The impact of different aerosol layering conditions on the high-resolution MODIS/MAIAC AOD retrieval bias: The uncertainty analysis

The Eastern Mediterranean/Middle East (EMME) region belongs to one of the most polluted and vulnerable to climate change related areas in the world. To monitor these changes, comprehensive set of measurements need to be conducted. Ground-level monitoring sites provide continuous measurements, yet their spatial coverage in EMME is very limited. Satellite data largely expand spatial coverage, however, retrieval accuracy over arid regions with bright surface background, mixed with urban and industrial sources, remains challenging. In this study, we analyzed the agreement between aerosol optical depth (AOD) data from the Multi-Angle Implementation of Atmospheric Correction (MAIAC) algorithm at 1-km spatial resolution and corresponding ground-based Aerosol Robotic Network (AERONET) measurements, using data from September 2019 to October 2022 in Tel Aviv, Israel. We identified all overestimation and underestimation MAIAC AOD measurements, relative to AERONET, and used multiwavelength polarization lidar observations of vertical aerosol profiles to characterize these conditions. Our findings suggest that under clear atmospheric conditions (AOD<0.12), MAIAC overestimation prevails, while under dusty (AOD>0.3), and partially cloudy conditions, MAIAC underestimation prevails. We found that dust component (high particle depolarization ratio, and low Angstrom exponent) is often presented in cases of underestimation, while a single layer overestimation cases typically related to anthropogenic pollution. The two-layers overestimation cases tend to have marine aerosols at the bottom and mixed pollution sources at the top. Our study highlights the importance of studying the different layering conditions that largely bias the MAIAC AOD retrieval accuracy. This knowledge is highly important since AOD is widely used as input variable in numerous modeling studies and air quality applications and rarely prepossessed for such a bias.