Methodical problems of magnetic field measurements in umbra of sunspots

Visual measurements of magnetic field strengths in sunspot umbra provide data on magnetic field strength modulus directly, i.e., irrespective from any solar atmosphere model assumptions. In order to increase the accuracy of calculation of the solar magnetic indexes, such as B‾max or Bsp, the inclusion of all available data from different observatories is needed. In such measurements some methodical problems arise, which bring about inconsistency of the data samples combined from different sources; this work describes the problems at hand and proposes solutions on how to eliminate the inconsistencies. Data sets of sunspot magnetic field strength visual measurements from Mt. Wilson, Crimea and Kyiv observatories in 2010–2012 have been processed. It is found that two measurement modes of Zeeman split, σ → σ and σ → π, yield almost the same results, if data rows are long enough (over ∼100 sunspots in central area of Sun, r < 0.7 R). It is generally held that the most reliable measurement results are obtained for magnetic fields that exceed 2400 G. However, the empirical comparison of the internal data consistency of the samples produced by different observers shows that for reliable results this limit can be lowered down to 1100 G. To increase the precision of measurements, empirical calibration of the line-shifter is required by using closely positioned telluric lines. Such calibrations have been performed at Kyiv and Crimea, but as far as we know, it has not been carried out at Mt. Wilson observatory after its diffraction grate was replaced in 1994. Taking into consideration the highest quality and coverage of Mt. Wilson sunspot observational data, the authors are convinced that reliable calibration of its instrument by narrow telluric lines is definitely required.