Rapid urbanization has negative effects on the ecological environment, especially the air pollution in some areas of China. Land use regression (LUR), which has been widely used in predicting spatio-temporal variation of air pollution concentrations and estimating population exposure in recent years, is applied in this paper to explore the relationship between land use and air pollution level in Shanghai, China. PM2.5 concentrations of each monitoring locations are selected as dependent variations and a series of geographic factors are collected as independent variables. The built LUR is then used to produce continuous concentration fields for PM2.5 to describe the spatio-temporal variation of air pollution in Shanghai. In addition, taking population density extracted by mobile phone Signaling data into consideration, PM2.5 exposure intensity index is built to objectively reflect the impact of population factors on pollution exposure and human health risk assessment. Keywords: Land use regression, spatio-temporal variability, population exposure, air pollution, Shanghai

CICTP