Predicting typhoon induced storm surges using the operational ocean forecast system

Abstract

This study was performed to compare storm surges simulated by the operational storm surges/tide forecast system (STORM : Storm surges/Tide Operational Model) of the Korea Meteorological Administration (KMA) with observations from 30 coastal tidal stations during nine typhoons that occurred between 2005 and 2007. The results (bias) showed that for cases of overestimation (or underestimation), storm surges tended to be overestimated (as well as underestimated) at all coastal stations. The maximum positive bias was approximately 6.92 cm for Typhoon Ewiniar (2006), while the maximum negative bias was approximately -12.06 cm for Typhoon Khanun (2005). The maximum and minimum root mean square errors (RMSEs) were 14.61 and 6.78 cm, which occurred for Typhoons Khanun (2005) and Usagi (2007), respectively. For all nine typhoons, total averaged RMSE was approximately 10.2 cm. Large differences between modeled and observed storm surges occurred in two cases. In the first, a very weak typhoon, such as Typhoon Khanun (2005), caused low storm surges. In the other, exemplified by Typhoon Nari (2007), there were errors in the predicted typhoon strength used as input data for the storm surge model.