Development of land use regression models for PM2.5, SO2, NO2 and O3 in Nanjing, China

    loading  Checking for direct PDF access through Ovid


Ambient air pollution has been a global problem, especially in China. Comparing with other methods, Land Use Regression (LUR) models can obtain air pollutant concentration distribution at finer scale without the air pollution source data based on a few monitoring sites and predictors. However, limited LUR studies have been conducted on the basis of regular monitoring networks. Thus, we explored the applicability of conducting LUR models for four key air pollutants: PM2.5, SO2, NO2 and O3, on the basis of national monitoring networks which have good representation of areas with different characteristics in Nanjing, China. Fifty-nine potential predictor variables were considered, including land use type, population density, traffic emission, industrial emission, geographical coordinates, meteorology and topography. LUR models of these four air pollutants were with good explained variance for four key air pollutants. Adjusted explained variance of the LUR models was highest for NO2 (87%), followed by SO2 (83%), and was lower for PM2.5 (72%) and O3 (65%). Annual average distributions of pollutants in 2013 were obtained based on predicted values, which revealed that O3 in Nanjing was more heavily impacted by regional influences. This study would not only contribute to the wider use of LUR studies in China but also offer important reference for the application of regular monitoring network with high representativeness in LUR studies. These results would also support for air epidemiological studies in the future.HighlightsLUR models for four key air pollutants performed well in Nanjing, China.It is feasible to conduct LUR models based on national monitoring network in China.The influence of land use and traffic emission vary at different geographic scales.LUR models based on linear regression sometimes may not reflect the direct influence.

    loading  Loading Related Articles