Land use regression modelling of PM2.5 spatial variations in different seasons in urban areas

Abstract

As one of the principal components of haze, fine particulate matter (PM2.5) has potential negative health effects, causing widespread concern. Identification of the pollutant spatial variation is a prerequisite of understanding ambient air pollution exposure and further improving air quality. Seven urban built-up areas in Liaoning central urban agglomeration (LCUA) were used for land use regression (LUR) modelling of PM2.5 concentrations using small amounts of spatially aggregated data and to assess the model's seasonal consistency. LUR models explained 52–61% of the variation in the PM2.5 concentrations at urban scales. The average building floor area was the key predictor in each model, and the percent water area was predictor with a negative coefficient. Good seasonal consistency was observed between the heating-seasonal model and annual average model, showing that the annual average PM2.5 pollution in the LCUA was mainly influenced by pollution during the heating season. Extending the linear LUR model with regression kriging improved the model's explanatory ability and predictive performance. The predicted PM2.5 concentrations in Shenyang and Anshan were the highest and that in Yingkou was the lowest. The building three-dimensional variables played important roles in the urban spatial modelling of air pollution.