Satellite-Based Assessment of Air Pollution Dynamics and Land Surface Characteristics in Riyadh, Saudi Arabia Using Sentinel-5P, MODIS, and Google Earth Engine

<title>Abstract</title> <p>Air pollution has become a major environmental concern in rapidly urbanizing arid regions, where anthropogenic emissions, land-use change, and climatic conditions collectively influence atmospheric quality. This study investigated the spatiotemporal dynamics of atmospheric pollutants and land surface characteristics in Riyadh, Saudi Arabia, during 2018–2025 by integrating Sentinel-5P TROPOMI atmospheric observations, MODIS-derived Land Surface Temperature (LST) and Normalized Difference Vegetation Index (NDVI), Sentinel-2 Land Use/Land Cover (LULC), and Google Earth Engine (GEE). Annual and seasonal variations of NO₂, SO₂, CO, O₃, the Absorbing Aerosol Index (AAI), LST, NDVI, and LULC were analyzed using spatial analysis, Pearson correlation, Principal Component Analysis (PCA), and Multiple Linear Regression (MLR). The results revealed pronounced spatial and seasonal variability in all pollutants. NO₂ and CO were primarily concentrated over urban and industrial areas, whereas O₃ exhibited higher concentrations in downwind rural regions due to photochemical production. LST reached a maximum of 43.10°C in 2021 and gradually declined to 37.07°C in 2025, while NDVI showed a gradual increase after 2022, reflecting localized urban greening. Pearson correlation revealed a strong positive relationship between NO₂ and CO (r = 0.81), and PCA explained 55.8% of the total variance, emphasizing the dominant role of seasonal meteorological conditions in pollutant variability. Furthermore, the MLR models produced R² values ranging from 0.124 for O₃ to 0.231 for SO₂, with intermediate values of 0.155 for CO and 0.219 for NO₂, indicating that land surface characteristics (LST, NDVI, and AI) partially explain pollutant variability, whereas emission intensity and atmospheric processes remain the dominant controlling factors. Overall, the integration of multi-source satellite observations with cloud-based geospatial analysis and statistical modelling provides a robust framework for long-term environmental monitoring and supports sustainable urban planning and air quality management in rapidly developing arid cities.</p>