Measurement and modeling of particulate matter concentrations: Applying spatial analysis and regression techniques to assess air quality

Abstract

This paper presented the levels of PM2.5 and PM10 in different stations at the city of Sabzevar, Iran. Furthermore, this study was an attempt to evaluate spatial interpolation methods for determining the PM2.5 and PM10 concentrations in the city of Sabzevar. Particulate matters were measured by Haz-Dust EPAM at 48 stations. Then, four interpolating models, including Radial Basis Functions (RBF), Inverse Distance Weighting (IDW), Ordinary Kriging (OK), and Universal Kriging (UK) were used to investigate the status of air pollution in the city. Root Mean Square Error (RMSE), Mean Absolute Error (MAE) and Mean Absolute Percentage Error (MAPE) were employed to compare the four models. The results showed that the PM2.5 concentrations in the stations were between 10 and 500μg/m3. Furthermore, the PM10 concentrations for all of 48 stations ranged from 20 to 1500μg/m3. The concentrations obtained for the period of nine months were greater than the standard limits. There was difference in the values of MAPE, RMSE, MBE, and MAE. The results indicated that the MAPE in IDW method was lower than other methods: (41.05 for PM2.5 and 25.89 for PM10). The best interpolation method for the particulate matter (PM2.5 and PM10) seemed to be IDW method.•The PM10 and PM2.5 concentration measurements were performed in the period of warm and risky in terms of particulate matter at 2016.•Concentrations of PM2.5 and PM10 were measured by a monitoring device, environmental dust model Haz-Dust EPAM 5000.•Interpolation is used to convert data from observation points to continuous fields to compare spatial patterns sampled by these measurements with spatial patterns of other spatial entities.