Air pollution and atmospheric variables in Türkiye: A comprehensive analysis
Abstract
This study examines particulate matter (PM 10 ) and sulphur dioxide (SO 2 ) concentrations in Türkiye, one of the world's 20 largest economies with a population of more than 86 million people. The analysis covers from 2015 to 2019, a period during which pollutant levels frequently exceeded international air quality standards. While anthropogenic emissions are significant contributors, atmospheric variables also play a crucial role in pollutant dispersion and dilution. The research investigates how temperature, relative humidity, wind speed, precipitation, and boundary layer height affect pollutant concentrations in this country. Using a generalised additive model (GAM) and conducting a sensitivity analysis, the study explores seasonal and regional variations in the influence of these atmospheric variables across Türkiye, offering a national, multivariable perspective. The analysis relies on hourly data to capture diurnal variations during summer and winter. Additionally, it provides projections of pollutant concentrations for 2030 and 2050 under two IPCC climate scenarios: one assuming strong mitigation measures (RCP 4.5) and the other representing high-emission conditions (RCP 8.5). The GAM outputs show that temperature, humidity, wind speed, and precipitation affect PM 10 and SO 2 differently across the country, with marked regional and seasonal variations. Winter PM 10 commonly peaks near ∼10 °C, whereas in several regions summer concentrations decline once temperatures exceed ∼30 °C; relative humidity is generally positively associated with PM 10 yet negatively with SO 2 ; stronger winds and precipitation coincide with lower pollutant levels (with PM 10 highest at low winds <5 m/s and notably reduced on days with ∼25 mm rainfall). Coastal regions show the strongest wind- and rain-related reductions in concentrations; interior continental and semi-arid regions display a weaker precipitation signal with a tighter humidity–PM 10 association; and high-elevation basins and valleys are characterised by heightened sensitivity to boundary-layer stability. Sensitivity analysis indicates that boundary-layer height and wind are the primary atmospheric drivers for PM 10 , whereas SO 2 shows a weaker overall sensitivity to atmospheric factors. Projections about the influence of atmospheric variables on pollution for 2030 and 2050 suggest that PM 10 levels will generally decline across both seasons, while SO 2 concentrations are expected to rise during summer. The study highlights the increasing influence of climate change on air pollution, stressing the importance of region-specific air quality management strategies. By integrating mitigation and adaptation measures, the findings offer valuable insights for policymakers aiming to improve air quality and reduce health risks across Türkiye. • A comprehensive analysis of PM 10 and SO 2 concentrations across Türkiye was conducted at a regional level. • Generalised Additive Models with sensitivity analysis captured nonlinear effects of atmospheric variables on pollutant levels. • Distinct seasonal and regional patterns in pollutant–atmosphere relationships were identified. • Air pollution projections for 2030 and 2050 were produced under two climate scenarios. • The modelling framework supports targeted mitigation strategies and long-term air quality planning.