Upgrades And Application Of Storm Surge Forecasting System

China is one of the countries suffering from typhoons with high-frequency impacts and severe disasters. Storm-induced disaster ranks first in marine disasters in China. Southeast China coasts with large population and increasing economics are frequently attacked by tropical and sub-tropical cyclones during summertime, which caused severe damages. Therefore, the studies of the numerical storm-surge forecasting are of great importance for the security and economy.This thesis focuses on the major upgrading of an integrated storm surge forecasting system for the Changjiang Estuary and adjacent waters with the improvement of the forecasting efficiency. The main three improvements are listed as follows:1) invoking SWAN under OpenMP-supported parallel framework; 2) optimizing the source code for hydrodynamic model; 3) designing a new software interface. The comparison between the original and upgraded version is given with five storm-surge hind-castings, with an aim to show the system improvement on forecasting efficiency and precision. It indicates the model precision of upgraded version is guaranteed. The efficiency is quantitatively increased by 70% and 60% with consideration and neglecting of wave effect, respectively. Total forecasting time with consideration of wave effect could be limited less than 45 minutes. Thus it solves the original problem that wave effect was not considered in fast storm surge forecasting due to its poor efficiency, although the original storm surge forecasting system had included the wave interaction module.A 2-D storm surge forecasting model system is preliminary set up for the Zhejiang and Fujian provinces which significantly suffer from the storm-induced disasters. The verification of astronomical tide and wave processes shows this model system is robust to reveal the major oceanic and coastal dynamics structures. Six storm surges in recent years were simulated and analyzed with the hind-casting precision guaranteed. The results show that the revised coefficient of Typhoon in Zhejiang and Fujian should be significantly smaller than the empirical values in Changjiang Estuary and Hangzhou Bay. In the cases of this thesis, cl ranges 0.6～0.7 and c2 equals 0.7. These parameter configurations have best modeling performance around the Zhejiang and Fujian coasts. The hind-casting results agree well with observational data.The works in this thesis laid a solid ground for future research and forecasting of the storm impacts at the Zhejiang and Fujian regions.