Retrieval of aerosol single scattering albedo using joint satellite and surface visibility measurements

Aerosol single scattering albedo is a critical optical parameter that determines aerosol radiative effect. However, most existing passive satellite sensors such as MODIS and VIIRS only measures the intensity of reflected solar radiation and can only retrieve aerosol optical depth, while aerosol single scattering albedo needs to be assumed in the retrieval algorithm. On the other hand, if aerosol optical depth is known, it would be possible to retrieve aerosol single scattering albedo using satellite sensors.  In this study, we develop a machine learning based algorithm that retrieves aerosol single scattering albedo using joint visibility and satellite measurements. Combined with meteorology variables including relative humidity and boundary layer height, surface visibility can be converted to column aerosol optical depth. Then combining this converted aerosol optical depth with VIIRS measured TOA apparent reflectance, we retrieve aerosol single scattering albedo at over 2000 stations worldwide. The results compare well with AERONET retrieved SSA. However, compared with AERONET, visibility is recorded at every WMO meteorology station and has much more extensive coverage. We also applied our method to surface PM2.5 measurements obtained satisfactory results. Our work provides an aerosol single scattering albedo dataset with extensive coverage over land, which can be used for aerosol radiative forcing calculations and model validation.