Analysis of observations with dual sensor superconducting gravimeters

Among the 21 superconducting gravimeters presently operating worldwide four instruments exist that are equipped with two vertically aligned sensor units. Three of the instruments are installed in Germany (Bad Homburg, Moxa, Wettzell) and one in South Africa (Sutherland). Comparisons of the data sets obtained with the dual sensor systems yield information on instrumental effects and sensitivity as well as on the efficiency of reductions of environmental effects applied to the data. The latter is an important constraint when looking for small geodynamic signals like Slichter and core modes or aperiodic variations.From analyses of the two data sets of each instrument a small but significant difference of 1-3% in the response of the sensor units on barometric pressure variations is found. Likewise, the records of lower and upper sensor vary slightly but not systematically with regard to the noise levels in the different frequency ranges. The tidal analyses yield an agreement of the tidal parameters generally well within the standard deviations determined from the least squares adjustment in the tidal analysis. The deviations are in the range between 0×10⁻⁴ and 3×10⁻⁴ for the amplitude factor and the phases differ between 0.0005° and 0.01° for the four main tidal constituents O1, K1, M2, and S2.The comparison of the gravity residuals of the two sensors with each other as well as with their sum and difference in the time and frequency domain shows the existence of identical signals in the records of the two sensors in the whole range of observation. This probably means that either the environmental reductions applied are not sufficient or there are additional disturbing effects in the data which have not been taken care of yet. From the study it emerges that it is not possible to get entirely rid of the tidal signals in the data. This is probably also due to the fact that despite reductions the data sets contain additional signals and slightly different noise at tidal frequencies which affect the result of the tidal analysis.