotheses can empower the regulators and citizens to not only identify the source of these changes but also implement changes that can positively impact the quality of life in the city.
However, we would like to end with a question to you. You live in a city that breathes in a manner similar to these cities. Maybe it breathes worse, maybe better. If you had the power, to assist with this knowledge, how would you try to influence how the city targets the dynamics of air in your city?