AI-driven digital twinning quantifies the tradeoff between flood conveyance and riparian stability under climate extremes

This research paper presents an AI-driven digital twin framework that helps quantify the trade-off between flood conveyance and riparian vegetation stability under increasingly extreme climate conditions. Combining UAV imagery, deep learning, and hydrodynamic modelling, the study demonstrates how seasonal changes in vegetation influence river hydraulics and erosion risk, offering a more dynamic alternative to conventional flood models.

The findings show that complete vegetation removal may improve water conveyance but substantially increases bank erosion risk, while selective nature-based interventions can enhance bank stability with only minimal impacts on flood conveyance. The study highlights the potential of AI-enabled digital twins to support evidence-based, climate-adaptive river management by integrating ecological resilience with flood risk management. It will be of interest to researchers, engineers and practitioners working on flood modelling, nature-based solutions and climate adaptation.