Abstract
Climate change perspectives intensified investigations of the radiative balance of the Earth‐atmosphere system. At the top of the atmosphere, solar irradiance is known with absolute uncertainty of 0.3% and theoretical models agree with albedo measurements, but solar shortwave radiation observations at Earth's surface are less than those calculated by radiative‐transfer models. This model observation discrepancy (10–25 Wm−2) led to a decade‐long controversy on unexplained enhanced absorption of shortwave radiation in clear sky atmospheres as well as in clouds. Here we show evidence for underestimation of surface shortwave irradiance by traditional “unconditioned” global and diffuse pyranometer measurements. Reinvestigations of pyranometer calibration in conjunction with thermal offsets and pyranometer thermal control demonstrate an underestimation of clear sky solar global, as well as diffuse irradiance by 8–20 Wm−2, caused by pyranometer differential cooling. Field measurements with “conditioned” and “unconditioned” pyranometers demonstrate that the so‐called night offset is present and considerably larger during daytime measurements, and this not only for diffuse but also for global pyranometer measurements. Long‐term comparisons between traditional unconditioned and well‐conditioned pyranometer measurements at Davos (midlatitude, 1580 m a.s.l.) show differences of several percent on the annual mean of global irradiance. Even though we are aware that measurements at higher altitudes are subject to larger thermal offsets and not representative for the global average, the results of our experiment lead us to believe that surface solar irradiance, measured in the past throughout the globe by traditional unconditioned pyranometers, is underestimated.