Author: Marcin Szwagrzyk
- There is a positive impact of green areas on the air quality
- 1% more of urban forest area in the surrounding of a sensor was accompanied by decrease of pollution by 0,1 to 0,5 µg/m³ PM2.5
- There is a positive relationship between the share of forests in sensor’s neighbourhood and the quality of the air measured at this sensor.
- Different types of urban green areas can be responsible for up to 50% of pollution variation within the city.
It is widely known and it has been proven many times that urban green areas perform many key environmental services for urban residents. These services significantly improve the quality of life in cities, for example they reduce surface heating, which lower the air temperature, and they reduce the surface flow, which mitigates the risk of flooding. These services seem to be important especially nowadays – in the times of climate change and rapid urban development. Urban development was particularly strong in recent years in Cracow. The population and economy have increased rapidly, which caused a huge development of settlements in the city and its surroundings. This led to a decrease of the green areas in the city.
But the urban green poses one more exceptional property. According to many research conducted across the globe, there is a positive impact of green areas on the air quality. It is happening mostly because of the fact that plants acts as natural filters which absorbs pollutants from the air.
However studies about green areas and air pollution in cities were usually the general estimates of pollution remediation, showing the overview amount of pollution that can be mitigate in each city. Such analysis were conducted for example, in Florence, Berlin, Munich and Rome. Results of those studies showed that role of urban forest in remediating of air pollution depends on many factors and therefore it can be – comparing to the pollution level in a city – modest or substantial.
Other studies were based on comparing of urban forest areas shares and the pollution levels at particular points, but due to the lack of spatially dense data about the air pollution, such studies were limited to singular air monitoring stations, without showing of the spatial variability of pollution within particular cities. Such studies were conducted for example in Beijing or in peri-urban area of South India and proved high levels of correlation between share of forests and the air quality.
As one can notice that above examples of scientific studies had limited utility both for media and for the public, because they are hard to understand and their results are not straightforward. But thanks to dense network of affordable sensors with known and adequate quality, like the Airly sensors, what was impossible up to this day, can now be done quickly and not expensive in every city around the globe.
As an example we examined if and what is the impact of urban green on the air quality, and checked how the air quality can vary within a city, regarding the distribution of urban green areas. We performed analysis for the year 2017 with the use of our exceptional air quality database consisting of a dense time series of air pollutants from 56 sensors located in the second biggest and one most polluted cities in Poland, Cracow.
For decades Cracow residents have been living with the threat of air pollution. The exceeded levels of particulate matter concentrations, sometimes even 10 times and more, did not attract any interest or government actions.
It lasted till the 2010.’s, when the Cracow Smog Alert – a social initiative for battling the air pollution was founded. People in Cracow started to realize that the problem of air pollution is extremely serious and dangerous for their lives. However, in those times no data about pollution were available and Cracow residents could only judge the air quality based on its smell and color.
In 2017, Airly company began to build a network of sensors, providing citizens with accurate and actual data about the air pollution. This was the year with very serious air pollution, with almost 140 days that exceeded WHO norms of Particulate Matter 2.5 concentration. Since then, many things have changed in Cracow, especially in terms of the social perception of the air pollution problem. It was followed by administrative decisions, such as free public transport on the days of highly exceeded pollution levels, or the ban on coal and wood burning in the city. The latter were popular fuels for heating individual houses, being the main source of the air pollution in the city.
Data about the location and type of the urban green areas were obtained from the Sentinel-2 satellite, which is the European Space Agency satellite program providing accurate and up-to-date images of almost the whole planet every couple of days. For the sake of our analyses, we classified the satellite images into three categories of urban green areas: trees (forests and parks), bushes and low vegetation (meadows). Map of those areas is presented below (Figure 1).
Subsequently, we counted the area of urban green areas in the buffer zones around each sensor (circles with 100 meters radius).
Processed data were then analysed, by statistically examining each day separately and checking which parameters of urban landscape, including share of urban green areas and also other geographical features, such as elevations, contributed most to the air quality.
For the simplicity, in the next sections of this paper, the sensors with a high share of green areas are called the ‘green sensors’ and the sensors with the lowest shares of green areas in their surrounding will be called the ‘urban sensors’.
We found out that for a vast majority of days (over 300) there was a positive relationship between the share of forests in sensor’s neighbourhood and the quality of the air measured at this sensor. This relation was strongest during days with moderate and high air pollution, but it weakened in the days of the extreme pollution. In case of the low green areas, similar positive impact was observed, but it was limited to the warmer months, with moderate pollution only.
On the graph below (Figure 2) mean values for groups of networks are presented, for the set of the most green and the set of the least green sensors in the city, for the exemplary period of two weeks in 2017.
Figure 2: PM 2.5 concentration for groups of sensors
In this graph, the lower level of PM concentration at green sensors is clearly visible. Moreover, for some days the PM concentration exceeded the norms for urban sensors and did not exceeded them for the green ones!
Location of those 20 selected sensors is also presented on the map (Figure 1).
The animation below shows the distribution of pollution in Cracow, regarding the urban forests, for two exemplary days. Higher pollution is marked in darken colors, and the urban forest areas are marked with green diagonal lines. The animation shows how diversified in both space and time was the pollution within the city.
Statistical analyses also showed, that share of different types of urban green areas can be responsible for up to 50% of pollution variation within the city. The highest pollution variation were explained by green area shares during the days of low and moderate air pollution.
Results of our statistical analysis may be translated for concrete numbers: 1% more of urban forest area in the surrounding of a sensor was accompanied by decrease of pollution by 0,1 to 0,5 µg/m³ PM2.5.
In relative quantities, the 1% more of urban forest were accompanied by a decrease of the pollution by 0.3%, comparing to the average pollution in the city.
The study showed that urban green areas may play an important role in battling the air pollution. Statistical computations showed that even slight changes in urban green extent – like for example cutting trees for making place for investment, may have a negative, tangible impact on the air quality.
Due to the availability of spatially and temporally dense data from Airly sensors, as well due to accessibility of satellite imagery, those processes and phenomena may be tracked in almost a real-time, allowing evaluation of spatial planning policies in cities.