Spatiotemporal distributions of roadside PM_2.5 and CO concentrations based on mobile observations

This study proposed a data preprocessing method for mobile traffic pollution observation based on previous work. The model was validated using 7days' (26 runs) observations of PM_2.5 and CO concentrations collected in Shanghai. The data revealed the spatial distributions and temporal variations of PM_2.5 and CO concentrations. Results showed that, an objective and comparable air pollutant distribution was characterized by the methods selected to remove the abnormal samples of high values, to correct the pollution background, and to determine the spatio-temporal scale. The high pollutant concentrations along the busy road intersections and their adjacent road sections were attributed to factors including large traffic flows, high proportion of diesel vehicles, frequent congestion and poor air flow. At these locations, PM_2.5 and CO concentrations were 1.7~2.8 and 12~20 times larger than observations on the clean campus, respectively. The living or production area showed about 3-fold higher PM_2.5 concentrations when compared with the campus, while this for CO in the living area was not prominent. The averaged PM_2.5 concentration of the whole area had a descending order in early morning, morning, afternoon and noon during a day. The averaged CO concentration was close in early morning and morning, which was greater than noon and afternoon. High humidity and low wind speed were unfavourable to air pollutant diffusion, and led to an accumulation of high pollutant concentrations along arterial roads in early morning.