Impact of Aerosol Vertical Distribution on Satellite-Retrieved Aerosol Products

Till now, a number of remote sensing instruments have been providing aerosol datasets, such as historical sensors like Advanced Very High Resolution Radiometer (AVHRR), and modern sensors like Moderate Resolution Imaging Spectrometer (MODIS) and Visible Infrared Imaging Radiometer (VIIRS). Previous studies have found that the retrieved aerosol properties can have large bias especially in China, which is a densely-populated, heavily-polluted region in East Asia. This is mainly due to the uncertainties in the retrieval algorithm, such as surface reflectance parameterization, aerosol type assumption, aerosol vertical distribution and cloud screening strategy. In this study, we investigate the uncertainties associated with aerosol vertical distribution. We use the Second Simulation of the Satellite Signal in the Solar Spectrum (6S) atmospheric radiative transfer model to set up experiments and discuss the sensitivity of AOD error to the scale heights of aerosols by assuming three representative aerosol types, namely scattering fine particles, absorbing fine particles and dust. We also explore the effects of boundary layer in the AOD retrieval errors. The results show that AOD error is most sensitive to scale height for fine absorbing aerosols: ±1km scale height error can lead to ~40% AOD retrieval error. And the sign of AOD bias can change from negative to positive with increasing surface albedo. Also, ignoring the boundary layer can result in ~10% AOD error. After replacing the default aerosol scale heights with those calculated from lidar observations, we found that the accuracy of retrieved AOD greatly improved over North China, especially during the winter season when aerosol absorption is strongest. Furthermore, we explored the impact of diurnal change of aerosol vertical distribution on geostationary satellite retrieval. Our results show that with the development of the planetary boundary layer, the aerosol scale height can vary largely over the course of a day. If this effect is not taken into account, the retrieved AOD diurnal cycle can be significantly off or even reversed. Our study highlights the necessity of improving the aerosol vertical profiles in satellite retrieval algorithms, and the importance of 3 dimensional aerosol observation.