Laboratory studies of slip along faults as a physical basis for a new approach to short-term earthquake prediction

Abstract

The physical effects that may prove useful for developing a new approach to short-term earthquake prediction have been studied in laboratory conditions. In seismology and earthquake foci mechanics, one of the major challenges is searching for indicators of an upcoming seismic event and attempting to reliably record such indicators by available instruments. In this regard, the best result of the laboratory studies of dynamic slip along faults would be the identification of specific macroscopic parameters controlling the deformation process, which are measurable in field. Dynamic stiffness of a fault zone seems to be an appropriate parameter. The recent laboratory experiments have shown that the value of this parameter predetermines the slip mode along the fault (unstable slip, creep, tremor, etc.), and a radical decrease in shear stiffness takes place as the fault zone reaches the metastable state. The effect discovered in the laboratory conditions gives grounds to suggest that changes in the stress-strain state of the fault zone at the final stage of earthquake preparation are detectable from the parameters of microseismic noise in the low-frequency range. Apparently, the noise records during and after the arrival of surface waves from distant earthquakes can provide the best opportunity for determining the parameters characterizing the study area. The wave oscillations with a period of a few dozen seconds have significant amplitudes and duration, which contributes to the excitation of resonance oscillations of the blocks. There are problems requiring additional laboratory experiments: estimating the size of a fault, which predetermines regularities in decreasing of the own frequency of the block-fault system; determining the ratio of the mechanical parameters of the fault in the nucleation zone and on the periphery of the future rupture, etc. Having analyzed the results of experimental studies carried out by other researchers, we conclude that laboratory experiments under normal conditions and low pressures can successfully address a number of fundamental issues on the way to creating a new approach to short-term earthquake prediction. Increasing pressure and temperature to values characteristic of seismogenic depths does not lead to the occurrence of any fundamentally new features in the behavior of the block-fault system at the stage when dynamic slip is being prepared. During slip, friction reduces due to melting, physical and chemical transformations at the micro- and nanoscales and other processes on the slipping surface, but these effects play no role at the stage when dynamic rock failure and the onset of slip are being prepared.