M Memarianfard - Khaje Nasir Toosi University of Technology, Department of Civil Engineering, Tehran, Iran
A Mostafa Hatami - Khaje Nasir Toosi University of Technology, Department of Civil Engineering, Tehran, Iran
M Memarianfard - Khaje Nasir Toosi University of Technology, Department of Civil Engineering, Tehran, Iran
M Ajam - Khaje Nasir Toosi University of Technology, Department of Civil Engineering, Tehran, Iran

In many industrialized areas, the highest concentration of particulate matter, as a major concern on public health, is being felt worldwide problem. Since the air pollution assessment and its evaluation with considering spatial dispersion analysis because of various factors are complex, in this paper, GIS-based modeling approach was utilized to zoning PM2.5 dispersion over Tehran, during one year, from 21 March 2014 to 20 March 2015. The RBF method was applied to obtain the zoning maps and determining the highest concentration of PM2.5 in the 22 Tehran’s regions for each season. The RMSEmin values according to the number of neighbors and types of functions in the radial basis function method, including completely regularized spline, Spline with tension, Multiquadric function, Inverse multiquadric function, and Thin-plate spline for each month have been assessed. By performing analysis on the errors, the numbers of neighbors were estimated. The numbers of neighbors in the model for each function were varied from 2 to 30. The results indicate that the models with 3 and 4 neighbors have the best performance with the lowest RMSE values with using RBF method. The highest PM2.5 concentrations have been occurred in the summer and winter especially at the center, south, and in some cases at northeast of the city.

Air pollution; Completely regularized spline (CRS); Inverse multiquadric function (IMF); Multiquadric function (MF); PM2.5 concentrations; Radial basis function (RBF) method; RMSEmin; Spline with tension (SWT); Tehran’s regions; Thin-plate spline (TPS)