Using a Sky Imager for aerosol characterization

Abstract

The All-Sky Imager developed by the Atmospheric Physics Group has been tested for aerosol characterization. Different neural network-based models calculate the aerosol optical depth (AOD) for two wavelengths and the Ångström turbidity parameter α using as input parameters data extracted from the principal plane of sky images from the All-Sky Imager. The models use data from a CIMEL CE318 radiometer for training and validation. The deviations between model and reference values are in the range of uncertainties assigned to Aerosol Robotic Network (AERONET) network. © 2007 Elsevier Ltd. All rights reserved.