S-MESH: Surface PM2.5 daily 1km dataset over Central Europe using CAMS, Low-Cost Sensors and AOD

This dataset provides estimated daily surface PM2.5 concentrations over Central Europe at approximately 1 km spatial resolution for the years 2021–2022. It was generated with the S-MESH model, which combines CAMS forecasts, low-cost sensor data, satellite aerosol optical depth, and meteorological variables through a stacked XGBoost machine learning approach. The dataset includes several model versions to compare the role of low-cost sensors in training and prediction, offering a high-resolution resource for air quality analysis, exposure studies, and local decision support.