Study about Paris' air pollutants.

School project for the "Green AI" course

Here's the list of the studies we'll be conducting:

Temporal Analysis

• Trend Analysis: Examine the long-term trends in pollutant concentrations since 2018. Are levels of PM10, NO2, NO, NOX, and O3 increasing or decreasing over time?
• Seasonal Patterns: Investigate how concentrations vary by season, month, or day of the week. For example, are pollution levels higher in winter due to heating systems or lower in summer due to reduced traffic?
• Time of Day: Analyze daily cycles of pollutants to identify peak hours for pollution (e.g., during morning and evening commutes).

Policy Impact

• Effects of Regulations: Correlate changes in pollutant levels with specific policy implementations in Paris, such as restrictions on diesel vehicles or low-emission zones.
• COVID-19 Lockdowns: Study how pollution levels changed during the lockdown periods in 2020 and 2021, offering insights into the impact of reduced human activity.

Inter-Pollutant Relationships

• Study correlations between pollutants (e.g., NO2 and O3, or NO and NOX) to understand their chemical and atmospheric interactions.
• Identify patterns or thresholds where one pollutant's increase significantly influences another.

Comparative Studies

• Compare Paris's pollutant levels with other major cities to contextualize the data.

Event-based Analysis

• Investigate pollution spikes during major events (e.g., New Year's Eve, protests, or transportation strikes).
• Analyze the effects of special initiatives like car-free days or public transit campaigns.

Machine Learning and Predictive Modeling

• Use machine learning to predict pollutant levels based on historical data. (As we're lacking weather conditions, or traffic data).

Data from: https://data-airparif-asso.opendata.arcgis.com/

We're using Paris' 18th arrondissement measure station's data since 2018.