Characterisation of a smartphone image sensor response to direct solar 305nm irradiation at high air masses

This research reports the first time the sensitivity, properties and response of a smartphone image sensor that has been used to characterise the photobiologically important direct UVB solar irradiances at 305nm in clear sky conditions at high air masses. Solar images taken from Autumn to Spring were analysed using a custom Python script, written to develop and apply an adaptive threshold to mitigate the effects of both noise and hot-pixel aberrations in the images. The images were taken in an unobstructed area, observing from a solar zenith angle as high as 84° (air mass=9.6) to local solar maximum (up to a solar zenith angle of 23°) to fully develop the calibration model in temperatures that varied from 2°C to 24°C. The mean ozone thickness throughout all observations was 281±18 DU (to 2 standard deviations). A Langley Plot was used to confirm that there were constant atmospheric conditions throughout the observations. The quadratic calibration model developed has a strong correlation between the red colour channel from the smartphone with the Microtops measurements of the direct sun 305nm UV, with a coefficient of determination of 0.998 and very low standard errors. Validation of the model verified the robustness of the method and the model, with an average discrepancy of only 5% between smartphone derived and Microtops observed direct solar irradiances at 305nm. The results demonstrate the effectiveness of using the smartphone image sensor as a means to measure photobiologically important solar UVB radiation. The use of ubiquitous portable technologies, such as smartphones and laptop computers to perform data collection and analysis of solar UVB observations is an example of how scientific investigations can be performed by citizen science based individuals and groups, communities and schools.