利用GRACE、測高資料、海洋模型估計地心變動

Abstract

Geocenter variation, defined as the motion of center of mass (CM) of the Earth relative to the center of figure (CF), is usually represented by degree one spherical harmonic coefficients. The geocenter variation is mainly caused by the mass redistribution in the Earth system including the mass exchange between ocean, atmosphere, ground hydrology, mountain glacier and ice sheets, and the mass change inside of the Earth like Post-Glacial Rebound (PGR) and earthquakes. Accurate Estimation of geocenter variation has a contribution to understand the variation in the Earth since geocenter variation reflects the Earth mass balance and the interaction between the solid Earth and transported mass in the Earth system. Several studies for estimating geocenter variation have been done, but the results were limited as a result of uneven distribution of measurements or shortage of data in oceans. GRACE has provided monthly gravity solutions with high precision since launched in 2002, which can be used for estimating geocenter variation by combining with climate models or other satellite measurements, but the results are still limited because of the uncertainties of climate models. In this study, geocenter variation is determined by two novel methods using the combination of GRACE and altimetry data including the method used by Swenson et al. [2008] and an iterative algorithm. Both results agree well with previous researches and the accuracy of the result derived by the iterative method is higher. Since the altimetry measurements refer to ITRF (International Terrestrial Reference Frame), the displacement between the origins of ITRF and CF would affect the estimation of geocenter variation. Therefore, ocean models are used for geocenter variation determination instead of altimetry data. The result shows the displacement between the origins of ITRF and CF would not change the estimation of geocenter variation significantly but the different covering areas of altimeters or ocean models would change the estimation of geocenter variation in terms of trends and phases. In addition, leakage effects are also considered in the computation by removing ocean data near coastlines within 300 km distance. Different degree one coefficients of PGR models only change the trends of estimated geocenter variation with 80% of their values.