The effect of forest on PM2.5 concentrations: A spatial panel approach

Spatial relationships between forest and PM2.5 concentrations are of great policy implications in regional afforestation layout and air pollution control. This paper investigates the transboundary externality of a city's forest on the concentrations of PM2.5 in different city segments. Employing a mixed-regressive spatial panel model with data for 255 Chinese cities over 2000 to 2015, we find that the concentrations of PM2.5 tend to be substantially lower in cities with larger forest area and the depositing effect of forest spills over significantly to neighboring cities. A one percentage increase in forest area reduces the average annual concentrations of PM2.5 by 2.53%, of which 76% is contributed to the spillover effect. Moreover, the average marginal effect of forest on PM2.5 concentrations exhibits an inverted-U relationship with wind speed and the depositing effect minimizes (in magnitude) as the average annual speed of wind approaches to 23 kilometers per hour. These findings suggest that severe hazing cities with mild wind speed are priority afforestation areas for transboundary air pollution control.