Aerosol classification and bias-adjustment of global horizontal irradiance for Middle East-North Africa region

Abstract

Aerosol optical properties are considered as the most significant atmospheric constituent that affects global horizontal irradiance (GHI) in arid and semi-arid areas, like Middle East-North Africa (MENA) region. We examine the different types of aerosol in MENA region and the distribution of GHI using satellite-derived inputs to an Radiative Transfer Model (RTM) and ground-based observations at two reference stations located in MENA region, Tamanrasset, Algeria and Sede Boker, Israel. Regarding the aerosol clustering, coarse absorbing aerosols are mostly dominant in MENA sites covering a percentage of 99% and 55% respectively. The statistical verification of modeled GHI against ground-based observations, showed generally results. Two site-adaptation methods are applied for correcting the uncertainties delivered by the satellite-derived inputs and the GHI modeling process. Both methods eliminate the systematic errors providing MBE=0 and lower RMSE values. To combine site-adaptation with aerosol clustering results, an aerosol-conditioned linear correction is further performed. Sede Boker is used for the application of the methodology due to the more complicated aerosol pattern. According to results, RMSE is substantially improved. For the coarse absorbing cluster, RMSE value is 0.27 kWh/m2, in contrast to 0.33 kWh/m2 (RMSE independently of aerosol characterization). Similar RMSE pattern is found for the mixed aerosol cases.