Soil resources in tropical savannas are rapidly degrading, posing an imminent threat to food, water and livelihood security. Caveats in our understanding of geomorphological responses to extreme events are a major hindrance for attributing soil erosion and sediment flux dynamics to environmental drivers. Using the Burdekin and Makuyuni catchments as natural laboratories for semi-arid regions, the AsFoRESEEN project will assess feedback dynamics in soil erosion through the lens of variable sediment connectivity to test the hypothesis that extreme events can trigger regime shifts towards highly connected ephemeral gully networks. The project will develop novel approaches and integrate them with established techniques in an open-access diagnostic toolkit to support targeted soil- and water management interventions. Temporal dynamics in fine sediment and Phosphorous transport will be quantified using high-frequency sensors and sediment dating techniques. We will be the first to evaluate the use of secondary weathered metal species as tracers, providing a new pathway for attributing the contribution of gully erosion in deeply weathered or alluvials soils. Stream monitoring and sediment source tracing outputs will be integrated in a dynamic sediment budget to elucidate non-linear geomorphological responses to extreme events and land use changes. As a source of innovation, we will couple a machine-learning gully quantification tool with a dynamic catchment model, wherein gullies are both a direct source of sediment and a driver of changing sediment connectivity. The hybrid model will be used to test the efficacy of gully remediation strategies under current and future climatic conditions.