Interaction between climate factors and air quality in three Norwegian cities: A machine learning analysis

This study examines the interaction between climate factors and air quality in three Norwegian cities, addressing the gap where policymakers often analyze air quality and climate change in isolation. We investigate the association of specific climate variables on air pollution by comparing traditional regression models with machine learning techniques, including k-means clustering, hierarchical clustering, random forest, and recursive feature elimination. The models used are based on Europe's standard environmental policy frameworks, and the analysis draws on a decade's worth of daily data on traffic, weather, and air pollution from three major cities in Norway (2009–2018). Our findings highlight a strong correlation between Heating Degree Days (HDD) and elevated levels of pollutants like PM 2.5 and NO x , indicating that increased heating demand and traffic volume contribute significantly to worsening air quality. This research provides valuable insights into the seasonal dynamics of air pollution and offers a robust data-driven framework to help policymakers develop more effective and integrated urban climate and air quality policies. The research emphasizes the necessity of accounting for the interplay between climate change and air quality in the development of strategies to mitigate the health hazards linked to air pollution.

Cong Cao: Writing – review & editing, Writing – original draft, Visualization, Methodology, Data analysis, Data curation, Conceptualization. Ramit Debnath: Writing – review & editing, Resources, Project administration, Conceptualization. R. Michael Alvarez: Writing – review & editing, Resources, Supervision, Project administration, Funding acquisition, Conceptualization.

The traffic flow, air pollutants, and meteorological data used in this work are publicly available and freely available from the Norwegian Public Road Administration, Statens Vegvesen (SVV), Norwegian Institute for Air Research (NILU), and the Norwegian Meteorological Institute (MET). The descriptive statistics for this data can be located in section 2.

The code is made available from Github repository.