PRECURSOR-BASED EARTHQUAKE PREDICTION: WISHFUL THINKING OR REAL POSSIBILITY?

Abstract

Scientists attitude towards the delicate precursor-based earthquake prediction subject was widely oscillating during the last half century, from optimism to deep pessimism according to milestone seismic events, claims of successful predictions and published expert debate results. Despite the current official/mainstream opinion according to which the precursor-based earthquake prediction is impossible in principle, new ideas and claimed positive results emerged in the last few decades allowing potentially paradigm-shifting new research strategies to be envisaged and followed. Such a strategy is proposed here based on the principles of 1) coupled and interacting geospheres, 2) uniqueness of the seismic structures, hence of their precursory fingerprints, and 3) non-equivalence of Earth surface measuring sites from the perspective of signal-reception capability. Such a strategy aims at discovering the particular precursory fingerprint of individual seismogenic structures instead of looking for universally valid precursory signals. The precursory fingerprint of a particular seismogenic structure is the assemblage of a number of pre-seismic signals of physical, chemical and biological nature detectable in advance of impending large-magnitude earthquakes using a matrix of high-accuracy sensors emplaced at pre-selected “sensitive” locations at Earth surface and on Earth-orbiting satellites. It has a pattern component (configuration of the above-threshold values of the monitored parameters) and a time component (sequence, succession and lead time of emergence of the anomalous signal parameters). Due to its complex nature, the assemblage of detected precursory signals has to be processed and evaluated by pattern recognition algorithms of Artificial Intelligence systems and validated by human experts before final conclusions to be drawn. Due to its particular setting in Europe’s geodynamically most active area, and its long-known seismic history including recurrent high-magnitude destructive events, the Vrancea seismic nest in Romania appears as an excellent experimental target for the implementation of a novel, concept-based long-term research strategy in the field of earthquake prediction.