中文摘要：

Zero drift and solid Earth tide corrections to static relative gravimetric data cannot be ignored. In this paper, a new principal component analysis(PCA) algorithm is presented to extract the zero drift and the solid Earth tide, as signals, from static relative gravimetric data assuming that the components contained in the relative gravimetric data are uncorrelated. Static relative gravity observations from Aug. 15 to Aug. 23, 2014 are used as statistical variables to separate the signal and noise with PCA to obtain desired signals. The results of the linear drift extracted by PCA are consistent with those calculated by the least squares linear fitting, and the differences only reach to 10 2mG al/day order of magnitude.Furthermore, PCA is used to estimate the solid Earth tide from the relative gravimetric data corrected by the zero drift. The statistical results are consistent with the results derived from the solid Earth tide correction provided by the internal software of the CG-5 gravimeter(SCINTREX Limited Ontario Canada). The statistical results of the differences between the two methods are both less than 8 m Gal, and the RMSs for 9 days are all less than5 m Gal.

英文摘要：

Zero drift and solid Earth tide corrections to static relative gravirnetric data cannot be ignored. In this paper, a new principal component analysis （PCA） algorithm is presented to extract the zero drift and the solid Earth tide, as signals, from static relative gravimetric data assuming that the components contained in the relative gravimetric data are uncorrelated. Static relative gravity observations from Aug. 15 to Aug. 23, 2014 are used as statistical variables to separate the signal and noise with PCA to obtain desired signals. The results of the linear drift extracted by PCA are consistent with those calculated by the least squares linear fitting, and the differences only reach to 10-2μGal/day order of magnitude. Furthermore, PCA is used to estimate the solid Earth tide from the relative gravimetric data corrected by the zero drift. The statistical results are consistent with the results derived from the solid Earth tide correction provided by the internal software of the CG-5 gravimeter （SCINTREX Limited Ontario Canada）. The statistical results of the differences between the two methods are both less than 8 ,Gal, and the RMSs for 9 days are all less than 5 μGal.