Multi-Century Erosion and Landscape Evolution of Ephemeral Catchments in Response to Sub-Daily Rainfall Distribution Changes

Changes in the properties of rainfall distributions at sub-daily scales are key to assessing soil erosion rates under climate transition. However, such changes are difficult to detect and model, especially over landscape evolution timescales. In this contribution, we validate a new catchment-scale landscape evolution model against event-scale runoff and sediment records. Through multi-century numerical experiments, we also show that changes in the sub-daily rainfall distribution, like those observed under modern climate change, can increase soil erosion rates by $\sim$40% but cannot be accurately inferred from changes in the average event properties and total rainfall. We quantify erosion and topographic trajectories associated with plausible changes in the sub-daily rainfall distribution, highlighting scenarios in which shifting tail properties impact landscape evolution, at times, contrary to expectations based on changes in total rainfall.