Active faults, Paleoseismological trenching and seismic hazard assessment in the Northern Mygdonia Basin, Northern Greece: The Assiros-Krithia fault and the Drimos fault zone

The Plio-Quaternary Mygdonia Basin in Northern Greece is bounded by active normal faults that can host significant earthquakes and produce considerable damage to the city of Thessaloniki (the second biggest city in Greece – 1 million population), located only 15–20 km away. In 1978, a damaging M = 6.5 earthquake occurred along a fault in the southern part of the basin, generating primary and secondary surface ruptures. This manuscript offers the results of the first detailed geomorphological and paleoseismological investigations in the northern part of the basin. As a result: i) a new active normal fault has been mapped, the Assiros-Krithia fault (AKF); and ii) another fault zone that has been assessed with probable activity in the literature, the Drimos fault zone (DFZ), has been confirmed active and mapped in detail. Paleoseismological trenching investigations supported by radiometric dating were carried out along both structures. The AKF and DFZ form the northern and northwestern boundaries of the Mygdonia Basin, respectively. The AKF deflects the local drainage network and is expressed as a single north dipping localised fault trace, gently uplifting the northern part of the Mygdonia Basin. The last event produced 48 cm of vertical displacement around 4540 cal. yrs BP, and the fault has an estimated slip rate of ∼0.1 mm/yr. The DFZ is an 8-10 km-long NW-SE trending normal fault that dips and downthrows to the NE (045ᵒ-050ᵒ). It is characterised by distributed deformation across eight major and secondary fault planes. Four of these planes have been activated over the last 9000 years, therefore the structure is Holocene active. The DFZ has an estimated slip rate of 0.1–0.3 mm/yr and may merged with the Liti-Gerakarou fault zone in multi-segment ruptures. The DFZ highlights that NW-SE trending faults are potentially active in Northern Greece. Their long recurrence intervals justify why they were not activated during the period covered by the historical record and emphasise, once again, the necessity of paleoseismological studies and the implementation of fault specific approaches in seismic hazard assessment.