Review of air pollutants modeling techniques and methods.

Air pollution modeling has attracted experts in the field of the atmospheric environment due to its priority in future studies. In general, many studies conducted to model air pollutants such as proximity models, geostatistical interpolation, dispersion models, land use regression models, remote sensing, and artificial intelligence models. In this study, we review studies that are relevant to the modeling of air pollutants. We qualitatively and quantitatively evaluated these studies. Proximity models are considered the most basic methods to assess the quantitative, temporal, and spatial variation of air pollutants in a specific area. The main drawbacks of this technique, proximity models often use a limited number of parameters like traffic and roads neglecting important parameters such as weather information, and topography. In addition, these models neglect the type of vehicles like trucks and vehicles. Moreover, these models need self-reported approaches that represent the exposure to pollutants emitted from traffic, which may have been affected by biased information. On the other hand, this approach is usually performed based on the geostatistical interpolation techniques or GIS modeling, the main advantage of this method is; the precision of maps mainly relies on the accuracy of the used dataset and the interpolation algorithm used in GIS. Land use regression models are well-known models that use multiple linear regression algorithms to model the relationship between the dependent variable and independent variables. The main constraint of these models that they are difficult to be generalize for other study areas that include different natures. Dispersion models are important models that used various kinds of parameters such as emission data, topography, and meteorological data to model pollutants emissions in ambient air. The most recent approaches are the remote sensing methods that are used to detect and monitor air pollutants concentrations over large areas. The main drawbacks of remote sensing methods are that they are not suitable for high-resolution predictions in urban scales such as micro-scales studies. The most advanced techniques are artificial intelligence algorithms due to their ability to allow computer applications through learning from experience based on algorithmic training and iterative processing. This study can be a good guideline for researchers who are interested in the modeling of air pollutants. [ABSTRACT FROM AUTHOR]