Statistical significance of geomagnetic diurnal variation anomalies prior to worldwide earthquakes

Abstract

The anomalous behavior of geomagnetic diurnal variation was reported by prior studies several months before the great 2011 Tohoku earthquake. In order to be further developed for a reliable earthquake prediction system, the statistical significance of such anomalies needs to be verified through a multiple-events study. In this paper, 157 past earthquakes that happened from the year 2000 until 2019 around the world were studied by utilizing vast low-resolution (1-min sampling period) geomagnetic field data. The Diurnal Variation Range Ratio (DVRR) method was employed to identify the disappearance of typical diurnal variation at magnetometer stations near epicenters, whereas superposed epoch analysis was the statistical tool used to reveal the periodicity of anomaly appearances. This research found that the number of anomalies in all three geomagnetic field components (i.e., northward, eastward, and vertical) increased significantly only before the earthquakes starting from around one month prior. The significant increments were not simultaneous, instead there were temporal lags between the components. Several mechanisms were proposed to elucidate the generation of the anomalies as well as to explain the observed temporal lags. Through the analysis, it could be concluded that the statistical significance of the anomalies could be verified. It was possible that the anomalies were the earthquake precursors.