A self-correcting stochastic model of earthquake occurrence enhanced by geological fault data

We analyse sequences of earthquakes of Mw >= 5.3 drawn from the Parametric Catalog of Italian Earthquakes CPTI04 and associated with the composite seismogenic sources of the Database of Individual Seismogenic Sources (DISS). According to the elastic rebound theory we assume that these events are modelled by a self-correcting stochastic model characterized by an intensity function (rate) depending on the level of the process through an exponential function. The rate of the process is typically defined as ?(t) = exp{? + ? X(t)} where ? and ? are parameters, and X(t) is a proxy measure of the state of the system. Key elements of this analysis are: 1) the occurrence probability which increases with the level X(t) and, 2) the choice of the quantity S which forms the underlying system variable, or level, of the stochastic process. In the literature (Zheng and Vere-Jones (1991), (1994)) the evolution of the level is given by the expression X(t) = X(0) + ? t - , where X increases linearly at a constant loading rate ? and then decreases abruptly at the instants ti when an earthquake occurs. The quantity S used as measure of the earthquakes is generally an approximation of the strain and of the seismic moment, both derived from the moment magnitude that is the only available estimation of the size of our historical earthquakes. In our analysis we add geological information about the type of the causative fault for each earthquake by first estimating area A and average displacement D and then determining the seismic moment by M0 = ? D A with ? shear modulus. The different versions of the stochastic model are compared on the basis of the Bayes factor. Drawing inspiration from Jaumé and Bebbington (2004), we propose a geophysical interpretation of the results by relating the different versions of the stochastic model and the occurrence of accelerating seismic release. Moreover we examine which is the role of the geological information in this context. References Jaumé S.C. and Bebbington M.S. (2004) Accelerating seismic release from a self-correcting stochastic model, J. Geophys. Res., 109, B12301, doi:10.1029/2003JB002867 Zheng X. and Vere-Jones D. (1991) Applications of stress release models to earthquakes from North China, Pure Appl. Geophys., 135, 559-576. Zheng X. and Vere-Jones D. (1994) Further applications of the stochastic stress release model to historical earthquake data, Tectonophysics, 229, 101-121.