HECTOR for Analysis of GPS Time Series

摘要

High precision GPS coordinate time series has become a rich source of data in many fields of research, such as studying the slow deformation of the earth's surface, establishing and maintaining a regional or global reference frame, and studying the deformation process of earthquake pregnancy. In view of the current GPS time series analysis software in the processing of data is slow, inefficient, less choice of combination model, this paper studies and analyzes a new time series analysis software HECTOR. Firstly, the function and characteristics of HECTOR software are described in detail; Then, the software is used to obtain the periodic items and trend items in the three directions of the GPS time series, and compared with the parameters obtained by the CATS software. Secondly, the data of GPS time series in the study area are analyzed by using different noise models; Finally, BIC numerical analysis based on the maximum likelihood estimation and spectrum analysis are used to compare the results of different combinations of noise models. The results show that the HECTOR software can be used to obtain the parameters quickly and further. At the same time, GPS data for most of the selected study areas, white noise + power law noise for the optimal noise model. For GPS data of a few study areas, white noise + generalized Gaussian Markov noise, white noise + ARMA (5) noise are better noise models. Finally, it is found that the best noise model obtained in different directions of the same site is not the same, which can provide certain reference meaning for the future research in this direction.