Seismic measurements to reveal short-term variations in the elastic properties of the Earth crust

Since the late 1960s - early 1970s, seismologists started studying the elastic properties of the Earth crust looking for signals from the Earth interior indicating that a large earthquake is coming. To be useful for prediction a signal needs to: 1) occur before most large earthquakes and 2) occur only before large earthquakes. Up to now, no one has ever found such a signal, but since the beginning of the search, seismologists developed theories that included variations of the elastic property of the Earth crust prior to the occurrence of a large earthquake. The most popular is the theory of the dilatancy: when a rock is subject to stress, the rock grains are shifted generating micro-cracks, thus the rock itself increases its volume. Inside the fractured rock, fluid saturation and pore pressure play an important role in earthquake nucleation, by modulating the effective stress. Thus, measuring the variations of wave speed and of anisotropic parameter in time can be highly informative on how the stress leading to a major fault failure builds up. In 1980s and 1990s such kind of research on earthquake precursors slowed down and the priority was given to seismic hazard and ground motions studies, which are very important since these are the basis for the building codes in many countries. Today, we have dense and sophisticated seismic networks to measure wave-field characteristics: we archive continuous waveform data recorded at three components broad-band seismometers, we almost routinely obtain highresolution earthquake locations. Therefore, we are ready to start to systematically look at seismic-wave propagation properties to possibly reveal short-term variations in the elastic properties of the Earth crust. One seismological quantity which, since the beginning, is recognized to be diagnostic of the level of fracturation and/or of the pore pressure in the rock, hence of its state of stress, is the ratio between the compressional (P-wave) and the shear (S-wave) seismic velocities: Vp/Vs. Variations of this ratio have been recently observed and measured during the preparatory phase of a major earthquake. In active fault areas and volcanoes, tectonic stress variation influences fracture field orientation and fluid migration processes, whose evolution with time can be monitored through the measurement of the anisotropic parameters. Through the study of S-waves anisotropy it is therefore potentially possible to measure the presence, migration and state of the fluid in the rock traveled by seismic waves, thus providing a valuable route to understand the seismogenic phenomena and their precursors. On the other hand, only in the very recent times with the availability of the continuous seismic records, many authors have shown how it is possible to estimate the relative variations in the wave speed through the analysis of the crosscorrelation of the ambient seismic noise. In this paper we first analyze in detail these two seismological methods: shear wave splitting and seismic noise cross-correlation, presenting a short historical review, their theoretical bases, the problems, learning, limitations and perspectives. We, then, compare the main results in terms of temporal trends of the observables retrieved applying both methods to the Pollino area (southern Apennines, Italy) case study.