Abstract: Clock accuracy is one of the most important parameters in seismic observation, which is usually guaranteed by satellite timing signal. When the satellite signal is shielded or the time-keeping components of the instrument have problems, the internal clock of seismometer will drift gradually, which brings great affect to the data processing. The empirical Green's function extracted from cross-correlation of seismic background noise stations can not only be used for structural imaging, but also can be used to detect whether the waveform clock has drift and how many times drifted. In this paper, the cross-correlation monitoring method of seismic background noise is used to detect the clock drift of four mobile stations in Yongsheng area of Yunnan Province from 2017 to 2019. The results show that some mobile stations have different clock drift modes in different time periods, and the maximum amplitude can reach 1.75 s. Furthermore, the above results are verified by using the repeated signals from the Fixed Airgun Signal Transmission Stations (FASTS) of Binchuan in Yunnan province. It is found that the two results have a good consistency. Our research shows that the cross correlation of background noise has a high sensitivity to the clock drift of seismic waveform, which can effectively detect the clock problem of waveform, prevent the waveform from being misused, and provide reference information for the subsequent clock correction of waveform.

Key words: Clock drift detection, Seismic ambient noise, Airgun repeated source