Air pollution is a global issue affecting both the environment and human health. While geographical and temporal data are fundamental to studies on pollution and monitoring emissions, there appears to be a lack of observational data, most of which is focused on populated areas. While monitoring stations provide highly accurate data on pollutant concentrations at high temporal resolution, spatial distribution is poor due to the high construction, operation, and maintenance costs associated with running each station. This instigates a large portion of the territory without measurements of pollutant concentrations, mainly in inhabited areas. The chemical transport models supplement monitoring-station data, providing extended temporal and spatial coverage. However, they also have their limitations: they require a great deal of input data, typically unavailable or unreliable for developing countries, such as meteorological and emissions data. Besides, some complex physical and chemical processes are simulated using simpler numerical and empirical equations; also, the evaluation of each region is required before use, which can be very challenging due to the lack of ground measurements. The limitations of ground-based measurements and chemical transport models are addressed by integrating remote sensing data. Although its temporal resolution and cloud cover are lower, it provides strong worldwide coverage with good spatial resolution. Furthermore, because remote sensing obtains only the complete column of each material, estimating surface-level concentrations requires the application of statistical models or machine learning algorithms.