Analysis of high rate superconducting gravimeter data

The aim of this thesis is to investigate various procedures for the processing of high rate and high resolution superconducting gravimeter (SG) data. First, two SG gravity data sets from Cantley, Canada and Strasbourg, France are processed by following exactly the same steps. After we manually pre-process the recorded gravity, we perform a least squares procedure (HYCON package) to fit tides, atmospheric pressure and long term drifts. We obtain the gravity residuals and the corresponding power spectral density (PSD) using a Fast Fourier Transform. Subsequently a slew rate method is applied to automatically detect and correct the spikes and offsets in the Cantley data. A least squares procedure is again performed. The PSD of the gravity residuals show that there are no significant geophysical signals except harmonics of the daily atmospheric pressure in the sub-tidal frequency bands. In particular, we do not find evidence of the reported Slichter triplet.In order to fully exploit the various components of the gravity data and determine what factors affect the determination of the pressure admittances, tidal parameters ($ delta, kappa$) and signal levels, we have simulated the SG data by adding together of a synthetic tide, a brown noise, an actual atmospheric pressure record, a polynomial instrumental drift, artificial spikes and offsets, and a theoretical model for the Slichter triplet. The slew rate method and a more precise least squares procedure (ETERNA package) were applied to process the combined synthetic data. Using the PSD of the residual gravity signals in the sub-tidal bands, the theoretical Slichter triplet constructed in the data could be well recovered. Therefore we are confident the reported Slichter triplet should have been seen in our observed data