Data: The  study uses the official pollution and weather data from the RK Puram pollution monitoring station of Delhi Pollution Control Committee, an autonomous agency of the Government of National Capital Territory of Delhi, India. The monitoring station is the only one located in South Delhi which has a population of more than 2.7 million people with approximate population density of 28,000 people per square mile. Landuse is mostly residential and institutional.  

The dataset includes hourly concentrations of ozone, NOx and PM 2.5 (particulate matter of size 2.5 micrometer or lesser) and hourly measurement of ambient temperature, solar radiation, wind speed and relative humidity from April 2015 to November 2016. The data was downloaded from the website of the Central Pollution Control Board (CPCB) of India (equivalent of the US Environmental Protection Agency). 

Data wrangling:  CPCB website allows to download only excel files containing monitored data for one parameter and three months at a time. Therefore, the collected data is spread across 49 excel files. All the information in the excel files were collated, cleaned, merged  and written onto a CSV file for this study (Click on the figure 3 to access them). The original data had dd/mm/yyyy date-time format which was converted to a format compatible with Pandas plotting package.