基于集合经验模态分解法的重力极潮提取与研究

重力极潮，即极移引起的重力变化，是地球由于惯性离心力变化导致地球形变的综合反映，其观测和研究有助于了解地球在长周期频段的响应，约束地球形变、地球内部构造和物理参数.区别于以往研究，本文采用了集合经验模态分解方法（Ensemble Empirical Mode Decomposition，EEMD）从全球5个超导重力仪台站连续重力观测资料中提取重力极潮、消除仪器漂移及极潮频段以外的大部分噪声信号，该方法所用的基函数是基于自身信号获得的；在此基础上，利用最小二乘匹配法分离极潮中的周年项和Chandler项，估算Chandler周期的极潮潮汐因子.结果表明，基于EEMD得到的极潮潮汐因子与前人基于其他消除仪器漂移方法（数学模型或小波分析）得到的结果相符合，精度相当，但由于这种方法是自适应的，因此本文结果更能反映实际物理过程.

Etraction and study of the gravity pole tide using EEMD

The gravity pole tide, i.e. the polar-motion-induced gravity variations, is the comprehensive reflection of the Earth's deformation caused by the perturbations in the inertial centrifugal force. The observation and study of the gravity pole tide are helpful to understand the Earth's long-period response and to constrain the Earth's deformation, interior structure and physical properties. Different from the previous studies, in this paper the EEMD (Ensemble Empirical Mode Decomposition) method, the basis function of which is derived from the data themselves, is used to process continuous gravity observation data recorded by 5 global superconducting gravimeter to extract the gravity pole tide, to eliminate the instrumental drift and noise outside the pole tidal band; on this basis, the annual and the Chandler wobbles are separated by least-squares fitting a combination of 2 sinusoidal functions with periods of 365.25 and 432 days to the extracted pole tide, and then the gravimetric factors of the pole tide associated with the Chandler wobble are estimated. Our results show that the gravimetric factors of the pole tide associated with the Chandler wobble obtained through EEMD agree with previous results (based on drift models or wavelet approach) and have comparable internal precision. However, considering that the EEMD is an adaptive process, the results in this study can better reflect the real physical process.