Correct Off-Site Determination of Seismic Sensor Orientation from Combined Analyses of Earthquake and Microseism Records

Correct sensor orientation is vital for seismological analysis. However, seismic sensors including both borehole and surface seismometers are often installed in incorrect orientations. Individual methods proposed for sensor-orientation determination suffer from their own limitations and uncertainty, leaving the estimates in question before on-site verification. We introduce a method to combine a set of seismic phase analyses, yielding accurate sensor-orientation estimates. The method determines the sensor orientations by weighted-averaging independent estimates from three individual sensor-orientation analyses that are based on earthquake-origin P waves, earthquake-origin Rayleigh waves, and microseism-origin Rayleigh waves. The earthquake-origin seismic phase analyses may suffer from seismic anisotropy along ray paths even with accurate source-location information. On the other hand, the microseism-origin Rayleigh-wave analysis is hardly affected by seismic anisotropy along ray paths, being applicable to any seismic station with a couple of hour-long records. The three analyses complement each other, which enables us to determine representative sensor orientations correctly. We apply the proposed method to densely deployed 377 seismometers in South Korea, examining the sensor orientations. The representative sensor orientations are determined stably with standard errors less than 1°, supporting the accuracy of results. Borehole seismometers are poorly oriented relative to surface seismometers. The proposed method is useful for instant examination of sensor orientations of seismometers in remote regions and borehole seismometers in which physical accessibility is highly limited.