Dead time effect on the Brewer measurements: correction and estimated uncertainties

Abstract

Abstract. Brewer spectrophotometers are widely used instruments which perform spectral measurements of the direct, the scattered and the global solar UV irradiance. By processing these measurements a variety of secondary products can be derived such as the total columns of ozone (TOC), sulfur dioxide and nitrogen dioxide and aerosol optical properties. Estimating and limiting the uncertainties of the final products is of critical importance. High-quality data have a lot of applications and can provide accurate estimations of trends.The dead time is specific for each instrument and improper correction of the raw data for its effect may lead to important errors in the final products. The dead time value may change with time and, with the currently used methodology, it cannot always be determined accurately. For specific cases, such as for low ozone slant columns and high intensities of the direct solar irradiance, the error in the retrieved TOC, due to a 10 ns change in the dead time from its value in use, is found to be up to 5 %. The error in the calculation of UV irradiance can be as high as 12 % near the maximum operational limit of light intensities. While in the existing documentation it is indicated that the dead time effects are important when the error in the used value is greater than 2 ns, we found that for single-monochromator Brewers a 2 ns error in the dead time may lead to errors above the limit of 1 % in the calculation of TOC; thus the tolerance limit should be lowered. A new routine for the determination of the dead time from direct solar irradiance measurements has been created and tested and a validation of the operational algorithm has been performed.Additionally, new methods for the estimation and the validation of the dead time have been developed and are analytically described. Therefore, the present study, in addition to highlighting the importance of the dead time for the processing of Brewer data sets, also provides useful information for their quality control and re-evaluation.

Temperature dependence of the Brewer global UV measurements

“…However, insufficient correction for the effects of individual constructional and operational characteristics -e.g., stray light (Karppinen et al, 2014), dead time (Fountoulakis et al, 2016b), cosine response (Antón et al, 2008;Bais et al, 1998) and temperature dependence (Garane et al, 2006;Lakkala et al, 2008) may lead to even larger uncertainties (Gröbner et al, 2006). Thus, better understanding of the instrument's characteristics and improvement of the characterization methods are necessary for keeping the uncertainties within acceptable limits (Seckmeyer et al, 2001).…”

Response to comments of Referee #1

“…The most important corrections, which need to be applied after raw data measurements are corrections for dark counts, dead time and stray light, temperature correction and cosine error correction (Bais, 1997;Bernhard and Seckmeyer, 1999). Generally, the corrections for dark counts and dead time (Fountoulakis et al, 2016) are done using common practices described by the manufacturer (Kipp & Zonen, 2015), while corrections for stray light, temperature dependence and non-ideal angular response are more operator dependent. A usual way to correct for stray light is to consider that all counts which are measured at wavelengths shorter than 292 or 293 nm are stray light, and can be subtracted from the counts measured at other wavelengths (Mäkelä et al, 2016).…”

EuBrewNet – A European Brewer network (COST Action ES1207), an overview

“…Individual instruments must therefore be characterised to establish how component characteristics, e.g. slit function, wavelength calibration (Redondas et al, 2014 , Photomultiplier tube (PMT) linearity (Fountoulakis et al, 2016), spectral responsivity, temperature 25 dependence Fountoulakis 2017), stray light (Karppinen et al, 2015, Puli et al 2016, field of view, angular response, polarisation (Carreño et al, 2016) etc., uniquely affect measurement outcomes and how potential errors may be avoided or corrected for.…”