CAREER: Uncertainty-Informed Design of Regional Natural Hazard Engineering Simulations for Infrastructure Risk and Recovery Assessment

This Faculty Early Career Development Program (CAREER) award will support the development of new methods for simulation-driven regional-scale risk and recovery assessment of civil infrastructure systems subjected to extreme and cascading natural hazards. Evaluating the vulnerability of infrastructure to extreme event scenarios is crucial for protecting lives, preserving economic security, and accelerating post-event community recovery. However, designing informative regional-scale computer simulations of such scenarios is challenged by the spatial and temporal variability of extreme events and their cascading threats to infrastructure. These challenges are compounded by the computational costs of high-resolution regional simulations and lack of scientific mechanisms to identify the most informative simulation scenarios. This project will address these challenges by developing a computational paradigm for designing uncertainty-informed regional-scale simulations, which will enable engineers to understand the consequences of cascading natural hazards and optimize the allocation of computational resources toward reliable infrastructure risk and recovery assessment. The project outcomes will advance national welfare and contribute to the National Science Foundation’s role in supporting the National Earthquake Hazards Reduction Program by improving earthquake risk assessment and recovery planning methods and understanding of the consequences of earthquake sequences on infrastructure. The research outcomes will be translated into computational tools for the natural hazard engineering community and educational materials on uncertainty quantification for engineering students and professionals. This CAREER award will address the fundamental gap between low-resolution probabilistic modeling and high-resolution physics-based simulations in the context of regional-scale analysis of infrastructure systems subjected to high-consequence low-frequency fault-rupture events and their cascading threats. The main contribution is a computational method for quantifying and reducing uncertainty in natural hazard engineering through adaptive acquisition of the most informative simulations or field data. The project will develop schemes for coupling physics-based and probabilistic simulations using the concepts of optimal design of computer experiments, information theory, and artificial intelligence models. The project has three objectives: (1) to optimize the design of physics-based simulations of cascading seismic events and distributed infrastructure systems using an uncertainty-informed regional sequential Bayesian experimental design framework; (2) to enable learning of infrastructure recovery trajectories and dependencies by integrating regional simulations, stochastic recovery analysis, and Bayesian network models; and (3) to optimize feedback loops between field-data acquisition and natural-hazard simulation using an adaptive spatiotemporal optimization scheme. The project outcomes will enable scientific discovery and engineering assessments in typically cost-prohibitive contexts, specifically: measuring through simulation the regional vulnerabilities and recovery trajectories of multiple infrastructure systems under cascading extreme events. 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. NSF Award ID: 2541092 | Program: 01002627DB NSF RESEARCH & RELATED ACTIVIT | Principal Investigator: Maha Kenawy | Institution: Oklahoma State University, STILLWATER, OK | Award Amount: $629,101 View on NSF Award Search: https://www.nsf.gov/awardsearch/show-award/?AWD_ID=2541092 View on Research.gov: https://www.research.gov/awardapi-service/v1/awards/2541092.html