Mapping urban ultrafine particle exposure in Guangzhou, China: Development of a high-resolution land-use regression...

Ultrafine particles (UFPs) pose elevated health risks, yet high-resolution exposure models remain limited in rapidly urbanizing megacities of developing countries. This study developed a seasonal land-use regression (LUR) model for UFPs in Guangzhou, China, using a dense monitoring network of 96 sites to capture spatial and temporal variations. Particle number concentrations (30-120 nm) were measured across dry and wet seasons (2021-2022) using portable monitors, with predictor variables derived from transportation networks, land use, socioeconomic factors, emission inventories, and points of interest across buffer zones (100-1000 m). A supervised forward stepwise regression identified key predictors, including trunk road length within 100 m, restaurant density within 100 m, industrial land area, and green spaces. The final model explained 76 % of spatial variability (adjusted R² = 0.76, RMSE = 2784 particles/cm³), validated by leave-one-out cross-validation (LOOCV R² = 0.69). Results highlighted vehicular emissions and cooking activities as dominant contributors, with trunk roads and restaurants explaining 52.4 % of variability. Seasonal analysis revealed higher UFPs concentrations during the dry season, linked to lower humidity and reduced dispersion. The 1-km resolution exposure map identified urban hotspots near major roads and commercial zones. This study establishes the first high-resolution LUR model for UFPs in the Pearl River Delta region, offering a transferable framework for exposure assessment in megacities and supporting targeted mitigation strategies and epidemiological research on UFPs-related health impacts.