Urban paleoseismology: case studies from Thessaloniki, Greece

Faults in urban areas pose a real danger for buildings and infrastructures, not only in the case of an earthquake, but also in such cases as differential sediment compaction, water overpumping, etc. Their importance is often underestimated with sometimes severe consequences.
   Detection of faults is not easy in urban environments, as usually outcrops have been covered by built structures, prohibiting direct observations. Geophysical surveys are rare, but even in those cases the exact location of a surface-intercepting fault is not clear.
   One of the most promising methods for acquiring quantitative information about faults in urban areas is paleoseismological investigation. It consists of an integrated set of methodologies that can provide hands-on data for displacement, timing, etc. In this paper we present two cases of paleoseismological applications in the metropolitan area of Thessaloniki.
   Two faults, one evident and one unknown, have been studied in Peraia and Kalamaria respectively.
   Peraia fault: This fault defines the contact between the footwall Pliocene sandstonemarl series and the hangingwall loose Holocene deposits. It is a fault that coincides with the well known Anthemountas fault zone, a roughly E-W trending normal fault zone that is associated with several historical earthquakes. Its exact location through Peraia town was not known in much detail due to the lack of outcrops. Nevertheless, it forms a well defined scarp that divides the town into an upper (Ano) and lower (Kato) part. In 2005 and 2006 a set of surface ruptures along this fault caused significant damage on buildings and roads. Paleoseismological investigation with two trenches along the fault showed that faulting was not random, as a large displacement was detected, with successive steps of cumulative faulting. Borehole data confirmed that the total displacement was indeed large (35 m). Trenching showed that the fault has been continuously active during the Quaternary, with all of its displacement on the same surface, posing thus a severe danger for the area in general and specifically for the buildings that are built along its trace. The 2005-06 surface ruptures are interpreted as a combination of overpumping, compaction and fault creep. The contribution of each factor is not possible to be calculated, as there are too many uncertainties concerning the deformation model.
   Kalamaria fault: This fault was exposed during the construction of a multi-stored residence building in Kalamaria, a town located next to Thessaloniki city. This fault is displacing marls and a paleosoil that is located on top of the sedimentary sequence. Morphologically it is manifested as a gentle scarp, observable in roads that cut through the fault along at least 500 m. Paleoseismological analysis showed that the fault has been inactive during Upper Pleistocene – Holocene as there are no indications for recent reactivations. However, the existence of the morphological scarp suggests that it has probably been active during that period, but microstratigraphical evidence for this activity has been destroyed by anthropogenic factors. Even if it is not active, the fault zone exists and it can act as a weakness zone during a distant earthquake or in response to water level fluctuations.
   In conclusion, paleoseismological techniques can be of great effectiveness in the study of urban faults, either active or not. Planners should take into account this methodology, because it can greatly enhance the understanding of ground response in abnormal conditions.