Quantifying the contribution of periodicity and national holidays to air pollution levels in the United Kingdom using a decomposable time series model

This paper quantifies the impact of periodicity and national holidays on air pollution levels in the United Kingdom using a decomposable time series forecasting model. The analysis focuses on nitrogen dioxide ( N O 2 ) concentrations, with data sourced from the Automatic Urban and Rural Network and Air Quality England networks between January 2017 and December 2023. The Prophet model developed by Meta is used to identify, quantify, and appropriately remove the temporal periodicities in air pollution concentration, demonstrating how annual holidays such as Christmas, and one-off events, such as the state funeral of Elizabeth II and the London Marathon, influence local air pollution in isolation. The findings provide empirical evidence supporting widely held assumptions around national holidays and show some localised reductions in N O 2 concentrations during major events, with contextual variation also observed. For example, the state funeral of Elizabeth II shows a reduction in 21 . 15 μ g m − 3 compared to a median reduction of − 2 . 43 μ g m − 3 outside of London for urban traffic sites. This paper emphasises the need for localised air pollution mitigation policies and demonstrates the utility of large, complete, and publicly available datasets coupled with modern forecasting tools in environmental research. • Traffic patterns and seasonal variations strongly influence air pollution levels. • National holidays events are associated with reductions in NO 2 concentrations. • The effects of holidays on air pollution are highly localised. • There is a need for localised, data-driven air pollution mitigation strategies.