Academic Journal
Προβολή σε EDS
Συνδεδεμένο Πλήρες Κείμενο

Machine-learning-based estimation of ground-level NO2 concentrations across Southeastern Europe using multi-source satellite, reanalysis, and emission data.

Λεπτομέρειες βιβλιογραφικής εγγραφής
Τίτλος: Machine-learning-based estimation of ground-level NO2 concentrations across Southeastern Europe using multi-source satellite, reanalysis, and emission data.
Συγγραφείς: Bilgiç E; Department of Environmental Engineering, Faculty of Engineering, Dokuz Eylul University, Izmir, Türkiye. Electronic address: efem.bilgic@deu.edu.tr., Elbir T; Department of Environmental Engineering, Faculty of Engineering, Dokuz Eylul University, Izmir, Türkiye; Environmental Research and Application Center (CEVMER), Dokuz Eylul University, Izmir, Türkiye.
Πηγή: Environmental pollution (Barking, Essex : 1987) [Environ Pollut] 2026 Aug 15; Vol. 403, pp. 128461. Date of Electronic Publication: 2026 May 28.
Τύπος έκδοσης: Journal Article
Γλώσσα: English
Στοιχεία περιοδικού: Publisher: Elsevier Applied Science Publishers Country of Publication: England NLM ID: 8804476 Publication Model: Print-Electronic Cited Medium: Internet ISSN: 1873-6424 (Electronic) Linking ISSN: 02697491 NLM ISO Abbreviation: Environ Pollut Subsets: MEDLINE
Imprint Name(s): Original Publication: Barking, Essex, England : Elsevier Applied Science Publishers, c1987-
Ιατρικοί όροι (MeSH): Air Pollutants*/analysis , Environmental Monitoring*/methods , Nitrogen Dioxide*/analysis , Boosting Machine Learning Algorithms*, Air Pollution/statistics & numerical data ; Europe ; Satellite Imagery
Περίληψη: Reliable estimation of ground-level nitrogen dioxide (NO2) remains challenging due to limited monitoring coverage and complex interactions among emissions, meteorology, and land-use factors. This study develops a machine-learning framework to estimate surface NO2 concentrations across Southeastern Europe, covering thirteen countries with diverse emission sources and sparse monitoring networks. Daily NO2 observations from 317 monitoring stations were integrated with multiple satellite and ancillary datasets, including tropospheric NO2 from Sentinel-5P TROPOMI, vegetation indices from MODIS, and nighttime light data from VIIRS. Additional inputs included meteorological variables from ERA5, anthropogenic emissions from EDGAR v8.1, land-use data from Copernicus Corine Land Cover, elevation from the Copernicus Digital Elevation Model, and population density from the Global Human Settlement Layer. All datasets were processed and harmonized using Google Earth Engine. Three Gradient Boosting Decision Tree models (XGBoost, LGBM, and CatBoost) were trained and optimized with Optuna. Model performance was assessed using 10-fold cross-validation and independent test sets with R2, MAE, and RMSE metrics. All models performed consistently, with LGBM achieving the best results (R2 = 0.86; MAE = 5.34 μg/m3; RMSE = 8.32 μg/m3). SHapley Additive exPlanations (SHAP) identified key predictors and improved interpretability. Seasonal analysis showed higher accuracy in summer and autumn than in winter and spring. These results demonstrate that integrating multi-source data with advanced machine learning enables reliable surface NO2 estimation in regions with limited monitoring, supporting air quality assessment, exposure analysis, and evidence-based policymaking in Southeastern Europe.
(Copyright © 2026 Elsevier Ltd. All rights reserved.)
Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Contributed Indexing: Keywords: Air quality modeling; Machine learning; Nitrogen dioxide; Satellite data integration; Southeastern Europe
Substance Nomenclature: 0 (Air Pollutants)
S7G510RUBH (Nitrogen Dioxide)
Entry Date(s): Date Created: 20260529 Date Completed: 20260610 Latest Revision: 20260611
Update Code: 20260611
DOI: 10.1016/j.envpol.2026.128461
PMID: 42214547
Βάση Δεδομένων: MEDLINE
Περιγραφή
ISSN:1873-6424
DOI:10.1016/j.envpol.2026.128461