The uplift of the southern African Plateau is often attributed to mantle processes, but there are conflicting theories for the specific timing and drivers of topographic development. Evidence for most proposed plateau development histories is derived from continental erosion histories, marine stratigraphic architecture, or landscape morphology. Here we use a landscape evolution model to integrate these three types of data for southern Africa, including a large dataset of low temperature thermochronology, sediment flux rates to surrounding marine basins. We explore three main hypotheses for surface uplift: 1) southern Africa was already elevated at the time of Gondwana breakup, 2) uplift and continental tilting occurred in the mid-Cretaceous, or 3) uplift occurred in the mid to late Cenozoic. We test which of these three intervals of plateau development are plausible by using an inversion method to constrain the range in erosional and uplift model parameters that can best reproduce the observed data. Results indicate two families of uplift histories are most compatible with the data. Both have limited initial topography with some topographic uplift and continental tilting starting in the east at ~95 Ma. In one acceptable scenario, nearly all of the topography, ~1400 m, is created at this time with little Cenozoic uplift. In the other acceptable scenario, only ~500 m of uplift occurs in the mid-Cretaceous with another ~850 m of uplift in the mid-Cenozoic. The two model scenarios have different geodynamic implications, which in the future could be evaluated by direct comparison between geodynamic and landscape model predictions.

We use a Landscape Evolution Model (FastScape S2S) to explore the impact of inherited topography in the foreland domain of a rising mountain range on its stratigraphic architecture and sediment accumulation history, inspired by the northern Pyrenean foreland. We simulate an uplifting half mountain range, its foreland basin and forebulge, and beyond, an open marine domain. We ran models with 4 different initial reliefs in the foreland domain: an initially flat foreland domain at sea-level, an elevated flat continental foreland (+300 m), a pre-existing 1 km-deep and 100 km-wide bathymetry at the location of the future foreland basin associated with a forebulge domain either at sea-level or elevated at +300m. All models show a prograding mega-sequence associated with building of mountain topography and development of the flexural foreland basin and forebulge, coalescence of alluvial fans at the foot of the range, progressive continentalization of the foreland domain, and burial of the forebulge. An initially elevated foreland domain ultimately produces a thinner foreland basin while an initially deep foreland basin produces a thicker one. After 10-13 Myr, the initial relief of foreland domain is smoothed out and the landscape does not exhibit a record of pre-existing relief. In contrast, the stratigraphic architecture of the foreland basin allows to trace inherited relief with deep marine sediments in the initially deep foreland basin, marine sediments onlapping and then burying the forebulge initially at sea-level, and continental sediments onlapping and burying the initially elevated foreland domain. We compare these interpretations to the Pyrenean retro-foreland.