U.S. Pollution

A Python notebook for analyzing pollution CSV data.

About the Data:

Aims to analyze the change in air quality index (AQI) for 4 major pollutants (Nitrogen Dioxide, Sulphur Dioxide, Carbon Monoxide, Ozone) that cover all 50 states of the United States from 2000 to 2016
Procured from the Environmental Protection Agency (EPA)
Consists of daily pollution data for the 4 major pollutants over 16 years for all 50 states (382 MB, 1746661 lines)
Key points:
- Organized by Date, dates repeat per county where the source data was acquired
- NO2 and SO2 are in parts per billion while O3 and CO are in parts per million
- The max hour describes what hour of the day the AQI was highest

Reading Data:
- Pandas to read the CSV
- Data relatively unorganized¹ but contained diverse information
  - Grouped data by date
  - Calculated the means for each state
Visualization:
- Matplotlib - Graphs
- Cartopy - Maps
- Types of Visualizations:
  - Multi-line charts
  - Bubble maps
  - Multivariate linear regressions
  - Heat maps

Line Charts	Map Visualizations

CO and NO₂ have decreased while O₃ and SO₂ have largely remained stagnant
O₃ and NO₂ are recently the more prevalent pollutants compared to CO and SO₂
When looking at results, be aware of:
- Results may be affected by holes in the data as shown in the multi line charts and the heat maps
- Differentiating between interpreting results on actual AQI index versus the change on various factors
- Analyzing on case-by-case basis per pollutant is necessary to understand trends and how factors such as laws may affect pollution
Awareness of air pollution is one big factor in the slow decline in some pollutants