The present experiment investigates the effect of the green belt structure on the pollutant concentration above roads for non-motorized vehicles and provides a basis for designing a road green belt pattern. We simulated green belts separated by 10, 20, and 30 m intervals with shade cloth. The three simulations were of 10 m roads alternating with 10 m green belts, 10 m roads alternating with 12.5 m green belts, and no roads in an 80 m green belt. Microclimate (wind) conditions and the concentrations of five major pollutants, namely SO₂, NO_x, NH₃, total suspended particles (TSP) and respirable particulate matter (PM10), were observed for all types of road. The observation periods were from 7 to 8 am, from 9 to 10 am, from 11 am to 12 pm, from 1 to 2 pm, and from 3 to 4 pm on the same day. The results show that when the wind speed was less than 2 m/s, the green belts separated by 10 m and 20 m intervals had a "micro-canyon effect", and the wind speed above the green belt interval increased by 23.33% and 20.22%, respectively, whereas the wind speed above green belts separated by 30 m reduced by 3.33%. When the wind speed exceeded 2 m/s, none of the green belts displayed a micro-canyon effect. There was a nonlinear relationship between increasing the interval width and the percentage removal of pollutants, with the green belts at 10 m intervals having the greatest effect. When the wind speed dropped below 2 m/s, for the 10 m interval, the daily average percentage removals of the five pollutants SO₂, NO_x, NH₃, TSP and PM10 were 32.01%, 29.43%, 7.66%, 54.27% and 53.90%, respectively. When the wind speed exceeded 2 m/s, for the 10 m interval, the daily average percentage removals of the pollutants were 16.50%, 16.90%, 12.02%, 32.86% and 25.14%, respectively. It was also found that the percentage removal of the five pollutants had positive correlation with the wind speed for the 10 m interval but not for the 20 m and 30 m intervals.
The structure of 12.5 m green belts at 10 m intervals can more effectively reduce the pollutant concentration of roads for non-motorized vehicles; the daily average percentage removals of the pollutants were 8.74%, 9.13%, 7.63%, 13.01% and 15.56% when the wind speed was less than 2 m/s and 7.03%, 3.51%, 10.4%, 24.35% and -14.28% when the wind speed exceeded 2 m/s, respectively. The positive correlations between the removal percentages of the pollutants and the wind speed were the most significant for this green belt mode. The removal percentages of pollutants were small or negative for other green belt modes. The green belt structure affects the micro-climatic conditions of the road and thus the pollutant concentration over roads for non-motorized vehicles. Applying a reasonable structure for the urban road green belt, we can improve air quality over roads for non-motorized vehicles.