Urban Air Pollution in Lagos – SQL Project

This project looks at simulated air quality data from a few urban areas in Lagos, Nigeria. I used SQLite to clean the data and ran my analysis in Jupyter Notebook using ipython-sql. It’s part of my process of learning SQL while building a portfolio around environmental data.

Project Overview

Goal: Clean and analyze air quality data to explore pollution trends in Lagos. Tools Used:

• SQLite (for data cleaning)
• Jupyter Notebook + ipython-sql (for analysis)

Dataset Description

The dataset includes 60+ records collected over 10 days from 6 Lagos locations.

Column	Description
Date	Day of measurement
Location	Lagos area (e.g., Ikeja, Lekki)
PM2.5 (µg/m³)	Fine particulate matter — a harmful pollutant linked to respiratory issues
CO (ppm)	Carbon monoxide — a gas from cars, generators, etc.
Temperature (°C)	Affects how pollutants behave in the air
Wind Speed (km/h)	Influences how pollution spreads or stagnates

This data is fictional but modeled on realistic environmental behavior.

Data Cleaning (SQLite)

See: cleaning_script.sql

Key steps:

• Fixed inconsistent casing in Location
• Handled missing CO values
• Removed duplicate rows using rowid
• Removed outliers in pollutant, temperature, and wind speed values

Output file: cleaned_air_pollution_lagos.csv

Data Analysis (Jupyter + ipython-sql)

Questions explored:

• Which location had the highest average PM2.5?
• Which locations consistently stay below the WHO PM2.5 limit (15 µg/m³)?
• What’s the average CO level across all locations?
• Do higher temperatures correlate with higher PM2.5 levels?
• Rank all locations by average pollution level

See: air_pollution_analysis.ipynb

Only SQL concepts used:

• SELECT, GROUP BY, ORDER BY
• CASE statements
• CTEs (Common Table Expressions)
• DENSE RANK window function