Search

Abstract: A long-term decrease in downward surface solar radiation from the 1950s to the 1980s (“global dimming”) followed by a multi-decadal increase up to the present (“brightening”) has been detected in many regions worldwide. In addition, some researchers have suggested the existence of an “early brightening” period in the first half of the 20th century. However, this latter phenomenon is an open issue due to the opposite results found in literature and the scarcity of solar radiation data during this period. This paper contributes to this relevant discussion analyzing, for the first time in Southern Europe, the atmospheric column transparency derived from pyrheliometer measurements in Madrid (Spain) for the period 1911–1928. This time series is one of the three longest datasets during the first quarter of the 20th century in Europe. The results showed the great effects of the Katmai eruption (June 1912, Alaska) on transparency values during 1912–1913 with maximum relative anomalies around 8%. Outside the period affected by this volcano, the atmospheric transparency exhibited a stable behavior with a slight negative trend without any statistical significance on an annual and seasonal basis. Overall, there is no evidence of a possible early brightening period in direct solar radiation in Madrid. This phenomenon is currently an open issue and further research is needed using the few sites with available experimental records during the first half of the 20th century. [Copyright &y& Elsevier]

Some works have reported variations in direct solar irradiance at surface with solar activity one order of magnitude higher than changes in total solar irradiance at the top of the atmosphere (TSI) over the solar cycle. In the present study, daily pyrheliometer records from the Astronomical Observatory of Madrid for the period 1910–1929 have been recovered. Metadata have been used to remove measurements affected by clouds, local pollution or any other factor dimming solar radiation noticed by means of visual observations of the state of the sky near the solar disk. Furthermore, corrections for volcanic eruptions, outliers and seasonal variations have been applied. When long-term changes in atmospheric transparency are taken into account in the analysis, non-significant variations in direct solar irradiance at surface with Sunspot Number (of the same order of magnitude as the variations in TSI) are found. Otherwise, a spurious high anticorrelation is obtained. This indicates, firstly, the influence of solar activity on solar irradiance at surface is so weak that atmospheric factors mask it, and secondly, careful analyses in order to infer this relationship have to be carried out not to reach results leading to wrong interpretations. • Non-significant variations in direct solar irradiance at surface with solar activity. • Spurious high correlation when changes in atmospheric transparency are not considered. • The solar signal in surface solar radiation is masked by atmospheric factors. [ABSTRACT FROM AUTHOR]