Contents lists available at ScienceDirect Atmospheric Environment journal homepage: www.elsevier.com/locate/atmosenv Development of spatiotemporal models to predict ambient ozone and NO x concentrations in Tianjin, China Shahir Masri a , Haiyan Hou b,c,1 , Andy Dang a , Ting Yao d , Liwen Zhang e , Tong Wang e , Zhe Qin f , Siyu Wu c , Bin Han g , Jiu-Chiuan (JC) Chen h , Yaqiong Chen c , Jun Wu a,∗ a Program in Public Health, Susan and Henry Samueli College of Health Sciences, University of California, Irvine, CA, USA b Chinese Academy of Medical Sciences & Peking Union Medical College, Peking Union Medical College Hospital, Beijing, China c Department of Obstetrics and Gynecology, Affiliated Hospital of the Chinese People's Armed Police Force Logistics College, Tianjin, China d Department of Obstetrics and Gynecology, Tianjin Hospital of Integrated Chinese and Western Medicine Tianjin Nankai Hospital, Tianjin, China e Department of Occupational and Environmental Health, College of Public Health, Tianjin Medical University, Tianjin, China f Department of Obstetrics and Gynecology, General Hospital of the Chinese People's Armed Police Force, Beijing, China g State Key Laboratory of Environmental Criteria and Risk Assessment, Chinese Research Academy of Environmental Sciences, Beijing, China h Department of Preventive Medicine, University of Southern California, Keck School of Medicine, Los Angeles, CA, USA ARTICLE INFO Keywords: Exposure model Exposure assessment Air pollution Ozone China ABSTRACT Nitrogen oxides (NO x ) and ozone (O 3 ) are important air pollutants that are associated with adverse health efects. Land-use regression (LUR) models have been widely developed to estimate air pollution concentrations. Due to data availability, however, such models are usually not applied in developing countries. We aimed to characterize NO x and O 3 concentrations and develop LUR models to predict their spatial and temporal dis- tributions using publicly-available data in Tianjin, a heavily polluted city in China. Seasonal samples were collected across Tianjin at 29 locations for O 3 and 49 locations for NO x . Heavy-duty vehicle counts estimated from 0.5 m × 0.5 m satellite images correlated well with feld-measured counts, thus supporting the use of high- resolution satellite images to assess vehicle trafc. Concentrations of NO x were highest in winter, while the opposite pattern was observed for O 3 . The majority of the variance in NO x was explained by season (36.2%) and heavy vehicle trafc (19.8%). For O 3 , the variance was explained by season (80.7%) in a pooled model, and by distance to roads (43.4%) and distance to coal plants (26.2%) in a summer model. Cross-validation showed reasonable practicability for NO x (R 2 = 0.53 with feld-measured heavy-duty vehicle count; R 2 = 0.46 with satellite-based heavy-duty vehicle count) and O 3 (R 2 = 0.90 for pooled model; R 2 = 0.70 for summer model) models. This study provides utility for researchers investigating air pollution in regions where feld-measured vehicle trafc data are not available, as well as for policy makers and public health ofcials seeking to under- stand the sources and spatial distribution of air pollution in Tianjin. 1. Introduction Worldwide, outdoor air pollution is among the top 15 causes of death (top 10 for high-income nations), leading to an estimated three million deaths annually (WHO, 2016). Nitrogen oxides (NO x ) and ozone (O 3 ) are important components of ambient air pollution and have been widely associated with a variety of adverse health efects. Epidemio- logical studies have associated exposure to NO 2 (a main constituent of NO x ) with cardiopulmonary mortality, lung cancer, and asthma ex- acerbations (Beelen et al., 2008; Brauer et al., 2008; Castellsague et al., 1995; Filleul et al., 2006; Kim et al., 2004). Similarly, O 3 exposure has been associated with respiratory mortality, cardiovascular mortality, and asthma-related emergency room visits (Anenberg et al., 2010; Peng et al., 2013; Shefeld et al., 2015). NO x in the atmosphere primarily originates from fossil fuel com- bustion. Major sources therefore include emissions from automobiles and power plants, as well as coal and biomass burning for heating and cooking. In contrast, ground-level O 3 is a secondary pollutant that forms through a complex series of photochemical reactions involving sunlight and NO x . Therefore, components of NO x act as a precursor to ground-level O 3 formation. China represents a particularly important case with regard to air https://doi.org/10.1016/j.atmosenv.2019.05.060 Received 11 September 2018; Received in revised form 24 April 2019; Accepted 25 May 2019 ∗ Corresponding author. Program in Public Health, Anteater Instruction & Research Bldg (AIRB) # 2034, University of California, Irvine, CA, 92697-3957, USA. E-mail address: junwu@uci.edu (J. Wu). 1 First authors with equal contribution. Atmospheric Environment 213 (2019) 37–46 Available online 27 May 2019 1352-2310/ © 2019 Elsevier Ltd. All rights reserved. T