Research on the horizontal reduction effect of urban roadside green belt on atmospheric particulate matter in a semi-arid area

In this study, the horizontal abatement effects of green belts on atmospheric particulate matter at different horizontal distances and plant community structures were investigated in urban roadside green belts in semi-arid areas.We collected mass concentrations of six types of atmospheric particulate matter (PM) per unit time of PM0.3, PM0.5, PM1.0, PM2.5, PM5.0, and PM10 and various meteorological indices, to compare the horizontal reduction efficiency of different distances and plant community structures on different particle sizes, and to establish a support vector machine model. The results showed that 1) the horizontal abatement efficiency of green belts was different for six particle sizes, while the horizontal abatement rate strengthened as the particle size increased. The abatement rate was significantly correlated with microclimatic factors such as temperature and humidity, but less correlated with wind speed. 2) The horizontal abatement rate of roadside green belts on atmospheric particulate matter varies with the increase of horizontal distance in a "single-peak" or "double-peak" pattern, with the best abatement effect of green belts on each particle size at a horizontal distance of 45m. Among the four types of plant community structures, the strongest abatement ability was in the arbor-shrub-herb structure. 3) The correctly tuned prediction model, based on Support Vector Machines, resulted in better horizontal abatement ability of green belts on atmospheric PM. The prediction results showed that the average horizontal abatement rate has the best abatement effect at 45–55 m, peaked at 50 m, and formed stagnant dust at 65 m. In the design of urban road green spaces in semi-arid areas, to achieve the best dust retention effect, the green belt width should be ≥40 m and it is desirable to choose arbor-shrub-herb structures. This study provides a design basis and theoretical support for urban road green space planning in semi-arid areas.