Seminar: Dr. Nadal-Caraballo, U.S. Army Corps of Engineers, Research and Development Center

UNC Marine Sciences Professor and UNC Institute of Marine Sciences Director Rick Luettich hosts a seminar presented by Norberto Nadal-Caraballo, Ph.D..

Presenter Affiliations: Leader, Coastal Hazards and Stochastic Processes Group, Coastal and Hydraulics Laboratory, U.S. Army Engineer Research & Development Center

Title: Probabilistic Coastal Hazard Assessment and Uncertainty Quantification

Abstract: The methodology adopted by the U.S. Army Engineer Research & Development Center (ERDC) for estimating coastal hazards in areas susceptible to the effects of tropical or extratropical cyclones is termed Probabilistic Coastal Hazard Assessment (PCHA). The PCHA framework relies on probabilistic models for the estimation of storm-induced hazards, such as storm surge, waves and wind, along with their corresponding annual exceedance probabilities. StormSim (Storm Simulation System), developed by the Coastal and Hydraulics Laboratory (CHL) is the primary tool used for the joint probability analysis of storm forcing and responses. In the case of tropical cyclones, StormSim employs standard JPM-parameterization (i.e., reference location, heading direction, central pressure deficit, radius of maximum winds, translational speed, Holland B) and high-resolution meteorological and hydrodynamic modeling for the simulation of storm responses. To maximize coverage of the multivariate probability space, a Gaussian process metamodel (StormSim-GPM) is used to augment the number of storms in the synthetic tropical cyclone suite to the order of 10⁵ to 10⁶.

PCHA results are deployed via the Coastal Hazards System (CHS) website. CHS is unique in that it is currently the only database providing coastal storm hazard results on a national scale. Uncertainty quantification is an integral part of the PCHA framework. Epistemic uncertainty, in particular, can be characterized by the application of different, technically defensible, data, methods, and models relevant to hazard assessment, resulting in families of hazard curves. The quantification of epistemic uncertainty associated with the computation of storm recurrence rate, definition of univariate distributions and joint probability of storm parameters, development of the synthetic storms, and storm suite augmentation methods can be achieved through numerical experimentation. Findings from uncertainty quantification experiments are used to inform and improve the PCHA methodologies. Ongoing CHL efforts include the update of PCHA results for North Atlantic Coast Comprehensive Study (Virginia to Maine), the Coastal Texas Study, and the South Atlantic Coast Comprehensive Study Phase I (North Carolina to South Florida + Caribbean) and Phase II (South Florida to Mississippi).