Sunspot Group Numbers Since 1900 and Implications for the Long-term Record of Solar Activity

Recent work on improving and revising estimates of solar activity [Clette et al., 2014] has resulted in renewed interest in what has been called the longest running 'Science Experiment'. We compare four reconstructions of solar activity as reflected in the number of sunspot groups ('active regions') constructed by different authors using very different methods. We concentrate on the period since AD 1900 where the underlying solar and geomagnetic data are plentiful and of sufficient quality and find that all four methods yield essentially the same Sunspot Group Number series. We take that as indicating that protracted and pernicious criticisms of the individual methods are neither fruitful nor helpful and we suggest that future efforts be directed towards understanding the specific reasons why the methods give discordant results for centuries prior to the 20th. The main area of disagreement occurs during the last 25 years of the 19th century and feeds back into the time prior to that. The solar Extreme Ultraviolet flux can be reconstructed since the 1740s [Svalgaard, 2016] and with suitable scaling fits the Svalgaard & Schatten [2016] Sunspot Group Number series since 1865 very well, so we argue that the discordant group series have problems once we move out of the 20th century, and that the community should concentrate on finding out what those are, so a true and useful consensus can emerge.