Processing Of Gravity Tides With Singular Spectrum Analysis Based On Relative Gravimetric Data

The earth tide is the deformation of the earth under the force of the moon,the sun and other celestial bodies.The study of the earth tide is a discipline closely related to astronomy,geophysics,geodesy and oceanography.By comparing the observations of the earth tide with the theoretical values calculated from the earth model,the internal structure of the earth can be studied.Besides,by observing and analyzing the spatial distribution characteristics of the earth tide on the earth surface,the ocean tide structure,the crust and upper mantle structure,and the tectonic movement of the earth's crust can be studied.Moreover,at present,gravity measurement and distance measurement have been able to achieve high accuracies,and the gravity of any point on the earth surface and the distance between the point and geocenter can be expressed as functions of time.Therefore,in precise gravity and distance measurement,the influence of the earth tide must be considered.According to representations,the earth tide can be divided into gravity tides,ocean tides,strain tides,etc.In this paper,we mainly studied gravity tides and proposed to apply singular spectrum analysis(SS A)to gravity measurement data processing.In this paper,the SSA method was firstly applied to extract gravity tides without any priori information(e.g.station coordinate)from relative gravimetric data.The SSA and Wavelet results of extracting gravity tides from simulated signals were compared.The RMS of residuals between gravity tides extracted with SSA and theoretical values is 0.3?Gal,and the RMS of residuals between gravity tides extracted with Wavelet and theoretical values is 1.6 ?Gal.Based on the relative gravimetric data observed by CG-5 Autograv,a new method of extrapolating time series with SSA to weaken the impact of SSA boundary effect is putted forward.The experimental results indicate that the RMS and STD of residuals between the gravity tides and the model theoretical values are all reduced.Based on the above processing,the gravity tides extracted with SSA are consistent with model theoretical values calculated by Tamura table,and the RMS of residuals between them is only 2.2 ?Gal.In addition,the zero drift obtained with SSA is basically consistent with the value obtained with the least squares fitting method.The absolute difference between them is less than 0.4?Gal/d.For the long-term superconducting gravity data obtained from superconducting gravity measurement,the SSA method was applied for interference signal correction and denoising on the original gravity observation data.The accuracy of harmonic analysis was improved based on that processing.The superconducting gravity data before and after SSA denoising were processed with harmonic analysis by Baytap-G,respectively.The results show that after denoising with SSA,the accuracies of amplitude factor and phase lag of most wave groups obtained by Baytap-G harmonic analysis were improved significantly.The residual standard deviations between synthetic tide and theoretical gravity tide are 0.1775 ?Gal and 0.0098 ?Gal respectively.The ABIC(Akaike's Bayesian Information Criterion)values of harmonic analysis for denoised data are better than those before denoising,which indicates that that SSA denoising on gravity data can improve the accuracy of harmonic analysis.In addition,the RMS values of gravity residuals obtained by two harmonic analyses are 0.1646 ?Gal and 0.0013 ?Gal respectively,which indicates that SSA method can effectively eliminate most of the background noise existing in the original gravity observation data,and proves that the SSA denoising method is effective.