Drainage reversal of the Amazon River due to the coupling of surface and lithospheric processes

- Numerical models of surface processes and lithospheric flexure in northern S. America.
- Amazon drainage reversal due to surface processes dynamics during Andean formation.
- Dynamic topography plays a secondary role in the drainage reversal.
- Timing of reversal controlled by sediment transport efficiency and Andean uplift rate.
- Prediction of sedimentation increase in the Amazon fan after the drainage reversal.

The uplift of central and northern Andes occurred concomitantly with an important reorganization of the drainage pattern in the northern South America. In Early Miocene, the fluvial systems that flowed from the Amazonian craton towards the sub-Andean basins and northward to the Caribbean were replaced by a megawetland, the Pebas system, covering more than 106 km2 in western Amazonia. By Late Miocene the Pebas system progressively disappeared and gave place to the transcontinental Amazon River, connecting the Andes and the equatorial Atlantic margin. A previous work suggested that the reversal of the Amazon River and the disappearance of the Pebas system were driven by dynamic topography promoted by mantle convection. Based on a three-dimensional numerical model that couple surface processes, flexural isostasy and crustal thickening due to orogeny, here I propose that the response of the surface processes to the uplift of the central and northern Andes, along with the flexural isostasy of the lithosphere, can explain the drainage reversal of the Amazon River during Miocene time with no need to invoke dynamic topography induced by mantle convection. Moreover, according to this new numerical model, the timing of the drainage reversal in the northern South America is mainly controlled by the efficiency of the sediment transport within the drainage basins and the rate of the Andean uplift. Also, the present numerical experiments were able to predict the increase in sedimentation rate in the Amazon fan after the drainage reversal of the Amazon River as observed in Late Miocene-Pliocene sedimentary record. However, the proposed model fails to fully reproduce the spacial and temporal evolution of the Pebas system as observed in geological data. Further investigation is therefore needed in order to understand the dynamic interaction between surface and tectonic processes and their implications on the development of a megawetland that preceded the reversal of Amazon River drainage. Nevertheless, this work can give new insights into the landscape and stratigraphic evolution of northern South America during Andean orogeny, providing quantitative constraints for the paleogeography reconstructions of Amazonia.