Cooking fumes are an important source of urban air pollution, for which the ambient impact near exhaust outlets remains insufficiently characterized. This study aimed to systematically characterize the volatile organic compounds (VOCs), odor pollution, health risks, and comprehensive impacts from five selected catering businesses as exploratory case studies in Shanghai: barbecue (BBQ), hot pot, Xiang cuisine, institutional canteen, and mixed emission. Field sampling and analysis were performed during peak cooking hours between September 2024 and January 2025. Results showed that BBQ emissions exhibited the highest measured pollution levels, with 81 VOC species detected, a total VOCs concentration of 6968.57 μg/m3, an odor concentration of 1513, and higher screening-level carcinogenic and non-carcinogenic indices than the other cases. Aldehydes including hexanal, acetaldehyde, acrolein and pentanal were the primary odor contributors, and hexanal presented the highest odor contribution rate of 74.04 ± 16.83%. Under a source-proximate screening scenario, formaldehyde, acetaldehyde, benzene, and ethylbenzene were the main contributors to carcinogenic and non-carcinogenic indices, and the cumulative non-carcinogenic hazard index exceeded unity at all five cases. A Comprehensive Evaluation Index (CEI) integrating sensory and health evidence ranked hexanal first and formaldehyde second in all five cases, with aldehydes dominating the leading group. Read together with a reaction-derived atmospheric lifetime, the ranking separated the aldehydes, which were removed close to the source, from carcinogenic chlorinated and aromatic compounds that persist long enough to be transported beyond the outlet surroundings; near-source control of aldehydes and regional coordination for the persistent carcinogens were therefore proposed as distinct control strategies. Because each category was represented by a single case and the indices were upper bounds from undiluted exhaust-outlet concentrations, these results identified priority substances and a reproducible prioritization procedure rather than quantified risk to nearby residents.
The spatial distribution and temporal variability of aerosol components are crucial for understanding aerosol effects on climate, ecosystems, and public health. MERRA-2 and CAMS are two widely used reanalysis datasets, but the differences in quantifying aerosol components have not been sufficiently investigated. In this study, MERRA-2 and CAMS reanalysis datasets are used to compare the global spatial distribution and long-term trends of major aerosol components during 2003-2022. Trends are estimated using the Theil-Sen estimator and tested for significance via the trend-free pre-whitening Mann-Kendall test. The Generalized Retrieval of Atmosphere and Surface Properties (GRASP)/Component dataset, an AERONET-based aerosol component retrieval product, serves as independent observational data for component-level comparison. Results show that MERRA-2 and CAMS are generally consistent in spatial distributions and long-term trends of aerosol optical depth (AOD) at 550 nm, with high AOD values located in Asia and Africa. Aerosol components show substantial differences in both spatial distributions and long-term trends between the two reanalysis datasets. For carbonaceous aerosols, MERRA-2 suggests increases in black carbon (BC) AOD and organic aerosol AOD over Southern Africa (SAF), Ocean (OCN), North America (NAM), and Central America (MCA), whereas CAMS shows negative trends or weaker positive trends in these regions. Both MERRA-2 and CAMS are evaluated against GRASP/Component for BC column mass concentration (R = 0.59 and R = 0.52, respectively) and organic matter (OM) column mass concentration (R = 0.74 and R = 0.68, respectively). In addition, trends of AOD and BC column mass concentration derived from both MERRA-2 and CAMS also show good agreement with GRASP/Component data. These findings help clarify the uncertainties associated with using reanalysis datasets to investigate aerosol component distributions and long-term trends.
Rare earth elements (REEs) are increasingly recognized as emerging atmospheric contaminants, yet diurnally resolved data on PM2.5-bound REEs remain scarce, particularly for the karst regions of southwestern China. In this study, 85 p.m.2.5 samples were collected in Xingbin District, Laibin, Guangxi, during winter 2025 using a paired day-night design, and 14 lanthanides were quantified together with co-occurring pollutants. The mean total REE concentration was 22.08 ± 16.89 ng m−3. Upper continental crust (UCC)-normalized patterns were strongly inclined toward the heavy REEs, with a (La/Yb)N of 0.433, lower than that of every coal- and dust-derived end-member compiled from the literature. REE concentrations did not differ significantly between day and night, and an ensemble of five machine-learning methods failed to separate the two periods, indicating that the REE source structure remained invariant at the diurnal scale during the campaign. Non-negative matrix factorization resolved three sources: a crust-like mixed dust source (68.3%), a middle-REE-enriched mixed anthropogenic source (15.9%), and a heavy-REE-enriched source of undetermined origin (15.8%). A cross-validated R2 of −0.55 and only 13 significant rank correlations out of 84 show that PM2.5-bound REEs are decoupled from the co-pollutants. REEs in PM2.5 at this karst site are therefore controlled by primary mechanical particle emissions rather than by gas-phase combustion chemistry, and controls on gaseous precursors are unlikely to reduce REE exposure.
Volatile organic compounds (VOCs) are emitted into the atmosphere from natural and anthropogenic sources and are precursors of tropospheric ozone and secondary organic aerosols (SOA). This is particularly relevant in Mediterranean environments, where frequent anticyclonic conditions and high solar radiation promote photochemical activity. In this study, over 100 VOCs were analyzed at the molecular level across rural, suburban and industrial environments in Catalonia (NE Spain). Source apportionment identified seven emission sources and atmospheric processes: i) industrial emissions, ii) traffic, iii) biomass burning, iv) temperature-dependent emissions, v) gasoline evaporation, vi) pesticide formulations and vii) diesel-oil exhaust. Their contributions varied markedly among sites and seasons. In winter, industrial emissions dominated the industrial site (76% of TVOC), whereas traffic (37%) and biomass burning (25%) were the main contributors at the suburban site. In spring, agricultural activities, particularly diesel-related emissions (47%), dominated at the rural site. In summer, temperature-dependent biogenic emissions accounted for 77% and 68% of TVOC at rural and suburban sites, respectively. OH reactivity contributors exhibited marked seasonal and spatial variability: anthropogenic VOCs such as toluene and m-xylene were dominant during winter, whereas isoprene, formaldehyde and acetaldehyde were more important during the warm seasons. Pent-1-ene was a significant contributor throughout most seasons and sites. In contrast, SOA formation potential was related with a few anthropogenic precursors (toluene, benzene, m-xylene and C13-C15 n-alkanes). These findings indicate that a small group of VOCs are relevant for the generation of secondary pollution and should be prioritized in air quality management strategies in similar Mediterranean environments.
Airborne pollen concentrations are used as a proxy for ragweed allergen exposure, yet their relationship with the major allergen Amb a 1 remains poorly understood. We combined simultaneous monitoring of Ambrosia pollen and Amb a 1 with HYSPLIT back-trajectory analysis and regional land-use mapping to examine whether both share source and transport signatures across six seasons (2019–2024) in Bratislava. Both variables peaked under south-easterly airflow, but their seasonal and directional patterns diverged. Pollen-rich seasons (2023–2024) did not coincide with allergen-rich seasons (2019, 2022), and the seasonal Amb a 1 content per pollen grain varied 15-fold (0.17–2.40 pg/pollen), being lowest in the two pollen-richest seasons. High-allergen/low-pollen episodes were absent or under-represented under the pollen-rich SE, E and S flows. The proportion of top-25% pollen days differed strongly among transport sectors (χ2 = 59.8, p < 0.001), whereas that of top-25% Amb a 1 days did not (χ2 = 8.7, p = 0.28). An exploratory positive effect of the SW sector was also noted. Correlations between high-concentration days per sector and agricultural land cover within a 300 km buffer were non-significant and not robust to normalisation by sector sample size. A multiple linear regression explained only 19% of daily Amb a 1 variance, indicating that factors other than pollen abundance substantially influence allergen exposure. These findings indicate a partial decoupling of airborne Ambrosia pollen and Amb a 1, associated with variability in pollen potency and atmospheric transport, and highlight the need to supplement pollen monitoring with direct allergen measurements.
