A comprehensive review delineates advancements in retrieving particulate matter utilising satellite aerosol optical depth: Parameter consideration, data processing, models development and future perspectives

Particulate matter (PM), one of the major air pollutants, is generated by variety of natural or man-made sources, leading to acute and chronic diseases in humans since the last few decades. Employing satellite-derived aerosol optical depth (AOD) data, a wide technique that allows the retrieval of the PM concentration to support diverse on-field applications, including eliminating regular monitor stations and enhancing quality of breathing air. In this current review, the collection of data like ground-level particulate matter concentration, AOD, and metrological data, along with data screening, model grid and grid alignment, have been discussed. This paper endeavours the various development of models like single-stage, multi-stage, hierarchical, and ensemble models for PM prediction. Furthermore, the multicollinearity problem in developing the optimization model has been addressed. In order to evaluate the model efficiency, different performance evaluation techniques have been explored. Finally, the paper presents future directions for PM retrieval with the help of AOD data, such as the need for increased spatial and temporal resolution satellite data, incorporating more sophisticated machine learning models, and developing a global PM product from satellite AOD. The present review also accentuated the determination of PM in the air, which might be advantageous for the environment and human beings.