Siemens: Detect air quality violations using rolling mean algorithms and anomaly detection. Python, Pandas, Matplotlib, Scipy.
Analyze and compare the air pollution data from 2 stations: one in New York City (US) and one in Bogota (Columbia). Extract useful insights and include visualizations that support your findings.
You can start by analyzing the data from each station individually. Perform a time series analysis on the pollution levels. When is pollution the highest and the lowest? Can you see any trends or seasonal patterns?
Next, you may compare the PM2.5 pollution in both cities. Using a visualization, show at what times is the pollution level in New York higher than in Bogota. Can you see any trends or seasonal patterns? Are the two time series correlated?
You are also encouraged to explore the pollution limits set by the World Health Organization (WHO), for example, based on this document. Is the annual mean pollution level in the two cities below the acceptable value defined by the WHO? Write an algorithm that detects the time when the 24-hour mean pollution level exceeds the acceptable value and use a visualization to demark these periods.
Above are just suggestions of what you can extract from these data. Still, after analyzing the datasets, feel free to use your imagination and draw additional insightful conclusions!
The data are given in the txt format and using | as the column separator. There are two files, one for each station, as described below.
StationData-NY_QueensCollege.txt
StationData-Bogota_SanCristobal.txt
Analyze the provided data and answer the questions to the best of your abilities. Include the relevant tables/graphs/visualization to explain and support your findings. Make sure that the solution reflects your entire thought process including the preparation of data - it is more important how the code is structured rather than just the final result or plot. You are expected to spend no more than 1-3 hours on this project.