ISSS608 2018-19 T1 Assign Choo Mei Xuan task2
Air Quality in Sofia
|
|
|
|
|
|
Task 2
Spatio-temporal Analysis of Citizen Science Air Quality Measurements Using appropriate data visualisation, you are required will be asked to answer the following types of questions:
Part 1-
Characterize the sensors’ coverage, performance and operation. Are they well distributed over the entire city? Are they all working properly at all times? Can you detect any unexpected behaviors of the sensors through analyzing the readings they capture? Limit your response to no more than 4 images and 600 words.
Figure 1: Distribution of sensors
Citizen Science sensors are distributed across Bulgaria hence only the sensors located within Sofia are selected. It is observed that the sensors are mostly clustered in the centre of the Sofia city, with a few of the sensors spread out to the city borders These sensors are installed by the public outside their home to generate a continuously updated particulate matter map. [1] They collect P1 which is PM10, P2 which is PM2.5, temperate, humidity and pressure. [2] [3]
Hence, it is inferred that the residential areas are concentrated in the centre of Sofia city.
Figure 2: Boxplot of air measurements in Sofia city
Observations:
- There are multiple especially high readings in January 2018 for median level of P1.
- Likewise, this was observed for median level of P2.
- There are multiple especially low and high readings in end February 2018 to Early March 2018 for median level of temperature.
- There are multiple especially low readings in April 2018 and June 2018 for median level of pressure.
With a 3 standard deviation range set, we can also identify the unexpected readings of the sensors and exclude them.
Figure 3: Median concentration of P1 (left) and P2 (right) over time
Using figure 3, it was observed that the sensors are not working during the 2nd half of August 2017 and the first 5 days of September 2018. It was also observed there are higher readings of P1 and P2 during the winter months from November to March. This could possibly be due to the increased in energy consumption during the winter months which has caused an increase in air pollutants of P1 and P2. Similarly, it was observed that there are higher readings of P1 and P2 starting in the evening hours of 1700 through the night to 0800 in the morning. It is especially so during 0300 to 0700 and 1900 to 2300. This could possibly be due to the release of P1 and P2 from the surrounding industrial areas, as well as the increased burning of fossil fuels by the households during the winter months at night.
Part 2-
Now turn your attention to the air pollution measurements themselves. Which part of the city shows relatively higher readings than others? Are these differences time dependent? Limit your response to no more than 6 images and 800 words.
Figure 4. Spatial point patterns
From the 2 charts above, it was observed that the right side of Sofia experienced higher P1 and P2 readings than others. This could be due to the nearby streets and highway where air pollutants are produced by vehicle emissions.
Figure 5. Calendar heatmap and line chart for P1
It was observed that there is difference in the median concentration of P1 over time.
Across the months, it was observed that the winter months from November to January are higher, especially towards the end of the months. And over the other months from February to October, the P1 readings are kept relatively constant.
Likewise, it is observed that the night hours from 1700 to 0700 are higher. However, there are no readings recorded from mid Aug to 5th of September.
Figure 6. Calendar heatmap and line chart for P2
Similar patterns have been observed for P2. However, the readings for P2 are lower than that of P1.
Figure 7. Density heatmap and Correlation between P1 and P2 with air measurements (left: January 2018, Right: November and December 2017)
Using this interactive dashboard would allow users to find out the correlation between the pollutants and air measurements. Noticeably, it was observed that there is no correlation between them.
Additionally, users can look into the specific months of the year to see the concentration of the median P1 and P2 at the different part of Sofia. In comparing the density heatmap of P1 and P2 during January 2018 with November to December 2017, the following observations can be made:
- Both comparison period is during the winter months.
- However, in January 2018 the colour density appears to concentrate in the centre of Sofia city while the colour density appears to be more spread out.
