This series was calibrated by iteratively determining the personal k coefficients between parallel observers, working backward in time. The starting point of the calibration was the photographic catalog of the Royal Greenwich observatory, covering the modern period 1874-1976, instead of visual observers. The reconstruction goes back to the first telescopic observations in 1610.
Recent analyses have shown that the early Greenwich data (1874-1915) suffered from a progressive upward drift, which the first data before 1880 begin 40% lower than the sacle of the 20th century. This bias was propagated to the entire series before 1874 through the daisy-chaining of k coefficients. Moreover, the series also suffers from a ~10% underestimate over the last period 1976-1995, when a few alternate stations were susbtituted after the end of the Greenwhich catalogue.
Those large inhomogeneities and the inconsistency with the parallel sunspot number series motivated the revision of the group number, making this Version 1.0 deprecated. It is thus provided here only for historical reference when consulting past publications that used this version of the group number.
The files provided here are derived from the original files distributed by NOAA/NCEI and converted to a standard format more adapted for import into computer programs.
or by FTP at:
(NB: Some slight differences exist between the numbers contained in those files and the values given in the paper publications)
Note that we list here the un-scaled group numbers, while the original numbers were multiplied by the constant factor 12.08 to match the scale of the sunspot number. However, as this factor was affected by the above mentioned inhomogeneities in the data, it is unreliable and anyway artificial. Therefore, we dropped it to return to the actual basic group counts.
The series stops in 1995, the final year contained in the 1998 publication.