Our latest research has been featured in Imperial News. In collaboration with colleagues at Aarhus University, we developed a physics-informed AI framework that challenges a long-standing assumption about how water is retained in soil. By combining machine learning with physical principles, the study reveals previously hidden processes governing soil water retention, with potential implications for hydrology, agriculture, flood prediction, and climate modelling. More broadly, the work demonstrates how AI can help uncover new scientific understanding rather than simply fit existing models.
Read the full Imperial News story: AI tackles 50-year-old challenge in soil science – a new approach to discovering complex Earth processes
