Reproducibility package for Tracking Air Pollution From Space: High-Resolution Daily Estimates For PM2.5, O3, NO2 And CO

Air pollution imposes severe health and economic costs worldwide, yet ground-based monitoring data are difficult and expensive to collect, limiting information available for policy analysis. This paper develops machine learning models to estimate daily air quality using satellite observations, meteorological data, emissions inventories, and geographic information. Data sources include IQAir ground station reports, atmospheric pollutant concentrations from ESA's TROPOMI platform, and meteorological data from the Copernicus ERA5 Land database. We estimate separate random forest models for PM2.5 (fine particulates), O3 (ozone), NO2 (nitrogen dioxide), and CO (carbon monoxide), using a random forest algorithm that quantifies variable importance.

Results are robust: TROPOMI dominates predictions for PM2.5, NO2, and CO, with EDGAR emissions and meteorological variables providing secondary contributions. For O3, meteorological variables are most influential, followed by TROPOMI and EDGAR. Geographic features matter across all pollutants — elevation, latitude, and coastal proximity consistently rank high, while Köppen–Geiger climate zones vary in importance by pollutant.

We validate temporal tracking by correlating daily monitored versus predicted pollutant levels across urban areas using IQAir ground station data. Median intertemporal correlations exceed 0.69 for all pollutants, with upper quartiles above 0.82 and lower quartiles above 0.53. Although the pilot uses April–November 2025 data, TROPOMI has been available since 2018 and ERA5 Land provides matching coverage, enabling daily estimates from April 2018 onward on a global terrestrial grid at 0.1 degree resolution. The database could also link our models to the ESA Copernicus reporting system to deliver daily air quality updates and forward-looking trend estimates.