Abstract
We present two applications of Polynomial Chaos methods to quantify the uncertainty in the oceanic response to hurricane forcing stemming from parametric uncertainties in the mixed-layer and the wind drag parameterizations. The first study centered on hurricane Ivan 2004 in the Gulf of Mexico and included mixed-layer and wind-drag uncertainties. Sensitivity analyses revealed that the uncertainty in sea surface temperature was dominated by uncertainties in the wind drag coefficient. The second study leveraged the Fanapi 2010 ITOP data in the Pacific to infer, via a Bayesian formalism, wind drag parameters that are poorly known, namely: the peak wind-drag coefficient and the wind speed at which it occurs, and the slope of the wind-drag coefficient after the peak. The efficiency of our approach is enhanced by the use of adaptive polynomial expansions to build an inexpensive surrogate for the high-resolution numerical model coupling the simulated winds and the sea surface temperature, and by the surrogate which is relied upon to dramatically reduce the computational burden of the Markov Chain Monte-Carlo sampling. Our results indicate that the most likely values for the drag coefficient saturation and the corresponding wind speed are about 2.3$\times10^{ -3}$ and 34~m/s, respectively; the data was not informative regarding the drag coefficient behavior at higher wind speeds.
Biography
Dr. Iskandarani's educational background include a BE in 1984 from the American University of Beirut where he studied civil engineering, and MSc (1987) and PhD (1991) in Civil Engineering from Cornell University. Upon graduation he became an Assistant Research Professor at Rutgers University where he worked on the Spectral Element Ocean Model. In 2000 Dr. Iskandarani joined the faculty of the University of Miami where he spends his time teaching and advancing the causes of Ocean Modeling and Uncertainty Quantification.