Final Report Summary - ANDYN (the andes-nazca dynamic topography experiment)
Project context and objectives
Crustal tectonics and mantle flow area the main driving mechanisms of Earth's topography. Crustal tectonics is associated with isostatically (equilibrium) topography and mantle convection with dynamically supported long wavelength deflections that explain geoid anomalies. We term the latter dynamic topography. South America is an ideal natural laboratory to analyse and quantify these two contrasting components. The South American Andes are a still active orogenic system affected by complex geodynamic processes like the subduction of oceanic ridges with thicker than average oceanic curst (Nazca, Juan Fernandez and Chile, among others) (see Figure 1). These subducting features affect not only the convergence dynamics and stresses along the entire margin but also the distribution of mass anomalies in the mantle, which drive sublithospheric flow and generate dynamic topography.
Project work
In this project we demonstrated using instantaneous flow models and flexural calculations that most of South America has been uncompensated since the beginning of the Cenozoic (when the Andes started to grow) and that additional forces, such as mantle down-wellings and up-wellings, are required to account for the observed topography. Although earlier dynamic topography models have predicted large negative and subsiding areas along most of South America (~1000 km across, tens of kilometres along and thousands of metres in amplitude), associated with the subduction history of the extinct Farallon Plate, no such subsidence is evident in the geological record. These previous studies did not account for the large along-strike variations of the slab dip angle in South America or the supracrustal loading effects on the total topography, especially close to the Cordillera. These large along-strike dip changes manifest themselves as two large flat segments, one in Peru (14° S) and one in Argentina (32 °S). Previous work on the effects of flat subduction on dynamic topography suggested that slab flattening induced greater subsidence than normal subduction. The results were used to suggest that slab flattening was one of the major driving mechanisms of long-wavelength subsidence and intracontinental marine flooding during the Cretaceous-Early Cenozoic of North America. We found the exact opposite, in good agreement with the geological record of South America.
We focused our dynamical models on the instantaneous viscous flow and stresses induced by subducted slabs. The density anomaly associated with cold dense slabs in an average mantle, excites down-welling flow that produces surface deformation, explains plate motions, and geoid anomalies. An important advantage of focusing on subducted heterogeneity is that slab morphology is known in some detail. Consequently we designed a present-day model (PDM), representative of the evolution from the Mio-Pliocene to the present-day (Figure 2), which is derived from constraints on slab morphology from regional seismic and magneto-telluric surveys. The subduction angle varies along strike from flat (0°) to normal (~30°) and the density contrast (??) of the slab (with respect to the mantle) is controlled by the slab age, the inferred crustal thickness and the Benioff zone geometry. In order to explore time variation of dynamic topography, we also construct a simplified Cenozoic model (CM) in which the slab dips 15° (Southern Andes) to 30° (Central and Northern Andes) eastward homogeneously along strike.
Project results
Our main results show that the geometry and density distribution of the Nazca slab had a strong effect on the pattern and amplitude of dynamic topography. The PDM reproduced a significant migration of the topographic wave position to the east with respect to the CM results, especially along ridge segments, and a strong reduction in the subsidence amplitude. The latter is the direct result of a using a ?? = 0 over the subducted aseismic ridges, which simulates the expected neutrally buoyant oceanic lithosphere of segments with thicker than average oceanic crust. Particularly, across the flat slab segments, dynamic subsidence (or negative dynamic topography) is reduced by ~400 m in Peru and not recorded at all in Argentina. In the latter we observe only a minor signal (~100 m) above and surrounding the plunging part of the slab leading edge >800 km away from the trench.
We also examined the temporal evolution of dynamic topography induced by progressive flattening of the slab by comparing the PDM and CM, and computing the change in dynamic topography after the break-up of the Farallon Plate and the creation of Nazca (Neogene). Flattening produces positive signals, consistent with dynamic uplift, especially over subducting ridges. The dynamic topography evolution shows that the proximal forelands (foredeeps) of South America would have profoundly subsided in its early stages and then uplifted and rotated clockwise (<0.5°) as a consequence of a progressive slab shallowing only, during the Miocene to Present. This overlaps with the standard flexural history of the Andean foreland driven by Cordilleran shortening (loading).
We compared our predicted dynamic topography with geological and geophysical observations that are good proxies of the topographic evolution and, hence, of the observed dynamic signals in the Andean foreland. Non-isostatic residual topography calculations across a representative swath in the flat-slab segment of Argentina agree with the general trend of our models, i.e. positive residual topography on the flat slabs and negative where slabs plunge with 30 ° dip. Oceanic trench bathymetry (corrected topography to the top of the basement) and a total relief analysis in the stable foreland of Argentina (outside the influence of Andean tectonics) also support our conclusions. Moreover, the morphology of the distal foreland helps us in calibrating our models and proposing an "observed dynamic topography" value of =300 m. This observed value correlates remarkably well with the 'predicted' dynamic topography calculated across the South American plains.
Observations (geological and geophysical) across the still active flat-slab segments of South America agree with our model predictions. Dynamic uplift is more consistent than subsidence across segments affected by ridge dynamics. In fact, in the southern part of the Peruvian foreland up to >700 km from the trench, we find an elevated and long-wavelength bulge: the Fitzcarrald Arch. Further south, above the Argentine-Chilean flat-slab segment, the distal foreland (>600 km eastward) also shows a high-elevation system: the Sierras Pampeanas broken foreland. These feature date to the latest Miocene-Pliocene, coeval with the advent of flat subduction. The same occurs in Patagonia, where an emerging plateau has developed since the arrival of the Chile seismic ridge.
In Southern Patagonia, we estimated dynamic topography on the basis of detailed plate reconstructions in three tectonic stages 19-12 and 12-0 my and present-day. These stages correspond to key episodes of the Chile ridge dynamics. The uplift of the Patagonian plateau (from hinterland to the coast) is associated with the subduction of the ridge since the early Miocene (dynamic uplift). This model does not reproduce a key feature of the SW Atlantic margin of South America, the very deep Argentine abyssal basin. Neither do models with accurate reconstructions of the slab morphology in the upper mantle. Models that use long-wavelength seismic tomography that produce subsidence over all of South America do show strong negative dynamic topography over the Argentine basin but without the right geometry and only as a result of very large regional subsidence that does not match the geological and geophysical data. Our current working hypothesis is that instead this area is an over-thickened oceanic lithospheric region, perhaps prone to oceanic lithospheric delamination or small-scale convection.
Project impact
Our results have broader implications for understanding the subduction-induced topography in foreland, platform and oceanic regions. The "anomalous" subsidence pattern of the Western Interior basins in the US is not related to flat subduction but rather by the subsidence induced by the leading edge of the slab. Our work allowed initiating a cooperation project with the National Petroleum Company of Argentina, YPF. We started to design a model to understand the subsidence and exhumation of the pericratonic foreland of South America, where dynamic topography is likely the main control on the generation of sedimentary accommodation spaces. Our new ideas have also formed part of teaching and short courses since the beginning of the project in the UK and Argentina. We also started new projects transferring our new hypotheses on unexplored regions of northern South America and Patagonia, which will allow a more fluent link between industry and science. A postdoc student started work in Patagonia to test our results and a Masters thesis is being finished in Peru, based on our ideas.
Crustal tectonics and mantle flow area the main driving mechanisms of Earth's topography. Crustal tectonics is associated with isostatically (equilibrium) topography and mantle convection with dynamically supported long wavelength deflections that explain geoid anomalies. We term the latter dynamic topography. South America is an ideal natural laboratory to analyse and quantify these two contrasting components. The South American Andes are a still active orogenic system affected by complex geodynamic processes like the subduction of oceanic ridges with thicker than average oceanic curst (Nazca, Juan Fernandez and Chile, among others) (see Figure 1). These subducting features affect not only the convergence dynamics and stresses along the entire margin but also the distribution of mass anomalies in the mantle, which drive sublithospheric flow and generate dynamic topography.
Project work
In this project we demonstrated using instantaneous flow models and flexural calculations that most of South America has been uncompensated since the beginning of the Cenozoic (when the Andes started to grow) and that additional forces, such as mantle down-wellings and up-wellings, are required to account for the observed topography. Although earlier dynamic topography models have predicted large negative and subsiding areas along most of South America (~1000 km across, tens of kilometres along and thousands of metres in amplitude), associated with the subduction history of the extinct Farallon Plate, no such subsidence is evident in the geological record. These previous studies did not account for the large along-strike variations of the slab dip angle in South America or the supracrustal loading effects on the total topography, especially close to the Cordillera. These large along-strike dip changes manifest themselves as two large flat segments, one in Peru (14° S) and one in Argentina (32 °S). Previous work on the effects of flat subduction on dynamic topography suggested that slab flattening induced greater subsidence than normal subduction. The results were used to suggest that slab flattening was one of the major driving mechanisms of long-wavelength subsidence and intracontinental marine flooding during the Cretaceous-Early Cenozoic of North America. We found the exact opposite, in good agreement with the geological record of South America.
We focused our dynamical models on the instantaneous viscous flow and stresses induced by subducted slabs. The density anomaly associated with cold dense slabs in an average mantle, excites down-welling flow that produces surface deformation, explains plate motions, and geoid anomalies. An important advantage of focusing on subducted heterogeneity is that slab morphology is known in some detail. Consequently we designed a present-day model (PDM), representative of the evolution from the Mio-Pliocene to the present-day (Figure 2), which is derived from constraints on slab morphology from regional seismic and magneto-telluric surveys. The subduction angle varies along strike from flat (0°) to normal (~30°) and the density contrast (??) of the slab (with respect to the mantle) is controlled by the slab age, the inferred crustal thickness and the Benioff zone geometry. In order to explore time variation of dynamic topography, we also construct a simplified Cenozoic model (CM) in which the slab dips 15° (Southern Andes) to 30° (Central and Northern Andes) eastward homogeneously along strike.
Project results
Our main results show that the geometry and density distribution of the Nazca slab had a strong effect on the pattern and amplitude of dynamic topography. The PDM reproduced a significant migration of the topographic wave position to the east with respect to the CM results, especially along ridge segments, and a strong reduction in the subsidence amplitude. The latter is the direct result of a using a ?? = 0 over the subducted aseismic ridges, which simulates the expected neutrally buoyant oceanic lithosphere of segments with thicker than average oceanic crust. Particularly, across the flat slab segments, dynamic subsidence (or negative dynamic topography) is reduced by ~400 m in Peru and not recorded at all in Argentina. In the latter we observe only a minor signal (~100 m) above and surrounding the plunging part of the slab leading edge >800 km away from the trench.
We also examined the temporal evolution of dynamic topography induced by progressive flattening of the slab by comparing the PDM and CM, and computing the change in dynamic topography after the break-up of the Farallon Plate and the creation of Nazca (Neogene). Flattening produces positive signals, consistent with dynamic uplift, especially over subducting ridges. The dynamic topography evolution shows that the proximal forelands (foredeeps) of South America would have profoundly subsided in its early stages and then uplifted and rotated clockwise (<0.5°) as a consequence of a progressive slab shallowing only, during the Miocene to Present. This overlaps with the standard flexural history of the Andean foreland driven by Cordilleran shortening (loading).
We compared our predicted dynamic topography with geological and geophysical observations that are good proxies of the topographic evolution and, hence, of the observed dynamic signals in the Andean foreland. Non-isostatic residual topography calculations across a representative swath in the flat-slab segment of Argentina agree with the general trend of our models, i.e. positive residual topography on the flat slabs and negative where slabs plunge with 30 ° dip. Oceanic trench bathymetry (corrected topography to the top of the basement) and a total relief analysis in the stable foreland of Argentina (outside the influence of Andean tectonics) also support our conclusions. Moreover, the morphology of the distal foreland helps us in calibrating our models and proposing an "observed dynamic topography" value of =300 m. This observed value correlates remarkably well with the 'predicted' dynamic topography calculated across the South American plains.
Observations (geological and geophysical) across the still active flat-slab segments of South America agree with our model predictions. Dynamic uplift is more consistent than subsidence across segments affected by ridge dynamics. In fact, in the southern part of the Peruvian foreland up to >700 km from the trench, we find an elevated and long-wavelength bulge: the Fitzcarrald Arch. Further south, above the Argentine-Chilean flat-slab segment, the distal foreland (>600 km eastward) also shows a high-elevation system: the Sierras Pampeanas broken foreland. These feature date to the latest Miocene-Pliocene, coeval with the advent of flat subduction. The same occurs in Patagonia, where an emerging plateau has developed since the arrival of the Chile seismic ridge.
In Southern Patagonia, we estimated dynamic topography on the basis of detailed plate reconstructions in three tectonic stages 19-12 and 12-0 my and present-day. These stages correspond to key episodes of the Chile ridge dynamics. The uplift of the Patagonian plateau (from hinterland to the coast) is associated with the subduction of the ridge since the early Miocene (dynamic uplift). This model does not reproduce a key feature of the SW Atlantic margin of South America, the very deep Argentine abyssal basin. Neither do models with accurate reconstructions of the slab morphology in the upper mantle. Models that use long-wavelength seismic tomography that produce subsidence over all of South America do show strong negative dynamic topography over the Argentine basin but without the right geometry and only as a result of very large regional subsidence that does not match the geological and geophysical data. Our current working hypothesis is that instead this area is an over-thickened oceanic lithospheric region, perhaps prone to oceanic lithospheric delamination or small-scale convection.
Project impact
Our results have broader implications for understanding the subduction-induced topography in foreland, platform and oceanic regions. The "anomalous" subsidence pattern of the Western Interior basins in the US is not related to flat subduction but rather by the subsidence induced by the leading edge of the slab. Our work allowed initiating a cooperation project with the National Petroleum Company of Argentina, YPF. We started to design a model to understand the subsidence and exhumation of the pericratonic foreland of South America, where dynamic topography is likely the main control on the generation of sedimentary accommodation spaces. Our new ideas have also formed part of teaching and short courses since the beginning of the project in the UK and Argentina. We also started new projects transferring our new hypotheses on unexplored regions of northern South America and Patagonia, which will allow a more fluent link between industry and science. A postdoc student started work in Patagonia to test our results and a Masters thesis is being finished in Peru, based on our ideas.
