Parameterization and Application of Storm Surge/Tide Modeling Using a Genetic Algorithm for Typhoon Periods

A genetic algorithm was used to optimize the parameters of thetwo-dimensional Storm Surge/Tide Operational Model (STORM) to improve sealevel predictions. The genetic algorithm was applied to nine typhoons thataffected the Korean Peninsula during 2005--2007. The following modelparameters were used: the bottom drag coefficient, the background horizontaldiffusivity, Smagorinski's horizontal viscosity, and the sea level pressurescaling. Generally, the simulation results using the optimized, mean, andmedian parameter values improved sea level predictions. The four estimatedparameters improved the sea level prediction by 76% and 54% in thebias and root mean square error for Typhoon Kalmaegi (0807) in 2008,respectively. One-month simulations of February and August 2008 were alsoimproved using the estimated parameters. This study demonstrates thatparameter optimization on STORM can improve sea level prediction.