Applications of Artificial Intelligence and Remote Sensing for Air Quality Analysis in Dammam and Jubail, Saudi Arabia

Abstract

Dammam and Jubail face significant environmental and public health challenges due to elevated air pollution resulting from industrial activities and traffic emissions. This study aimed to assess air quality in both cities during the period 2019–2024 by utilizing satellite data (Sentinel-5P and MODIS) in conjunction with artificial intelligence models. The results revealed that Jubail recorded higher concentrations of fine particulate matter (PM₂.₅), averaging 45.28 μg/m³, compared to Dammam (42.04 μg/m³). Additionally, NO₂ and SO₂ levels were higher in Jubail, reflecting the pronounced impact of industrial activities, while no significant difference was observed in CO concentrations between the two cities. The XGBoost model was employed to predict air quality, achieving high accuracy (R² between 0.96 and 0.98), with prediction errors being greater in Jubail. The study confirmed that industrial emissions and traffic are the primary sources of pollution, although some improvement in certain pollutants was observed due to environmental policies. The study recommends enhancing vehicle efficiency, expanding green spaces, and developing stricter environmental policies to mitigate rising NO₂ and O₃ levels and to achieve a more sustainable environment.