The topographic growth of a mountain belt is commonly attributed to isostatic balance in response to crustal and lithospheric thickening. However, deeper mantle processes may also influence the topography of the Earth. Here, we discuss the role of these processes in the Eastern Cordillera (EC) of Colombia. The EC is an active, double-vergent fold and thrust belt that formed during the Cenozoic by the inversion of a Mesozoic rift, and topography there has risen up to ∼5,000 m (Cocuy Sierra). The belt is located ∼500 km away from the trench where two separate portions of the Nazca plate subduct below the South American plate. North of 5°N, the EC rises above a flat-slab subduction region. Volcanic arc migration implies slab shallowing by ∼10 Ma and flattening up to the present-day configuration at ∼6 Ma. The occurrence of a high vP/vS anomaly and clustered seismicity below the belt at ∼160 km depth delineates the slab geometry and has been related to dehydration of the slab, suggesting the presence of a hydrated mantle wedge.
We compiled thermochronologic data and inverted for the exhumation history of the chain over the last 20 Ma using the age-elevation relationship and the different closure temperatures of multiple thermochronologic systems. Results indicate that exhumation rates increased during the Plio–Pleistocene at different wavelengths and amplitudes. The small wavelength and large amplitude signals could be related to shallow crustal deformation, whereas the source of the long wavelength and moderate amplitude signal has yet to be identified. Pulses of fast exhumation are found to be concomitant with the uplift that occurred from ∼7 Ma to the present-day.
Previous studies suggested that the high topography of the chain cannot be achieved solely through isostatic adjustment. The highest residual topography is centered on the highest elevations of the EC, whereas the lowest residual topography corresponds to the Magdalena Valley, following the regional slab geometry. We propose that the recent uplift and exhumation events were triggered by the transition from regular to flat-slab subduction, along with the hydration of the mantle wedge above the slab. We test the dynamic feasibility of our hypothesis with a series of numerical models for the present-day state. Predicting the correct trends in elevation requires a flat-slab geometry, and a weak and buoyant mantle wedge.