Reconstructed Zurich Air Pollution Data (2015–2019) with Lag Structure for Urinary Metal Toxicokinetic Analysis

This dataset contains reconstructed daily air pollution data from the Zurich NABEL monitoring station (“Kaserne”) between 2015 and 2019, including NO₂, SO₂, CO, PM₁₀, PM₂.₅, and derived PM₁₀–₂.₅ values. A few missing values for all pollutants except PM₂.₅ were imputed using linear interpolation, while missing PM₂.₅ values were reconstructed using an XGBoost-based machine learning model trained on the remaining pollutants. All imputed values are flagged accordingly.To enable toxicokinetic modeling of airborne metal exposure, the dataset includes:14-day average concentrations for each pollutant preceding the urine test,27 lagged weekly averages (lag –1 to –26) aligned with the dates of Chelate-Evoked Metal Excretion Testing (CEMET) conducted at a clinical center in Pfäffikon SZ, Switzerland.The lag structure was designed to explore the delay between environmental exposure and urinary excretion of heavy metals. These lag times are still largely unknown, which may explain why many environmental impact studies fail to detect associations between pollutant exposure and metal biomarkers. This dataset provides a novel foundation for investigating these temporal dynamics.