A new model for seismicity parameterization and predictive aspects of its application to the Sakhalin region

Seismicity is generally treated as an example of self-organized criticality. One alternative to this treatment may possibly be furnished by a model of seismicity that is thought of as a set of episodes of avalanching relaxation that occur randomly on a set of metastable subsystems. This model is defined by the parameter r, which specifies the hierarchical divisibility of the block structure of the medium, and by the parameter p, which specifies the probability for the incipient relaxation of a metastable state to continue. These two parameters together define the modeled value of the slope of the recurrence curve, or b-value, thus determining the mode of seismicity occurrence. This model is used to describe the seismicity in southern Sakhalin Island. In this modeling, the coefficient of hierarchical divisibility r in the block structure of the medium was assumed to be stationary and the disequilibrium parameter p was assumed to describe time-dependent variations of seismicity. We calculated models for the spatial variability of r and the time variability of p. We found abnormal growth in p during the Gornozavodsk earthquake (2006, M _w = 5.6) and the Nevel??sk earthquake (2007, M _w = 6.2). At present, we report values of p that are high (and increasing over time) in the wide area of the Poyasok isthmus. The results derived here are compared with other ideas on seismicity and with the experience that was previously gained in the area of earthquake prediction.