Search for a command to run...
Landslide sediment in small mountain rivers (SMRs), particularly in East Asia, is a major source of the sediment exported from land to ocean. These landslides are usually triggered by earthquakes or rainstorms, at different locations on the hillslope: earthquake-induced landslides tend to occur in hillslope crest regions, whilst rainstorm-induced landslides tend to occur at the hillslope base. These characteristic landslide locations affect the timescales over which the sediment is transported to and through the river network. Previous studies found that landslides closer to the river network will have a shorter sediment residence time, whilst earthquake-driven landslides in hillslope crest regions have a longer residence time. Our earlier work has shown that earthquakes have a legacy impact on the location of the subsequent rainstorm-induced landslides, potentially increasing the sediment residence time of these events, compared to rainstorm-driven landslides that are not shaped by previous earthquakes . However, the effects of these legacy earthquake impacts on controlling sediment export from SMRs during successive rainstorm-triggered landslide events are not well understood, yet are likely to be important in countries such as Taiwan that are exposed to the combined effects of earthquakes and tropical rainstorms. In this study, we used the MassWastingRouter (MWR) model to simulate landslide sediment transport from the landslide source location to the river outlet in the Nei-Mao-Pu catchment, Choshui River, Taiwan, for the 2013 Nan-Tou earthquake and three subsequent rainstorm events, each with a reduced legacy impact of the Nan-Tou earthquake: Typhoon Soulik (2013), an extreme rainfall event (2015), and Typhoons Lekima and Bailiu (2019). We simulated landslide movement on hillslopes using the MassWastingRunout(MWRu) submodel, and then simulated the sediment transport from hillslopes to the river network using the MassWastingEroder(MWE) submodel. Next, the NetworkSedimentTransporter (from Landlab) was used to simulate fluvial sediment transport process to characterize the spatial and temporal dynamics of sediment transport from landslide locations to the river outlet. We then applied a functional connectivity-based analysis to explore time and space scales over which landslide-derived sediment from landslide source locations is connected to downstream locations within the river network. This approach enables us to better understand how sediment from different landslide locations contributes to overall sediment residence time within the system. The results demonstrate how the interaction between earthquakes and subsequent rainstorms ultimately controls sediment transport, providing crucial knowledge of sediment transport regimes and sediment source management in SMRs.