UV-A (315–400nm) irradiance from measurements at 380nm for solar water treatment and disinfection: Comparison between model and measurements in Buenos Aires, Argentina and Almería, Spain

Abstract

A linear correlation between UV-A and 380 nm was developed by means of the TUV 4.1 radiative transfer model. The prediction error of the correlation was evaluated with data from Buenos Aires, Argentina, 2001, and from 2006, Almería, Spain. Percent random mean square error (RMSE%) was calculated for intervals of 10° of solar zenith angles, ranging 4.75% at 20° to 37.70% at 90° in clear days and 22.16% at 20° to 26.17% at 90° for cloudy days in Buenos Aires Argentina, and 1.27% at 20° to 11.27% at 90° for clear days in Almeria, Spain. Clouded days were not assessed with the data from Spain. In Argentina, the UV-A radiometer is located in a rural area and the 380 nm radiometer is located in an urban area 6 km away. Hence the real error of the proposed model is closer to that found in Spain were both measurements were performed at the same site. The objective of the work is to achieve a simple and precise method to assess UV-A availability for environmental applications of solar energy, particularly for solar water treatment, at any desired latitude.