Natural and anthropogenic aerosol emissions play a fundamental role both in directly modulating the incoming solar radiation and affecting air quality in the planetary boundary layer. Likewise, their indirect effects impact cloud lifetime, atmospheric column thermodynamics and precipitation patterns. For this reason, it is of crucial importance to assess aerosol spatial and temporal variability to reduce the uncertainty in forecasting future scenarios by the climatological models. In this study we developed an image based robust methodology that permits to retrieve the atmospheric path radiance and then the Aerosol Optical Depth (AOD) using satellite high-resolution spatial images paired with the Fu-Liou-Gu radiative transfer model. We applied our methodology to study aerosol variability in the PO valley (Northern Italy), one of the most polluted region in Europe.