Inter-Site Correlation of Ambient Air Pollutants Across Swiss Plateau NABEL Stations (2015–2019)

This dataset presents inter-site Spearman correlation matrices for six ambient air pollutants—PM10, PM2.5, elemental carbon (EC), sulfur dioxide (SO2), nitrogen dioxide (NO2), and carbon monoxide (CO)—as recorded by the Swiss NABEL monitoring network between 2015 and 2019.<br>The analysis is limited to NABEL stations located within the Swiss Plateau. Monitoring sites situated at high altitudes or south of the Alps were excluded for two primary reasons:(1) none of the patients in the associated clinical dataset reside in those regions, and(2) these areas represent distinct meteorological and pollution zones that are not representative of typical outpatient exposure in the Swiss Plateau.<br>The resulting Spearman correlation matrices reveal strong coherence in pollutant levels across all included sites. Within this context, the Zurich–Kaserne monitoring site stands out as both the most representative and comprehensive dataset available. It is geographically close to the clinical setting in Pfäffikon SZ—approximately 27 kilometers in straight-line distance—with the patient population largely confined to a radius of 60 kilometers, and the majority residing within 25 kilometers of the practice.<br>Zurich–Kaserne provides long-term, near-continuous monitoring for all six pollutants. While PM2.5 shows a moderate gap in data coverage (missing 186 out of 1,645 possible daily values), a robust imputation strategy was employed to reconstruct missing PM2.5 data. Specifically, missing values were predicted using a gradient-boosted regression tree model (XGBoost) trained on other pollutants (PM10, NO2, SO2, CO) measured concurrently at the same site. Predicted values were constrained to remain below or equal to PM10 to ensure physical plausibility. The remaining five pollutants (PM10, EC, NO2, SO2, CO) exhibited minimal missingness and required only occasional linear interpolation, which is flagged in the data.<br>In addition to the Excel file, a map of the included NABEL measuring sites is provided. This visual reference shows the geographic distribution and names of all stations considered in the analysis. It helps illustrate the spatial coherence of pollutant trends across the Swiss Plateau and highlights the central location of the Zurich–Kaserne site relative to the clinical population.<br>These findings justify the use of Zurich–Kaserne as the primary reference station for environmental exposure modeling in toxicokinetic analyses, particularly in studies correlating air pollutant exposure with urinary metal excretion in outpatient settings.<br>