LAUSR

Broadband seismic networks are becoming more intensive, generating a large amount of data in the long-term collection process. When processing the data, the researchers rely almost on instrument response files to understand the information related to the instrument. Aiming at the process of instrument response recording and instrument response correction, we identify several sources of the instrument response phase error, including pole–zero change, the causality difference in instrument correction method, and the problem of filter coefficient recording. The data time offset range from the instrument response phase error is calculated from one sample point to several seconds using the ambient noise data recorded by multiple seismic stations. With different data delays, the time offset of the noise correlation function is estimated to be 74% to 99% of the data delay time. In addition, the influence of instrument response phase error on the measurement of seismic velocity change is analyzed by using ambient noise data with pole–zero change, and the results show that the abnormal wave velocity with exceeding the standard value is exactly in the time period of the instrument response error, which indicates that the instrument response error affects the study of seismology.