Storm surge is one of the most severe natural hazards that can lead to significant flooding in coastal areas and severe damages to the life and property from a hurricane. Since post-Katrina coastal flood hazard studies, a key technique called Joint Probability Method (JPM) has become the gold standard to compute annual exceedance probabilities (AEP) levels at certain frequencies by federal agencies such as Federal Emergency Management Agency, private sectors, and academic researchers in coastal engineering. However, the JPM suffers several disadvantages including excessive usage of computing resources, inappropriate uncertainty quantification, and lack of optimal statistical modelling, which make the JPM based coastal flood hazard studies prohibitively costly and unrealistic. To address these issues, we employ a new risk assessment framework to assess storm surges hazards in Southwest Florida (SWFL), where we develop an emulator - a fast approximation to the surge prediction model, based on which, we use an efficient Monte Carlo sampling technique to enable fast computation of AEP. Our methodology provides rigorous uncertainty quantification for risk assessment of storm surges.