A hybrid model of environmental impact assessment of PM2.5 concentration using multi-criteria decision-making (MCDM) and geographical information system (GIS)—a case study

Abstract

In the last decades, Tehran, as the capital of Iran, has faced severe air pollution. Considering the importance of air pollution, particularly PM2.5, the study assesses the environmental impacts (EIA) of PM2.5 in the Tehran metropolis using 4-year consecutive data (2014–2017). For this purpose, the inverse distance weighting model, geographically weighted regression (GWR) model, and fuzzy analytic network process (FANP) for interpolation of PM2.5 concentration, measuring the relationship between main factors and PM2.5 concentration, and EIA were utilized, respectively. The research findings showed that the highest PM2.5 concentration was 56 μg/m3 from 2014 to 2015, then with a decreasing trend in 2016 and 2017. Moreover, the local R2 ranges obtained from the GWR model between PM2.5 concentration and traffic, population distribution, land use, gas, and fuel consumption are 30–69%, 14–23%, 44–76%, 31–42%, and 1–35%, respectively. In the following, in order to study the EIA of PM2.5 using the FANP model, four alternatives, strict environmental management in the surrounding neighborhoods, change in the country’s economic capital, change in the urban management, and NO action, were proposed. Then, the relationships of dependence and pairwise comparisons among criteria and alternatives were performed through grades of membership for linguistic variables of fuzzy sets. Next, the unweighted fuzzy supermatrix was constructed, and the weighted fuzzy supermatrix and limit supermatrix were calculated, respectively. Finally, the centroid of area technique was applied for the defuzzification process and prioritization of the alternatives. The EIA results of the FANP model demonstrated that strict environmental management in surrounding neighborhoods by having the highest weight (38%) is the best alternative.