Clustering and machine learning techniques identify air pollution regimes in Greater Cairo

Area of Science:

• Environmental Science
• Public Health
• Data Science

Background:

• Greater Cairo faces significant air pollution challenges from urbanization, traffic, and dust.
• Understanding distinct air pollution patterns is crucial for effective public health strategies.

Purpose of the Study:

• To develop and validate a data-driven framework for identifying and interpreting recurring air pollution regimes in Greater Cairo.
• To assess the effectiveness of machine learning models in classifying these pollution patterns.

Main Methods:

• Utilized unsupervised K-means clustering and supervised Decision Tree (DT) and Random Forest (RF) models.
• Employed atmospheric reanalysis data from Copernicus Atmosphere Monitoring Service (CAMS) for 2023-2024.
• Analyzed feature importance (NO₂, PM₁₀) for regime differentiation.

Main Results:

• Identified four distinct air pollution regimes, with low pollution dominating (75.1%).
• Higher pollution regimes linked to traffic emissions and dust events.
• Random Forest achieved high accuracy (97.43%), outperforming Decision Tree (93.10%).

Conclusions:

• Integrating clustering with tree-based classification offers an effective approach for urban air pollution analysis.
• The framework supports regime-based air quality interpretation and targeted management strategies.
• Nitrogen dioxide (NO₂) and particulate matter (PM₁₀) are key indicators for traffic and dust events, respectively.