Iezzi, Francesco

We have mapped and constrained the timing of tectonically deformed uplifted Late Quaternary palaeoshorelines in the Messina Strait, southern Italy, an area above a subduction zone containing active normal faults. The palaeoshorelines are preserved from up to thirteen Late Quaternary sea-level highstands, providing a record of the deformation over this timescale (~500 ka) for the Messina-Taormina Fault, the Reggio Calabria Fault and the Armo Fault. The palaeoshorelines reveal spatial patterns of uplift through time along the strike of these normal faults, and, given the across strike arrangement of the faults, also reveal how the contribution of each fault to the regional strain-rate progressed through time. The results reveal that the uplift rates mapped within the fault hangingwalls and footwalls were not constant through time, with a marked change in the location of strain accumulation at ~50 ka. The uplift rates, once converted into throw-rates, imply that the three faults comprised similar throw-rates prior to ~50 ka (in the range 0.77–0.96 mm/yr), with the Armo and Reggio Calabria faults then switching to lower rates (0.32 mm/yr and 0.33 mm/yr respectively), whilst the Messina-Taormina Fault accelerated to 2.34 mm/yr. The regional extension rate, gained by summing the implied heave rates across the three faults, was maintained through time despite this re-organisation of local strain accumulation at ~50 ka. We explain these out-of-phase fault throw-rate changes during the constant-rate regional extension conditions as due to interactions between these upper plate normal faults. We finally discuss how fault throw-rates changing through time may affect a long-term seismic hazard assessment within active normal fault systems.

Surface faulting earthquakes are known to cluster in time from historical and palaeoseismic studies, but the mechanism(s) responsible for clustering, such as fault interaction, strain-storage, and evolving dynamic topography, are poorly quantified, and hence not well understood. We present a quantified replication of observed earthquake clustering in central Italy. Six active normal faults are studied using 36Cl cosmogenic dating, revealing out-of-phase periods of high or low surface slip-rate on neighboring structures that we interpret as earthquake clusters and anticlusters. Our calculations link stress transfer caused by slip averaged over clusters and anti-clusters on coupled fault/shear-zone structures to viscous flow laws. We show that (1) differential stress fluctuates during fault/shear-zone interactions, and (2) these fluctuations are of sufficient magnitude to produce changes in strain-rate on viscous shear zones that explain slip-rate changes on their overlying brittle faults. These results suggest that fault/shear-zone interactions are a plausible explanation for clustering, opening the path towards process-led seismic hazard assessments.

The geometry, rates and kinematics of active faulting in the region close to the tip of a major crustal-scale normal fault in the Gulf of Corinth, Greece, are investigated using detailed fault mapping and new absolute dating. Fault offsets have been dated using a combination of 234U/230Th coral dates and in situ 36Cl cosmogenic exposure ages for sediments and wave-cut platforms deformed by the faults. Our results show that deformation in the tip zone is distributed across as many as eight faults arranged within ~700 m across strike, each of which deforms deposits and landforms associated with the 125 ka marine terrace of Marine Isotope Stage 5e. Summed throw-rates across strike achieve values as high as 0.3–1.6 mm/yr, values that are comparable to those at the centre of the crustal-scale fault (2–3 mm/yr from Holocene palaeoseismology and 3–4 mm/yr from GPS geodesy). The relatively high deformation rate and distributed deformation in the tip zone are discussed in terms of stress enhancement from rupture of neighbouring crustal-scale faults and in terms of how this should be considered during fault-based seismic hazard assessment.