| Popis: |
Understanding the behaviour of faults over thousand year timescales is key to assessing seismic hazard. It is becoming apparent that rates of earthquake activity on active normal faults over thousand year timescales vary throughout this time period. This thesis investigates the impacts and potential causes of variable fault activity over the last 15 thousand years in the central Apennines, Italy. The central Apennines are an extensional mountain chain that contains a network of active normal faults. Rates of slip on the faults are up to c.1 mm/yr. The region has exceptional published constraints on the rates and kinematics of fault slip. In this thesis, I use terrestrial laser scanning, combined with ground penetrating radar, to add to the dataset of fault slip rates. I combine these new measurements with existing studies to bring together a dataset that contains measurements of slip rates for the vast majority of faults in the region. The central Apennines has one of the longest available historical earthquake catalogues for any tectonic region around the world. The catalogue extends reliably back to 1349 AD for earthquakes with M≥5.8. I analyse the records to identify the individual fault ruptures for 27 earthquakes extending back over 600 years. I use this database of historical earthquake ruptures to compare geodetic observations of interseismic strain accumulation with the historical record of faulting and the geological record of active faults. I model interseismic strain accumulation using a novel method that adapts an existing model of shear zones and incorporates geological observations of fault slip rates, fault kinematics and shear zone depth. I show how GPS observations struggle to resolve the location of all active faults due to possible transient fault activity. I then compare the historical and geological record of earthquake activity with the seismic hazard map in central Italy. As part of this analysis, I show how earthquakes in the historical catalogue have clustered along faults on the northeastern side of the central Apennines. I provide a critique of the historical data used to create the seismic hazard map of the region to show how geological records of active faults should be the primary driver of seismic hazard assessment as fault slip rates are variable through time. To investigate the causes of the clustered earthquake behaviour in the central Apennines, and the cause of variable slip rates on normal faults, I model the co-seismic Coulomb stress interactions between active faults in the region for each of the earthquakes in the historical database. Interseismic stresses are included by adapting the shear zone model to resolve stresses on upper crustal faults in the central Apennines. I show how earthquake inter event times can be perturbed by thousands of years due to the cumulative effects of Coulomb stress changes over 27 earthquakes. I also show that these perturbations are greater on shorter faults. The results presented in this thesis show how the transient behaviour of faults affects methods used to observe fault systems that are the basis of seismic hazard assessment. I illustrate how fault interaction can affect the behaviour of faults over thousand year timescales. |
