Predicting flood-induced flow and sediment dynamics using data-driven physics-informed models

Suny At Stony Brook, Stony Brook NY

Investigators

Abstract

It is vital to understand the physics of flood flow in rivers. Such understanding can help practicing engineers, researchers, and stakeholders (a) appropriately design infrastructures along and across rivers and (b) better protect the river environment. To understand the flood flow in rivers, this project will develop and utilize artificial intelligence to produce mean flow characteristics of the flood flow and riverbed topography. The developed artificial intelligence algorithms will enable reliable flood flow prediction at a small fraction of the computational cost of existing models. Thus, this research will benefit society by providing practicing engineers and stakeholders with modeling tools to (a) predict flood impacts on the stability of infrastructure in natural waterways and (b) evaluate the efficiency of flood mitigation strategies. The project will support two graduate students for four years and engage 12 undergraduates in summer research activities relevant to machine learning and flood prediction. As a part of this project, underserved adolescents in a non-secure detention center in New York will be trained in computer programing and simple machine learning. The research goals of this project include (a) improving the existing capabilities of high-fidelity numerical modeling tools to enable physics-based simulations of floods in natural waterways, (b) applying the high-fidelity model to evaluate flood impacts on the stability of infrastructure installed in large-scale waterways, and (c) employing the high-fidelity flood simulation results of large-scale rivers to inform and develop machine learning algorithms for efficient and affordable prediction of flood impacts on infrastructure and potentially transform the field of flood prediction research. This project will involve a group of practitioners from private, local, and federal agencies to learn how to use the developed tools. The research and educational objectives of this proposal will result in (a) advancement of knowledge regarding the complex interactions among flood flow, waterway, vegetation, and infrastructure; (b) promotion of educational performance by (i) engaging underserved adolescents of a non-secure detention center in STEM, (ii) encouraging the participation of underrepresented groups in research, (iii) involving graduate and undergraduate students in multidisciplinary research, and (iv) developing a new interdisciplinary graduate course; and (c) broad dissemination of high-fidelity models and machine learning through online repositories and journal publications for free access by practitioners and the scientific community. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.