Gravity tides extracted from SSA-denoised superconducting gravity data with the harmonic analysis: a case study at Wuhan station, China

Superconducting gravimetry is one main technique to determine the gravity tides. In superconducting gravimeter (SG) measurement, the gravity tides is affected by the gross errors (including spikes and severe deviations and mainly caused by instrument problems and earthquakes) and noise (caused due to the internal structure of SG instrument and measurement environment) in SG data. The singular spectrum analysis (SSA) method was proposed to correct and denoise SG data. Then the harmonic analysis method is used to precisely extract the gravity tides from SSA-denoised SG data. The SG data at Wuhan station of China are used in the case study. The results show that the accuracies of tidal amplitude factors, phase differences and amplitudes of most wave groups determined from SSA-denoised SG data by the harmonic analysis were improved significantly. The standard deviation of residuals between the synthetic gravity tides obtained by harmonic analysis of SSA-denoised SG data and theoretical gravity tides is 0.0096 mu Gal (1 mu Gal = 10(-8) m s(-2)), and is 0.1772 mu Gal while replacing the SSA-denoised SG data from the un-denoised SG data. The Akaike's Bayesian Information Criterion value of harmonic analysis on SSA-denoised SG data is better than that from un-denoised SG data, which indicates that SSA denoising on SG data can improve the accuracy of harmonic analysis results. The root mean square of gravity residuals obtained by harmonic analysis of SSA-denoised SG data is 0.0019 mu Gal, and is 0.2275 mu Gal when replacing the SSA-denoised SG data from the un-denoised SG data. The result shows that SSA can effectively eliminate most of gross errors and noise existing in the SG data, and improve accuracy of the gravity tides with the harmonic analysis.