Validation of the GUM uncertainty framework and the Unscented transformation for Brewer UV irradiance measurements using the Monte Carlo method

Abstract

The uncertainty evaluation of spectral ultraviolet (UV) irradiance measurements is usually performed following the recommendations set by the GUM, the standard for uncertainty evaluation in metrology. In particular, the GUM uncertainty framework (GUF) is applied. Due to the difficulties in the propagation of uncertainties, it is standard practice to neglect the non-linearity of the irradiance model. To verify this assumption, the GUF is validated using the Monte Carlo method (MCM), as recommended by the GUM. As an alternative to these methods, an Unscented transformation (UT) has also been implemented and validated with the MCM. The data used are the UV scans recorded by a Brewer MKIII spectroradiometer at El Arenosillo Observatory. Evaluation of the spectral ratios between the combined standard uncertainties provided by the GUF, the UT, and the MCM showed that the agreement between the three methodologies is satisfactory.