Short-term earthquake prediction by reverse analysis of lithosphere dynamics

Short-term earthquake prediction, months in advance, is an elusive goal of earth sciences, of great importance for fundamental science and for disaster preparedness. Here, we describe a methodology for short-term prediction named RTP (Reverse Tracing of Precursors). Using this methodology the San Simeon earthquake in Central California (magnitude 6.5, Dec. 22, 2003) and the Tokachi-Oki earthquake in Northern Japan (magnitude 8.1, Sept. 25, 2003) were predicted 6 and 7 months in advance, respectively. The physical basis of RTP can be summed up as follows: An earthquake is generated by two interacting processes in a fault network: an accumulation of energy that the earthquake will release and a rise of instability triggering this release. Energy is carried by the stress field, instability is carried by the difference between the stress and strength fields. Both processes can be detected and characterized by “premonitory” patterns of seismicity or other relevant fields. Here, we consider an ensemble of premonitory seismicity patterns. RTP methodology is able to reconstruct these patterns by tracing their sequence backwards in time. The principles of RTP are not specific to earthquakes and may be applicable to critical transitions in a wide class of hierarchical non-linear systems.