The non-collisional subduction margin of South America is characterized by different geometries of the subduction zone and upper-plate tectono-magmatic provinces. The localization of deformation in the southern Central Andes (29°S–39°S) has been attributed to numerous factors that combine the properties of the subducting oceanic Nazca plate and the continental South American plate. In this study, the present-day configuration of the subducting oceanic plate and the continental upper plate were integrated in a data-driven geodynamic workflow to assess their role in determining strain localization within the upper plate of the flat slab and its southward transition to a steeper segment. The model predicts two fundamental processes that drive deformation in the Andean orogen and its foreland: eastward propagation of deformation in the flat-slab segment by a combined bulldozing mechanism and pure-shear shortening that affects the broken foreland and simple-shear shortening in the fold-and-thrust belt of the orogen above the steep slab segment. The transition between the steep and subhorizontal subduction segments is characterized by a 370-km-wide area of diffuse shear, where deformation transitions from pure to simple shear, resembling the transition from thick to thin-skinned foreland deformation in the southern Sierras Pampeanas. This pattern is controlled by the change in dip geometry of the Nazca plate and the presence of mechanically weak sedimentary basins and inherited faults.

The crustal seismogenic thickness (CST) has direct implications on the magnitude and occurrence of crustal earthquakes, and therefore, on the seismic hazard of high-populated regions. Amongst other factors, the seismogenesis of rocks is affected by in-situ conditions (temperature and state of stress) and by their heterogeneous composition. Diverse laboratory experiments have explored the frictional behavior of the most common materials forming the crust and upper most mantle, which are limited to the scale of the investigated sample. However, a workflow to up-scale and validate these experiments to natural geological conditions of crustal and upper mantle rocks is lacking. We used NW South America as a case-study to explore the spatial variation of the CST and the potential temperatures at which crustal earthquakes occur, computing the 3D steady-state thermal field taking into account lithology-constrained thermal parameters. Modelled hypocentral temperatures show a general agreement with the seismogenic windows of rocks and mineral assemblies expected in the continental crust. A few outliers in the hypocentral temperatures showcase nucleation conditions consistent with the seismogenic window of olivine-rich rocks, and are intepreted in terms of uncertainties in the Moho depths and/or in the earthquake hypocenters, or due to the presence of ultramafic rocks within the allochthonous crustal terranes accreted to this complex margin. Our results suggest that the two largest earthquakes recorded in the region (Murindo sequence, in 1992) nucleated at the lower boundary of the seismogenic crust, highlighting the importance of considering this transition into account when characterizing seismogenic sources for hazard assessments.

The dynamics of the Alps and surrounding regions is still not completely understood, partly because of a non-unique interpretation of its upper-mantle architecture. In this respect, it is unclear if interpreted slabs are consistent with the observed surface deformation and topography. We derive three-end member scenarios of lithospheric thickness and slab geometries by clustering available shear-wave tomography models into a statistical ensemble. We use these scenarios as input for geodynamic simulations and compare modelled topography, surface velocities and mantle flow to observations. We found that a slab detached beneath the Alps, but attached beneath the Northern Apennines captures first-order patterns in topography and vertical surface velocities and can provide a causative explanation for the observed seismicity.