ISSS608 2018-19 T1 Assign Stanley Alexander Dion Task 3
|
|
Help'em Breathe in Sofia
|
|
|
|
|
|
|
Contents
Interaction with local energy source
To understand the relationship between local energy sources and the pollution, we should acknowledge nearby power plants to the city. Not all kind power plants will cause harm to the weather. According to US Energy Information, power plants such as Nuclear, given that it operates normally (without any leaks), won’t emit Carbon Dioxide or any other pollutants. However, fuel-fired plants such as some thermal plants in Bulgaria emits air pollutants heavily. Here is the list of some nearby plants in Sofia:
|
We will annotate the locations in the map to understand the potential correlation of high peaks due to the presence of these plants. |
|
The two closest powerplants, the Iztok and Sofia plant is located close to region Mladost and Nadezhda correspondingly. Coincidentally the two powerplants also close to our first and second sensor clusters defined in the 2nd Task. Third powerplant is located further to the west of Sofia in the nearby province called Pernik. Although the third powerplant seems located further from Sofia, the capacity is the biggest out of three and can potentially disrupt Sofia’s weather. This third powerplant is also located closest to our third clusters we have discussed previously. Since our clusters are located closely to the three powerplants, we created a hypothesis that the activity of these powerplants might have successfully affected local pollution sensors’ readings such that each sensor in the cluster we found are having similar pattern. However, we need further evidence to proof the hypothesis since we don’t have complete information whether the pattern of the sensors' clusters found in Task 2 coincide with the plant’s operation. |
Bringing Topology into View
|
Using TOPOData.csv, we could visualise various locations’ altitude within Sofia City with approximation of 20m horizontal accuracy, 10m vertical accuracy based on NASA SRTM digital elevation model. The locations form a grid of dots that covers the city central and some rural areas. Look that the South-West part of Sofia has higher altitude (darker green) than the rest location point due to the presence of Vitosha and Lyulin mountains (varies around 1000 M ASL). We can interactively highlight the line chart to know exactly the location where there is a sudden drop or rise in altitude. |
Complex Interaction of Topology and Meteorogical Factor
Here is the link to the visualisation: Correlation Dashboard
In order to correlate different meteorological information with the pollution concentration, we are plotting the average parameter estimates for different meteorological factors affecting each of the sensors. The model is calibrated on each of the observation using GWR model. We extended the analysis to also include neighbouring sensors in order to see certain meteorological pattern affecting the city skirt of Sofia.
Suspecting Republica Power Plant
|
Starting from the correlation with wind, we could spot that there is negative relationship happening between the speed of the wind and the PM concentration read majority of the sensor. The lower the wind speed, the higher the concentration. This is particularly the nature of the city where the pollution will only get washed away if there is strong wind coming. The negative relationship is particularly strong at the around the city center (toward north-west region). Interestingly, just beyond the city’s skirt, there are some sensors detecting high pollution whenever there is a stronger wind (positive relationships). We suspect that this due to the cross-border activity of Republica Power Plant, whereby whenever wind is coming, the pollution rises up and goes entering the city from the north-west of Sofia (through the gap of the mountains cut by highway). Once the pollution is trapped in the city, there is a need of high-speed wind to blow the particles away. |
Citizen's Claim toward the Cause
Correlating with temperature, we could see another negative relationship between temperature and the concentration. The lower the temperature, the higher the PM 10 concentration. Interestingly, we could see the strongest negative relationship is happening in the same spot where we have the strongest negative relationship with the wind speed. We could also spot there are certain places where we have positive relationship between temperature and PM10 concentration. This further explains not all pollutions are caused by houses’ burning of fuel due to the cold winter.
The Last Stab
Among the four factors that we bring in, the pressure variable has the least effect on PM 10 concentrations as we can see most of the coefficients in different regions are close to 0. From visualisation, we could observe most of the stations have a slight tendency to increase the pollution concentration. Notice that this is the inverse case on what happened in PM 10 concentration with the temperature, where an increase in temperature will mostly lead to a reduction of PM 10 concentration. The finding aligns with the law of physics: Constant=Pressure x Volume / Temperature , given that the volume or the valley where Sofia lies is a constant.
Lastly, we could observe an interesting insight that only the west-central-east region of Sofia’s PM 10 concentration is negatively affected by humidity. The higher the humidity in these regions, the lesser the pollution. One hypothesis generated here is the humidity can be partially caused by local precipitations, in which the west-central-east region is a trail of cloud movement.
