Storm surge, a type of coastal flooding caused by tropical cyclones and a main component of extreme sea level change, is a major hazard for many population centers, especially highlighted by recent disasters such as Typhoon Haiyan (2013) and Hurricane Sandy (2012). Furthermore, the impacts of climate change on storm frequency, intensity, and trajectory make storm surge an important uncertainty in long-term hazard planning. Numerical simulation of storm surge can be highly demanding due to large computational domains and the need for fine resolution along coastlines, but a recent approach incorporates adaptive mesh refinement (AMR) to reduce computational cost, allowing for dynamic local refinement to track features of interest while less active regions can be coarsely resolved. One such numerical model is GeoClaw, a finite volume solver for the nonlinear shallow water equations with AMR incorporating elevation data and atmospheric forcing. In this talk, I'll discuss an overview of storm surge modeling followed by results for the 2017 Atlantic hurricane season. Finally, I'll present results of an impact assessment of climate change on coastal hazards in the Northwest Pacific Ocean, based on ensemble atmospheric downscaling experiments for Haiyan-like super storms in the Philippines.