Incorporation of aerosol effects in a clear-sky semi-empirical model of UVER radiation for Valencia, Spain

The study develops a simple but physically based model for estimating erythemal irradiance [(UV erythemal radiation (UVER)] for cloudless urban regions with substantial aerosol load, such as that of the Valencia environment in Spain. The model estimates direct and diffuse UVER for cloudless aerosol-free conditions using the LibRadtran model with a DISORT algorithm. Aerosol optical thickness for UVER (tau(aER)) is obtained as a residual term by comparing measured direct UVER against modelled direct UVER in the absence of aerosol load. On average, tau(aER) is 18% higher than tau data extrapolated to 305 nm using the CIMEL CE-318. Comparison of measured direct and diffuse UVER with model estimates provide the single scattering albedo (omega(ER)), which on average is 0.88 for Burjassot, Valencia. When tested against independent data, hourly root mean square (RMS) errors are 6.2% if both aerosol properties are known independently. The error is 7.6% when monthly averages of these aerosol properties are used to estimate hourly erythemal irradiance in other years. However, errors are larger, at 18.0%, when monthly estimates of these aerosol properties are used at other experimental sites in the region with no aerosol data. Copyright (c) 2010 Royal Meteorological Society