A machine learning-based framework for high resolution mapping of PM2.5 in Tehran, Iran, using MAIAC AOD data

This paper investigates the possibility of high resolution mapping PM2.5 concentration over Tehran city using satellite AOD (MAIAC) retrievals. For this purpose, a framework including three main stages, data preprocessing; regression modeling; and model deployment was proposed. The output machine learning trained to predict from MAIAC retrievals meteorological data. results testing revealed efficiency capability developed for PM2.5, which not realized in former investigations performed city. Thus, study, first time, daily, 1 km with R2 around 0.74 RMSE better than 9.0 μgm3.