New key publication on global UFP concentrations
The study Global high-resolution ultrafine particle number concentrations through data fusion with machine learning by Pantelis Georgiades, Matthias Kohl, Mihalis A. Nicolaou, Theodoros Christoudias, Andrea Pozzer, Constantine Dovrolis and Jos Lelieveld (MARKOPOLO partners from The Cyprus Institute and the Max Planck Institute for Chemistry) presents an important scientific advancement in the field of atmospheric pollution.
Air pollution continues to cause millions of excess deaths each year, with particulate matter playing a central role in these impacts. While most research has focused on larger particle fractions, ultrafine particles (UFPs, smaller than 100 nanometers) pose a distinct health risk due to their ability to penetrate deep into the lungs, enter the bloodstream and contribute to systemic health effects.
This publication introduces the first global, high-resolution (1 km) maps of UFPs by integrating ground-based measurements with a machine-learning model (XGBoost). The model demonstrates robust performance and captures striking global contrasts, ranging from only a few thousand particles per cubic centimeter in clean environments to more than 40,000 particles per cubic centimeter in heavily polluted urban areas.
Because UFPs account for approximately 91% of all airborne particle numbers, the dataset produced by this work represents an essential resource for advancing exposure research and supporting more accurate assessments of the health risks associated with particulate air pollution.
Full publication: Global high-resolution ultrafine particle number concentrations through data fusion with machine learning by Georgiades et al., Scientific Data 2025
