Analyzing spatiotemporal concentration of PM2.5 in Isfahan city using random forest methods
Nowadays, industrialization, urbanization and population growth are known as the main causes of air pollution in different cities. Due to the limited accessibility to air pollution monitoring stations because of their cost, access to high spatial and temporal coverage of air pollutants and their distribution is very difficult. To address this issue, the main purpose of this study was to estimate daily PM2.5 concentration in Isfahan city from March 2018 to March 2020. For this purpose, we integrated random forest (RF) algorithm with the remotely sensed aerosol optical depth and meteorological data. In order to evaluate the accuracy of the estimated PM2.5, the root mean square error (RMSE), mean absolute error (MAE), and coefficient of determination (R2) were used as the evaluation metrics. The obtained results showed that the employed methodology performed well in estimating PM2.5 through the study period in Isfahan city. The averaged RMSE, MAE, and R2 were 4.15 (µg/m3), 3.27 (µg/m3), and 0.73, respectively. The PM2.5 trend analysis showed an uneven distribution of PM2.5 during the study period. The results of this study can be beneficial for decision makers and city planners to provide useful strategies to alleviate the negative impact of PM2.5 concentrations and variations.
Copyright (c) 2023 Soolmaz Shamsaei, Mozhgan Ahmadi Nadoushan (Author)
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.