We present new tomographic models of the Nazca slab under South America from 6 degrees S to 32 degrees S, and from 95 km to lower mantle (895 km) depths. By combining data from 14 separate networks in the central Andes, we use finite-frequency teleseismic P-wave tomography to image the Nazca slab from the upper mantle into the mantle transition zone (MTZ) and the uppermost lower mantle on a regional scale. Our tomography shows that there is significant along-strike variation in the morphology of the Nazca slab in the MTZ and the lower mantle. Thickening of the slab in the MTZ is observed north of the Bolivian orocline, possibly related to buckling or folding of the slab in response to the penetration of a near-vertical slab into the higher-viscosity lower mantle, which decreases the sinking velocity of the slab. South of the orocline, the slab continues into the lower mantle with only minor deformation in the MTZ. In the lower mantle, a similar difference in morphology is observed. North of 16 degrees S, the slab anomaly in the lower mantle is more coherent and penetrates more steeply into the lower mantle. To the south, the slab dip appears to be decreasing just below the 660 km discontinuity. This change in slab morphology in the MTZ and lower mantle appears to correspond to the change in the dip of the slab as it enters the MTZ, from steeply dipping in the north to more moderately dipping in the south.

We present new tomographic models of the Nazca slab under South America from 6 degrees S to 32 degrees S, and from 95 km to lower mantle (895 km) depths. By combining data from 14 separate networks in the central Andes, we use finite-frequency teleseismic P-wave tomography to image the Nazca slab from the upper mantle into the mantle transition zone (MTZ) and the uppermost lower mantle on a regional scale. Our tomography shows that there is significant along-strike variation in the morphology of the Nazca slab in the MTZ and the lower mantle. Thickening of the slab in the MTZ is observed north of the Bolivian orocline, possibly related to buckling or folding of the slab in response to the penetration of a near-vertical slab into the higher-viscosity lower mantle, which decreases the sinking velocity of the slab. South of the orocline, the slab continues into the lower mantle with only minor deformation in the MTZ. In the lower mantle, a similar difference in morphology is observed. North of 16 degrees S, the slab anomaly in the lower mantle is more coherent and penetrates more steeply into the lower mantle. To the south, the slab dip appears to be decreasing just below the 660 km discontinuity. This change in slab morphology in the MTZ and lower mantle appears to correspond to the change in the dip of the slab as it enters the MTZ, from steeply dipping in the north to more moderately dipping in the south.

Subduction of the Nazca and Caribbean Plates beneath northwestern Colombia is seen in two distinct Wadati Benioff Zones, one associated with a flat slab to the north and one associated with normal subduction south of 5.5 degrees N. The normal subduction region is characterized by an active arc, whereas the flat slab region has no known Holocene volcanism. We analyze volcanic patterns over the past 14 Ma to show that in the mid-Miocene a continuous arc extended up to 7 degrees N, indicating normal subduction of the Nazca Plate all along Colombia's Pacific margin. However, by similar to 6 Ma, we find a complete cessation of this arc north of 3 degrees N, indicating the presence of a far more laterally extensive flat slab than at present. Volcanism did not resume between 3 degrees N and 6 degrees N until after 4 Ma, consistent with lateral tearing and resteepening of the southern portion of the Colombian flat slab at that time.

Subduction of the Nazca and Caribbean Plates beneath northwestern Colombia is seen in two distinct Wadati Benioff Zones, one associated with a flat slab to the north and one associated with normal subduction south of 5.5 degrees N. The normal subduction region is characterized by an active arc, whereas the flat slab region has no known Holocene volcanism. We analyze volcanic patterns over the past 14 Ma to show that in the mid-Miocene a continuous arc extended up to 7 degrees N, indicating normal subduction of the Nazca Plate all along Colombia's Pacific margin. However, by similar to 6 Ma, we find a complete cessation of this arc north of 3 degrees N, indicating the presence of a far more laterally extensive flat slab than at present. Volcanism did not resume between 3 degrees N and 6 degrees N until after 4 Ma, consistent with lateral tearing and resteepening of the southern portion of the Colombian flat slab at that time.

Seismic anisotropy has been documented in many portions of the lowermost mantle, with particularly strong anisotropy thought to be present along the edges of large low shear velocity provinces (LLSVPs). The region surrounding the Pacific LLSVP, however, has not yet been studied extensively in terms of its anisotropic structure. In this study, we use seismic data from southern Peru, northern Bolivia and Easter Island to probe lowermost mantle anisotropy beneath the eastern Pacific Ocean, mostly relying on data from the Peru Lithosphere and Slab Experiment and Central Andean Uplift and Geodynamics of High Topography experiments. Differential shear wave splitting measurements from phases that have similar ray paths in the upper mantle but different ray paths in the lowermost mantle, such as SKS and SKKS, are used to constrain anisotropy in D ''. We measured splitting for 215 same station-event SKS-SKKS pairs that sample the eastern Pacific LLSVP at the base of the mantle. We used measurements of splitting intensity(SI), a measure of the amount of energy on the transverse component, to objectively and quantitatively analyse any discrepancies between SKS and SKKS phases. While the overall splitting signal is dominated by the upper-mantle anisotropy, a minority of SKS-SKKS pairs (similar to 10 per cent) exhibit strongly discrepant splitting between the phases (i.e. the waveforms require a difference in SI of at least 0.4), indicating a likely contribution from lowermost mantle anisotropy. In order to enhance lower mantle signals, we also stacked waveforms within individual subregions and applied a waveform differencing technique to isolate the signal from the lowermost mantle. Our stacking procedure yields evidence for substantial splitting due to lowermost mantle anisotropy only for a specific region that likely straddles the edge of Pacific LLSVP. Our observations are consistent with the localization of deformation and anisotropy near the eastern boundary of the Pacific LLSVP, similar to previous observations for the African LLSVP.

Flat or near-horizontal subduction of oceanic lithosphere has been an important tectonic process both currently and in the geologic past. Subduction of the aseismic Nazca Ridge beneath South America has been associated with the onset of flat subduction and the termination of arc volcanism in Peru, making it an ideal place to study flat-slab subduction. Recently acquired seismic recordings for 144 broadband seismic stations in Peru permit us to image the Mohorovicic discontinuity (Moho) of the subducted oceanic Nazca plate, Nazca Ridge, and the overlying continental Moho of the South American crust in detail through the calculation of receiver functions. We find that the subducted over-thickened ridge crust is likely significantly eclogitized similar to 350 km from the trench, requiring that the inboard continuation of the flat slab be supported by mechanisms other than low-density crustal material. This continuation coincides with a low-velocity anomaly identified in prior tomography studies of the region immediately below the flat slab, and this anomaly may provide some support for the flat slab. The subduction of the Nazca Ridge has displaced most, if not the entire South American lithospheric mantle beneath the high Andes as well as up to 10 km of the lowermost continental crust. The lack of deep upper-plate seismicity suggests that the Andean crust has remained warm during flat subduction and is deforming ductilely around the subducted ridge. This deformation shows significant coupling between the subducting Nazca oceanic plate and overriding South American continental plate up to similar to 500 km from the trench. These results provide important modern constraints for interpreting the geological consequences of past and present flat-slab subduction locations globally.

Flat or near-horizontal subduction of oceanic lithosphere has been an important tectonic process both currently and in the geologic past. Subduction of the aseismic Nazca Ridge beneath South America has been associated with the onset of flat subduction and the termination of arc volcanism in Peru, making it an ideal place to study flat-slab subduction. Recently acquired seismic recordings for 144 broadband seismic stations in Peru permit us to image the Mohorovicic discontinuity (Moho) of the subducted oceanic Nazca plate, Nazca Ridge, and the overlying continental Moho of the South American crust in detail through the calculation of receiver functions. We find that the subducted over-thickened ridge crust is likely significantly eclogitized similar to 350 km from the trench, requiring that the inboard continuation of the flat slab be supported by mechanisms other than low-density crustal material. This continuation coincides with a low-velocity anomaly identified in prior tomography studies of the region immediately below the flat slab, and this anomaly may provide some support for the flat slab. The subduction of the Nazca Ridge has displaced most, if not the entire South American lithospheric mantle beneath the high Andes as well as up to 10 km of the lowermost continental crust. The lack of deep upper-plate seismicity suggests that the Andean crust has remained warm during flat subduction and is deforming ductilely around the subducted ridge. This deformation shows significant coupling between the subducting Nazca oceanic plate and overriding South American continental plate up to similar to 500 km from the trench. These results provide important modern constraints for interpreting the geological consequences of past and present flat-slab subduction locations globally.

Current end-member models for the geodynamic evolution of orogenic plateaus predict (a) slow and steady rise during crustal shortening and ablative subduction (i.e., continuous removal) of the lower lithosphere or (b) rapid surface uplift following shortening, which is associated with punctuated removal of dense lower lithosphere and/or lower crustal flow. This review integrates results from recent studies of the modern lithospheric structure, geologic evolution, and surface uplift history of the Central Andean Plateau to evaluate the geodynamic processes involved in forming it. Comparison of the timing, magnitude, and distribution of shortening and surface uplift, in combination with other geologic evidence, highlights the pulsed nature of plateau growth. We discuss specific regions and time periods that show evidence for end-member geodynamic processes, including middle-late Miocene surface uplift of the southern Eastern Cordillera and Altiplano associated with shortening and ablative subduction, latest Oligocene-early Miocene and late Miocene-early Pliocene punctuated removal of dense lower lithosphere in the Eastern Cordillera and Altiplano, and late Miocene-early Pliocene crustal flow in the central and northern Altiplano.

Current end-member models for the geodynamic evolution of orogenic plateaus predict (a) slow and steady rise during crustal shortening and ablative subduction (i.e., continuous removal) of the lower lithosphere or (b) rapid surface uplift following shortening, which is associated with punctuated removal of dense lower lithosphere and/or lower crustal flow. This review integrates results from recent studies of the modern lithospheric structure, geologic evolution, and surface uplift history of the Central Andean Plateau to evaluate the geodynamic processes involved in forming it. Comparison of the timing, magnitude, and distribution of shortening and surface uplift, in combination with other geologic evidence, highlights the pulsed nature of plateau growth. We discuss specific regions and time periods that show evidence for end-member geodynamic processes, including middle-late Miocene surface uplift of the southern Eastern Cordillera and Altiplano associated with shortening and ablative subduction, latest Oligocene-early Miocene and late Miocene-early Pliocene punctuated removal of dense lower lithosphere in the Eastern Cordillera and Altiplano, and late Miocene-early Pliocene crustal flow in the central and northern Altiplano.

Foreland deformation has long been associated with flat-slab subduction, but the precise mechanism linking these two processes remains unclear. One example of foreland deformation corresponding in space and time to flat subduction is the Fitzcarrald Arch, a broad NE-SW trending topographically high feature covering an area of > 4 x 10(5) km(2) in the Peruvian Andean foreland. Recent imaging of the southern segment of Peruvian flat slab shows that the shallowest part of the slab, which corresponds to the subducted Nazca Ridge northeast of the present intersection of the ridge and the Peruvian trench, extends up to and partly under the southwestern edge of the arch. Here, we evaluate models for the formation of this foreland arch and find that a basal-shear model is most consistent with observations. We calculate that similar to 5 km of lower crustal thickening would be sufficient to generate the arch's uplift since the late Miocene. This magnitude is consistent with prior observations of unusually thickened crust in the Andes immediately south of the subducted ridge that may also have been induced by flat subduction. This suggests that the Fitzcarrald Arch's formation by the Nazca Ridge may be one of the clearest examples of upper plate deformation induced through basal shear observed in a flat-slab subduction setting. We then explore the more general implications of our results for understanding deformation above flat slabs in the geologic past.