Ionospheric precursors of earthquakes from satellites

In-situ magnetic field and electron density, as observed by Swarm and CSES satellites, are analyzed to identify possible anomalies in geomagnetic quiet time with respect to the ionospheric background. To avoid detecting possible anomalies induced by auroral activity we investigate regions between +50 and -50 degrees in magnetic latitude. Then a superposed epoch and space approach is applied to this anomaly dataset with respect to their time and space distance from shallow M5.5+ earthquakes occurred in about last 6 years. A comparison with analogous homogeneous random distribution of anomalies shows that the real anomaly concentrations found before the occurrence of earthquakes are statistically significant. In addition, we find that, in general, the anticipation times of the ionospheric precursors scale with the earthquake magnitude, confirming the validity of the Rikitake law for ionospheric signals, previously valid for ground precursors. We also find that the anomaly duration seems to depend on the magnitude of the impending earthquake. Finally, we propose a simple scheme of potential earthquake forecast on the base of the previously mentioned characteristics.