By Claudia Sgambato, on 30 July 2020
A new study by IRDR PhD student Claudia Sgambato, Dr Joanna Faure Walker, Dr Zoe Mildon and Prof Gerald Roberts, has identified a novel aspect of fault interaction that links the stress loading to the geometry of the fault system, which has important implications for fault-based earthquake hazard modelling.
By analysing the earthquake sequence of the Southern Apennines (Italy) in the last 600 years, the study presents a comparison of the stress loading history of “isolated” faults and multiple faults across strike, and shows that the stress evolution is not the same for all faults, but is influenced by the way faults are arranged in the system.
In the Southern Apennines the fault system geometry is relatively simple, with most of the structures aligned along strike. This area was the location of some of the strongest earthquakes ever recorded in Italy, such as the Mw 7.1 1857 event on the Val di Diano and Val d’Agri faults, that caused ~10,000 deaths and the more recent Mw 6.8 1980 earthquake on the Irpinia fault, that caused 3,000 deaths. For these reasons, this area provides an ideal place to investigate the role of fault geometry and fault interaction in the historical earthquake sequence.
Fault interaction during earthquakes is usually calculated through Coulomb stress transfer. Using data from historical earthquakes, the authors have calculated the coseismic stress changes on the active faults. Deformation rates measured in the field have been used to derive the annual rate of stress loading. The combined coseismic and interseismic stress components allow calculation of the stress present on the fault prior to an earthquake, or pre-stress.
The analysis of pre-stress on all the faults before each historical earthquake has shown that 94% of the earthquakes occurred on faults that were positively stressed, and that where earthquakes occurred on an isolated fault, this had the highest pre-stress of all the faults at the time of the event. This is due to the fact that when a fault is isolated, the build-up of stress is not influenced by earthquakes occurring on other faults across strike, and the stress is distributed homogeneously across their surface, compared to faults that are across strike.
This suggests that isolated faults have fewer areas of negative stress, which can promote the propagation of the rupture, generating earthquakes with similar magnitude. An example of this can be seen for the Irpinia fault, with the earthquakes in 1694 and 1980, which share similar magnitude and damage distribution, and a similar value of pre-stress.
This means that for studies of seismic hazard it is important to consider the pre-stress on the faults in order to understand which fault is likely to rupture, and to take into account that the fault system geometry influences the way the stress is accumulated on faults.
The article is open access and available here: Sgambato et al. (2020)