An empirical model of erythemal ultraviolet radiation in the city of Valencia, Spain

Abstract

This paper presents an improved empirical model that predicts ultraviolet erythemal radiation (UVER) and considers all aspects of atmospheric conditions in Valencia, Spain. The analyzed model is a potential function whose dependent variable is UVER radiation and independent variables are the clearness index and slant ozone column. A potential regression function with all the information contributed a small coefficient of determination and one chose to use a regression potential–exponential mathematical form which improved the coefficient of similar determination. A study was carried out on the influence of season on the regression parameters. This was found to be considerable due to the clearness index. The convergence between the values calculated by the model and the experimental values was analyzed using the mean bias error (MBE) and mean absolute bias error (MABE) statistical parameters. The clearness index and ozone column intervals were analyzed and found to give an improved prediction of the UVER clearness index using regression analysis. Also, a sensitivity analysis was performed on the regression coefficients and parameters. It is important to study the effects of UVER radiation predicted by the model on human health or on agriculture crop growth and yield.