Nonextensive models for earthquakes

We have revisited the fragment-asperity interaction model recently introduced by Sotolongo-Costa and Posadas [Phy. Rev. Lett. 92, 048501 (2004)] by considering a different definition for mean values in the context of Tsallis nonextensive statistics and introducing a scale between the earthquake energy and the size of fragment epsilon proportional to r3. The energy-distribution function (EDF) deduced in our approach is considerably different from the one obtained in the above reference. We have also tested the viability of this EDF with data from two different catalogs (in three different areas), namely, the NEIC and the Bulletin Seismic of the Revista Brasileira de Geofísica. Although both approaches provide very similar values for the nonextensive parameter , other physical quantities, e.g., energy density, differ considerably by several orders of magnitude.