Aixpoll — What you did not know about air pollution in Aachen (North Rhine Westfalia, Germany)
This project was carried out as part of the TechLabs “Digital Shaper Program” in Aachen (Summer Term 2020).
1 PROJECT SUMMARY
Within the framework of the project, data on pollutant concentrations (NO, NO2 and PM10) were analyzed from the measuring station in Wilhelmstraße (VACW/DENW207) in Aachen since 2011 as a function of various meteorological parameters such as temperature or precipitation. It was investigated whether and if so, how different weather conditions affect the pollutant concentrations. The results are integrated into a dashboard.
2 INTRODUCTION
Overall, 92% of the world’s population is exposed to high concentrations of pollutants in the air (UNEP 2020). Air pollution is therefore the greatest health risk facing people worldwide. Especially in megacities such as Beijing, Delhi, or Cairo. In Europe, considerable success has been achieved in recent decades and air quality has been steadily improved. Nevertheless, the limit values are still being exceeded — especially in conurbations (EEA 2017). According to current data from the European Environment Agency, air pollution leads to about 400,000 premature deaths annually (EEA 2020:8). Exceeding the limit values repeatedly leads to disputes in court in Germany. As a result, the State Office for Nature, Environment and Consumer Protection (LANUV) in North Rhine-Westphalia has installed measuring stations which record meteorological data and in addition to various pollutants. Fine dust (PM 2.5 and PM 10) and nitrogen oxides were considered for the analysis. These originate from different anthropogenic sources. Both pollutants are mainly caused by traffic and industrial plants such as power plants or high-temperature combustion. The health effects are manifold. Depending on their size, particulate matter reaches the lungs, can cause heart attacks or strokes, or lead to premature death from other lung diseases. Nitrogen oxides lead in particular to asthma and bronchitis or increase the risk of developing heart disease. The idea of the project was to visualize the existing data and then to find correlations between meteorological data and the pollutant data. Finally, these data are presented in a dashboard. The measuring station in Wilhelmstraße (50.773126° N, 6.09576° E) was used for the air pollution data and the measuring station in Aachen Orsbach (50.798666° N, 6.023861° E) from the German meteorological service for the precipitation data.
3 Methods and Results
For the analysis, the pollutants were compared with the individual meteorological parameters. Figure 1 shows the course of the pollutants over the investigation period from 2011 to 2019. A slight decrease in pollutant levels in Aachen was recorded over the past years. For the pollutants, NO2, and PM10, the limit values for human health are also shown by the red horizontal lines.
3.1 Pollutants and temperature
To assess the concentration of pollutants as a function of temperature, the year was first divided into the four seasons. On this basis, boxplots were created for the individual pollutants (Figure 2 and Figure 3). For NO (Figure 2), the medians are often higher in autumn and winter compared to the other seasons. It is also visible that NO pollution has been decreasing since 2017.
For NO2 (Figure 3) it is visible that the median is often increased in spring. For the first time during the investigation period, the medians were particularly low in 2019.
On the other hand, no patterns can be detected based on the boxplots for PM10F.
3.2 Pollutants and precipitation
In order to investigate the correlation between the air pollution and the precipitation, it was first necessary to merge the air pollution data provided by the measuring station “Aachen Wilhelmstraße” and the precipitation data provided by the weather station “Aachen Orsbach”. Furthermore, three different categories of precipitation were created consisting of “None”, “Low” (< 2mm), and “High” (>= 2mm). The boxplots (Figure 4) show the effect of rain on the air pollutants.
The higher position of many red boxes compared to the blue and green ones as well as the outliers indicate a higher air pollutant concentration when there is no rain, especially for PM10. To further underline this, the mean air pollution (Figure 5) values over the entire time range is calculated for each precipitation category.
The bar chart shows the mean air pollution values on the y-axis and the different air pollutants on the x-axis grouped by the precipitation categories, each of the pollutants having the highest bar for the category “None”. Again, the clearest negative correlation between precipitation and air pollution can be seen for PM10.
3.3 Pollutants and wind speed
To further evaluate the distribution of the pollutant rates regarding wind speed, a scatter density plot (Figure 6) is drawn. In the graphic, the lowest to the highest density is characterized by the darkest to the brightest color. As the brightest part of the diagram appears on the bottom left, it is observed that the most occurring case is relatively low pollutant rates at low wind speeds of < 2 m/s. The diagrams are close to a triangle-shape, where high pollutant concentrations are found at low wind speed and the pollutant concentrations are decreasing as the wind gets stronger. It indicates that wind carries the pollutants away and disperse them over a wider area, which causes lower concentration.
For further observation, the wind speed is classified into 5 groups based on Beaufort Scale: Level 0 (< 0.5 m/s), level 1 (0.5–1.5 m/s), level 2 (1.6–3.3. m/s), level 3 (3.4–5.5. m/s), and level 4 (5.5–7.9 m/s). The bar chart below (Figure 7) depicts the mean value of each pollutant based on the mentioned classification. As can be seen, high wind speeds of level 3 including 4 have lower mean value in comparison to levels 0, 1, and 3. This validates the dispersion of air pollutants in windy situations and the accumulation of air pollutants in less windy situations.
3.4 Pollutants and humidity
The air humidity is categorized into four groups “Low” (0–50%), “Medium” (50–70%), “High” (70–85%), and “Very high” (85–100%) and the following box plots (Figure 8) show how the air pollutants behave depending on the different humidity categories.
It seems that there is a significant negative correlation between the amount of NO2 in the air and the air humidity as the red boxes representing low humidity are always positioned highest. For NO and PM10 no clear trend can be observed from this figure.
The following bar chart (Figure 9) shows the mean air pollution values on the y-axis and the different air pollutants on the x-axis grouped by the four humidity categories.
Interestingly, the mean NO values increase with higher air humidity. It is the other way round for the mean NO2 values which was already shown in the box plots. The mean PM10 value is almost constant in overall humidity categories.
3.5 Pollutants and wind direction
To further analyze the effects of the wind direction on the pollutant concentration, first, the detailed degree measurements for the wind direction were grouped into North (315°-45°), East (45°-135°), South (135°-225°), and West (225°-315°). After grouping the measurements, it became apparent that for colder months the wind tends to blow more often from the east compared to the warmer months. Other than that, no significant differences in wind direction between months could be identified. It should be noted that no wind coming from the north was detected.
Looking at the statistics for the pollution values (Figure 10) it can be seen that there is a clear trend for a higher pollution concentration when the wind is blowing from the east. Especially for NO and NO2 this correlation is very pronounced. For PM10F there is also an increased
concentration when the wind is coming from the west. For a south wind, the pollutant concentration is always the lowest.
Looking at the surroundings of the measurement station it can be seen that south of the station there is a rather large area that is mostly populated by trees. Furthermore, in the eastern region next to the station many industrial regions can be found. Therefore, the lower values for the pollutants for a south wind might be due to the large area of trees the wind passes through. However, no clear statement based on one station can be made, since the effects of the distance on a macroscopic level as well as the direction of the street (channel effect due to high buildings on the sides of the street) can be made. No immediate assumptions for higher values of PM10F for west winds were identified.
4 Summary
In conclusion, an overview of the most interesting findings of our data analysis is presented.
Regarding temperature, it should be noted that NO concentrations are higher in autumn and winter. On the other hand, the NO2 concentration is higher more often in spring. A significant negative correlation between precipitation and PM10 can be found whereas the correlation between precipitation and the other pollutants NO and NO2 is less clear but also slightly negative. This is in line with the intuition that rain “washes” air pollution away. According to the result obtained from the analysis of wind speed, the higher the wind speed, the more the pollutants are dispersed so that its concentration is lower. Furthermore, the analysis has shown significantly lower NO2 concentrations for higher air humidity but rather negligible dependence between the other pollutants and air humidity. To sum up the findings from the analysis of wind direction, the data shows a clear dependence of the wind direction and the pollutant concentration. In general, pollutant concentration is lowest for south winds. There is a slight tendency towards more south than east winds during cold months.
The dashboard can be found under: https://aixpoll.pythonanywhere.com
Team:
Patrick Bi https://www.linkedin.com/in/patrick-bi-1b7486160/
Azka Firdaus http://www.linkedin.com/in/azka-firdaus
Martin Pillich
Hendrik Hamacher www.linkedin.com/in/hendrik-hamacher (Data Science Track)
Mentor: Tim Sandermann https://www.linkedin.com/in/tim-sandermann
The project’s repository can be found here: https://github.com/TechLabs-Aachen-e-V/SoSe20_Team6_Main
5 References
United Nations Environmental Programme (2020): Air. <https://www.unenvironment.org/explore-topics/air> abgerufen am 06.09.2020.
European Environmental Agency (2017): Luftverschmutzung. <https://www.eea.europa.eu/de/themes/air/intro> abgerufen am 06.09.2020.
Landesamt für Natur, Umwelt und Verbraucherschutz Nordrhein-Westfalen (LANUV) (2018): Bewertungsmaßstäbe wichtiger Luftschadstoffe für Genehmigungsverfahren. <https://www.lanuv.nrw.de/umwelt/umweltmedizin/genehmigungsverfahren-uvp/bewertungsmassstaebe-wichtiger-luftschadstoffe> abgerufen am 06.09.2020
TechLabs Aachen e.V. reserves the right not to be responsible for the topicality, correctness, completeness or quality of the information provided. All references are made to the best of the authors’ knowledge and belief. If, contrary to expectation, a violation of copyright law should occur, please contact aachen@techlabs.org so that the corresponding item can be removed.
