Predicting earthquakes through digging satellite data

Abstract

Earthquake is, perhaps, still considered as the foremost unpredictable natural disaster experienced to date. Even though, predicting an earthquake remains an elusive goal to scientists over the last several decades, we are still to achieve a good prediction mechanism for earthquakes. To this extent, in this study, we propose a new mechanism for predicting earthquakes. Our mechanism enables prediction around four months prior to happening of an earthquake. To do so, we analyze open satellite data and nd gravity gradient anomaly as a prominent pre-seismic activity. To the best of our knowledge, we are rst to consider and demonstrate gravity gradient anomaly as a pre-seismic activity for predicting earthquakes. Accordingly, using this anomaly, we formulate necessary mathematical models and illustrate how we can predict earthquakes having magnitude greater than or equal to 6.0 in moment magnitude scale. We analyze e ciency of our proposed mechanism in predicting several earthquakes (magnitude 6.0) occurred in di erent parts of the world. Our analysis con rms that we can predict more than 90% of the earthquakes using our proposed method while having a very low false alarm rate.