Spatial Simulation of Black Carbon Concentrations Based on a Land Use Regression Model and Mobile Monitoring over Shanghai, China

Black carbon (BC) is an important component of atmospheric pollution and has significant impacts on air quality and human health. Choosing Shanghai city for a case study, this paper explores the statistical characteristics and spatial patterns of BC concentrations using a mobile monitoring method, which differs from traditional fixed-site observations. Land use regression (LUR) modeling was conducted to examine the determinants for on-road BC concentrations, e.g. population, economic development, traffic, etc. These results showed that the average on-road BC concentrations were (9.86±8.68) μg·m^-3, with a significant spatial variation. BC concentrations in suburban areas [(10.47±2.04) μg·m^-3] were 32.03% (2.54 μg·m^-3) higher than those in the city center [(7.93±2.79) μg·m^-3]. Besides, meteorological factors (e.g. wind speed and relative humidity) and traffic variables (e.g. the length of roads, distance to provincial roads, distance to highway) had significant effects on on-road BC concentrations (r: 0.5-0.7, P<0.01). Moreover, the LUR model, including meteorological and traffic variables performed well (adjusted R²: 0.62-0.75, cross validation R²: 0.54-0.69, RMSE: 0.15-0.20 μg·m^-3), which demonstrates that on-road BC concentrations in Shanghai are mainly affected by these factors and traffic sources to some extent. Among them, the most accurate LUR model was developed with a 100 m buffer, followed by the LUR model with a 5 km buffer. This study is of great significance for the identification of spatial distribution patterns for on-road BC concentration and exploring their influencing factors in Shanghai, which can provide a scientific basis and theoretical support for simulating and predicting the response mechanisms of BC on human health and the natural environment.