Modeling of earthquake chronology from paleoseismic data: Insights for regional earthquake recurrence and earthquake storms in the Central Apennines

Abstract

Among the geologic data, trenching records of paleoearthquakes represent an important input for the development of seismic hazard evaluations and, specifically, for the characterization of fault rupture behavior in time. Based on the available paleoseismological trenching data in the Central Apennines we have compiled a new database of surface faulting earthquakes for 10 faults. The compilation contains 109 surface faulting events, occurred in the past ~28 kyr. Events younger than 7 kyr (59 dated events) are much better constrained and, therefore, are the basis for most of the analyses. Through a quantitative multistep method, we integrate paleoseismic trench data and treat them statistically, providing relevant improvement needed for trenching-based seismic hazard evaluation. Indeed, the combined analysis of trenching data from multiple sites on the same fault led to a reduction of the approximations and uncertainties in the rupture history of individual faults. The procedure was also applied on specific fault systems and indicated that the single faults may have occasionally ruptured simultaneously, or close in time, in the past. The whole set of age ranges was also statistically analyzed to produce regional earthquake scenarios for a period much longer than the millennial historical record. The built scenarios for the last 7 kyr define a regional mean inter event time (IET) of 230-240 yr, comparable with the average recurrence time of about 200 yr, considering the number of events in time. We also identify the possible occurrence of earthquake storms, i.e., concentration of surface faulting earthquakes in the region within time periods of 200 yr, suggesting a variability in time of the seismic behavior of the faults, with alternation of peaks of activity with more “quiet” periods. Even though the paleoseismic data from the Apennines contains uncertainties, the results appear quite stable and promising for future applications in earthquake-hazard assessment.