Background noise and data anomalies analysis of strong earthquakes

The project of National Earthquake Intensity Rapid Reporting and Early Warning undertakes the mission of earthquake early warning and intensity rapid reporting in China. With the construction of the project, Inner Mongolia has established a strong earthquake observation station network consisting of 193 strong earthquake observation stations and a data center. In this paper, the power spectral density of the strong motion observation data of 193 stations is calculated by using the power spectral probability density function method. By drawing the power spectral density and power spectral probability density function maps and RMS value distribution maps of different time and space, different strong seismometers, different frequency bands, the background noise of strong motion stations in our region is evaluated, and the emerged spectral anomalies are studied. Accordingly, the constructive suggestions are put forward for the anomaly detection and processing of strong motion data. The research results show that the diurnal difference of background noise at different strong seismometers, strong earthquake stations and at different time and space is mainly reflected in the frequency band above (1~2) Hz, and the diurnal PSD difference in different spaces is quite large; The background noise value of the data recorded by QA-2g, JS-A2 and TDA-33M strong seismometers at 2s frequency point shows little change with space and direction; By monitoring the background noise of 2s frequency points of strong earthquake stations, the anomalies of strong earthquake observation data occur, and the present results firstly concluded can be extended to the data quality monitoring of thousands of strong earthquake stations nationwide; The change of background noise in high frequency band of strong earthquake stations mainly results from the change of environmental noise, but the influence of instrument self noise cannot be completely ignored; Anomaly of strong earthquake data is mainly caused by large zero point of strong seismometer, failure of strong seismometer, pulse interference and error of system parameters. The results in the present paper can be used to optimize the observation environment of the station, further standardize the installation, commissioning of instruments and JOPENS system parameter configuration and review.