River long profile modelling since the Mid-Pleistocene for the Ro Santa Cruz, Southern Patagonia.

Alluvial rivers connect sediment sources in mountain belts to depositional basins. They not only transport water and sediment, but also adjust to changing forcing conditions by aggrading or incising their beds, a process potentially recorded by fluvial terraces. The formation of stepped terrace sequences is commonly thought to be driven by cyclic changes in sediment and water supply, but may be modulated by rock uplift or sea-level changes. The Ro Santa Cruz in Southern Patagonia, Argentina, flows ca. 250 km from its glacial headwaters in the Andes eastward to the Atlantic Ocean. There are no major tributaries or substantial anthropogenic impacts along its course. A set of at least six exceptional fluvial terraces stretches along ca. 230 km and rise up to 110 m above the river. Our preliminary 10Be cosmogenic nuclide exposure dates show that river incision began at ca. 1 Ma, and that terrace formation proceeded in roughly 100-ky intervals, suggesting control by orbital climate cycles, likely through their impacts on the sediment-to-water supply ratio. However, a step in the terrace age-elevation sequence between 700 and 300 ky points to a change in net incision rate at that time. Particularly at the upstream end of the river, terraces have been uplifted at a rate nearly twice as high compared to the rest of the river. While it is likely that multiple factors affected the evolution of the Ro Santa Cruz over the last 1 Myr, the magnitude and spatial pattern of impacts from these different drivers is unclear. We apply a recent numerical model (GRLP), implemented here as a simple single-thread channel, to solve for the channel long profile evolution under different forcing scenarios. This approach allows us to test the impacts of individual, or combinations of, drivers on river-profile evolution. Our results suggest that Late Pleistocene 100 ky climate cycles have had the main impact on long profile evolution, especially along the upper 70 km of the river, with aggradation-incision cycles of up to an order of 10 m in magnitude. In contrast, sea-level change does not seem to influence significantly long profile evolution, as the exposed offshore slope does not change significantly compared to that onshore. To match the vertical distribution of terrace surfaces requires a long-term uplift rate of around 0.2 mm/yr, but with a hiatus between 700 and 300 ky. To accurately simulate the full terrace sequence, enhanced uplift is required upstream, decreasing exponentially towards the middle reaches.