Holocene slip rate variability along the Pernicana fault system (Mt. Etna, Italy): Evidence from offset lava flows

The eastern flank of the Mount Etna strato-volcano is affected by extension and is slowly sliding eastward into the Ionian Sea. The Pernicana fault system forms the border of the northern part of this sliding area. It consists of three E-W-oriented fault sectors that are seismically active and characterized by earthquakes up to 4.7 in magnitude (M) capable of producing ground rupture and damage located mainly along the western and central sectors, and by continuous creep on the eastern sector. A new topographic study of the central sector of the Pernicana fault system shows an overall bell-shaped profile, with maximum scarp height of 35 m in the center of the sector, and two local minima that are probably due to the complex morphological relation between fault scarp and lava flows. We determined the ages of lava flows cut by the Pernicana fault system at 12 sites using cosmogenic He-3 and Ar-40/Ar-39 techniques in order to determine the recent slip history of the fault. From the displacementage relations, we estimate an average throw rate of similar to 2.5 mm/yr over the last 15 k.y. The slip rate appears to have accelerated during the last 3.5 k.y., with displacement rates of up to similar to 15 mm/yr, whereas between 3.5 and 15 ka, the throw rate averaged similar to 1 mm/yr. This increase in slip rate resulted in significant changes in seismicity rates, for instance, decreasing the mean recurrence time of M >= 4.7 earthquakes from similar to 200 to similar to 20 yr. Based on empirical relationships, we attribute the variation in seismic activity on the Pernicana fault system to factors intrinsic to the system that are likely related to changes in the volcanic system. These internal factors could be fault interdependencies (such as those across the Taupo Rift, New Zealand) or they could represent interactions among magmatic, tectonic, and gravitational processes (e.g., Kilauea volcano, Hawaii). Given their effect on earthquake recurrence intervals, these interactions need to be fully assessed in seismic hazard evaluations.