Constraining 90 Ma landscape evolution model of Madagascar using erosional and sedimentary data
Résumé
Madagascar has its own geological and geomorphic histories since its separation from India at ~90 Ma. Since then despite the lack of intense tectonic activity, the island has undergone two phases of topographic uplift, which were driven by 1) flexural rebound due to sediment loading along the west coast and erosional unloading on the high plateau during the Late Cretaceous, and 2) dynamic doming supported by the convective upwelling of the sub-lithosphere mantle during the Late Cenozoic. The temporal and spatial variations of the uplift shaped the fluvial network, e.g. by modifying channel gradients and flow directions. The evolution of the drainage basins played an important role in creating the complex topography, which has been causally linked to the island's high-level of biodiversity and biotic endemism. Therefore, reconstructing the long-term landscape evolution and drainage network history is a key step in revealing the potential link between the Malagasy landscape and its biogeographic evolution.
We attempt to reproduce the landscape evolution of Madagascar by combining a wide range of observational constraints and using a numerical landscape evolution model, in which the surface evolution is predicted as a consequence of uplift, fluvial incision and hillslope procesess. Since the Early Cretaceous, several erosional surfaces were created by planation, but the ages of their formation are not well dated. We use the vertical offsets between the remnants of these surfaces as proxies for uplift magnitudes, and use an inversion approach (Neighbourhood Algorithm) to estimate the ages of the uplift as well as the uncertainties on these estimates. Simultaneously, parameters in the fluvial erosion and hillslope diffusion equations are inverted. Sediment flux data from the Morondava Basin along the west coast is used to constrain the inversion, assuming a general balance between the sediment volume preserved in the basin and the eroded mass from the source area. Other model constraints include the present topography and catchment geometry. Our current best-fit model shows that although the Cenozoic uplift started from ~30 Ma, most of the current elevation of the island was gained over the last ~10 Ma, during which period the drainage system of the island has been dynamically reorganized.