Department of Epidemiology and Environmental Health, School of Public Health and Health Professions, University at Buffalo, Buffalo, NY, USA.
Department of Civil, Structural and Environmental Engineering, University at Buffalo, Buffalo, USA; Research and Education in Energy, Environment and Water Institute, University at Buffalo, Buffalo, NY, USA.
Environ Pollut. 2022 Jan 1;292(Pt A):118285. doi: 10.1016/j.envpol.2021.118285. Epub 2021 Oct 8.
Air pollution has become a major issue in China, especially for traffic-related pollutants such as nitrogen dioxide (NO). Current studies in China at the national scale were less focused on NO exposure and consequent health effects than fine particulate exposure, mainly due to a lack of high-quality exposure models for accurate NO predictions over a long period. We developed an advanced modeling framework that incorporated multisource, high-quality predictor data (e.g., satellite observations [Ozone Monitoring Instrument NO, TROPOspheric Monitoring Instrument NO, and Multi-Angle Implementation of Atmospheric Correction aerosol optical depth], chemical transport model simulations, high-resolution geographical variables) and three independent machine learning algorithms into an ensemble model. The model contains three stages: (1) filling missing satellite data; (2) building an ensemble model and predicting daily NO concentrations from 2013 to 2019 across China at 1×1 km resolution; (3) downscaling the predictions to finer resolution (100 m) at the urban scale. Our model achieves a high performance in terms of cross-validation to assess the agreement of the overall (R = 0.72) and the spatial (R = 0.85) variations of the NO predictions over the observations. The model performance remains moderately good when the predictions are extrapolated to the previous years without any monitoring data (CV R > 0.68) or regions far away from monitors (CV R > 0.63). We identified a clear decreasing trend of NO exposure from 2013 to 2019 across the country with the largest reduction in suburban and rural areas. Our downscaled model further improved the prediction ability by 4%-14% in some megacities and captured substantial NO variations within 1-km grids in the urban areas, especially near major roads. Our model provides flexibility at both temporal and spatial scales and can be applied to exposure assessment and epidemiological studies with various study domains (e.g., national or citywide) and settings (e.g., long-term and short-term).
空气污染已成为中国的一个主要问题，特别是对于交通相关污染物，如二氧化氮（NO）。当前，中国的研究主要集中在细颗粒物暴露及其对健康的影响上，而对 NO 暴露及其后果的研究较少，这主要是由于缺乏高质量的暴露模型，无法对长时间内的 NO 进行准确预测。我们开发了一个先进的建模框架，该框架结合了多源、高质量的预测因子数据（例如卫星观测[臭氧监测仪器 NO、对流层监测仪器 NO 和多角度大气校正气溶胶光学深度]、化学输送模型模拟、高分辨率地理变量）和三种独立的机器学习算法，形成一个集合模型。该模型包含三个阶段：（1）填补卫星数据的缺失值；（2）建立集合模型，预测 2013 年至 2019 年中国各地 1×1 公里分辨率的每日 NO 浓度；（3）将预测结果下推到城市尺度的更精细分辨率（100 米）。我们的模型在交叉验证中表现出很高的性能，评估了整体（R=0.72）和空间（R=0.85）NO 预测与观测之间的一致性。当预测结果在没有任何监测数据的情况下外推到前几年（CV R>0.68）或远离监测器的区域（CV R>0.63）时，模型性能仍然保持较好。我们发现，2013 年至 2019 年期间，全国范围内的 NO 暴露水平呈明显下降趋势，郊区和农村地区的降幅最大。我们的细化模型在一些特大城市进一步提高了 4%-14%的预测能力，并在城市地区的 1 公里网格内捕捉到了大量的 NO 变化，特别是在主要道路附近。我们的模型在时间和空间尺度上都具有灵活性，可应用于具有各种研究领域（如全国或全市范围）和设置（如长期和短期）的暴露评估和流行病学研究。
