Estimation of PM2.5 from Satellite Observations with a Physical-informed Method

Satellite observations can be successfully used to increase the spatial and temporal resolution of ground-based data. Recently, many studies exploited satellite observations to estimate the Particulate Matter (PM) mass concentration near the ground from Aerosol Optical Depth (AOD) data. This is a promising field of research, which can expand the coverage of local and regional ground-based networks. However, significant gaps remain in the correct representation of local aerosol properties such as the Effective Radius, and of the aerosol growth due to high humidity conditions (i.e., hygroscopic growth). Standard PM measurements are taken at dry conditions, thus needing a corrective function to successfully correlate with satellite AOD. This study exploits ground-based measurements taken at ambient conditions by an Optical Particle Counter (OPC) in the city of Bologna, in the Po Valley. In this way, the need for corrective functions is overcome. The OPC used in this study, the LOAC (Light Optical Aerosol Counter), measures Particle Number concentration and provides information on particle typology, which are used to compute the Effective Radius and the PM at ambient Relative Humidity (RH). A linear regression between AOD and PM is performed by exploiting the theoretical relationship which depends on the local aerosol characteristics estimated from the available LOAC data. Results show that using all the auxiliary data (instead of the PM alone) improves the linear regression parameters: as an example, the Pearson's correlation coefficient increases from 0.55 to 0.76. The increase is independent on the season. The linear function between AOD and PM has a slope of 0.81 or 1.04 depending on the estimate of the Effective Radius considered.