Integrating Sentinel-5P Satellite Data and Machine Learning Algorithms for Air Quality Index Prediction in Tehran: A Comprehensive Study on Factors Influencing Air Quality

Abstract

Frequent air pollution episodes pose severe health and environmental challenges in Tehran, Iran. Despite recent efforts, pollutant levels often exceed WHO-based national standards. This study addresses the pressing need for accurate air quality prediction by leveraging advanced satellite data and machine learning techniques. Our methodology integrates Sentinel-5P satellite data with optical depth remote sensing information. We systematically evaluated five machine learning algorithms to identify the most effective approach for AQI prediction. This study aims to advance air quality prediction in Tehran by integrating Sentinel-5P satellite data with machine learning algorithms. We examined the efficacy of various algorithms, including Decision Tree, K-Nearest Neighbors, Random Forest, Support Vector Machine, and Logistic Regression, in correlating air pollutant levels with the Air Quality Index (AQI). The selection criteria focused on algorithmic efficiency and accuracy in handling diverse environmental datasets. The Random Forest algorithm, utilizing Sentinel-5P and optical depth data, achieved a remarkable accuracy of 74% in predicting AQI. Further enhancement was observed by incorporating climatic data, COVID-19 status, and environmental parameters; the model achieved a significant predictive accuracy of up to 75.6%. These findings underscore the critical impact of nitrogen dioxide, ozone, and aerosol optical depth on Tehran's AQI, with notable variations observed post-COVID-19 restrictions. The increase in AQI following the lifting of COVID-19 restrictions suggests a significant correlation between human activity and air quality. These insights can inform targeted environmental policies in Tehran. We demonstrate the potential of integrating satellite data with machine learning to predict AQI accurately. Our approach offers a scalable model for urban air quality management with implications for environmental policy and public health initiatives.