Ladislav Budík

Method of estimating solar UV radiation in high-latitude locations based on satellite ozone retrieval with an improved algorithm

The effects of the cloudiness and satellite-based ozone measurements on erythemally effective ultraviolet (EUV) radiation were examined using a non-linear regression model. Instead of the widely used ozone transmissivity exponential function, we proposed a new approach based on a quantum transmission model using hyperbolic attenuation of the EUV radiation. The radiation data were collected at the Czech Johann Gregor Mendel Station, James Ross Island, Antarctica (63° 48′ S, 57° 53′ W), between 14 March 2007 and 3 March 2009. The total ozone content and effective surface reflectivity at 360 nm were obtained from the Ozone Monitoring Instrument on board the EOS-Aura spacecraft for the geographical coordinates of the J. G. Mendel Station. The model predicted 98.6% variability of the EUV radiation. The residuals between the measured and predicted EUV radiation intensities were evaluated separately for the ranges of solar elevation angle, total ozone content and surface reflectivity. The results of this study were compared to previous findings where the influence of ground-based and satellite-based ozone measurements and model usefulness was discussed.

Estimation of solar UV radiation in maritime Antarctica using a nonlinear model including cloud effects

A new approach to the estimation of erythemally effective ultraviolet (EUV) radiation for all sky conditions that occur in maritime Antarctica is reported. The spatial variability of the total ozone content (TOC) and attenuation of the EUV radiation in the atmosphere are taken into consideration. The proposed nonlinear regression model of EUV radiation is described by a hyperbolic transmission function. The first results and the model validation for Vernadsky Station (formerly the British Faraday Station) during the period 2002–2005 show very good agreement with the measured values (R 2 = 99.2). The developed model was evaluated using daily doses of EUV radiation with respect to solar elevation angle and cloudiness. The mean average prediction error (MAPE) for cloudy (4.1–7.0 oktas) and overcast skies (7.1–8.0 oktas) varied between 4.0% and 4.3%, while for partly cloudy days (0–4.0 oktas) with high variability of cloud types during a day, MAPE reached 5.9%.