AUTHOR=Karimian Shamsabadi Milad , Yeganeh Mansour , Pourmahabadian Elham 

TITLE=Urban buildings configuration and pollutant dispersion of PM 2.5 particulate to enhance air quality

JOURNAL=Frontiers in Sustainable Food Systems

VOLUME=Volume 6 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/sustainable-food-systems/articles/10.3389/fsufs.2022.898549

DOI=10.3389/fsufs.2022.898549

ISSN=2571-581X

ABSTRACT=A pivotal element for metropolitan planning and an important component describing the urban design is Block typology, affecting the pollution concentration. Consequently, this research points to examine the influence of various urban block typologies in urban pollutant distribution. Four typologies are simulated by ENVI-MET software. These are Separated single bloke, separated L-shaped, C _Shaped, horizontal Rows. Urban air quality was assessed using three parameters relative humidity, temperature, concentration pollution. 
The results showed that in the pattern one of group 4 in point A has the most pm.2.5 value (0.265), and after that single block group in pattern one has pm2.5 (0.240). In describing this behavior, the highest concentration can be said that increasing the opening of the winding path that moves through the pollution path, when the height of the building or by definition the height of the green walls increases, has an adverse effect causes pollution deposition. Horizontal Rows in pattern six revealed a minimum pollutant concentration at point A. 
Regression models propose a more reliable prediction of PM2.5 when the independent variables are temperature, relative humidity, and height of buildings, among various groups. The dependent variable is PM.2.5 in two different points, A and B. Hence, this article suggests a machine learning approach. Regression analysis suggests a better prediction of PM2.5. In the model evaluation, the results show that the Polynomial Linear Regression model (PLR) model is an excellent model for measuring at point A and B R-square is 0.942 and 0.967, respectively.