School of Public Health, Key Lab of Public Health Safety of the Ministry of Education, & Key Lab of Health Technology Assessment of the Ministry of Health, Fudan University, Shanghai, China; Shanghai Key Laboratory of Atmospheric Particle Pollution and Prevention (LAP3), Fudan University, Shanghai, China.
College of Urban and Environmental Science, Tianjin Normal University, Tianjin, China.
Environ Res. 2015 Feb;137:308-15. doi: 10.1016/j.envres.2015.01.003. Epub 2015 Jan 16.
Limited by data accessibility, few exposure assessment studies of air pollutants have been conducted in China. There is an urgent need to develop models for assessing the intra-urban concentration of key air pollutants in Chinese cities. In this study, a land use regression (LUR) model was established to estimate NO2 during 2008-2011 in Shanghai. Four predictor variables were left in the final LUR model: the length of major road within the 2-km buffer around monitoring sites, the number of industrial sources (excluding power plants) within a 10-km buffer, the agricultural land area within a 5-km buffer, and the population counts. The model R(2) and the leave-one-out-cross-validation (LOOCV) R(2) of the NO2 LUR models were 0.82 and 0.75, respectively. The prediction surface of the NO2 concentration based on the LUR model was of high spatial resolution. The 1-year predicted concentration based on the ratio and the difference methods fitted well with the measured NO2 concentration. The LUR model of NO2 outperformed the kriging and inverse distance weighed (IDW) interpolation methods in Shanghai. Our findings suggest that the LUR model may provide a cost-effective method of air pollution exposure assessment in a developing country.
受数据可获取性的限制，在中国，针对空气污染物的暴露评估研究较少。因此，非常有必要开发模型来评估中国城市中关键空气污染物的城市内浓度。本研究建立了一个用于估算 2008-2011 年上海二氧化氮浓度的基于土地利用回归（LUR）模型。最终的 LUR 模型中保留了四个预测变量：监测点周围 2 公里缓冲区的主要道路长度、10 公里缓冲区的工业源数量（不包括电厂）、5 公里缓冲区的农业用地面积和人口数量。NO2 LUR 模型的 R²和留一法交叉验证（LOOCV）R²分别为 0.82 和 0.75。基于 LUR 模型的 NO2 预测浓度具有较高的空间分辨率。基于比和差方法的一年预测浓度与实测 NO2 浓度拟合良好。在上海，LUR 模型在预测 NO2 浓度方面优于克里金插值法和反距离权重插值法。我们的研究结果表明，LUR 模型可能为发展中国家的空气污染暴露评估提供一种具有成本效益的方法。
