ARIMA analysis of the effect of land surface coverage on PM10 concentrations in a high-altitude megacity

This paper uses ARIMA models for daily temporal analysis of the effect of land surface coverage (LSC) on PM10 concentrations in a high-altitude megacity. Bogota, the capital of Colombia, is the urban center with the greatest population density and third-highest air pollution levels in Latin America. Six automatic monitoring stations were used; they were equipped with measurement instruments for PM10, temperature and solar radiation as well as wind speed and wind direction. The duration of the sampling period was 6 years. The hourly PM10 sampling system included continuous-monitoring equipment that used beta ray attenuation. We analyzed atmospheric stability and the spatial distribution of LSC (vegetated, non-vegetated, impervious and water bodies) before applying the iterative process of Box-Jenkins for ARIMA models. ARIMA analysis indicates greater persistence in PM10 pollution in the presence of increased vegetated LSC (trees and grasslands); persistence decreased in the presence of more impervious LSC (roofs, pavements and footpaths). PM10 persistence is found to be 2 days (48 h). The best distance to demonstrate these findings is between 50 and 100 m, with respect to the monitoring stations' physical location. Urban areas with a predominance of vegetated LSC register lower PM10 concentrations than urban areas with a predominance of impervious LSC (average daily difference = 42.7%). This study's findings serve as a reference point for the development of differentiated strategies for air pollution control in line with urban LSC.