Seismotectonics review and 3D rheological modelling of the Durres area (Albania)

The Durres sector of the Adria-Eurasia collision was recently affected by the 2019 seismic sequence. The seismogenic volume belongs to the active portion of the accretionary wedge representing the external part of the Albanides fold-and-thrust belt. In this sector, the front is kinematically partitioned by the ENE-WSW Shkodra (to the North) and Lushnje (to the South) transfer faults (Caputo and Pavlides, 2013). This region shows a documented seismic activity (Baker et al., 1997; Louvari et al., 2001; Muco et al., 1994) associated to both thrust and back-thrust structures (Muceku et al., 2006; Roure et al., 2004; Skrami et al., 2001; Velaj et al., 2001), though the orientation of the causative source of the 26 November, 2019 Durres mainshock (Mw=6.4) is still debated. In order to better characterize the seismogenic sources recognised in the area (from GreDaSS database; Caputo and Pavlides, 2013), we performed a seismotectonic review and created a 3D rheological model for constraining their (maximum) seismogenic potential. The BDT (Brittle Ductile Transition) represents a good proxy of the seismic/aseismic transition and fundamentally corresponds to a mechanical boundary for coseismic rupture propagation processes (Maggini and Caputo, 2021). Close to this mechanical and behavioural transition, quasi-continuum deformation within the viscous crustal body generally induces elastic deformation within the contiguous elasto-brittle body; therefore, when sufficient energy is accumulated it is commonly released seismically. Following a similar approach by Maggini and Caputo (2020, 2021) for the broader Aegean Region, we compared the seismicity cut-off depth of the investigated sector with the modelled BDT depth, thus confirming that the rheological and seismological transitions are tightly correlated. Our model, based on dedicated MatLab scripts and proper calibrated input parameters, evidence the presence of different superposed brittle and ductile layers (clearly evident in transversal reconstructed 2D sections) within the modelled volume. The maximum seismogenic depth occurs at values ranging between 10 and 25 km. Finally, based on the inferred source parameters and applying appropriate empirical relationships (Wells and Coppersmith, 1994; Leonard, 2014), we estimate the maximum expected magnitude for each source and hence their contribution for the Seismic Hazard Assessment of the area.