Extracting Air Quality Insights with Data Exploration and Visualization
Analyzing air quality trends through responses from an air quality multi-sensor device located in a city in Italy
The human need of air is a constant condition until death. Living beings rely on the respiration process; the intake of oxygen fuels our cells, breaks down the food we eat into energy, kills bacteria, etc.
That being said, it goes without saying that the quality of air we breathe is also critical to our health and well-being. Air ridden with pollutants can be detrimental to our health. Upon intake, the air travels through our bloodstream and could expose key internal organs to pollutants. With increased anthropogenic activities both from mobile sources like operating vehicles, buses, and other modes of transportation and stationary sources like power plants, industrial facilities, factories emitting air pollutants like Ozone (at ground level; O3), Nitrogen Oxides (NOx), Carbon Monoxide (CO), Sulfur Dioxide (S02), and Particulate matter (PM10, PM2.5) which can cause long-term impacts on health. The statistics behind air pollution are enough to provide clear insight into the scale of this issue: according to data from the World Health Organization, nine out of ten people around the world are breathing air with high quantities of air pollutants, every year, seven million deaths are caused by exposure to outdoor and indoor air pollution and air pollution is linked to strokes, heart diseases, cancers and both chronic and acute respiratory diseases.
The air we breathe is the fuel to the engine that is our body, but if that fuel is knocking off years from life spans and increasing risk of diseases and health conditions, it becomes inimical to our best interests. Using a data visualization-based approach to draw out trends hourly air quality concentrations of Carbon Monoxide, Sulfur Dioxide, ground level Ozone, etc can mitigate the impact these harmful pollutants have on health by providing us with key insights into how external factors contribute to increases and decreases to air quality in our environment.
In this article, I’ll be drawing out trends through analyzing air quality collected from in an undisclosed city in Italy. Values were collected as responses from a gas multi sensor device deployed at ground level. Prior to data visualizations and creating conclusions, I’ll run through the dataset’s attributes to provide additional context. For the dataset I worked with which you can find here, there were 16 attributes.
Date and time; self-explanatory in dd/mm/yyyy and hh/mm/ss format to represent the date and time of the responses. True hourly concentration CO in mg/m³ (reference analyzer); CO also stands for Carbon Monoxide; an gas that’s difficult to be picked up by the senses due to its lack of colour, taste, and smell. Carbon Monoxide can be found in the Earth’s atmosphere but at very low concentrations and is naturally formed through combustion. Sources of outdoor Carbon Monoxide are cars, trucks, trains, vehicles or other human activities that burn fossil fuels. PT08.S1 (tin oxide) hourly averaged sensor response; tin oxide is created by burning tin metal (nominally CO targeted). 
True hourly averaged overall Non-Metanic HydroCarbons concentration in microg/m³ (reference analyzer); gases such as ethane, ethene, propane, propene, and isoprene are encompassed under NMHCs. Non-Metanic HydroCarbons are naturally released through sources such as fossil carbon depositis, volcanic activity and wildfires, however, anthropogenic activities such as biomass burning and industrial emissions are also causes.  True hourly averaged Benzene concentration in microg/m³ (reference analyzer); Benzene can be found in quantities in outdoor air to due emissions from tobacco smoke, gas stations, motor vehicle exhaust, and industrial activities. 
PT08.S2 (titania) hourly averaged sensor response (nominally NMHC targeted), true hourly averaged NOx concentration in ppb (reference analyzer); NOx or Nitrogen Oxides are a branch of poisonous, highly reactive gases. Nitrogen Oxide gases encourage the formation of smog and acid rain as well as play a role in influencing tropospheric ozone.  Nitrogen Oxide is formed through the reaction of Nitrogen and Oxygen gases and is released from sources such as vehicle emissions.  PT08.S3 (tungsten oxide) hourly averaged sensor response (nominally NOx targeted), True hourly averaged NO2 concentration in microg/m³ (reference analyzer); NO2 (Nitrogen Dioxide) falls under the group of NOx or Nitrogen Oxides but is greater cause for concern due to its aggressive impact on health and the environment. PT08.S4 (tungsten oxide) hourly averaged sensor response (nominally NO2 targeted), PT08.S5 (indium oxide) hourly averaged sensor response (nominally O3 targeted), Temperature in °C, relative humidity (%), and absolute humidity.
With that additional attribute context in mind, it’ll be easier to perform data visualization and extract key insights.
You can find the Kaggle Notebook complete with data visualizations here.