Abstract:
According to the fact that there is no enough research of the noise characteristics in the study of earthquake prediction, this paper analyze the influence of large earthquake on the noise characteristics of time series. Focusing on the large earthquake and the corresponding station nearby, the GPS time series were separated to five periods, before the earthquake 4~3 years, 3~2 years, 1~0 years, after the earthquake 0~1 years and 1~2 years. The spectral index of each period. Noise analysis was implemented using different noise model combination. Analysis results shows that, the change of spectral index, velocity field, amplitude and phase before and after the earthquake is very big, and the noise characteristics of GPS time series are significant to seismic response; For part of the stations the spectral index, velocity field and anniversary amplitude before and after the earthquake is consistent compare with results of the 1~0 years before the earthquake. In earthquake incubation period, the proportion of flicker noise + white noise(FN+WN) increased. And the noise component FN in flicker noise + random walk noise+ white noise(FN+RWN+WN) increased among all the noise components. All the ratio decreased after the earthquake, which is in consistent of the earthquake energy releaseand restoration of the site more meat situation to normal status after the earthquake.