Monitoring-based assessment of spatial and seasonal air-quality associations with low-emission zones in South Korea

Abstract Regional assessments of vehicle restrictions rely on both the construction of concentration fields from monitoring stations and the comparability of policy and comparison areas. We integrate these two dimensions of assessment using monthly concentrations of five pollutants in 247 South Korean districts during 2012-2024. Quality-screened station observations are interpolated and area-weighted to districts, with precipitation reconciled against published monthly totals. Station-holdout validation and a common-grid nearest-station alternative assess measurement sensitivity. Spatial Durbin models assess pooled and seasonal associations while accounting for spatial dependence and district and year-month fixed effects. Baseline models associate permanent low-emission-zone rollout with lower nitrogen dioxide concentrations; trend-adjusted models show that regional trajectories are central to interpreting this finding. Associations also differ across pollutants and seasons, with some patterns depending on interpolation choice. Pre-adoption differences support interpreting these estimates as conditional associations rather than causal effects. This study provides an integrated, reproducible framework for evaluating low-emission zones, characterizing pollutant-specific spatial and seasonal associations while distinguishing sensitivity to measurement choices from sensitivity to regional-trend assumptions.